UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Hearing aid processing of auditory evoked potential stimuli : acoustic effects Huen, Myron 2016

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata

Download

Media
24-ubc_2016_may_huen_myron.pdf [ 15.2MB ]
Metadata
JSON: 24-1.0229467.json
JSON-LD: 24-1.0229467-ld.json
RDF/XML (Pretty): 24-1.0229467-rdf.xml
RDF/JSON: 24-1.0229467-rdf.json
Turtle: 24-1.0229467-turtle.txt
N-Triples: 24-1.0229467-rdf-ntriples.txt
Original Record: 24-1.0229467-source.json
Full Text
24-1.0229467-fulltext.txt
Citation
24-1.0229467.ris

Full Text

i     HEARING AID PROCESSING OF AUDITORY EVOKED POTENTIAL STIMULI: ACOUSTIC EFFECTS by MYRON HUEN B.Sc., The University of British Columbia, May 2013 A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF SCIENCE in The Faculty of Graduate and Postdoctoral Studies (Audiology and Speech Sciences) THE UNIVERSITY OF BRITISH COLUMBIA (Vancouver) April 2016 ©Myron Huen, 2016    ii  Abstract  Amplification is one of the main interventions used to manage hearing loss in people of all ages. Difficult-to-test populations fitted with amplification may be unable to respond reliably through behavioural methods to evaluate the quality of amplification. In these situations, the use of objective measures, such as auditory evoked potentials (AEP), has been suggested as a means to obtain more information regarding the fitting of the amplification.  Recent investigations on aided potentials have been inconsistent in showing more robust responses that would be expected after amplification. This may be in part due to the unpredictable changes that can happen with hearing aid processing. Thus, there is a need to conduct a systematic analysis of the acoustic effects hearing aid processing can have on AEP stimuli to offer possible explanations for the inconsistency. Data were collected using three hearing aids programmed for a mild to moderately-severe sloping hearing loss with linear gain and compression, and stimuli commonly used to elicit brainstem and cortical AEPs at two input intensities. Large changes were noted in stimulus rise times, intensities, signal-to-noise ratios, and modulation depth. The amount of change differed by hearing aid and gain settings (linear or compression). Changes to rise times were also noted with changes to stimulus duration based on the hearing aid as well as the gain setting. These results suggest hearing aid processing can cause many changes to the signal, some of which may affect the morphology of evoked potentials. The changes that occur can also vary widely based on the hearing aid through which the stimuli are processed. Thus, the interpretation of the responses elicited by processed stimuli must be done with caution, and should only be done after taking hearing aid changes into account.  iii  Preface   This thesis was based on research conducted in the Amplification Research Lab at the University of British Columbia. The methodology used in this study was created through discussions with L. Jenstad and A. Herdman based on typical clinical and research protocols with aided auditory evoked potentials. The author of this thesis, M. Huen, was responsible for set-up, data collection, data analysis, and interpretation.                   iv  Table of Contents Abstract ..................................................................................................................................... ii Preface ..................................................................................................................................... iii Table of Contents ..................................................................................................................... iv List of Tables ........................................................................................................................... ix List of Figures ......................................................................................................................... xv List of Abbreviations ............................................................................................................ xxv Acknowledgements............................................................................................................. xxvii Dedication .......................................................................................................................... xxviii Chapter 1: Literature Review.................................................................................................... 1 1.1 Introduction ......................................................................................................................1 1.2 Overview of existing applications of auditory evoked potentials ....................................3 1.2.1 Applications of auditory evoked potentials in amplification ................................... 4 1.3 Hearing aid processing .....................................................................................................7 1.4 Morphological changes of electrophysiological measures from changes in stimulus parameters ............................................................................................................................10 1.5 Inconsistent findings with aided cortical AEPs .............................................................18 Chapter 2: Methods................................................................................................................. 23 2.1 Hearing aids ...................................................................................................................23 2.2 Materials and equipment ................................................................................................23 2.2.1 Recording system ................................................................................................... 23 2.2.2 Stimuli .................................................................................................................... 25 v  2.3 Calibration procedures ...................................................................................................27 2.3.1 Sound field calibration ........................................................................................... 27 2.3.2 Stimulus Calibration ............................................................................................... 28 2.4 Experimental procedures ...............................................................................................29 2.4.1 Hearing aid programming ...................................................................................... 29 2.4.2. Stimulus presentation ............................................................................................ 30 2.5 Acoustic analysis ...........................................................................................................31 2.5.1 Initial treatment of stimuli ...................................................................................... 31 2.5.2 Measurement of acoustic parameters ..................................................................... 32 Chapter 3: Results and Interpretations .................................................................................... 38 3.1 Data collection summary ...............................................................................................38 3.2 Hearing aid processing delays........................................................................................38 3.3 Rise and fall time ...........................................................................................................38 3.3.1 SCP Tones .............................................................................................................. 38 3.3.2 SCP Noise............................................................................................................... 48 3.3.3 ACC tonal stimuli ................................................................................................... 57 3.3.4 SCP Speech and ACC Speech ................................................................................ 58 3.3.5 HEARLab™ stimuli ............................................................................................... 66 3.3.6 ABR Tone bursts and click .................................................................................... 69 3.3.7 MLR Tone bursts and click .................................................................................... 74 3.3.8 MMN Speech ......................................................................................................... 79 vi  3.4 Duration .........................................................................................................................80 3.4.1 SCP Tones .............................................................................................................. 80 3.4.2 SCP Noise............................................................................................................... 84 3.4.3 ACC Tonal stimuli ................................................................................................. 86 3.4.4 SCP Speech and ACC Speech ................................................................................ 89 3.4.5 HEARLab™ stimuli ............................................................................................... 91 3.4.6 ABR Tone bursts and click .................................................................................... 94 3.4.7 MLR Tone bursts and click .................................................................................... 97 3.4.8 MMN Speech ....................................................................................................... 100 3.4.9 ASSR 40 Hz and 80 Hz AM Tones ...................................................................... 101 3.5 Intensity ........................................................................................................................101 3.5.1 SCP Tones ............................................................................................................ 101 3.5.2 SCP Noise............................................................................................................. 108 3.5.3 ACC Tonal stimuli ............................................................................................... 112 3.5.4 SCP Speech and ACC Speech .............................................................................. 115 3.5.5 HEARLab™ Stimuli ............................................................................................ 118 3.5.6 ABR Tone bursts and click .................................................................................. 121 3.5.7 MLR Tone bursts and click .................................................................................. 128 3.5.8 MMN Speech ....................................................................................................... 135 3.5.9 ASSR 40 and 80 Hz Amplitude Modulated (AM) Tones .................................... 137 vii  3.6 SNR ..............................................................................................................................141 3.6.1 SCP Tones ............................................................................................................ 141 3.6.2 SCP Noise............................................................................................................. 147 3.6.3 ACC Tonal stimuli ............................................................................................... 150 3.6.4 SCP Speech and ACC Speech .............................................................................. 152 3.6.5 HEARLab™ Stimuli ............................................................................................ 155 3.6.6 ABR Tone bursts and click .................................................................................. 159 3.6.7 MLR Tone bursts and click .................................................................................. 162 3.6.8 MMN Speech ....................................................................................................... 165 3.6.9 ASSR 40 and 80 Hz AM Tones ........................................................................... 167 3.7 Spectra ..........................................................................................................................170 3.7.1 SCP Tones ............................................................................................................ 170 3.7.2 SCP Noise............................................................................................................. 184 3.7.3 ACC Tonal stimuli ............................................................................................... 192 3.7.4 SCP Speech and ACC Speech .............................................................................. 196 3.7.5 HEARLab™ Stimuli ............................................................................................ 206 3.7.6 ABR Tone bursts and click .................................................................................. 215 3.7.7 MLR Tone bursts and click .................................................................................. 218 3.7.8 MMN Speech ....................................................................................................... 223 3.7.9 ASSR 40 and 80 Hz AM Tones ........................................................................... 226 viii  3.8 Difference between Formant 2 and Formant 1 frequencies .........................................229 3.8.1 SCP Speech and ACC Speech - /a/ and /i/ ........................................................... 229 3.8.2 MMN Speech ....................................................................................................... 231 3.9 VOT .............................................................................................................................234 3.9.1 SCP Speech - /da/ and /ta/ .................................................................................... 234 3.9.2 MMN Speech - /ba/ and /da/ ................................................................................ 236 3.10 Fricative Duration ......................................................................................................238 3.10.1 SCP Speech - /sa/ and /ʃa/ .................................................................................. 238 3.11 Modulation Depth ......................................................................................................240 3.11.1 ASSR 40 and 80 Hz AM Tones ......................................................................... 240 3.12 F-ratio .........................................................................................................................243 3.12.1 ASSR 40 and 80 Hz AM Tones ......................................................................... 243 Chapter 4: General Discussion and Conclusions .................................................................. 246 References............................................................................................................................. 251 Appendices ........................................................................................................................... 273 Appendix A: Hearing Aid Settings ....................................................................................273 Starkey Z series mini BTE i70 ...................................................................................... 273 Siemens Motion SX Micon 5mi .................................................................................... 276 Phonak Bolero Q90 ....................................................................................................... 280 Appendix B: Averaged waveform and spectra of all recorded stimuli ..............................284 Appendix C: Ear canal resonances of KEMAR .................................................................286  ix   List of Tables Table 1.1. Effects of simple stimulus parameter changes on AEP morphology .................... 17 Table 1.2. Effects of speech stimuli parameter changes on SCP morphology. ..................... 18 Table 2.1. Description of stimuli used ................................................................................... 27 Table 2.2. Measured compression thresholds, attack and release times using 250 Hz to 4000 Hz input stimuli for all hearing aids ....................................................................................... 30 Table 2.3. Amplification applied to each set of stimuli for visualization and analysis ......... 32 Table 2.4. Praat settings for optimized formant measurements ............................................. 36 Table 3.1. Hearing aid processing delays measured on the Fonix 7000 ................................ 38 Table 3.2. Measured rise times, fall times, and transition rise times, ms, in all hearing aid and intensity conditions. ................................................................................................................ 57 Table 3.3. Measured rise times, ms, of speech sounds in the unaided condition and the differences calculated between aided and unaided conditions, with speech sounds presented at 65 dB SPL and 45 dB SPL.................................................................................................. 60 Table 3.4. Measured fall times, ms, of speech sounds in the unaided condition and the differences calculated between aided and unaided conditions, with speech sounds presented at 65 dB SPL and 45 dB SPL.................................................................................................. 61 Table 3.5. Measured rise times, ms, of HEARLab™ speech sounds in the unaided condition and the differences calculated between aided and unaided conditions, with speech sounds presented at 65 dB SPL and 45 dB SPL. ................................................................................ 67 x  Table 3.6. Measured fall times, ms, of HEARLab™ speech sounds in the unaided condition and the differences calculated between aided and unaided conditions, with speech sounds presented at 65 dB SPL and 45 dB SPL. ................................................................................ 67 Table 3.7. Measured rise times, cycles, of 2-1-2 tone bursts of each frequency presented in all hearing aid and intensity conditions. ................................................................................. 71 Table 3.8. Measured fall times, cycles, of 2-1-2 tone bursts of each frequency presented in all hearing aid and intensity conditions. ................................................................................. 71 Table 3.9. Measured rise times, ms, of rarefaction clicks presented in all hearing aid and intensity conditions. ................................................................................................................ 72 Table 3.10. Measured fall times, ms, of rarefaction clicks presented in all hearing aid and intensity conditions. ................................................................................................................ 72 Table 3.11. Measured rise times, cycles, of 2-1-2 tone bursts of each frequency presented in all hearing aid and intensity conditions. ................................................................................. 75 Table 3.12. Measured rise times, ms, of 4-2-4 ms tone bursts of each frequency and the rarefaction click presented in all hearing aid and intensity conditions. .................................. 75 Table 3.13. Measured fall times, cycles, of 2-1-2 tone bursts of each frequency presented in all hearing aid and intensity conditions .................................................................................. 76 Table 3.14. Measured fall times, ms, of 4-2-4 ms tone bursts of each frequency and the rarefaction click presented in all hearing aid and intensity conditions. .................................. 76 Table 3.15. Difference between the rise time, ms, measured from /ba/ and the rise time, ms, measured from /da/ in all hearing aid and intensity conditions. ............................................. 79 Table 3.16. Difference between the fall time, ms, measured from /ba/ and the fall time, ms, measured from /da/ in all hearing aid and intensity conditions. ............................................. 79 xi  Table 3.17. Trend line equations fit to recorded 500 Hz tone durations as a function of stimulus durations. .................................................................................................................. 81 Table 3.18. Trend line equations fit to recorded 1000 Hz tone durations as a function of stimulus durations ................................................................................................................... 81 Table 3.19. Trend line equations fit to recorded 2000 Hz tone durations as a function of stimulus durations ................................................................................................................... 82 Table 3.20. Trend line equations fit to noise burst durations as a function of generated durations. ................................................................................................................................ 85 Table 3.21. Measured durations, ms, for ACC tonal stimuli in all hearing aid and intensity conditions ................................................................................................................................ 86 Table 3.22. Duration difference, ms, between /ba/ and /da/ in all hearing aid and intensity conditions. ............................................................................................................................. 101 Table 3.23. Gain relative to unaided, dB, in all hearing aid and intensity conditions for 500 Hz tones. RF = rise and fall time. ......................................................................................... 102 Table 3.24. Gain relative to unaided, dB, in all hearing aid and intensity conditions for 1000 Hz tones. RF = rise and fall time. ......................................................................................... 103 Table 3.25. Gain relative to unaided, dB, in all hearing aid and intensity conditions for 2000 Hz tones. RF = rise and fall. ................................................................................................. 104 Table 3.26. CORFIG values for KEMAR............................................................................ 106 Table 3.27. Gain relative to unaided, dB, in all hearing aid and intensity conditions for noise bursts. RF = rise and fall time. .............................................................................................. 109 Table 3.28. Gain relative to unaided, dB, of ACC tonal stimuli in all hearing aid and intensity conditions. .............................................................................................................. 113 xii  Table 3.29. Measured difference between intensities of the 1705 Hz segment and the 1680 Hz segment, dB, in all hearing aid and intensity conditions. ............................................... 113 Table 3.30. Gain relative to unaided, dB, for speech stimuli in all aided conditions and both intensity conditions. .............................................................................................................. 116 Table 3.31. Gain relative to unaided, dB, for HEARLab™ speech sounds in all aided conditions and both intensity conditions. ............................................................................. 119 Table 3.32. Gain relative to unaided, dB, for 2-1-2 tone bursts of all presented frequencies and rarefaction clicks in all aided conditions and both intensity conditions. ....................... 123 Table 3.33. Intensity difference between the first presentation and third repetition, dB, of the full recording of 2-1-2 tone bursts of each frequency and rarefaction clicks in all hearing aid conditions and both intensity conditions. ............................................................................. 123 Table 3.34. Intensity difference between the first and six seconds, dB, of the full recording of 2-1-2 tone bursts of each frequency and rarefaction clicks in all hearing aid conditions and both intensity conditions. ...................................................................................................... 124 Table 3.35. Gain relative to unaided, dB, for all 2-1-2 and 4-2-4 ms tone bursts and rarefaction clicks in all aided conditions and both intensity conditions. .............................. 130 Table 3.36. Intensity difference between the first presentation and third repetition, dB, of the full recording of all 2-1-2 and 4-2-4 ms tone bursts and rarefaction clicks in all hearing aid conditions and both intensity conditions. ............................................................................. 131 Table 3.37. Intensity difference between the first and six seconds, dB, of the full recording of all 2-1-2 and 4-2-4 ms tone bursts and rarefaction clicks in all hearing aid conditions and both intensity conditions. ...................................................................................................... 132 xiii  Table 3.38. Gain relative to unaided, dB, of the syllables /ba/ and /da/ for all aided conditions in both intensity condition. ................................................................................. 136 Table 3.39. Intensity difference between /ba/ and /da/, dB, in all hearing aid and intensity conditions. ............................................................................................................................. 136 Table 3.40. Gain relative to unaided, dB, for all AM tones in all aided conditions and both intensity conditions. .............................................................................................................. 138 Table 3.41. Intensity difference between the first seven seconds and the rest of the recording, dB, for all AM tones in all hearing aid and intensity conditions. ......................................... 138 Table 3.42. SNR, dB, of 500 Hz tones in all hearing aid and intensity conditions. RF = rise time. ...................................................................................................................................... 142 Table 3.43.43SNR, dB, of 1000 Hz tones in all hearing aid and intensity conditions. RF = rise and fall time. .................................................................................................................. 143 Table 3.44. SNR, dB, of 2000 Hz tones in all hearing aid and intensity conditions. RF = rise and fall time. ......................................................................................................................... 144 Table 3.45. SNR, dB, of noise bursts in all hearing aid and intensity conditions. .............. 148 Table 3.46. SNR, dB, of ACC tonal stimuli in all hearing aid and intensity conditions. .... 151 Table 3.47. SNR, dB, of all speech stimuli measured in all hearing aid and intensity conditions. ............................................................................................................................. 153 Table 3.48. SNR, dB, of all HEARLab™ speech sounds measured in all hearing aid and intensity conditions. .............................................................................................................. 156 Table 3.49. SNRs, dB, of 2-1-2 tone-bursts and rarefaction clicks recorded in all hearing aid and intensity conditions. ....................................................................................................... 160 xiv  Table 3.50. SNRs, dB, of 2-1-2 and 4-2-4 ms tone bursts as well as clicks recorded in all hearing aid and intensity conditions. .................................................................................... 163 Table 3.51. SNRs, dB, of /ba/ and /da/ recorded in all hearing aid and intensity conditions. .............................................................................................................................................. 166 Table 3.52. SNRs, dB, of 40 Hz and 80 Hz AM tones recorded in all hearing aid and intensity conditions. .............................................................................................................. 168 Table 3.53. Differences between Formant 1 and Formant 2 in /a/ in /ba/ and /da/ recorded in all hearing aid and intensity conditions. ............................................................................... 233 Table 3.54. VOTs measured for all consonant-vowel syllables containing a stop consonant for all hearing aid and intensity conditions .......................................................................... 235 Table 3.55. VOTs measured for all syllables used in the odd-ball paradigm for all hearing aid and intensity conditions. ................................................................................................. 237 Table 3.56. Fricative durations measured for all consonant-vowel syllables containing a fricative for all hearing aid and intensity conditions ............................................................ 239 Table 3.57. Modulation depths measured and calculated from the envelopes of the AM modulated tones in all hearing aid and intensity conditions. ................................................ 241 Table 3.58. F-ratios measured for all AM tones in all hearing aid and intensity conditions. .............................................................................................................................................. 244 Table C.1. Ear canal resonances of KEMAR ...................................................................... 286     xv    List of Figures Figure 2.1. Diagram of system used in simultaneous recording and presentation of stimuli. 24 Figure 3.1. Rise times, ms, in unaided and aided conditions for all tones of different durations and stimulus rise times presented at 65 dB SPL; linear gain is in the top panels, and compression is in the bottom panels. Rise times were plotted as a function of duration; logarithmic trend lines were plotted for each data series. The grey dashed horizontal line at 30 ms indicates the point above which rise time is expected to affect SCP amplitude and latency. .................................................................................................................................... 40 Figure 3.2. Rise times, ms, in unaided and aided conditions for all tones of different durations and stimulus rise times presented at 45 dB SPL; linear gain is in the top panels, and compression is in the bottom panels. Rise times were plotted as a function of duration; logarithmic trend lines were plotted for each data series. The grey dashed horizontal line at 30 ms indicates the point above which rise time is expected to affect SCP amplitude and latency. .................................................................................................................................... 41 Figure 3.3. Fall times, ms, in unaided and aided conditions for all tones of different durations and stimulus fall times presented at 65 dB SPL; linear gain is in the top panels, and compression is in the bottom panels. Fall times were plotted as a function of duration; logarithmic trend lines were plotted for each data series. ...................................................... 42 Figure 3.4. Fall times, ms, in unaided and aided conditions for all tones of different durations and stimulus rise times presented at 45 dB SPL; linear gain is in the top panels, and xvi  compression is in the bottom panels. Fall times were plotted as a function of duration; logarithmic trend lines were plotted for each data series. ...................................................... 43 Figure 3.5. Waveforms of recorded stimuli of 60 ms duration processed by Starkey (compression). The stimulus waveform is in blue and the envelope of the stimulus is in black. Different panels show recordings with different stimulus rise and fall times: (A) 5 ms; (B) 20 ms. Arrowheads denote points between which fall times were measured. ................. 44 Figure 3.6. Additional fall times, ms, of stimuli presented at 65 dB SPL with compression active in hearing aids. ............................................................................................................. 45 Figure 3.7. Additional fall times, ms, of stimuli presented at 45 dB SPL with compression active in hearing aids. ............................................................................................................. 46 Figure 3.8. Rise times, ms, of noise bursts of different durations and stimulus rise times, presented at 65 dB SPL in unaided and aided conditions. The grey dashed horizontal line at 30 ms indicates the point above which rise time is expected to affect SCP amplitude and latency. .................................................................................................................................... 50 Figure 3.9. Rise times, ms, of noise bursts of different durations presented at 45 dB SPL in unaided and aided conditions. The grey dashed horizontal line at 30 ms indicates the point above which rise time is expected to affect SCP amplitude and latency................................ 51 Figure 3.10. Fall times, ms, of noise bursts of different durations presented at 65 dB SPL in unaided and aided conditions. ................................................................................................ 52 Figure 3.11. Fall times, ms, of noise bursts of different durations presented at 45 dB SPL in unaided and aided conditions. ................................................................................................ 53 Figure 3.12. Additional fall times, ms, of noise bursts of different durations presented at 65 dB SPL from Starkey (compression). ..................................................................................... 54 xvii  Figure 3.13. Averaged stimulus waveforms of /sa/ in the 65 dB SPL condition. Different panels show different hearing aid conditions: (A) Siemens (compression); (B) Siemens (linear); (C) Unaided............................................................................................................... 65 Figure 3.14. Measured durations, ms, of 500 Hz tones plotted as a function of stimulus durations in all hearing aid conditions, with stimuli presented at 65 dB SPL. ....................... 83 Figure 3.15. Stimulus waveforms of ACC tonal stimuli in the 65 dB SPL condition. The pre-transition segment was 1680 Hz and 300 ms in duration; the post-transition segment was 1705 Hz and 300 ms in duration; thus, the total stimulus duration was 600 ms. Different panels show different hearing aid conditions: (A) Siemens (linear); (B) Phonak (linear); (C) Starkey (linear); (D) Unaided. Transition points, denoted by the arrow heads, are all in the 13 230 sample (i.e., 300 ms) region............................................................................................. 88 Figure 3.16. Duration, ms, of all syllables and phonemes recorded in all hearing aid and intensity conditions. ................................................................................................................ 90 Figure 3.17. Duration, ms, of all HEARLab™ phonemes recorded in all hearing aid and intensity conditions. ................................................................................................................ 92 Figure 3.18. Duration, ms, of the most intense segment of the ABR tone bursts and rarefaction click for all hearing aid and intensity conditions. ................................................ 95 Figure 3.19. Duration, ms, of the most intense segment of the MLR tone bursts and rarefaction click for in all hearing aid and intensity conditions. ............................................ 98 Figure 3.20. Duration, ms, of the full MLR tone bursts and rarefaction click, including reverberation effects, in all hearing aid and intensity conditions. .......................................... 99 xviii  Figure 3.21. Waveform of the full recording of 4000 Hz in Unaided (A), Phonak (linear) (B), and  Starkey (linear) (C) in the 85 ppe SPL condition. Onset effect, observed only in Phonak, lasts approximately 6 seconds. ............................................................................... 125 Figure 3.22. Waveform of the full recording of 4000 Hz 2-1-2 tone burst in Unaided (A), and Phonak (linear) (B) in the 85 ppe SPL condition. Onset effect is not observed with Phonak in this set of stimuli. ................................................................................................ 133 Figure 3.23. Waveform of the 40 Hz AM tone (A), and 80 Hz AM tone (B) processed by Phonak (linear) in the 65 dB SPL condition. The black outline is the envelope of the recording. .............................................................................................................................. 139 Figure 3.24. FFT of 500 Hz tone bursts of 30 ms duration with stimulus rise and fall times of 5 ms, presented at 65 dB SPL. Different panels show different hearing aid conditions: (A) Unaided; (B) Starkey Linear; (C) Siemens Linear; (D) Phonak Linear. .............................. 171 Figure 3.25. FFT of 500 Hz tone bursts of 30 ms duration with stimulus rise and fall times of 5 ms, presented at 65 dB SPL. Different panels show different hearing aid conditions: (A) Unaided; (B) Starkey Compression; (C) Siemens Compression; (D) Phonak Compression. .............................................................................................................................................. 172 Figure 3.26. FFT of 500 Hz tone bursts of 480 ms duration with stimulus rise and fall times of 5 ms, presented at 65 dB SPL. Different panels show different hearing aid conditions: (A) Unaided; (B) Starkey Linear; (C) Siemens Linear; (D) Phonak Linear. .............................. 173 Figure 3.27. FFT of 500 Hz tone bursts of 480 ms duration with stimulus rise and fall times of 5 ms, presented at 65 dB SPL. Different panels show different hearing aid conditions: (A) Unaided; (B) Starkey Compression; (C) Siemens Compression; (D) Phonak Compression. .............................................................................................................................................. 174 xix  Figure 3.28.  FFT of 1000 Hz tone bursts of 30 ms duration with stimulus rise and fall times of 5 ms, presented at 65 dB SPL. Different panels show different hearing aid conditions: (A) Unaided; (B) Starkey Linear; (C) Siemens Linear; (D) Phonak Linear. .............................. 175 Figure 3.29. FFT of 1000 Hz tone bursts of 30 ms duration with stimulus rise and fall times of 5 ms, presented at 65 dB SPL. Different panels show different hearing aid conditions: (A) Unaided; (B) Starkey Compression; (C) Siemens Compression; (D) Phonak Compression. .............................................................................................................................................. 176 Figure 3.30. FFT of 1000 Hz tone bursts of 480 ms duration with stimulus rise and fall times of 5 ms, presented at 65 dB SPL. Different panels show different hearing aid conditions: (A) Unaided; (B) Starkey Linear; (C) Siemens Linear; (D) Phonak Linear. .............................. 177 Figure 3.31. FFT of 1000 Hz tone bursts of 480 ms duration with stimulus rise and fall times of 5 ms, presented at 65 dB SPL. Different panels show different hearing aid conditions: (A) Unaided; (B) Starkey Compression; (C) Siemens Compression; (D) Phonak Compression. .............................................................................................................................................. 178 Figure 3.32. FFT of 1000 Hz tone bursts of 30 ms duration with stimulus rise and fall times of 5 ms, presented at 45 dB SPL. Different panels show different hearing aid conditions: (A) Unaided; (B) Starkey Linear; (C) Siemens Linear; (D) Phonak Linear. .............................. 179 Figure 3.33. FFT of 1000 Hz tone bursts of 30 ms duration with stimulus rise and fall times of 5 ms, presented at 45 dB SPL. Different panels show different hearing aid conditions: (A) Unaided; (B) Starkey Compression; (C) Siemens Compression; (D) Phonak Compression. .............................................................................................................................................. 180 xx  Figure 3.34. FFT of 1000 Hz tone bursts of 60 ms duration with 5 ms (A), 10 ms (B), and 20 ms (C) rise and fall time presented at 65 dB SPL recorded after being processed by Starkey (Linear). Notice the change in width of the largest peak with an increase in rise time. ....... 181 Figure 3.35. FFT of noise bursts of 30 ms duration with stimulus rise and fall times of 5 ms, presented at 65 dB SPL. Different panels show different hearing aid conditions: (A) Unaided; (B) Starkey Linear; (C) Siemens Linear; (D) Phonak Linear. .............................................. 185 Figure 3.36. FFT of noise bursts of 30 ms duration with stimulus rise and fall times of 5 ms, presented at 65 dB SPL. Different panels show different hearing aid conditions: (A) Unaided; (B) Starkey Compression; (C) Siemens Compression; (D) Phonak Compression. .............. 186 Figure 3.37. FFT of noise bursts of 480 ms duration with stimulus rise and fall times of 5 ms, presented at 65 dB SPL. Different panels show different hearing aid conditions: (A) Unaided; (B) Starkey Linear; (C) Siemens Linear; (D) Phonak Linear. .............................. 187 Figure 3.38. FFT of noise bursts of 480 ms duration with stimulus rise and fall times of 5 ms, presented at 65 dB SPL. Different panels show different hearing aid conditions: (A) Unaided; (B) Starkey Compression; (C) Siemens Compression; (D) Phonak Compression. .............................................................................................................................................. 188 Figure 3.39. FFT of noise bursts of 30 ms duration with stimulus rise and fall times of 5 ms, presented at 45 dB SPL. Different panels show different hearing aid conditions: (A) Unaided; (B) Starkey Linear; (C) Siemens Linear; (D) Phonak Linear. .............................................. 189 Figure 3.40. FFT of noise bursts of 30 ms duration with stimulus rise and fall times of 5 ms, presented at 45 dB SPL. Different panels show different hearing aid conditions: (A) Unaided; (B) Starkey Compression; (C) Siemens Compression; (D) Phonak Compression. .............. 190 xxi  Figure 3.41. FFT of noise bursts of 60 ms duration with 5 ms (A), 10 ms (B), and 20 ms (C) rise and fall time presented at 65 dB SPL recorded after being processed by Starkey (linear). .............................................................................................................................................. 191 Figure 3.42. FFT of ACC tonal stimuli – a 1680 Hz tone transitioning to a 1705 Hz tone, each 300 ms long – presented at 65 dB SPL. Different panels show different hearing aid conditions: (A) Unaided; (B) Starkey Linear; (C) Siemens Linear; (D) Phonak Linear. ..... 194 Figure 3.43. FFT of ACC tonal stimuli – a 1680 Hz tone transitioning to a 1705 Hz tone, each 300 ms long – presented at 45 dB SPL. Different panels show different hearing aid conditions: (A) Unaided; (B) Starkey Linear; (C) Siemens Linear; (D) Phonak Linear. ..... 195 Figure 3.44. FFT of /a/ cut pre-recording in unaided and aided conditions where hearing aids were set to linear gain, with stimuli presented at 65 dB SPL. Different panels show different hearing aid conditions: (A) Unaided; (B) Starkey Linear; (C) Siemens Linear; (D) Phonak Linear. ................................................................................................................................... 198 Figure 3.45. FFT of /i/ in unaided and aided conditions where hearing aids were set to linear gain, with stimuli presented at 65 dB SPL. Different panels show different hearing aid conditions: A) Unaided; (B) Starkey Linear; (C) Siemens Linear; (D) Phonak Linear. ...... 199 Figure 3.46. FFT of /sa/ presented at 65 dB SPL. Different panels show different hearing aid conditions: (A) Unaided; (B) Starkey Linear; (C) Siemens Linear; (D) Phonak Linear. ..... 200 Figure 3.47. FFT of /sa/ presented at 65 dB SPL. Different panels show different hearing aid conditions: (A) Unaided; (B) Starkey Compression;  (C) Siemens Compression; (D) Phonak Compression. ........................................................................................................................ 201 Figure 3.48. FFT of /da/ presented at 45 dB SPL. Different panels show different hearing aid conditions: (A) Unaided; (B) Starkey Linear;  (C) Siemens Linear; (D) Phonak Linear. .... 202 xxii  Figure 3.49.  FFT of /da/ presented at 45 dB SPL. Different panels show different hearing aid conditions: (A) Unaided; (B) Starkey Compression;  (C) Siemens Compression; (D) Phonak Compression. ........................................................................................................... 203 Figure 3.50. FFT of /a/ cut pre-recording (A), and /a/ cut post-recording (B), recorded from Phonak (compression) with stimuli presented at 65 dB SPL condition. .............................. 204 Figure 3.51. FFT of /m/ (A), /s/ (B), /g/ (C), /t/ (D) presented at 65 dB SPL and recorded in the unaided condition. ........................................................................................................... 207 Figure 3.52. FFT of /m/ (A), /s/ (B), /g/ (C), /t/ (D) presented at 65 dB SPL and recorded from Starkey (Linear). .......................................................................................................... 208 Figure 3.53. FFT of /m/ (A), /s/ (B), /g/ (C), /t/ (D) presented at 65 dB SPL and recorded from Starkey (Compression)................................................................................................. 209 Figure 3.54. FFT of /m/ (A), /s/ (B), /g/ (C), /t/ (D) presented at 65 dB SPL and recorded from Siemens (Linear). ......................................................................................................... 210 Figure 3.55. FFT of /m/ (A), /s/ (B), /g/ (C), /t/ (D) presented at 65 dB SPL and recorded from Siemens (Compression). .............................................................................................. 211 Figure 3.56. FFT of /m/ (A), /s/ (B), /g/ (C), /t/ (D) presented at 65 dB SPL and recorded from Phonak (Linear). .......................................................................................................... 212 Figure 3.57. FFT of /m/ (A), /s/ (B), /g/ (C), /t/ (D) presented at 65 dB SPL and recorded from Phonak (Compression). ................................................................................................ 213 Figure 3.58. FFT of 4000 Hz 2-1-2 tone bursts presented at 85 ppe SPL. Different panels show different hearing aid conditions: (A) Unaided; (B) Starkey Linear; (C) Siemens Linear; (D) Phonak Linear. ............................................................................................................... 216 xxiii  Figure 3.59. FFT 4000 Hz 2-1-2 tone bursts presented at 85 ppe SPL. Different panels show different hearing aid conditions: (A) Unaided;  (B) Starkey Compression; (C) Siemens Compression; (D) Phonak Compression. ............................................................................. 217 Figure 3.60. FFT of 2000 Hz 2-1-2 tone bursts presented at 85 ppe SPL Different panels show different hearing aid conditions: (A) Unaided; (B) Starkey Linear; (C) Siemens Linear; (D) Phonak Linear. ............................................................................................................... 219 Figure 3.61. FFT of 2000 Hz 2-1-2 tone bursts presented at 85 ppe SPL. Different panels show different hearing aid conditions: (A) Unaided; (B) Starkey Compression; (C) Siemens Compression; (D) Phonak Compression. ............................................................................. 220 Figure 3.62. FFT of 2000 Hz 4-2-4 ms tone bursts presented at 85 ppe SPL. Different panels show different hearing aid conditions:  (A) Unaided;  (B) Starkey Linear; (C) Siemens Linear; (D) Phonak Linear. ................................................................................................... 221 Figure 3.63. FFT of 2000 Hz 4-2-4 ms tone bursts presented at 85 ppe SPL. Different panels show different hearing aid conditions: (A) Unaided;  (B) Starkey Compression; (C) Siemens Compression; (D) Phonak Compression. ............................................................................. 222 Figure 3.64. FFT of /ba/ (A) and /da/ (B) from the unaided condition, and /ba/ (C) and /da/ (D) from the Starkey (compression). All stimuli were presented at 65 dB SPL. ................. 224 Figure 3.65. FFT of /ba/ (A) and /da/ (B) from Siemens (compression), and /ba/ (C) and /da/ (D) from Phonak (compression). All stimuli were presented at 65 dB SPL. ....................... 225 Figure 3.66. FFT of 80 Hz AM tones presented simultaneously at 65 dB SPL. Different panels show different hearing aid conditions: (A) Unaided;  (B) Starkey Linear; (C) Siemens Linear; (D) Phonak Linear. ................................................................................................... 227 xxiv  Figure 3.67. FFT of 80 Hz AM tones presented simultaneously at 65 dB SPL. Different panels show different hearing aid conditions: (A) Unaided;  (B) Starkey Compression; (C) Siemens Compression; (D) Phonak Compression. ............................................................... 228 Figure 3.68. Differences between Formant 1 and Formant 2 in each vowel recorded in all hearing aid and intensity conditions. .................................................................................... 230 Figure A.1. Program settings printed from NOAH for Starkey........................................... 274 Figure A.2. Coupler measurements of hearing aid output with varying input intensities when in Memory 2. ........................................................................................................................ 275 Figure A.3. Coupler measurements of hearing aid output with varying input intensities when in Memory 3. ........................................................................................................................ 276 Figure A.4. Program settings printed from NOAH for Program 1 of Siemens ................... 277 Figure A.5. Program settings printed from NOAH for Program 2 of Siemens. .................. 278 Figure A.6. Coupler measurements of hearing aid output with varying input intensities when in Program 1. ........................................................................................................................ 279 Figure A.7. Coupler measurements of hearing aid output with varying input intensities when in Program 2. ........................................................................................................................ 280 Figure A.8. Program settings printed from NOAH for Phonak ........................................... 281 Figure A.9. Coupler measurements of hearing aid output with varying input intensities when in Calm Situation 1. .............................................................................................................. 282 Figure A.10. Coupler measurements of hearing aid output with varying input intensities when in Calm Situation 2. .................................................................................................... 283   xxv  List of Abbreviations Abbreviation Definition ABR Auditory brainstem response ACC Acoustic change complex  AEP Auditory evoked potential AM Amplitude Modulated ASSR  Auditory steady state response dB  Decibels  FFT Fast Fourier transform  FM Frequency Modulated HL Hearing level ISI Interstimulus interval LTASS Long-term average speech spectrum  KEMAR Knowles Electronics Manikin for Acoustic Research MLR Middle latency response MMN Mismatched Negativity nHL Normal hearing level  ppe Peak pressure equivalency  RMS Root mean squared  SCP Slow cortical potential SNR Signal-to-noise ratio  SOA Stimulus onset asynchrony xxvi  SPL Sound pressure level  THD Total harmonic distortion VOT Voice onset time  WAV Windows Wave (Audio file format) WDRC Wide dynamic range compression              xxvii  Acknowledgements First of all, I would like to thank my supervisory committee – Dr.  Lorienne Jenstad,  Dr. Anthony Herdman, and Dr. Anna Van Maanen – for all of your advice, your insightful input during our discussions, and help in shaping this thesis into what it has become. I want to specially thank my supervisor, Dr. Lorienne Jenstad, for all of her guidance, support, encouragement, and confidence in me. I have learnt so much more about being comprehensive, skeptical, and detail-oriented through this process and she was always been there for me, every step of the way! I would also like to thank Lorienne for tolerating my horrible jokes during our meetings!   Second of all, I would like to thank members of the Amplification Lab, past and present: Foong Yen Chong, Lise Gillen, Flora Pang, Albert Chau, Stephanie Kore, Lauretta Cheng, and Carney Zeng. Not only do I want to thank some of you for your direct help on this project, I also want to thank all of you for your support, encouragement, hugs, advice, jokes, laughter, and of course, your company!  Every lab meeting is like a mini family gathering with you all!   Third of all, I would like to thank my friends both inside and outside of SASS that have offered support and encouragement along the way! Special thanks to Anielle and Estephanie for sending me regular encourage “mints” and sisterly love, as well as Tivona, Brian, Kirsten and Martin, for always being there to offer support even though you weren’t sure what I was doing! I also want to give a special shout out to my thesis buddy in our cohort, Rachel Jordan, for sharing this experience with me, for being there during the rough patches, and for being there to celebrate one another’s milestones throughout!   Fourth of all, I would like to thank Dr. Cathy Rankin and Dr. Evan Ardiel from the Rankin Lab. My experience at the Rankin Lab was where my interest in research began!   Last but not least, I would like to thank my wonderful and beloved family. Thank you so much for your unconditional love and your spiritual and financial support. Thank you for providing a safe haven for me whenever I need it. Dad, Mom, Helena (Jie), and Kevin, I want to express my heartfelt gratitude to you!!  I could not have been where I am today without all of you. Love you!!   xxviii  Dedication To Dad, Mom, Helena, and Kevin.   1  Chapter 1: Literature Review  1.1 Introduction Hearing loss is an extremely prevalent chronic condition that affects people of all ages. An estimated three hundred twenty-eight million adults and thirty-two million children have a moderate hearing loss or greater (World Health Organization, 2015). Amplification is one of the main interventions prescribed by hearing health professionals to help improve communication, among other important interventions such as aural rehabilitation, and education on communication strategies.  Early amplification is extremely important for children to have normal language and behavioural development due to the abundance of learning that occurs during the preschool ages (Yoshinaga-Itano, 1998 & 2000). Similarly, the use of hearing aids is equally important in adults with hearing loss as untreated hearing loss has been associated with significant declines in physical and mental health-related quality of life (Chen et al., 2014; Chia, et al., 2006; Crews & Campbell, 2004). Hearing loss is associated with declines in cognitive function and various psychological effects including social isolation, symptoms of depression, anxiety, paranoia, decreased self-esteem, and feelings of insecurity (Harrison Bush, Lister, Lin, Betz, & Edwards, 2015; Lin et al., 2013; Mick, Kawachi, Lin, 2014; Nachtegaal et al, 2008; Strawbridge, Wallhagen, Shema, & Kaplan, 2000). Hearing loss is also associated with a decrease in cardiovascular health, increased risk of stroke, and an increased risk of injury due to an inability to hear environmental or vocal warnings (Campbell, Crews, Moriarty, Zack, & Blackman, 1999; Gates, Cobb, D’Agostino, & Wolf, 1993)  Although behavioural tests and questionnaires are available for use by hearing health professionals to evaluate the effectiveness of hearing aids in the everyday lives of clients, 2  difficult-to-test populations fitted with hearing aids, such as infants, older adults with dementia, or other uncooperative clients, may be unable to respond to these tools reliably. In these situations, the use of objective measures, such as auditory evoked potentials (AEP), may be needed to help provide more information in order to offer more effective amplification (Souza & Tremblay, 2006).  Several electrophysiological measures have been proposed for this use, including the auditory brainstem response (ABR) (e.g., Kileny, 1982; Hecox, 1983; Mahoney, 1985), auditory steady-state response (ASSR) (e.g., Picton, Dmitrijevic, & John, 2002; Picton et al., 1998; Stroebel, Swanepoel, & Groenewald, 2007), middle latency responses (MLR) (e.g., Kurnaz et al., 2009), slow cortical potentials (SCP) (e.g., Dillon, 2005, Korczak et al., 2005), acoustic change complex (ACC) (e.g., Tremblay, Kalstein, Billings, & Souza, & 2006; Miller & Zhang, 2014)  and mismatch negativity (MMN) (e.g., Korczak et al., 2005). Recent investigations on aided potentials have been inconsistent in showing more robust responses after amplification. Some studies show the expected increase in cortical response after amplification (e.g., Chang, Dillon, Carter, van Dun, & Young, 2012; Durante et al., 2014; Tremblay, Billings, Friesen, Souza, 2006; Korczak, Kurtzberg, & Stapells, 2005; Kuruvilla-Mathew, Purdy, & Welch, 2015) while others do not (e.g., Billings, Tremblay, Souza, & Binns, 2007; Billings, Tremblay, & Miller, 2011; Billings, Papesh, Penman, Batzell & Gallun, 2012; Marynewich, Jenstad, & Stapells, 2012). This may be in part due to the unpredictable changes that can happen with hearing aid processing. A major challenge with using AEP measures of aided hearing is that hearing aids alter the temporal and spectral characteristics of stimuli in unexpected ways that could affect the amplitude and presence or absence of an AEP (Hyde, 1997; Jenstad, Marynewich, & Stapells, 2012). To date, there has been no comprehensive investigation of how hearing-aid 3  processing affects the stimuli commonly used to generate AEPs.  Thus, there is a need for a systematic analysis of the stimuli used for AEP measures after hearing aid processing.   1.2 Overview of existing applications of auditory evoked potentials   Many auditory evoked potentials generated along the auditory pathway, from brainstem to cortex, are used for the purposes of establishing unaided thresholds. This includes the use of ABRs and ASSRs, both generated in the lower auditory pathway (brainstem and the thalamo-cortical pathways), and SCPs, generated in the cortex (Picton, 2011). SCPs can also be used in cases such as children or adults with auditory neuropathy spectrum disorder where auditory brainstem responses are unlikely or unhelpful (Picton, 2011; Sharma, Cardon, Henion, & Roland et al., 2011).    There are applications of auditory evoked potentials for other diagnostic purposes. For example, MLRs are sometimes used to assist in the diagnosis of central auditory processing disorders as a predictor of speech perception outcomes, and for intraoperative monitoring purposes, among others (Davies, Mantzaridis, Kenny & Fisher, 1996; Kileny, Paccioretti, & Wilson, 1987; Özdamar & Kraus, 1983; Weihing & Musiek, 2008). In relation to the slow cortical potential, the ACC is an AEP used to observe the detection of change within a continuous stimulus (Martin & Boothroyd, 1999; Martin & Boothroyd, 2000; Ostroff, Martin, & Boothroyd, 1998). The morphology of the ACC is the same as the SCP, with a P1-N1-P2 complex seen when the change in a continuous stimulus is detected (Martin, Tremblay, & Korczak, 2008). Although not widely used clinically yet, it can be an indication of the detection of change at the cortex (Martin et al., 2008). This provides some information on the difference limens of certain acoustic features for an individual, such as spectral resolution at the level of the cortex (Martin et al., 2008).  Similarly, MMN is a late 4  response correlated with the ability of the cortex to detect differences between stimuli at a pre-attentive level (Näätänen, Paavilainen, Rinne & Alho, 2007). These include both lower-order differences, such as changes to intensity or frequency, as well as higher-order differences, such as grammar violations (Näätänen et al., 2007). Thus, it has been used in studies exploring central auditory processing function (Näätänen et al., 2007).   1.2.1 Applications of auditory evoked potentials in amplification  ABRs (e.g., Kileny, 1982; Hecox, 1983; Mahoney, 1985), ASSRs (e.g., Picton et al., 1998 & 2002; Stroebel et al., 2007), MLRs (e.g., Kurnaz et al., 2009), SCPs (e.g., Chang et al., 2012; Korczak et al., 2005), ACCs (e.g., Miller & Zhang, 2004; Tremblay, Kalstein et al., 2006) and MMNs (e.g., Korczak et al., 2005), have all been proposed for measuring or validating amplification when behavioural responses cannot be obtained. Although these AEPs have different generators along the auditory pathway, and can be affected differently by the consciousness of the individual, none of these require active engagement in the task during the recording (Picton, 2011), making it possible to record from difficult-to-test subjects. Some measures (e.g., ASSRs, SCPs, and MMNs) are advocated because the common evoking stimulus can be speech or speech-like, and thus, can provide some information about speech processing (Shemesh, Attias, Magdoub, & Negeris, 2012).  The use of later potentials, such as MLRs, SCPs and MMNs, are supported because they provide information about cortical processing of signals that earlier potentials, such as ABRs and ASSRs at faster modulation rates, cannot provide (Kileny, 1991; Korczak, Kurtzberg, & Stapells, 2005; Naatenan et al., 2007).  Some researchers have also suggested the use of SCPs to provide evidence of speech discrimination because differences in the morphology of the SCP are observed when elicited by different speech stimuli, although it is agreed upon 5  that this use is not ready clinically yet (e.g., Kuruvilla-Mathew et al., 2015; Ostroff et al., 1998; Tremblay et al., 2006). Other researchers have identified the use of the SCPs for speech discrimination as problematic because they were unable to find robust amplification results for differentiating between SCPs that were elicited by suprathreshold speech stimuli when re-analyzing data from past studies (Billings et al., 2012). They attributed the problems to a combination of individual variability in stimulus parameters, which can be further affected by hearing aid settings, and subject variability (Billings et al., 2012). Instead, they suggested the use of SCPs for the purpose of physiological detection, identifying an SCP with hearing aids where none was identified previously, when studying SCPs in the context of amplification, due to significant and robust amplification effects (Billings et al, 2012). Researchers have used the physiological detection approach to study the effects of hearing aid features, such as Zhang et al. (2014), who detected a significant increase in the number of SCPs detected in children with hearing loss with the use of non-linear frequency compression, as compared to the number of SCPs detected with traditional pediatric hearing aid settings.  However, the recording time for later potentials can be much longer than earlier potentials, which can make them less clinically practical (Bellier et al., 2015). Some aided potentials, such as ABRs, ASSRs, and SCPs, can also provide a clinician with information on aided thresholds at different frequencies. Despite the ease of collecting such information, this use of these aided potentials provides little information on possible functional benefits obtained with hearing aids. (Schwartz & Larson, 1977; Seewald, Hudson, Gagné, & Zelisko, 1992).  Even when this information is collected behaviourally, aided thresholds provide limited information on speech understanding to help guide clinical decisions (Schwartz & Larson, 1977; Seewald, Hudson, Gagné, & Zelisko, 1992).   6  In relation to measuring functional benefits with the use of amplification, the MMN may be helpful as it can be elicited by both lower-order differences, such as physical intensities, and higher-order differences, such as grammar violations (Näätänen et al., 2007, Picton, 2011); the latter can provide information on preliminary language processing (Näätänen et al., 2007; Picton, 2011). For example, the MMN may help in determining whether there is an improvement in the ability to detect a phonemic difference with exposure to amplified speech if the MMN is more robust or present compared to a previous unaided measurement (Näätänen et al., 2007; Picton, 2011). However, the MMN may not be ideal because it is a small response that can be lost in the background electrical noise, and is also difficult to detect in some individuals, making it harder to interpret at an individual level (Picton, 2011). ACCs have been suggested as an alternative to MMNs for lower-order differences, such as spectral and intensity differences, due to their higher efficiency of recording and higher signal to noise ratios within the recording (Martin & Boothroyd, 1999 & 2000). It is important to note that the stimulus presentation parameters of the two have large differences as ACCs are present when an individual detects a change within a continuous stimulus, while MMNs are present when an individual detects a deviant amongst a standard stimulus that occurs at a much higher probability (Martin, Boothroyd, Ali, & Leach-Berth, 2010). ACCs also reflect stimulus detection while MMN reflects higher-order stimulus discrimination (Martin et al., 2010; Näätänen et al., 2007).  As evidenced, AEPs can be used for a variety of purposes, including those evaluative or diagnostic in nature. As the suggested AEPs are replicable and require little to no active involvement of the individual, they can be useful tools for obtaining information we would otherwise be unable to obtain without reliable behavioural responses. Even so, it is important 7  to be aware that the presence of an AEP does not necessarily mean that the individual perceives the evoking stimulus, or conversely, that the absence of a response does not mean the individual has not heard or detected the stimulus or stimulus change (e.g. Martin, Kurtzberg, & Stapells, 1999; Picton, 2011).  1.3 Hearing aid processing The unpredictable changes hearing aid processing can have on stimuli can make the resulting aided AEP more difficult to interpret. For example, when comparisons are made between AEP responses to stimuli presented at different levels, the assumption is that all properties of the stimulus remain the same other than the intensity (Picton, 2011). However, hearing aids not only increase intensity of stimuli, they also alter spectral and temporal properties of the stimulus (Chasin & Russo, 2004; Dillon, 2012).  Several factors affect the frequency response of hearing aids: the configuration of the hearing loss, the limited frequency output range by receivers, and properties of the external components of the hearing aid such as the tubing of behind-the-ear hearing aids (Dillon, 2012). Receiver limitations on high-frequency output can affect the spectral components that can affect both the perception of the stimulus and physical morphology of the AEP as well (Tremblay et al., 2006). These changes in frequency response can also affect timbre, which may in turn, affect the morphology of the AEP. Moreover, distortions to the stimulus can also occur, such as peak clipping, harmonic distortion, and the addition of internal noise (Hawkins & Naidoo, 1993; Tan, Chang, Chua, & Gwee, 2003; Macrae & Dillon, 1996). For example, in cases where the stimulus peak clips when it reaches the limits of the microphone on the hearing aid, the stimulus momentarily becomes a square wave and is processed as such, broadening the spectral properties of the stimulus.  Another example can be 8  demonstrated by the interaction of processed and unprocessed sound waves at the eardrum (Stone, Moore, Meisenbacher, & Derleth, 2008). With an open earmold or other open canal fittings, sounds, especially those of low frequency, are able to more easily travel unprocessed by the hearing aid into the ear canal (Stone et al., 2008). As they are of slightly different intensities and phase position due to differences in the time of arrival at the eardrum, they can interact constructively and destructively with the amplified signal, which is delayed by processing, affecting the frequency response (Stone et al., 2008). Hearing aids also have internal noise, known as equivalent input noise; this noise originates mainly from the microphone of the hearing aid (Macrae & Dillon, 1996).  The internal noise, in addition to amplified external background noise, can further distort the stimulus by adding broadband energy. It also decreases the signal to noise ratio of the evoking stimulus. This is important to note because the signal to noise ratio can have effects on both the amplitudes and latency of AEPs (Billings, Tremblay, Stecker, & Tolin, 2009; Borgmann, Roß, Dragnaova, & Pantev, 2001; Galambos & Makeig, 1992a & 1992b; Kuruvilla-Mathew et al., 2014; Papesh, Billings, & Baltzell, 2015; Shytrov et al., 1996).  For example, Billings et al. (2009) showed that the signal to noise ratio, regardless of the absolute intensity of the stimulus, was a major determinant of the amplitude and latency of the SCP when noise was present in the evoking stimulus.  Temporal properties of a stimulus may also be affected by several processing effects of the hearing aid: processing delays of the hearing aid, group delays, as well as attack and release times of the compressor of the hearing aid when compression is activated (Dillon, 2012). Processing delays occur with any digital systems; however the amount of processing delay can differ depending on the type of processing the hearing aid uses (Stone & Moore, 9  2003). Depending on the method used to separate and process signals over a given time point, delays from 2 ms to 7 ms are possible (Stone & Moore, 2003). Group delays are frequency-dependent delays in processing (Couch, 1990). As these are present in hearing aid processing as well, the short asynchronous onset of a click can, therefore, be changed into a frequency sweep. The attack and release times of the compressor can also cause an overshoot of the stimulus intensity before the compressor activates to increase or decrease the intensity (Stone, Moore, Alcántra, & Glasberg, 1999). In addition, hearing aids introduce phase distortions to the signal with processing (Chasin & Russo, 2004).  Thus, it is possible to observe changes to the temporal envelope of the stimuli with these internal characteristics of hearing aid processing.    Further changes from the original stimulus can happen with wide dynamic range compression (WDRC) in addition to the effects of the compressor mentioned above. WDRC increases the gain of low-level inputs and decreases the gain of high-level inputs (Jenstad, Pumford, Seewald, & Cornelisse, 2000). A ratio can be calculated to compare the magnitude differences between the input and output levels; the larger the ratio, the stronger the compression. Thus, although the intensity range of the input stimuli may be large, the change to the intensity range of the can be reduced. This can affect how the AEP is interpreted. For example, changing the input stimulus by 20 decibels should yield an AEP that has increased in amplitude and decreased in latency as we expect from a 20 decibel change; however, with a typical 2:1 compression ratio activated in the hearing aid, there may be only a 10 decibel change in the output of the stimulus. The ratio and the point at which compression is applied, known as the knee point, can differ at different frequencies, depending on how the hearing 10  aid is set. Changes that can be caused by compression will be investigated in this study, as measurements will be made both with and without compression activated.   Finally, there is a variety of advanced hearing aid processing, such as feedback cancellation and noise reduction for ongoing and impulse sounds. These features may introduce new signals in the output, or activate frequency-specific gain reduction, which further complicate and change output signals. Such features will therefore be turned off for this systematic analysis to prevent confounds to the data that can be caused by the activation of these features.  1.4 Morphological changes of electrophysiological measures from changes in stimulus parameters  Replicable morphological changes of AEPs in response to changes in stimulus parameters have been noted in many different studies. Many of these changes affect the latencies and amplitudes of the peaks or waves that define the AEP, which can in turn affect the interpretation of the AEP. For example, for many of the measures, we expect an increase in amplitude and decrease in latency with an increase in the intensity of the stimulus (Picton, 2011). With the provision of amplification, we would also expect an increase in amplitude and decrease in latency with aided AEPs. However, a multitude of other changes to the stimulus caused by hearing aid processing can combine to have the same effect, or dampen the effects of increasing the intensity, although the neural generator of these effects may differ. For instance, to replicate the same effect of an increase in amplitude and decrease in latency, a decrease in rise and fall time of the stimulus can be made. Similarly, an increase in rise and fall time of the stimulus can dampen the effects of an increase in the intensity. Such changes can easily occur following hearing aid processing. Table 1.1 outlines changes that 11  have been observed in response to changes in stimulus parameters for simple stimuli, such as tones, amplitude modulated stimuli, and broadband stimuli, such as clicks. 12   Electrophysiological Measure Stimulus Parameter Change ABR ASSR MLR  SCP  ACC MMN Rise/Fall time ↑ - increased absolute latency (Stapells & Picton, 1981) - Wave V: decreased amplitude near threshold, with the largest decrease when rise time is between 5 to 8 ms (Folsom & Aurich, 1987; Stapells & Picton, 1981) - analogous to wider modulation envelope (sinusoidal)   - decreased amplitude  (John et al., 2002)  - decreased amplitude of Na-Pa. (Xu et al., 1995); - trend of increase in latency of Na and Pa (Kodera et al., 1979).  - <30 ms: no effect on amplitude - 30-50 ms: gradual decline in N1 amplitude - >50 ms: no further decline in N1 amplitude (for stimuli of long durations); increases in latency, with smaller increments of increase with more intense stimuli (Alain et al., 1997; Kodera et al., 1979; Onishi & Davis, 1968).  N/A N/A ↓ - Wave V: decreased latency and increased amplitude (Stapells & Picton, 1981) - analogous to narrow modulation envelope (e.g., exponential sinusoidal function)  - increased amplitude  (John et al., 2002)              Opposite of increase Opposite of increase N/A      13   Electrophysiological Measure Stimulus Parameter Change ABR ASSR MLR SCP ACC MMN Duration                ↑ - Wave III: decrease in amplitude, no increase in latency.  - Wave V: latency increase (Hecox et. al., 1976); amplitude increase (Funasaka & Ito, 1986) - other studies note no significant effects  (Davids et al., 2008) N/A  No significant effect (Davids et al., 2008; McGee et al., 1988; Vivion et al., 1980) - up to 30 to 50 ms: increased amplitude  - when rise time is 3 to 5 ms: increased amplitude up to 70 ms duration.  - when rise time is long (≥ 30 ms), duration has no effect  (Gage & Roberts, 2000; Joutsiniemi et al., 1989; Kodera et al., 1979;  Onishi & Davis, 1968; Picton, 2011) - for duration of pre-transition stimulus: increased amplitude, decreased latency - increase between 100 ms to 150 ms: amplitude remains the same - minimum of 100 ms pre-transition stimulus duration to elicit tonal ACC and 80 ms pre-transition to elicit for speech ACC (Picton 2011; Ganapathy et al., 2013) - when increased from 4 to 300 ms: increased amplitude to intensity changes - smaller increments between 4 to 300 ms: no effect (Paavilainen et al., 1993) Duration   ↓ Opposite of increase N/A; No significant effect   (Davids et al., 2008) Opposite of increase  Opposite of increase  Opposite of increase.  Intensity/ Gain ↑ -Waves I-V: increased amplitude and decreased latency (Picton, 2011)  - increased amplitude and decreased phase delay  (Rodriguez et al.,1986) - up to 60 dB SL: increased amplitude  - amplitude effects saturates for Na, Pb  - decreased latency  (Borgmann et al., 2001) - increased amplitudes and decreased latencies for N1  - 60-80 dB nHL: amplitude effects saturate  (Davis et al., 1966, among others; Hari et al., 1982; Picton, 2011)   - for difference in intensity: increases amplitude and decreases latency  - Minimum difference of +2 dB (He et al., 2012; Martin & Boothroyd, 2000)  - +3 to +9 dB: increased amplitude  - +9 dB: decreased latency  - greater changes lead to larger amplitude, especially when near discrimination threshold of MMN (Horvath et al., 2008; Rinne et al., 2006) 14   Electrophysiological Measure Stimulus Parameter Change ABR ASSR MLR SCP ACC MMN Intensity/ Gain ↓ - Wave V: Amplitude stable when suprathreshold; decreases as nearing threshold - Wave I-IV: decreases begin significantly above threshold; - Waves I-V: increased latency (Picton, 2011) Opposite of increase - Po and Pa: most resilient in terms of amplitude to decrease in intensity - Po:  decreased latency (Maurizi et al., 1984)     Opposite of increase - for difference in intensity: decreased amplitude, and increased latency - latency increase plateaus when intensity changes are 4 dB or below  - minimum difference of -3 dB; response amplitude smaller to decreases in intensity (He et al., 2012; Martin & Boothroyd, 2000) - -3 to -9 dB: increased amplitude - -9 dB: decreased latency   - greater changes lead to larger amplitude, especially when near discrimination threshold of MMN (Horvath et al., 2008; Rinne et al., 2006)   Frequency ↑ - Decreased latency for all waveforms due to greater synchronicity  (Picton, 2011)  - for carrier frequencies: decreased amplitude and phase delay (Rodriguez, et al., 1986) No change in latency and amplitude of Na, Pa, Nb, Pb. (Oates & Purdy, 2001)  - Loudness kept constant: no change - Intensity kept constant across frequencies: N1 amplitude decreases, more so above 2000 Hz (Antinoro & Skinner, 1968).   - for difference in frequency:  increased amplitude and decreased latency  - minimum difference of 20 Hz - amplitude effect enhanced with increases in intensity, but not decreases (He et al., 2012; Martin & Boothroyd, 2000) - increases with increased frequency difference, up to 127 Hz, at which the effect reaches a ceiling (Hovath et al., 2008).  ↓ - Increased latency for all waveforms. Low frequency stimuli (e.g., below 1000 Hz), Wave V looks more broad  (Picton, 2011) - for carrier frequencies: increased amplitude and phase delay (Rodriguez, et al., 1986) Opposite of increase  - for difference in frequency: decreased amplitude and increased latency  - minimum difference of 20 Hz  (He et al., 2012).  15   Electrophysiological Measure Stimulus Parameter Change ABR ASSR MLR SCP ACC MMN Rate/ Inter-stimulus interval (ISI) ↓(ISI)/ ↑ (Rate)  - Wave V: slight increase of absolute latency and no effect on amplitude\ Waves I-IV: amplitude decreases   (Picton, 2011) - increased amplitudes and decreased latencies; neural generator closer to the brainstem (i.e., above 150 Hz)  - most robust responses at 40 and 80 Hz  (Picton, 2011) - up to 40/s: decreased amplitude  (Kraus, et al.,1987) (NB: animal data) - rapid decline in N1 amplitude (Davis et al., 1966)  - Increase in amplitude (Jones & Perez, 2001) Related: Change in rate of change (e.g. increased step size mid-way), elicited larger ACC - no particular trend for latency (Kohn et al., 1978 & 1980)  - <100 ms: no MMN occurs - ISI decreasing to 150 ms: increased amplitude (Picton, 2011).   ↑ (ISI) / ↓ (Rate) Opposite of increase; not usually presented lower than 10/s (Picton, 2011) - decreased amplitudes and increased latencies; neural generator being closer to the cortex (e.g., in the 40 Hz  modulation rate region)  - most robust responses at 40 and 80 Hz  (Picton, 2011)  Opposite of increase            - increased N1 amplitude (more rapidly up to ISIs of 3 s, and gradual increase up to 10 s) (Appleby, 1964; Davis et al, 1966).  - less obvious at lower intensities (Näätänen & Picton 1987)  - Decrease in amplitude (Jones & Perez, 2001).  Related: Change in rate of change: same as increase  - decreased amplitude. (e.g., ISI greater than 9 s) due to integration time (Sams et al., 1993).  - Increased amplitude if interdeviant interval is increased (Picton, 2011)  Click Polarity    Condensation - No effect  - some studies note small changes to latencies: usually earlier latencies for rarefaction  (Ballachanda, Moushegian, & Stillman, 1992, Picton 2011) N/A N/A N/A N/A N/A Rarefaction N/A N/A N/A N/A N/A Alternating  N/A N/A N/A N/A N/A 16   Electrophysiological Measure Stimulus Parameter Change ABR ASSR MLR SCP ACC MMN Spectrum  Broad  - decreased latency due to better synchronicity from greater number of neurons firing (Campbell et al., 1981).  (Picton, 2011) - increased amplitude with decreased frequency specificity (Dimitrijevic, et al., 2002) - increased amplitudes and decreased latencies for earlier waves (using clicks as evoking stimuli)  (Borgmann et al., 2001; Maurizi 1984) - Increased amplitude and decreased latency (Picton, 2011)  - increased amplitude and decreased latency with increased harmonics (Jones & Perez, 2001)   - increased amplitudes (spectrally rich stimuli as compared to pure tones) (Tervaniemi et al., 2000)  Narrow - latency changes depend on frequency (increased latency for narrow-band low frequencies) but not as frequency specific as tone-pips (Picton, 2011) Opposite of broad  - decreased amplitude and increased latency when tone-pip was used as evoking stimuli (Borgmann et al., 2001) Opposite of broad Opposite of broad Opposite of broad  Modulation depth  ↑ N/A - increased amplitude  (AM or FM)  (Dimitrijevic et al., 2001; Picton et al., 1987).  N/A N/A - increased amplitude when presented at modulation rates that elicit large ACC responses (i.e., slow rates such as 4 Hz to 40 Hz) ; near discrimination threshold, the response is less detectable (Dimitrijevic et al. 2008; Han & Dimitrijevic, 2015) N/A ↓ N/A Opposite of increase N/A N/A Opposite of increase N/A 17   Electrophysiological Measure Stimulus Parameter Change ABR ASSR MLR SCP ACC MMN Signal to Noise Ratio (ipsilateral)  ↑ Opposite of decrease - increased amplitude and phase coherence (Galambos & Makeig, 1992a) N/A -increased amplitude, decreased latency (Billings et al., 2009) N/A - processing mainly in left hemisphere (Shytrov et al.,1996) ↓ - increased latency and decreased amplitudes - >20 dB EM noise: increased latency and decreased amplitudes - 10 to 20 dB EM noise: decreased amplitude (Owen & Burkard, 1991) - at 40 dB SNR: similar in latency and amplitude as quiet - 10 to 0 dB SNR: no latency shifts, but decreases in amplitude observed  - no ABR detected at -10 dB SNR  (Hecox et al., 1989) - low-level background noise (i.e., not near the intensity of stimulus), increase in amplitude and increased phase coherence.  - background noise intensity near levels of stimulus, decreased amplitude (Galambos & Makeig., 1992a & 1992b) N/A - low-level background noise (i.e., not near the intensity of the stimulus with 900 ms ISI), increased N1 amplitudes (binaural presentation), decreased P1 & P2 amplitudes - longer ISIs:  decreased N1 amplitudes  - stimuli presented at higher rates and binaurally: increased latencies - background noise intensity near levels of stimulus, decreased amplitudes - general trend of increase in latency (N1, P1, and P2) and decrease in amplitude (P2). No significant latency effect for high-frequency emphasis speech stimuli (Billings et al., 2009; Davis et al., 1968; Kuruvilla-Mathew et al., 2015, Papesh et al., 2015).) N/A - processing in left hemisphere decreases (response amplitude decreases), and increased response amplitude in the right hemisphere (Shytrov et al., 1996)  Table 1.1. Effects of simple stimulus parameter changes on AEP morphology18  As speech is a complex type of stimulus with additional properties to those noted in Table 1.1, effects of changes to three different elements of speech are explained in Table 1.2. Most of the current research that has investigated changes to AEP morphology to changes in speech characteristics have focused on slow cortical potentials.  Table 1.2. Effects of speech stimuli parameter changes on SCP morphology.  1.5 Inconsistent findings with aided cortical AEPs Previous studies have shown interesting results with aided AEPs, albeit with some inconsistencies. With higher intensities, Chang et al. (2012) detected a larger number of present SCP responses in infants with confirmed sensorineural hearing loss when speech stimuli were presented at higher sensation levels, as compared to the amount SCP responses Speech Stimulus Parameter Change Effects on SCPs  Vowel Onset/Fricative Duration - Voice onset times (VOTs) are defined as the time between the release burst of a stop consonant and the beginning of the periodicity of the following vowel (Kent, 1992). Short vowel onset times (0-30 ms) evoke one N1 response, while longer vowel onset times (50-80 ms) evoke two N1 responses, one to the consonant, and the other to the vowel, expected from research on minimal pre-transition stimulus time before eliciting an ACC (Ganapathy et al, 2013) - N1 latency also had a high positive correlation with VOT, suggesting that longer VOTs lead to later N1 latencies (Sharma & Dorman, 1999). - P1 and N1 corresponding to acoustic onset of /s/ was greater than /ʃ/ in amplitude. As voice onset of /s/ is earlier, P1 and N1 also showed earlier latency, as expected. (Tremblay et al 2006, Agung-King et al., 2008) - Minimum VOT for a separation of two N1 responses is 40 ms (Sharma, Marsh, & Dorman, 2000). Tongue height and backness  - N1-P1 latencies are earlier with vowels with lower first formants and higher second formants. N1 amplitudes are larger for vowels with greater F1-F2 differences, such as /i/, than vowels with small F1-F2 differences, such as /a/. (Agung , Purdy, McMahon, & Newall, 2006; Johnson, 2011; Makela, Aku & Tiitinen, 2003; Obleser, Eulitz, & Lahiri, 2004). F1-F2 differences also influence generator source in left hemisphere, such that as F1-F2 differences increase, there is more dorsal activation along the iso-frequency contour (Ohl & Scheich, 1997) Rise Times ↑ - N1-P2 amplitude larger and shorter peak latencies in abrupt (e.g., word initial) vs slow / ʃ / - information from tones (i.e., shorter rise times lead to larger amplitudes and shorter latencies) can be generalized (e.g., neural synchrony) (Easwar, Glista, Purcell, & Scollie, 2012) ↓ 19  detected in the unaided condition. Using the same system and similar methodology, the effect was replicated in adults with sensorineural hearing loss; the researchers detected a significant increase in the number of SCP responses with amplification as compared to the number detected without amplification, for all of the presented speech sounds, /m/, /t/, and /g/ (Durante et al., 2014). A decrease in the latencies of the detected SCPs was also observed with amplification (Durante et al., 2014).  Tremblay, Billings, et al. (2006) also found a slightly larger SCP response at a central electrode placement, Cz, to an amplified speech signal, /si/, in normal hearing individuals. However, when the response at all electrodes was taken into account, the difference was not significant (Tremblay, Billings, et al., 2006).  Also using speech stimuli, Korczak et al. (2005), found larger effects with amplification in individuals with normal hearing as well as those with sensorineural hearing loss. They were generally able to obtain shorter latencies for recorded SCP responses, and larger responses for both SCP and MMN, as expected, although the amount of improvement of the responses was quite variable between individuals (Korczak et al., 2005). The magnitude of effects was also dependent on stimulus intensity (Korczak et al., 2005). Although non-significant, there was also a trend of larger responses for the P3b response that Korczak et al. (2005) measured.  However, they did find some inconsistencies with the MMN responses in some participants with hearing impairment, noting that aided responses were smaller than unaided responses (Korczak et al., 2005). They attributed this finding to a poorer signal to noise ratio in the recording due to variability in latencies (Korczak et al., 2005).  On the other hand, Billings et al. (2007 & 2011) did not show the expected increase in SCP amplitudes with hearing aid gain that had been verified with real ear measurements using tonal stimuli. Measured responses had similar amplitude and latencies to the unaided 20  response or even smaller and later responses than the unaided condition (Billings et al., 2007 & 2011). Billings et al. (2007 & 2011) suggested that this finding was due to the poorer signal to noise ratio with hearing aid processing. Billings et al. (2009 & 2012) subsequently found that signal to noise ratios were important in determining the amplitude of the response; that is, no matter the intensity of the stimulus, if the signal to noise ratio was the same, so were their SCP amplitudes. Similar results were replicated using speech sounds (Kuruvilla-Mathew et al., 2015).  Kuruvilla-Mathew et al. (2015) presented seven speech sounds, differing in place, manner, and voicing, in quiet and in noise, as well as in aided and unaided conditions, to normal hearing adults. They found a general trend of increased latencies for all peaks and decreased amplitudes for P2 in noise, although these effects were not significant for all speech sounds presented (Kuruvilla-Mathew et al., 2015). In addition to noise comparisons, Kuruvilla-Mathew et al. (2015) also observed the effects of amplification on the latencies and amplitudes of different peaks of the SCP to aided and unaided speech stimuli. Generally, they found a trend where amplification elicited a larger but later P2 and no significant differences for other peaks (Kuruvilla-Mathew et al., 2015). However, recent reports have suggested that rates of stimulus presentation and the effects of binaural and monaural stimulation, can interact with SNR to cause different changes to morphology, with the rate of stimulus presentation affecting N1 and N2 peak amplitudes of the SCP, and binaural and monaural stimulation affecting the P1and P2, responses (Papesh et al., 2015). That is, at faster rates (i.e., an interstimulus interval of 900 ms) with binaural stimulation, background noise can actually enhance N1 and N2 peaks, with a lower level of noise reaching significant differences from quiet, while slower rates (i.e., an interstimulus interval of 1900 ms) decrease the N1 and N2 peaks as shown previously (Papesh et al., 2015). This 21  was unexpected, given that slower rates lead to larger N1 peaks in general (Appleby, 1964; Davis et al., 1966). Although all N1 peaks were generally larger in the slow presentation rate condition than those acquired in the fast presentation rate condition, N1 amplitudes obtained in low-level background noise were smaller than those obtained in quiet conditions at a slower presentation rate, while N1 amplitudes were enhanced with low-level background noise at a fast presentation rate (Papesh et al., 2015). They suggested that this enhancement may be linked to the ability of the cortex to phase lock to predictable presentations of stimuli, indicating that the faster presentation rate was within the optimal range for the cortex to phase lock to the stimuli in order to improve the detection of them (Papesh et al., 2015). On the other hand, although P1 and P2 amplitudes decrease with background noise with both fast and slow presentation rates, binaural presentation of stimuli can enhance the amplitude of both the P1 and P2 responses (Papesh et al., 2015). Peak latencies, on the other hand, increased with increasing noise for all peaks; P1 latencies were additionally affected by the rate of presentation as latencies decreased with a slower presentation rate being earlier, while N1 latencies were only significantly shortened with a slower presentation rate when the stimuli were presented monaurally (Papesh et al., 2015). This may be an important consideration for a sound-field set up, in which aided AEPs are often recorded.  Marynewich et al. (2012) were also unable to find a reliable increase in amplitude and decrease in latency with 20 dB or 40 dB hearing aid gain as compared to the unaided condition. Subsequent acoustic measures of the stimuli used in their study revealed elongated rise times to in the aided condition, and less gain for the brief AEP stimuli than standard real ear pure tone stimuli (Jenstad et al., 2012). Thus, many studies note that the specific effects 22  of hearing aid processing need to be investigated in order to better understand aided evoked potentials and how to interpret them.  To date, there has been no comprehensive investigation of how hearing-aid processing affects the stimuli commonly used to generate AEPs.  Thus, there is a need for a systematic analysis of the stimuli used for electrophysiological measures after hearing aid processing and the purpose of the current study was to do so. The data gathered in this study may help improve our understanding of morphological changes that we may expect when obtaining aided AEPs.  This can hopefully guide further research investigating which AEP and stimuli may be of more suitable use to measure the functional benefits of amplification, such as an increased ability to detect language cues.   The purpose of the current study was to provide a systematic quantification of the effects of hearing aid processing on stimuli used to elicit AEPs. If changes were observed, then the magnitude was further interpreted based on the findings outlined in Table 1.1 for their possible effects on the morphology of AEPs.            23  Chapter 2: Methods 2.1 Hearing aids  Three digital mid-range to advanced behind-the-ear hearing aids from different manufacturers were used: Phonak Bolero Q90, Siemens Motion SX Micon 5mi, and the Starkey Z series mini BTE i70. Hereafter, the hearing aids were referred to by their respective manufacturer names. Three hearing aids were used to help explain any discrepancies between hearing aids. Advanced features such as feedback management, noise reduction, and directional microphones, were deactivated to observe the effects of hearing aid processing only and to minimize complications. The maximum power output was also set to maximum at each frequency band to prevent any unneeded output limiting. See Appendix A for the settings of each hearing aid and Speechmap™ measurements of each hearing aid in test box. All hearing aids were within tolerances (ANSI S3.22-2003).  2.2 Materials and equipment 2.2.1 Recording system The system that was used for the presentation and recording of stimuli is outlined in Figure 2.1. Stimuli were presented using a computer situated outside of the booth. Audacity® 1.3.14-beta (Unicode) was used to play the stimulus and record the hearing aid output simultaneously. The signal was routed through the RME Hammerfall DSP Multiface II soundcard to the loudspeaker, a Behringer Truth B2031A loudspeaker with a linear frequency response from 50 Hz to 21 kHz, an 8.75 inch woofer, a crossover at 2 kHz, and a 1 inch tweeter. This loudspeaker presented the stimulus directly in front of the 45BA Knowles Electronics Manikin for Acoustic Research (KEMAR; G.R.A.S. Sound and Vibration, Denmark). KEMAR was fitted with the IEC 60711 Coupler RA0045, a half-inch 24  microphone, and a large KEMAR pinna (No. KB0065). KEMAR was positioned at 0-degree azimuth and 0.53 m away from the loudspeaker, which was level with the head of KEMAR. Acoustic signals collected using KEMAR were sent to an amplifier and routed to the soundcard. The soundcard subsequently routed the signals to the computer running on a Windows operating system. The recording was sampled at 44 100 Hz. The recordings were stored as a single-channel 16-bit WAV file. Stimulus presentation was conducted in a double-walled sound attenuating booth with permissible ambient levels (ANSI S3 .1-1999). The behind-the-ear hearing aid connected to an ear mould simulator (Model: KB0111) was placed on the left ear of KEMAR.    Figure 2.1. Diagram of system used in simultaneous recording and presentation of stimuli.  25  To take into account the resonance effects of the ear included within the recording, the real ear unaided gain of KEMAR at 250 Hz to 6000 Hz was calculated by subtracting the measured intensities of stimuli at the mic location of a behind-the-ear hearing aid from the measured intensities of the same stimuli inside the ear canal of KEMAR across those frequencies.  2.2.2 Stimuli  Stimuli were created using Audacity® or MATLAB, using a custom-written amplitude-modulation tone generation code (© A. Herdman). The stimuli included were based on descriptions of those used clinically, such as by WorkSafeBC, and BC Early Hearing Program, among others, as well as those used in research studies that included the electrophysiological measures outlined for this inventory. We included pure tones and tone bursts of different rise and fall times, duration, and frequencies from 500 Hz to 4000 Hz. Stimuli containing multiple tones, such as simultaneously presented frequencies in ASSR stimuli, a continuous tone with a 25 Hz frequency change after 300 ms, and broadband stimuli such as clicks and noise bursts, with varying durations and rise and fall times, were also included. Speech stimuli, including consonant-vowel tokens as well as vowel tokens, presented in a repeated manner as well as in an odd-ball paradigm were used.  While tonal and non-speech broadband stimuli were generated from the outlined programs, speech tokens were extracted from speech spoken by a male speaker from the City University of New York (CUNY) Nonsense Syllable Test (NST) (Levitt & Resnick, 1978). The vowel tokens were extracted from vowel-consonant tokens using Praat and the consonant-vowel tokens were extracted from running speech (Boersma & David, 2013). Tokens used for the MMN were shortened from the original duration of the consonant-vowel 26  token by cutting cycles from the vowel portion. Speech stimuli from the HEARLab™ system (Frye Electronics, Tigard, USA; Golding, Pearce, Seymour, Cooper, Ching, & Dillon., 2012), which were more brief, were included. See Table 2.1 for the list of all presented stimuli and their detailed parameters included in this study. The stimulus onset asynchrony (SOA) in Table 2.1 was the time interval from the beginning of one presentation of the stimulus to the start of the next presentation of the stimulus.   AEP type Stimulus Description References or explanation SCP and ACC          Speech:   Consonants:  WORD INITIAL FRICATIVE /sa/  & /ʃa/, alternating polarity, 1125 ms SOA  WORD INITIAL STOP /da/ & /ta/, alternating polarity, 1125 ms SOA  Vowels:   /i/ (FRONT HIGH VOWEL), /a/ (MID LOW VOWEL), 1125 ms SOA, cut to 100 ms in duration by removing cycles in the vowel.   /a/ cut pre-recording from/ʃa/, 1125 ms SOA, no removal of cycles     /m/ (30 ms), /t/ (30 ms), /g/ (21 ms), /s/ (50 ms), alternating polarity, 1125 ms SOA   Tones:   60 ms duration, 5, 10, 20 ms rise/fall for .5, 1, and 2 kHz, at 0.9/s, alternating polarity  30 ms duration with 5 & 10 ms rise/fall for .5, .1, and 2 kHz at 0.9/s; and 120 ms & 480 ms duration with 5, 10, & 20 ms rise/fall each for .5, .1, and 2 kHz, at 0.9/s, alternating polarity  1 kHz tone 450 ms duration 10 ms rise/fall time;  & 1 kHz 757 ms duration, 7.57 ms rise/fall time, 1900 ms SOA, alternating polarity  1680 Hz tone, 300 ms duration, transition to 1705 Hz tone, 300 ms duration, with no overlap    Based on Easwar et al., 2006; Easwar et al., 2012; Tremblay et al., 2006   Based on Agung-King et al., 2008     Based on Agung et al.,  2006     To compare with /a/ cut post-recording from /ʃa/; for comparison of SCP and ACC paradigm of stimulus presentation     Based on HEARLab™ System, provided by H. Dillon, B. Van Dun, & T. Loi.      Based on WorkSafeBC stimuli, which are based on Hyde, 1994 & 1997  Systematic variation of duration and rise/fall time     Based on Billings et al., 2007 & 2012; simulate speech-like length with pure tone stimulus.   Based on Martin & Boothroyd, 2000   27  AEP type Stimulus Description References or Explanation SCP and ACC  Broadband stimuli:  Noise bursts, 30 ms, 60 ms, 120 ms, & 480 ms duration, 5, 10, & 20 ms rise/fall each for .5, 1, and 2 kHz, at 0.9/s; 30 ms noise bursts only have 5 and 10 ms rise times     Based on Joutsiniemi, Hari, & Vilkman, 1989; systematic variation of stimulus duration and rise/fall time ABR Tones: 2-cycle rise – 1-cycle plateau – 2-cycle fall (2-1-2) tones at .5, 1, 2 kHz, 39.1/s, alternating polarity  Clicks: 1 kHz square wave, 100 us duration, at 19/s; rarefaction and condensation  Based on Purdy & Abbas, 2002, BC Early Hearing Program, 2012    Based on Stapells, D.R. (2000). BC Early hearing Program, 2012   ASSR 0.50, 1.0, 2.0, and 4.0 kHz, 100% amplitude modulated (AM) at 78, 86, 93, and 101 Hz. Presented simultaneously (80 Hz response)  0.50, 1.0, 2.0, and 4.0 kHz carrier frequencies, 100% AM at 37, 39, 41, and 43 Hz. Presented simultaneously (40 Hz response) Based on Van Maanen & Stapells, 2005     Modulation rates based on John et al, 1998   MLR Tones:  4 ms rise/fall time, 2 ms plateau .5, 1 and 2 kHz 9.1/s, alternating polarity; 2-1-2 tones presented at 10/s  Clicks:  Rarefaction, 1 kHz square wave, 100 us duration, at 9.7/s    Based on Xu, De Vel, Vinck, and van Cauwenberge, 1997    Based on Schochat, Musiek, Alonso, & Ogata, 2010; Weihing & Musiek, 2008 MMN Speech:  /ba-da/ - 150 ms long; duration shortened by removing cycles from the vowel portion; 90% standard stimulus, /ba/, 10% deviant, /da/; 627 ms SOA.   Based on Korczak, Kurtzberg, Stapells, 2005 Table 2.1. Description of stimuli used   2.3 Calibration procedures 2.3.1 Sound field calibration   A calibrated position was found in sound-field using procedures outlined by Walker, Dillon, and Byrne (1984). The first step was to find a horizontal position at which there was a variation of less than 2 dB in the output measurements across an approximately 30 cm span. Using a Larson-Davis 824 Sound Level meter, a 1⁄2” microphone, and a microphone stand, the output of 250 Hz, 500 Hz, and 1000 Hz pure tone stimuli were measured. The 28  microphone was oriented such that the face plate of the microphone was perpendicular (i.e., facing the top of the sound booth) to the direction of the traveling sound source for the most accurate measurements. Subsequently, measurements were made at the edges of a sphere with a 30 cm diameter to determine whether the output was stable at the edges of head of KEMAR (i.e., measurements at the edge of the horizontal, vertical, and sagittal plane were measured). Generally, a difference of 2 dB or less from the edge to the center was preferred.  Although narrowband noise is often used for calibration purposes to minimize the effect of standing waves, the set of stimuli included pure tones. Thus, pure tones were used for calibration. For this reason, allowance for variations larger than 2 dB was given at several frequencies as long as the difference was within 10 dB, and that the variation stayed within 2.5 dB at most frequencies. 2.3.2 Stimulus Calibration Stimulus calibration was conducted using a Larson-Davis 824 sound level meter, a ½” microphone, and a microphone extension cable. A stand was used to hold the microphone close to the location of the hearing aid microphone above the pinna of KEMAR, in its calibrated position inside the sound booth. The sound level meter was operated outside of the sound booth to minimize any body reflections that may affect the measurement. The microphone was positioned such that the face of the microphone was facing the loudspeaker. Depending on the duration of the stimulus, the intensity measure and speed of the detectors differed: brief tones and clicks, such as those used to evoke ABRs and MLRs, were calibrated and measured using dB peak-to-peak equivalent (ppe) SPL, while all other stimuli were calibrated and measured in A-weighted dB SPL. All measurements were made using a fast detector, including peak measurements. The correction factor to convert the measured 29  flat dB Peak SPL to ppe SPL was -3 dB. This correction factor was used for both brief tone pips and clicks. This is standard for tones, based on their symmetry with reference to the zero line; this correction factor was also used for the rarefaction click in this study due to its symmetry about the zero line in sound field (Campbell et al., 1981, Stapells, Picton, & Smith, 1982). 2.4 Experimental procedures 2.4.1 Hearing aid programming The frequency response of the hearing aids were programmed, using NOAH module software provided by the manufacturer, for a standard audiogram of a sloping mild to moderately-severe hearing loss, to DSL v.5 adult targets at 65 dB SPL (audiogram “N3”, Bisgaard, Vlaming, & Dahlquist, 2010). These settings were verified by coupler measurements in the Audioscan Verifit 2 system using Speechmap™ (Audioscan, 2014). The long term average speech spectrum at 65 dB SPL was matched for both gain settings of the same hearing aid (i.e., linear and WDRC). The WDRC intended settings included a compression ratio of approximately 2:1 across the frequency channels, an attack time of less than 10 ms, a release time greater than 100 ms, and an input compression threshold of  50 dB SPL. An input compression threshold of 50 dB SPL was chosen based on the proposed compression thresholds recommended for a moderate hearing loss by Scollie et al. (2005).  Attack and release times were selected based on previous studies noting the importance of redundant temporal cues for phoneme recognition that could be distorted with shorter release times (e.g., Jenstad & Souza, 2005).  The compression threshold of 50 dB SPL could only be programmed in Siemens. Attack and release times were either not programmable or only customized with limited 30  options in Phonak and Starkey. Attack and release times were measured in a 2 cc coupler using the Fonix 7000 Hearing Aid Testing system using tonal stimuli. See Table 2.2 for compression thresholds, measured from input-output curves generated using 250 Hz to 4000 Hz tones, and attack and release times, also measured at those frequencies, for each hearing aid. The only instance where attack times were longer than release times was for a 500 Hz tone presented to Phonak. See Appendix A for hearing aid settings and test box measurements.  Table 2.2.2Measured compression thresholds, attack and release times using 250 Hz to 4000 Hz input stimuli for all hearing aids 2.4.2. Stimulus presentation  To observe effects of compression, stimuli were presented at two levels:  45 and  65 dB SPL. Brief stimuli were presented at 65 and 85 dB ppe SPL. Unaided recordings were made with KEMAR and the artificial ear. Recordings of processed stimuli were made with each of the three hearing aids on the left ear of KEMAR with WDRC inactivated. Finally, the same was done with WDRC activated in the hearing aids. Each stimulus recording contained multiple repetitions of the stimulus with recordings lasting up to one minute long. The number of repetitions within the recording differed for each set of stimuli due to the presentation rate, as outlined in Table 2.1. Each recording day included a recording of silence to measure the noise floor. The majority of the data collection occurred over four days, two days for each intensity condition containing unaided recordings as well as aided recordings Hearing aid Compression Threshold ( dB SPL) Attack Time (ms) Release Time (ms) Phonak  < 40 to 55 17.5 to 140  17.5 to 40 Siemens 50 to 65  12.5 to 15 50 to 150 Starkey < 40  27.5 to 32.5 250 to 369.4 31  from the three hearing aids in two hearing aid conditions. To prevent order effects, the order of the hearing aids was randomized for each intensity and hearing aid condition (i.e. for each hearing aid and each gain setting within the same hearing aid). Each stimulus set, categorized by AEP type, was also randomized for each recording day and each hearing aid condition. New batteries were used on each recording day for all hearing aids and the amplifier used with KEMAR.   2.5 Acoustic analysis Quantitative descriptive analyses of the acoustic changes after hearing aid processing were made. The acoustic parameters that were measured include: rise and fall times, duration, intensity, signal-to-noise ratio, and spectrum. For speech stimuli, the following were also measured: voice onset time (VOT) for stops, fricative duration for fricatives, and first and second formant frequencies for the vowels. For ASSR stimuli, the F-ratio at the modulation frequencies and modulation depths of the stimuli were measured.  Analyses were mainly done with MATLAB, Audacity®, and Praat.  2.5.1 Initial treatment of stimuli  A fast Fourier transform (FFT) of the recorded silence revealed high noise levels below 40 Hz and above 8000 Hz. Thus, all stimuli were passed through a high and low pass filter with a q-factor of 2, a slope of 48 dB per octave, and cut-off frequencies of 37.655 Hz,  8000 Hz for the high and low pass filters, respectively. The filtering was performed in Audacity®. Amplification of each stimulus set, grouped by AEP type and further by tonal, broadband, and speech material when applicable, was also done for better visualization of stimuli. Each stimulus set in each intensity condition was amplified by the same amount of gain to preserve relative differences between hearing aid conditions; all dB values were 32  calculated in reference to one volt. See Table 2.3 for details on the amount of amplification applied to each set.  Table 2.3.3Amplification applied to each set of stimuli for visualization and analysis  All stimuli, other than ASSR, were averaged over all repetitions within the recording for clearer visualization. Averaging was done separately for each polarity for stimuli presented in alternating polarity. Measurements of the acoustic parameters were made on the averaged stimuli, when possible, but observations of hearing aid effects that span over multiple stimuli were also made. Noise burst stimuli were created by stringing ten different noise bursts generated using the random function on MATLAB, at the rate identified in Table 2.1 and repeating this string over the one minute recording, alternating in polarity; thus, averages were made over each ten noise bursts, and only the fifth noise burst in each string was used for analysis.   2.5.2 Measurement of acoustic parameters   MATLAB codes were written to measure the acoustic parameters outlined in section 2.5, and to automate some of these measurements. Rise and fall times were measured using Set of Stimuli  Intensity Condition Amplification Applied  (dB) ABR tone bursts & click 85 ppe SPL  8 65 ppe SPL  33.5  ASSR 40 Hz and 80 Hz AM modulated tones  65 dB SPL  37.3 45 dB SPL  62.7 MLR tone bursts & click 85 ppe SPL  8 65 ppe SPL  32.7 SCP Speech & ACC Speech 65 dB SPL  27.8 45 dB SPL  59.4 SCP Noise 65 dB SPL  44 45 dB SPL  57.1 SCP Tones 65 dB SPL  54.6 45 dB SPL  70.3 HEARLab™ stimuli 65 dB SPL  40.7 45 dB SPL  59.4 ACC tonal stimuli 65 dB SPL  62.7 45 dB SPL  73.4 MMN Speech 65 dB SPL  50 45 dB SPL  65.2 33  functions within MATLAB (The MathWorks, Inc., 2015a, 2015c). To use these functions, the absolute value of each stimulus was taken to half-wave rectify the stimulus, and a Butterworth filter with a power of 2 and a cut off frequency of 88.2 Hz was applied to the stimulus to obtain its envelope. Due to the extraneous peaks and troughs in the recordings of the SCP noise data and their effects on the rise and fall time measurements, the envelopes of the SCP noise recordings were obtained by passing this set of stimuli through a Butterworth filter with a power of 2 and a cut off frequency of 44.1 Hz. The cut off frequency of the Butterworth filter for ASSR stimuli was adjusted to 352.8 Hz for the accurate formation of the envelope for modulation depth measurements. Due to the brief nature of the ABR and MLR stimuli, the cut off frequency of the Butterworth filter also needed to be adjusted for accurate rise and fall time measurements: tone bursts used a cut off frequency of 441 Hz and clicks used a cut off frequency of 2205 Hz. Envelopes were obtained because both the rise and fall time functions within MATLAB require a bi-level function with few oscillations in the stimulus: the rise time function measured the number of samples it took for the stimulus to reach 90% of the maximum plateaued amplitude from 10% of the total amplitude; the fall time function measured the opposite, which was the number of samples it took for the stimulus to reach 10% of the amplitude from 90% of the maximum amplitude (The MathWorks, Inc., 2015a, 2015b). The Butterworth filter was used due to its flat frequency response in the passband, to preserve the original frequency response as much as possible (Selesnick & Burrus, 1998). In stimuli where an additional fall was observed between the peak and the plateau of the stimulus, this fall time was measured by cutting the stimulus to focus on the region of the stimulus that contained this fall time. Hereafter, this fall time was referred to as the additional fall time.  34  Intensity was measured using the root-mean-squared (RMS) function within MATLAB (The MathWorks, Inc., 2015d). This value was then converted into dB volt in order to compare relative dB differences within each stimulus set (The MathWorks, Inc., 2015b).  Signal-to-noise ratio was calculated by subtracting the intensity measured from noise only, in dB volt, from the intensity level of the averaged stimulus, also measured in dB volt. The spectra of the stimuli were obtained by performing an FFT on each stimulus; this code was adapted and modified from the original code written by Zhiromirov (2012) in order to measure the amplitude at individual frequencies.  To measure modulation depth for ASSR stimuli, the envelope of the recording was taken. First, the absolute value of the stimulus was taken. Then, the stimulus was low-pass filtered using a Butterworth filter with a power of 2 and a cut off frequency of  88.2 Hz. The minimum and maximum amplitude were then measured from the envelope, and the percent modulation was calculated, per the method was explained by Frenzel Jr. (2008).  This was done by subtracting the minimum amplitude from the maximum amplitude, and dividing this number by the sum of the minimum and maximum amplitude, and then multiplying this value by 100 (Frenzel Jr., 2008). See Equation 1.   𝑀𝑜𝑑𝑢𝑙𝑎𝑡𝑖𝑜𝑛 𝑑𝑒𝑝𝑡ℎ =𝐴𝑚𝑎𝑥− 𝐴𝑚𝑖𝑛𝐴𝑚𝑎𝑥+ 𝐴𝑚𝑖𝑛× 100           (Equation 1) Duration was determined by identifying the start and end points of the signal by auditory and visual examination and calculating the difference between these two points. The end of the signal was determined visually in Audacity® at maximum zoom by locating the latest sample at which periodicity was still observed. This point was confirmed auditorily by checking that no part of the stimulus could be heard after the point that was determined 35  visually. For brief stimuli included in the ABR and MLR set of stimuli, the most intense part of the stimulus was measured for comparison between aided and unaided stimuli.  The duration was determined by calculating the difference between the start and the point at which at which the stimulus reached its lowest amplitude before the ringing of the stimulus began. Due to the short SOA and reverberation effects of the room, ABR stimuli continued to ring in the silence between stimuli, leading to no definitive end point before the beginning of the next stimulus.  For MLR stimuli, the SOA was long enough for the ringing to dampen; thus, the end point of the ringing was also identified to determine the full duration of the stimulus. Measurements of signal-to-noise ratio (SNR), intensity, and rise and fall times were automated using custom MATLAB code (©A. Chau) and modified as necessary. Any additional changes in intensity within the stimuli, or fall times from stimuli displaying an initial intensity overshoot, were measured using the MATLAB codes described in the beginning of section 2.5.2. The acoustic parameters pertaining to speech were also measured from the averaged signal in Praat. VOTs and fricative durations were determined by identifying the start point and the location at which periodicity begins, by visual and auditory examination of the stimuli. VOTs for stops were enlarged to a time window of approximately 0.07 s, allowing for visualization of the consonant-vowel transition. The time between the release burst and the closest zero-crossing to the beginning of periodicity, confirmed visually and auditorily, was taken as the VOT.  Fricative durations were measured in a similar manner as VOTs; the time between the beginning of the consonant and the closest-zero-crossing to the beginning of periodicity of periodicity was measured. Formant frequencies of vowels were also 36  measured in Praat, where the parameters were set to optimize the formant tracker in Praat. Formants were identified at the time point at which the formants were unchanging, near the mid-point of the stimulus. The same time point was used for the same vowels in each condition. The F1-F2 difference was subsequently calculated after these measurements. See Table 2.4 for details on the Praat parameters and the time point at which the formants were measured.  Table 2.4.4Praat settings for optimized formant measurements  ASSR stimuli were analyzed for additional energy at the modulation frequencies as a possible result of non-linearities of hearing aid processing; energy at the modulation frequencies would confound the physiological response. F-ratio calculations comparing the variance of the energy measured at the modulation frequencies to the variance of the energy measured in adjacent frequencies were calculated by separating the recordings of the ASSR stimuli into one-second segments. An FFT was performed on each segment, and the energy at the frequency bins of interest were measured: for stimuli modulated at frequencies centered at 80 Hz, information from frequency bins containing 76 Hz to 103 Hz were     Vowel Formant  and Spectrogram Settings Spectrogram settings:  window length  (s) Formant Settings: Maximum formant (Hz)  Formant Settings: Number of formants Formant settings: Window Length Measuring Time point (s) /a/  cut post-recording  0.008 4000  5 0.025 0.132 /a/ cut pre- recording 0.008 4000  5 0.025 0.134 /a/  0.008 4000  5 0.025 0.050 /i/  0.005 4000  5 0.01 0.028 /a/ in /ba/ *  0.008 4000  5 0.025 0.095 /a/ in /da/ * 0.008 4000  5 0.025 0.113 Other settings were left at default; *speech stimuli from the set of MMN stimuli   37  extracted; for stimuli modulated at frequencies centered at 40 Hz, information from frequency bins containing 35 to 45 Hz were extracted. As the bin width of the FFT was  0.2 Hz, the information from the five bins containing the modulation frequencies was used. The variance of the energy at the frequency bins containing the modulation frequencies was measured and compared to the variance of the frequency bins containing the noise within the range outlined. The F-ratio was calculated by dividing the variance of the noise bins by the variance of the bins containing the modulation frequency. Measurements and calculations were done in MATLAB.                 38  Chapter 3: Results and Interpretations 3.1 Data collection summary For this study, 518 recordings were collected in each intensity condition. As noted, all recordings other than stimuli in the ASSR and MMN set of stimuli were presented in alternating polarity; each polarity of each stimulus was separately analyzed. All the acoustic parameters, including rise and fall times, duration, intensity, signal-to-noise ratio, spectra, the difference between the first and second formant frequencies for vowels, VOTs for stop consonants, fricative duration for fricative consonants, modulation depth and F-ratios were measured as specified in section 2.5.2.   3.2 Hearing aid processing delays   Processing delays of the hearing aids were measured on the Fonix 7000, shown in Table 3.1. As all stimuli were manually aligned prior to the acoustic analysis, these delays are effectively removed from the data analysis.   Table 3.1. Hearing aid processing delays measured on the Fonix 7000  3.3 Rise and fall time  3.3.1 SCP Tones  To observe changes to rise and fall times as a function of tone duration, stimulus duration was systematically varied. Each tone of each frequency was generated to have rise and fall times of 5 ms, 10 ms, and 20 ms, all of which were below 30 ms, where N1 amplitudes of the SCP are not expected to be affected significantly, as outlined in Table 1.1. See Figures 3.1 and 3.2 for measured rise times in the 65 dB SPL and 45 dB SPL intensity  Hearing Aid Processing Delay (ms) Linear Compression Phonak  7.1 7.1 Siemens  6.4 6.4 Starkey  4.2 4.3 39  conditions, respectively, and Figures 3.3 and 3.4 for measured fall times in the 65 dB SPL and 45 dB SPL intensity conditions, respectively.  Both polarities of each stimulus and tones of the same frequency with stimulus rise times of 5 ms, 10 ms, and 20 ms, were all included in the same graph. Additional fall times from the peak of the stimulus to the plateau were noted in some conditions. See panel A in Figure 3.5 for an example of the additional fall time measured in a 60 ms duration tone. See Figures 3.6 and 3.7 for additional fall times of tones presented at  65 dB SPL and 45 dB SPL, respectively. Both polarities of each stimulus were included in each graph. In the 65 dB SPL condition, this phenomenon was noted for most 1000 Hz and  2000 Hz tones with durations of 60 ms or longer recorded from Starkey (compression) and Phonak (compression). This phenomenon did not occur for 60 ms duration tones with 20 ms rise and fall times as the additional fall time combined with the primary fall time to the end of the stimulus (see panel B in Figure 3.5). Thus, no data point is shown for this stimulus in Figure 3.6 and 3.7. See Appendix B for a visual representation of the envelope and waveform.  40   Figure 3.1.2 Rise times, ms, in unaided and aided conditions for all tones of different durations and stimulus rise times presented at 65 dB SPL; linear gain is in the top panels, and compression is in the bottom panels. Rise times were plotted as a function of duration; logarithmic trend lines were plotted for each data series. The grey dashed horizontal line at 30 ms indicates the point above which rise time is expected to affect SCP amplitude and latency.     Compression 41   Figure 3.2.3 Rise times, ms, in unaided and aided conditions for all tones of different durations and stimulus rise times presented at 45 dB SPL; linear gain is in the top panels, and compression is in the bottom panels. Rise times were plotted as a function of duration; logarithmic trend lines were plotted for each data series. The grey dashed horizontal line at 30 ms indicates the point above which rise time is expected to affect SCP amplitude and latency.   Compression 42   Figure 3.3.4Fall times, ms, in unaided and aided conditions for all tones of different durations and stimulus fall times presented at 65 dB SPL; linear gain is in the top panels, and compression is in the bottom panels. Fall times were plotted as a function of duration; logarithmic trend lines were plotted for each data series.   Compression 43   Figure 3.4.5Fall times, ms, in unaided and aided conditions for all tones of different durations and stimulus rise times presented at 45 dB SPL; linear gain is in the top panels, and compression is in the bottom panels. Fall times were plotted as a function of duration; logarithmic trend lines were plotted for each data series. Compression 44   Figure 3.5.6Waveforms of recorded stimuli of 60 ms duration processed by Starkey (compression). The stimulus waveform is in blue and the envelope of the stimulus is in black. Different panels show recordings with different stimulus rise and fall times: (A) 5 ms; (B) 20 ms. Arrowheads denote points between which fall times were measured.      A  B  45   Figure 3.6.7Additional fall times, ms, of stimuli presented at 65 dB SPL with compression active in hearing aids.       46         Figure 3.7.8Additional fall times, ms, of stimuli presented at 45 dB SPL with compression active in hearing aids. 47  3.3.1.1 Interpretation of rise and fall time effects on SCP Tones  The acoustic changes differed based on the frequency of the pure tones presented, duration of tones, the hearing aid processing, the stimuli, and the gain settings. As outlined in Table 1.1, measured rise times below 30 ms were not expected to have any significant effects, while measured rise times between 30 ms to 50 ms were expected to cause N1 amplitudes to decline; no further significant declines are expected with rise time elongations greater than 50 ms (Alain et al., 1997; Kodera et al., 1979; Onishi & Davis, 1968). The measured rise times depended on hearing aids and their gain settings across the durations. With linear gain, rise times measured from Starkey were similar to unaided recordings for tones of all durations, while recorded tones of 60 ms duration or more from Phonak and Siemens had longer rise times that increased with increasing stimulus duration. With compression, depending on frequency and the intensity condition, rise times of pure tones recorded from Phonak and Starkey were generally shortened as compared to those measured in unaided recordings, while, the elongations of rise times were shorter in recordings from Siemens (compression) as compared to Siemens (linear), with the exception of stimuli presented at 45 dB SPL to Siemens (compression) and 2000 Hz tones presented at 65 dB SPL to Siemens.   Unlike rise time, the fall times recorded in each hearing aid condition and intensity condition were similar to unaided, as shown in Figures 3.3 and 3.4.  The observation of shorter rise times in Starkey (compression), and additional fall times in stimuli processed by hearing aids with compression, was consistent with attack time overshoot of intensity seen with compression due to delays in the application of the compressor (Souza, 2002).   48  As evidenced, hearing aid processing can have large effects on rise times in the absence of large changes to fall times. Asymmetrical changes to rise and fall times observed with recordings from Siemens and Phonak can lead to unknown electrophysiological effects. For stimuli with longer durations, such as those greater than 100 ms, a separate offset response may be expected (Ganapathy, Narne, Kaliah, & Manjula, 2013; Onishi & Davis, 1968). As measured fall times were similar to unaided stimuli, the offset response would most likely not be affected by hearing aid processing, although this remains to be investigated.   Aided SCPs are mainly used in the context of onset responses. Hearing aid conditions that yield similar rise times should not affect SCP morphology.  In conditions where rise times are increased, especially for stimuli of 60 ms duration or longer, increased latency and decreased amplitudes may be expected, especially when rise times are measured between 30 to 50 ms. Decreased latencies and increased amplitudes may be expected when rise times became shorter with compression, although differences may not reach statistically significant levels, as outlined in Table 1.1.  The expected trends of increases in latency and decreases in amplitude would be different depending on the duration of the stimulus used, as the first  30 ms of the eliciting stimulus is the most influential part of the stimulus, in terms of latency and amplitude (Onishi & Davis, 1968).  3.3.2 SCP Noise  Rise and fall times were measured as function of noise burst duration. See Figures 3.8 and 3.9 for measured rise times in the 65 dB SPL intensity condition and 45 dB SPL intensity conditions, respectively, and Figures 3.10 and 3.11 for measured fall times in the 65 dB SPL and 45 dB SPL intensity conditions, respectively. Similar to tones, noise bursts were 49  generated to have rise and fall times of 5 ms, 10 ms, and 20 ms, all of which were below 30 ms, where N1 amplitudes of the SCP are not expected to be affected significantly, as outlined in Table 1.1. Both polarities of each stimulus and recordings with all presented stimulus rise times of 5 ms, 10 ms, and 20 ms, were included in the same graph. Additional fall times were only observed in the 65 dB SPL intensity condition and only for 120 ms and 480 ms noise bursts recorded from Starkey (compression). See Figure 3.12 for the measured additional fall times.50   Figure 3.8.9Rise times, ms, of noise bursts of different durations and stimulus rise times, presented at 65 dB SPL in unaided and aided conditions. The grey dashed horizontal line at 30 ms indicates the point above which rise time is expected to affect SCP amplitude and latency.     0204060801001201401601800 100 200 300 400 500 600Rise time (ms)Duration (ms)UnaidedStarkey LinearSiemens LinearPhonak LinearStarkey CompressionSiemens CompressionPhonak Compression51     Figure 3.9.10Rise times, ms, of noise bursts of different durations presented at 45 dB SPL in unaided and aided conditions. The grey dashed horizontal line at 30 ms indicates the point above which rise time is expected to affect SCP amplitude and latency.  0204060801001201401601800 100 200 300 400 500 600Rise time (ms)Duration (ms)UnaidedStarkey LinearSiemens LinearPhonak LinearStarkey CompressionSiemens CompressionPhonak Compression52      Figure 3.10.11Fall times, ms, of noise bursts of different durations presented at 65 dB SPL in unaided and aided conditions.   01020304050600 100 200 300 400 500 600Fall time (ms)Duration (ms)UnaidedStarkey LinearSiemens LinearPhonak LinearStarkey CompressionSiemens CompressionPhonak Compression53   Figure 3.11.12Fall times, ms, of noise bursts of different durations presented at 45 dB SPL in unaided and aided conditions.    01020304050600 100 200 300 400 500 600Fall time (ms)Duration (ms)UnaidedStarkey LinearSiemens LinearPhonak LinearStarkey CompressionSiemens CompressionPhonak Compression54   Figure 3.12.13Additional fall times, ms, of noise bursts of different durations presented at 65 dB SPL from Starkey (compression).  0510152025300 100 200 300 400 500 600 700 800Fall time (ms)Duration (ms) Starkey Compression55  3.3.2.1 Interpretation of rise and fall time effects on SCP Noise  The acoustic changes differed based on the duration of noise bursts, hearing aids, and the gain settings. As outlined in Table 1.1, measured rise times below 30 ms were not expected to have any significant effects, while measured rise times between 30 ms to 50 ms were expected to cause N1 amplitudes to decline; no further significant declines are expected with rise time elongations greater than 50 ms (Alain et al., 1997; Kodera et al., 1979; Onishi & Davis, 1968). In the 65 dB SPL condition, the rise times measured from Starkey (linear) were similar to unaided, while Siemens (linear) and Phonak (linear) produced measured rise times that increased with an increase in duration. With compression, Starkey and Phonak both produced measured rise times that were shorter than those measured from unaided recordings. The elongation of rise times was also shorter from Siemens (compression) as compared to Siemens (linear). In the 45 dB SPL condition, the rise times measured from Siemens with linear gain or compression and Starkey (linear) were similar to those noted for each hearing aid in the 65 dB SPL condition, while Starkey (compression) and Phonak (compression) produced rise times that were similar to unaided, suggesting that compression was not activated. Unexpectedly, Phonak with both gain settings in this intensity condition produced measured rise times that overlapped with unaided recordings. The difference in frequency composition of the noise burst may explain the different patterns observed with noise bursts processed by Phonak as compared to the processing behaviour observed in SCP Tones as hearing aids are known to process broadband and narrowband signals differently.  As shown in Figure 3.12, additional fall times were only observed and measured in the 65 dB SPL condition when stimuli of 120 ms and 480 ms duration were processed with Starkey (compression). Again, this fall time was expected and consistent with the activation 56  of the compressor in the hearing aid. Measured fall times also followed a similar pattern as observed for SCP Tones, although the spread of the measured fall times was also larger than those measured for SCP Tones. The longer fall time measured for noise bursts at 60 ms duration with stimulus rise and fall times of 20 ms processed by Starkey (compression) gain was likely due to the combination of the additional fall time and the fall time to end the stimulus. Most fall times were longer than unaided in all hearing aid conditions in the 45 dB SPL condition, possibly due to the improved SNRs in aided conditions, discussed in section 3.6.   As noted with SCP Tones, rise and fall times changed asymmetrically after hearing aid processing in some hearing aids although the stimuli had symmetrical rise and fall times. Based on Table 1.1, hearing aid processing that increased rise times in the eliciting stimuli would be expected to lead to decreased amplitudes of the SCP, up to rise times of approximately 50 ms (i.e., stimuli processed by Siemens), while decreases in rise times (i.e., stimuli presented at 65 dB SPL processed by Starkey (compression) and Phonak (compression) and stimuli presented at 45 dB SPL processed by Starkey (compression)) could lead to non-significant decreases in latency and increases in amplitudes. Increases in latencies are expected with increased rise times greater than 30 ms, with greater increases noted with a less intense presentation level. Decreases in SCP amplitude should not reach statistical significance until rise times are above 30 ms (Onishi & Davis, 1968). Therefore, the expected increases in latency and decreases in amplitude would be different depending on the duration of the stimulus used as well.   57  3.3.3 ACC tonal stimuli  ACC tonal stimuli were generated with a 10 ms rise and fall time. The rise and fall time was measured for stimuli in all hearing aid and intensity conditions. Rise times at the point at which the stimulus changes from 1680 Hz to 1705 Hz, hereafter noted as transition rise times, were also measured. See Table 3.2 for measured rise times, fall times, and transition rise times for all hearing aid and intensity conditions. Data for opposite polarity are not shown.   Table 3.2.2Measured rise times, fall times, and transition rise times, ms, in all hearing aid and intensity conditions.3.3.3.1 Interpretation of rise and fall time effects on ACC tonal stimuli recordings   The initial rise time, although measured and reported, is not relevant for the interpretation of ACC morphology, given the definition of the ACC as a response that is elicited by a change in a continuous stimulus. Thus, they are not interpreted further in this section. Hearing aid Intensity condition Gain setting Rise time (ms) Fall time (ms) Transition  rise time (ms) Unaided 65 dB SPL N/A  11.8 13.1 21.1 Phonak  65 dB SPL Linear 88.5 13.7 13.7 Compression 4.9 12.9 13.1 Siemens  65 dB SPL Linear 130.8 13.6 18.2 Compression 130.4 13.2 18 Starkey  65 dB SPL Linear 10.1 13.6 14.7 Compression 5.2 12.7 6.1 Unaided 45 dB SPL  N/A  11.2 10.9 N/A Phonak  45 dB SPL  Linear 80.2 11 9.3 Compression 16.2 12.5 8.4 Siemens  45 dB SPL  Linear 131.3 11.4 15.5 Compression 134.5 11.3 30.9 Starkey  45 dB SPL  Linear 8.5 11.9 14.7 Compression  7.8 11.7 6.9 58  Transition rise times were noted at the point when the stimulus frequency changes from 1680 Hz and 1705 Hz, although a transition rise time was not present in the stimulus prior to recording. Unlike patterns noted for initial rise times, transition rise times measured in all aided conditions were similar to, but shorter than, the unaided condition in the 65 dB SPL condition. Starkey (compression) and Phonak (compression) had transition rise times that were at least 5 ms shorter than unaided. These may have been the effects of the activation of the compressor, as observed with SCP Tones, and as noted by the shortening of rise times with intensity overshoots (Souza, 2002). It is important to note that no intensity overshoot was observed at the transition. However, as rise times effects have not been studied in a systematic manner for ACCs, the effects of these small changes remain to be investigated. Similar to the effects seen in SCP Tones, measured fall times from aided conditions were similar to unaided in both intensity conditions. It is unlikely that hearing aid processing would affect the ACC generated from the offset of the stimulus (Kodera et al., 1979; Onishi & Davis, 1968).  Similar to the discussion with SCP Tones, the effects of asymmetrical rise and fall times that can be introduced with hearing aid processing on aided ACC stimuli remain to be investigated.  3.3.4 SCP Speech and ACC Speech  Rise and fall times were measured for each syllable in each hearing aid condition. Possible onset effects that could affect ACC speech were studied by cutting /a/ from a recorded /ʃa/, referred to as /a/ cut post-recording, and comparing measured results with the recording of an /a/ that was cut from /ʃa/ prior to stimulus presentation, referred to as /a/ cut pre-recording. See Table 3.3 for measured rise times from speech sounds recorded in the 59  unaided condition and the difference between rise times measured from aided stimuli as compared to their unaided counterparts. The difference for each stimulus was calculated by subtracting the rise time measured in the unaided condition from those measured in the aided condition. See Table 3.4 for fall times of unaided recordings and the differences calculated between aided and unaided conditions for each stimulus. Opposite polarity data are not shown.  60  Table 3.3.3Measured rise times, ms, of speech sounds in the unaided condition and the differences calculated between aided and unaided conditions, with speech sounds presented at 65 dB SPL and 45 dB SPL.    .  Stimulus  Intensity Condition Unaided rise time (ms) Rise time difference relative to unaided (ms) Hearing Aid   Phonak Siemens Starkey Gain setting   Linear Compression Linear Compression Linear Compression /da/ 65 dB SPL 30.6 27.3 1.9 5.4 4.8 4.9 -28.5 /ta/ 16.3 -7 -8.7 2.1 -0.2 -3.4 -9.6 /sa/ 27 12.7 -6.8 17.7 17.4 27.3 -14.5 /ʃa/ 72.7 -12.2 -26.8 -3.5 14.3 15.4 -58.2 /a/ 39 -10.1 -11.1 -8.4 -6.7 -0.7 -22.3 /i/ 33.1 -15.1 -20.6 -7.6 0.2 -0.5 -20 /a/ cut pre-recording 32 19 -22.3 20.2 19.3 13.5 -24.8 /a/ cut post-recording 30.6 5.2 -26 12.7 13.5 14.2 -23.9 /da/ 45 dB SPL 17.4 30.8 11.2 21.6 21.6 6.5 17.1 /ta/ 17.1 -6.2 -4 -6 -5 -4.9 -10.2 /sa/ 22.6 46.1 4.5 20.2 19.5 26.4 -0.3 /ʃa/ 54.1 4 19.2 16 20.6 2.2 7.4 /a/ 29.7 0.9 11.4 14 10.8 -2 -9.4 /i/ 23.8 2.5 -1 9 9.3 1.9 1.5 /a/ cut pre-recording 35.2 17.7 -3 15.5 7.3 8.4 -22.9 /a/ cut post-recording 36 14.9 -2.1 7.2 8.2 7.8 -26.4 61   Table 3.4.4Measured fall times, ms, of speech sounds in the unaided condition and the differences calculated between aided and unaided conditions, with speech sounds presented at 65 dB SPL and 45 dB SPL.  Stimulus  Intensity Condition Unaided fall time (ms) Fall time difference relative to unaided (ms) Hearing Aid   Phonak Siemens Starkey Gain setting   Linear Compression Linear Compression Linear Compression /da/ 65 dB SPL 139 27 55.2 36 64.6 33.3 80.8 /ta/ 175.4 -66.7 -9.7 -77 -42.5 -45.3 -110.5 /sa/ 104.2 -77.7 42 238.2 238.9 37.6 52.9 /ʃa/ 158 45.1 20.8 52.8 56.5 46.3 88 /a/ 11.5 -0.8 0.2 0.7 0.5 0.1 -0.7 /i/ 29.3 -8.2 -2.2 -9 -9 -8.6 -1.6 /a/ cut pre-recording 157.9 -38.2 84 45.9 7.2 43.1 138.7 /a/ cut post-recording 158 -38.1 -0.5 52.7 56.9 43 77.2 /da/ 45 dB SPL 119.4 -64.4 38.6 -28.5 -38.1 14.9 40.1 /ta/ 97.3 6.7 22.6 -12.4 -2.6 20.7 -0.2 /sa/ 100.9 -77.4 242.7 217.1 220.5 26.2 71.3 /ʃa/ 119.1 46.2 63.9 56.8 62.1 65.5 87.6 /a/ 9.5 -2.6 1 -1.4 -0.9 0 0.5 /i/ 24.1 -4.7 -1.2 -5.8 -3.5 -6.2 -5.6 /a/ cut pre-recording 120.8 21 61.8 59.1 -18 61.1 66.7 /a/ cut post-recording 119.2 -21.8 63.8 63.8 69.8 67.6 81.3 62  3.3.4.1 Interpretation of rise and fall time effects on SCP and ACC Speech  Hearing aid processing can change the measured rise times, among other acoustic effects, as seen in previous studies indicating significant acoustic changes with hearing aid processing, such as with the application of compression (e.g., Jenstad & Souza, 2005; Tremblay, Billings, et al., 2006; Souza, 2002). Changes observed were similar between polarities unless otherwise noted.  Easwar et al. (2012) showed that an 85.6 ms difference between the rise times of a word-medial /ʃ/ and word-initial /ʃ/ was significant, with the latter eliciting larger and earlier SCPs, as expected from a stimulus with a shorter rise time. As the largest difference seen in the current study with hearing aid processing was less than 85.6 ms, the effects remain to be investigated. However, a shortening of the rise time will likely lead to an earlier SCP while elongating the rise time will likely lead to a later SCP. As seen in the data, changes that occur may include both the elongation and shortening of the rise time depending on the hearing aid, the gain setting, and the syllable. The point at which these changes can lead to a significant difference in SCP latency and amplitude cannot be determined as there have been no studies noting the changes in SCP morphology with a systematic variation of only the rise time of speech sounds without changing other characteristics of the speech sound.  To observe any changes that may occur when using an SCP paradigm for vowels (i.e., presenting the vowel alone) versus an ACC paradigm (i.e., presenting the vowel connected to another phoneme), data measured from /a/ cut pre-recording and /a/ cut post-recording were compared. Changes observed differed depending on the intensity conditions. In the 65 dB SPL condition, /a/ cut post-recording had smaller differences from unaided than /a/ cut pre-recording, with the exception of Phonak (compression) and Starkey (linear), 63  which had differences from unaided that were greater by 3.7 ms and 0.7 ms post-recording, respectively. The same pattern was observed in the 45 dB SPL condition, although the two aided conditions that had greater differences were Siemens (compression) and Starkey (compression), which had differences that were greater by 0.9 ms and 3.5 ms, respectively. Generally, there were similar or less variable onset effects seen from hearing aid processing when using the ACC paradigm.  Different hearing aids as well as gain settings can enlarge the differences on the onset of the vowel. Based on these data, the ACC paradigm may have stimuli that are more stable than the SCP paradigm, although it is important to note that differences from unaided can still reach 26.4 ms. The magnitude of difference that can elicit a significant difference in amplitude and latency in the SCP or the ACC remains to be investigated.  Isolated vowels had small differences in measured fall times between aided and unaided conditions in both intensity conditions. No hearing aid or gain setting had consistently larger differences from unaided for isolated vowels.  The largest differences noted in fall time were found for the syllable /sa/ in recordings from Siemens set in both gain settings for both intensity conditions, and for recordings from Phonak (compression) in the 45 dB SPL condition. The point at which the amplitude drops from 90% can be measured at an earlier point when consonant-vowel ratios are increased, causing longer fall times to be measured from /sa/ in some aided conditions. With a high-frequency emphasis, the consonant-vowel ratio was expected to increase. Even though all hearing aids were set to meet the same targets, small variations may have caused the consonant-vowel ratio to differ, leading to the observed variations among hearing aids. Hearing aids have been noted to affect consonant-vowel ratios, as noted by Tremblay, 64  Billings et al. (2006). See Figure 3.13 for a panel of waveforms to demonstrate how the fall time measurement began earlier in the 65 dB SPL condition. See Appendix B to see the waveforms in the 45 dB SPL condition for a visual of the interpretations above.  As noted in the previous section, effects of changes to fall time that were different from rise times have not been investigated, but changes of greater than 200 ms, which were measured for some processed stimuli, may lead to a later off-set or change response, such as the ACC, if the duration of the steady section is at least 50 to 80 ms (Sharma & Dorman, 1999). This remains to be investigated. However, the results illustrate once again the variety and diversity of the effects seen after hearing aid processing. 65  Figure 3.13.14Averaged stimulus waveforms of /sa/ in the 65 dB SPL condition. Different panels show different hearing aid conditions: (A) Siemens (compression); (B) Siemens (linear); (C) Unaided.  A B C 66  3.3.5 HEARLab™ stimuli Rise and fall times were measured in the recordings of phonemes used in the HEARLab™ system in all hearing aid and intensity conditions. See Table 3.5 for measured rise times from speech sounds recorded in the unaided condition and the difference between rise times measured from aided stimuli as compared to their unaided counterparts. The difference for each stimulus was calculated by subtracting the rise time measured in the unaided condition from those measured in the aided condition. See Table 3.6 for measured fall times of unaided recordings and the differences calculated between aided and unaided conditions for each stimulus. Opposite polarity data are not shown.  67  Table 3.5.5Measured rise times, ms, of HEARLab™ speech sounds in the unaided condition and the differences calculated between aided and unaided conditions, with speech sounds presented at 65 dB SPL and 45 dB SPL.  Table 3.6.6Measured fall times, ms, of HEARLab™ speech sounds in the unaided condition and the differences calculated between aided and unaided conditions, with speech sounds presented at 65 dB SPL and 45 dB SPL.    Stimulus  Intensity Condition Unaided rise time (ms) Rise time difference relative to unaided (ms) Hearing Aid   Phonak Siemens Starkey Gain setting   Linear Compression Linear Compression Linear Compression /m/ 65 dB SPL 14.4 6.1 7.3 8.1 -0.8 -4 -8.9 /t/ 20.7 -3.3 -0.9 0.9 0.7 1.9 -0.9 /s/ 6 2.2 -1.6 2.4 2.2 0.8 2 /g/ 5.7 1.4 -1.2 -1.2 -1.2 0.1 0 /m/ 45 dB SPL 14.9 2.9 -8.8 -4.3 -5 1.5 -4.7 /t/ 21.4 -8.1 -7.6 -0.4 -0.5 0.3 -2.5 /s/ 5.2 12.5 3.9 3.4 3 0.4 3.2 /g/ 5.3 0.4 0.5 -0.8 0.8 0.4 0.5 Stimulus  Intensity Condition Unaided fall time (ms) Fall time difference relative to unaided (ms) Hearing Aid   Phonak Siemens Starkey Gain setting   Linear Compression Linear Compression Linear Compression /m/ 65 dB SPL 13 -1.9 -1.2 -0.6 -1.4 -2.7 -2.8 /t/ 9.5 5.8 3.8 6.3 6.6 4.8 1.6 /s/ 12.2 -2.4 14.8 -1.7 -1.1 -1.9 29.2 /g/ 16.8 -5.2 -4.3 -5.1 -5.4 3.2 1.3 /m/ 45 dB SPL 10.6 -1.8 -5.1 -2.2 -2.8 -2.3 -2.6 /t/ 9 4.6 3.7 4.9 5.9 3.7 2 /s/ 11.9 -3.4 1.2 -1.8 -1.6 -1.4 32.2 /g/ 17.1 -10.4 -7.4 -8.3 0.8 1.7 1.6 68  3.3.5.1 Interpretation of rise and fall time effects on HEARLab™ stimuli  Although HEARLab™ stimuli are concentrated in specific regions in frequency, the point at which these changes can lead to a significant difference in SCP latency and amplitude cannot be determined as there have been no studies noting the changes in SCP morphology with a systematic variation of only the rise time of these speech sounds without changing other characteristics of the speech sound. For reference, the majority of the energy of /m/ is below 500 Hz; the peak energy of t/ is above 3000 Hz; the peak energy of /s/ is above 3600 Hz; the majority of the energy of /g/ is between 800 Hz and 1600 Hz (Carter, Golding, Dillon, & Seymour, 2010; Golding, Dillon, Seymour, & Carter, 2009; see section 3.7.5). Predictions of stability or possible directions of change were based on patterns noted in tonal stimuli (Kodera et al., 1979; Onishi & Davis, 1968).  From the four HEARLab™ stimuli, three of which are routinely used in their system, /g/ had the smallest differences in rise times compared to unaided conditions. Changes observed were similar between polarities unless otherwise noted. The small changes noted for /g/ are unlikely to cause significant unintended changes to the SCP morphology.  Measured rise times for /s/ showed that the aided conditions always yielded longer rise times than unaided. The longest rise times measured were from recordings processed by Phonak (linear) in the 45 dB SPL condition, with rise times that were 10.5 ms and 12.5 ms longer than unaided. Amplitudes and latencies are unlikely to be affected as the altered rise time is still below 30 ms (Onishi & Davis, 1968). No other hearing aids in either intensity condition would likely cause measurable effects on the SCP.   Measured rise times for /m/ differed by hearing aid and gain settings. Two of the three hearing aids produced rise times that were shorter than unaided, for which compression 69  seemed to shorten the rise times further in the 65 dB SPL condition.  Although the pattern of shortening and elongating rise times were not as clear between hearing aids in the 45 dB SPL condition, the activation of compression always yielded rise times that were shorter than their linear counterparts in both intensity conditions. This suggests that compression can play a factor in changing the rise time of the /m/ stimulus. If this is interpreted based on patterns noted in tonal stimuli, with differences in rise time ranging from 1.5 ms to 8.8 ms, changed rise times below 30 ms, and durations longer than 100 ms, SCP amplitude and latencies are unlikely to change significantly (Kodera et al., 1979; Onishi & Davis, 1968).  The phoneme /t/ had for a similar pattern as the phoneme /s/. The largest difference was also found from recordings measured from Phonak in either gain setting, with rise times that were shortened by 4.5 ms to 8.1 ms. These are not expected to change SCP amplitude and latencies as the resulting rise time was below 30 ms, if the data is interpreted using patterns noted in tonal stimuli.  Measured fall times were generally all similar to those measured from unaided for the four HEARLab™ stimuli. The only condition where fall times were at least 10 ms longer were in recordings from Starkey (compression) and Phonak (compression), which was expected with a higher peak of the stimulus from the attack time of the compressor. However, as all recorded durations were within 100 ms, an offset response is not expected.   3.3.6 ABR Tone bursts and click Rise and fall times were measured in the recordings of tone bursts and the rarefaction clicks in all hearing aid and intensity conditions.  See Table 3.7 for measured rise times of the tone bursts and Table 3.8 for measured fall times of the tone bursts.  See Table 3.9 for measured rise times of the rarefaction click and Table 3.10 for measured fall times of the 70  rarefaction click.  Rise and fall times were converted to cycles for tone-bursts, per the clinical definition.  Rise and fall times were represented in ms for the recorded rarefaction clicks. Opposite polarity data are not shown. 71  Table 3.7.7Measured rise times, cycles, of 2-1-2 tone bursts of each frequency presented in all hearing aid and intensity conditions.   Table 3.8.8Measured fall times, cycles, of 2-1-2 tone bursts of each frequency presented in all hearing aid and intensity conditions.   Stimulus  Intensity Condition  Rise time (cycles) Hearing Aid  Unaided Phonak Siemens Starkey Gain setting   Linear Compression Linear Compression Linear Compression 500 Hz 85 ppe SPL 1.8 1.9 1.5 1.6 1.5 1.8 1.6 1000 Hz 2 2.1 1.9 2 1.9 2 1.7 2000 Hz 2.4 3 2.8 2.6 2.6 2.6 2.6 4000 Hz 3.2 3.6 4 3.2 3.2 4 4 500 Hz 65 ppe SPL 1.7 1.5 1.6 1.9 1.6 1.8 1.4 1000 Hz 1.9 2 1.7 2.1 2 2 1.6 2000 Hz 2.4 3.2 2.6 2.4 2.6 2.6 2.6 4000 Hz 3.2 3.6 4 3.2 3.6 4 3.6 Stimulus  Intensity Condition  Fall Rise time (cycles) Hearing Aid  Unaided Phonak Siemens Starkey Gain setting   Linear Compression Linear Compression Linear Compression 500 Hz 85 ppe SPL 4.1 4.2 3.8 4.2 4.1 3.9 3.3 1000 Hz 2.6 2.6 2.2 3.5 2.9 2.7 2.8 2000 Hz 3.6 3.6 3.4 3.6 3.8 3.6 3.8 4000 Hz 6.8 8.4 7.6 6.8 6.8 7.2 6.8 500 Hz 65 ppe SPL 4.1 4.2 3.8 4.2 4.1 3.9 3.3 1000 Hz 2.6 2.6 2.2 3.5 2.9 2.7 2.8 2000 Hz 3.6 3.6 3.4 3.6 3.8 3.6 3.8 4000 Hz 6.8 8.4 7.6 6.8 6.8 7.2 6.8 72   Table 3.9.9Measured rise times, ms, of rarefaction clicks presented in all hearing aid and intensity conditions.  Table 3.10.10Measured fall times, ms, of rarefaction clicks presented in all hearing aid and intensity conditions.       Stimulus  Intensity Condition  Rise time (ms) Hearing Aid  Unaided Phonak Siemens Starkey Gain setting   Linear Compression Linear Compression Linear Compression Rarefaction Click 85 ppe SPL 0.3 0.2 0.3 0.2 0.3 0.2 0.2 Rarefaction Click 65 ppe SPL  0.3 0.2 0.7 0.3 0.2 0.2 0.1 Stimulus  Intensity Condition  Fall time (ms) Hearing Aid  Unaided Phonak Siemens Starkey Gain setting   Linear Compression Linear Compression Linear Compression Rarefaction Click 85 ppe SPL 0.8 1.6 1.6 1.2 1.1 1.3 1.3 Rarefaction Click 65 ppe SPL  0.8 0.8 0.8 0.8 1.1 0.8 0.4 73  3.3.6.1 Interpretation of rise and fall time effects on ABR Tone bursts and click  Changes between polarities were similar unless otherwise noted. Measured rise times for the unaided recordings demonstrated that room effects introduced by the sound field can elongate rise times for brief tones. In aided conditions, however, the rise times did not increase or decrease any more than one cycle from the rise times measured in unaided recordings. The largest difference noted was 0.8 cycles for the 4000 Hz stimuli recording from Starkey with both gain settings, and Phonak (compression) in both intensity conditions.  Siemens had measured rise times that were closest to unaided rise times in both intensity conditions. As none of the measured rise times were above 5 ms, a sharp decrease in the amplitude of Wave V of the ABR response is not predicted after hearing aid processing (Stapells & Picton, 1981). Changes from the unaided by one cycle would be expected to shift the amplitude only slightly without reaching a significant increase or decrease (Beattie & Torre, 1997; Stapells & Picton, 1981).   Measured fall times were longer than the 2-cycles programmed for pre-recorded stimuli in all unaided conditions, likely from the reverberation effects of a sound field set-up (Dillon & Walker, 1982). In aided conditions, differences in measured fall times from those measured in unaided recordings did not exceed more than one cycle across both intensity conditions with the exception of the 4000 Hz tone burst recorded from Phonak (linear) in both intensity conditions, where a difference of 1.6 cycles, or 0.4 ms, was noted. This difference is not likely to cause much effect (Stapells & Picton, 1981). Due to the reverberation effects, durations of the 500 Hz tone exceed 8 ms; this may elicit an offset ABR, although due to the longer fall time and short plateau time of each of the tone bursts, the probability of detecting an offset ABR is lower (Van Campen, Hall III, & Granthem, 74  1997). Van Campen et al. (1997) did not alter the fall time separately in a systematic manner to observe latency and amplitude trends; they investigated rise time effects with the fall time held constant. Thus, it is unknown whether the fall time differences measured will induce an offset response, but the small differences are unlikely to have an effect.   Finally, rise and fall times were also measured for rarefaction clicks. The stimulus was created using a square wave that rose and fell in 1 sample, which is 23 µs.  However, as expected from presenting in sound field using a loudspeaker, measured rise and fall times were longer in the unaided condition (Gorga & Thornton, 1989). Hearing aids tended to shorten rise time or left it unchanged, other than in Phonak (compression) where rise times were elongated. Measured fall times in aided conditions had larger elongations with hearing aid processing. Whether changes of several hundred microseconds  to rise and fall times will lead to significant changes to the morphology of the resulting ABR remains to be investigated, as ABR studies using clicks have traditionally used a square wave and transducers that introduce fewer room effects (Gorga & Thornton, 1989). These rise and fall times do not seem to have a large effect on intensity changes, as discussed in section 3.5.6.1.  3.3.7 MLR Tone bursts and click Rise and fall times were measured in the recordings of tone bursts and the rarefaction clicks in all hearing aid and intensity conditions.  See Table 3.11 for rise times, in cycles, of the 2-1-2 tone bursts. See Table 3.12 for rise times, in ms, of the rarefaction click and tone bursts generated to have a rise and fall of 4 ms and a plateau of 2 ms, hereafter referred to as 4-2-4 ms tone bursts.  See Table 3.13 for fall times, in cycles, of the 2-1-2 tone bursts, and Table 3.14 for fall times, in ms, of the rarefaction click and 4-2-4 ms tone bursts. Opposite polarity data are not shown.  75  Table 3.11.11Measured rise times, cycles, of 2-1-2 tone bursts of each frequency presented in all hearing aid and intensity conditions.  Table 3.12.12Measured rise times, ms, of 4-2-4 ms tone bursts of each frequency and the rarefaction click presented in all hearing aid and intensity conditions.    Stimulus  Intensity Condition  Rise time (cycles) Hearing Aid  Unaided Phonak Siemens Starkey Gain setting   Linear Compression Linear Compression Linear Compression 500 Hz  85 ppe SPL 1.8 4.6 1.85 1.85 1.35 1.8 1.6 1000 Hz 2 2.3 2.1 2.1 1.9 2 1.8 2000 Hz 2.4 3 2.4 2.4 2.8 3 2.6 4000 Hz 3.2 4 4 3.2 3.2 4 4 500 Hz 65 ppe SPL 1.75 1.9 4.6 1.85 1.9 1.45 1.45 1000 Hz 2 2.4 2.1 2.1 2 1.8 1.7 2000 Hz 2.4 3 2.4 2.6 2.6 2.8 3 4000 Hz 3.2 4 4.4 3.6 3.2 4 4 Stimulus  Intensity Condition  Rise time (ms) Hearing Aid  Unaided Phonak Siemens Starkey Gain setting   Linear Compression Linear Compression Linear Compression 500 Hz 85 ppe SPL 3.6 9.5 3.7 3.7 2.7 3.6 3.6 1000 Hz 3.5 4 2.9 3.6 2.9 3.4 3.3 2000 Hz 3.4 3.9 3 3.4 3.2 3.3 3.3 Rarefaction click 0.3 0.2 0.2 0.2 0.2 0.3 0.2 500 Hz 65 ppe SPL 3.4 2.5 8.5 3.9 3.7 2.9 2.8 1000 Hz 3.4 4.1 3.7 3.6 3.5 3.1 3.1 2000 Hz 3.3 3.9 3.9 3.2 3.2 3.3 3.2 Rarefaction click 0.2 0.2 0.2 0.2 0.2 0.2 0.2 76   Table 3.13.13Measured fall times, cycles, of 2-1-2 tone bursts of each frequency presented in all hearing aid and intensity conditions   Table 3.14.14Measured fall times, ms, of 4-2-4 ms tone bursts of each frequency and the rarefaction click presented in all hearing aid and intensity conditions. Stimulus  Intensity Condition  Fall time (cycles) Hearing Aid  Unaided Phonak Siemens Starkey Gain setting   Linear Compression Linear Compression Linear Compression 500 Hz 85 ppe SPL 4.05 4.6 3.6 4.15 3.3 4.1 3.55 1000 Hz 2.6 2.5 2.1 3.6 2.6 2.7 2.6 2000 Hz 3.8 4 4.2 3.6 4 3.6 3.6 4000 Hz 6.8 10.4 7.6 6.8 6.8 7.2 7.2 500 Hz 65 ppe SPL 3.8 4.65 4.65 3.05 3.85 2.2 2.25 1000 Hz 2.5 2.6 9.1 2.9 2.8 2.2 2.1 2000 Hz 3.2 4.4 4.4 3.4 3.4 3.2 3.2 4000 Hz 6.4 10.4 10.4 6.8 6.8 6.8 6.8 Stimulus  Intensity Condition  Fall time (ms) Hearing Aid  Unaided Phonak Siemens Starkey Gain setting   Linear Compression Linear Compression Linear Compression 500 Hz 85 ppe SPL 8.2 8.1 7.3 8.4 6.4 8.2 6.9 1000 Hz 7.2 9.2 4.3 9 4.9 8.9 7 2000 Hz 3.6 6.9 7.1 6.3 3.5 3.3 3.3 Rarefaction click 0.6 1.6 1.6 1.2 1.3 0.6 0.9 500 Hz 65 ppe SPL 7.6 8.6 9.7 6.1 7.8 5.3 4.3 1000 Hz 7.6 9.7 9 8.5 8.8 3.9 3.6 2000 Hz 3.4 7.3 7.2 6.2 5.6 2.8 2.9 Rarefaction click 0.7 1.5 1.6 1.1 1.3 0.7 0.8 77  3.3.7.1 Interpretation of rise and fall times on MLR Tone bursts and click  The stimuli used to elicit MLR and ABR can be the same, such as the 2-1-2 tone bursts; a slower presentation rate is required to elicit an MLR using these stimuli.  Other tonal stimuli used to elicit MLR may be longer in duration, such as the 4-2-4 ms tone bursts. Thus, the contributions of presentation rate and stimulus duration to possible changes with hearing aid processing can be investigated. Changes between polarities were similar unless otherwise noted.    With 2-1-2 tone bursts, unaided recordings had the same elongations in measured rise times as they did for ABR recordings. The differences from unaided were similar to those found for ABR recordings, where most aided conditions in both intensity conditions differed from unaided by less than 1 cycle. Phonak exceeded this difference in three instances in each polarity: the 500 Hz tone-burst recording from Phonak (linear) in both intensity conditions, and the 4000 Hz tone-burst recording from Phonak (compression) in the 65 ppe SPL condition. Not only were the largest differences observed from Phonak in the ABR set of stimuli, even larger absolute differences were noted from Phonak in the MLR stimuli set. The slower presentation rate may have allowed for measurable hearing aid processing effects to occur on the stimulus before the next stimulus began. Converting the differences back into milliseconds, rise times were increased by 4.7 ms to 5.7 ms; Xu, De Vel, Vinck, & Van Cauwenberge (1995) reported significant decreases in Na-Pa amplitudes, when rise times increased from 4 ms to 6 ms. Latencies are unlikely to be affected in a significant manner as non-significant increases of latency were seen when rise and fall times were increased from 5 ms to 20 ms (Kodera et al., 1979). Thus, processing with Phonak hearing aids may lead to an unaccounted decrease of Na-Pa amplitudes. 78   With 4-2-4 ms tone bursts, the measured rise times in the unaided condition were all within 4 ms. Most aided conditions did not lead to changes greater than 1 ms. The largest difference from unaided was found in the 500 Hz recordings from Phonak (linear) in the 85 ppe SPL condition and Phonak (compression) in the 65 ppe SPL condition, where differences of 5.1 ms to 5.9 ms were measured. As noted for 2-1-2 tone bursts, such increases in rise time can lead to significant, and unaccounted for, decreases to Na-Pa amplitude (Xu et al., 1995).  Measured fall times were longer than the 2 cycles generated for 2-1-2 tone bursts prior to recording in all unaided conditions, likely from the reverberation effects of a sound field set-up (Dillon & Walker, 1982). In aided conditions, differences in measured fall times from unaided recordings were generally within 1.6 cycles, most of which were elongations, with only Phonak exceeding this difference at two different frequencies in the 65 ppe SPL condition. Measured fall times for the unaided 4-2-4 ms tone bursts ranged from 3.6 ms in the 2000 Hz recording and 8.2 ms in the 500 Hz recording, likely due to frequency dependent reverberation effects of a sound field set-up as well (Dillon & Walker, 1982). Absolute differences from unaided in the aided conditions were generally within 3.8 ms. Effects of independent changes to fall times on MLR morphology remain to be investigated.  Finally, the same rarefaction click from the ABR stimuli set was recorded at a slower presentation rate in the MLR stimuli set. As expected from a sound-field set up, measured rise and fall times were longer (Gorga & Thornton, 1989). Hearing aids either shortened the measured rise time by 100 µs or left it unchanged in both intensity conditions. However, measured fall times in aided conditions had larger changes in most hearing aid conditions other than Starkey (linear) in both intensity conditions, where the measured fall times were unchanged. Differences ranged from 100 µs, measured from Starkey (compression) in the 65 79  ppe SPL condition, to 1000 µs, measured from Phonak with either gain setting in the 85 ppe SPL condition.  Whether these changes to rise and fall times will lead to significant changes to the morphology of the resulting MLR remains to be investigated as MLR studies using clicks have traditionally used a rectangular wave and transducers that introduce fewer room effects (Gorga & Thornton, 1989).  3.3.8 MMN Speech Rise and fall times were measured for the two syllables, /ba/, presented 90 percent of the time and /da/, presented 10 percent of the time. See Table 3.15 for the difference between the rise time measured from /ba/ and the rise time measured from /da/ in all hearing aid and intensity conditions. The difference was calculated by subtracting the rise time of /da/ from the rise time of /ba/. See Table 3.16 for the difference between the fall time measured from /ba/ and the fall time measured from /da/ in all hearing aid and intensity conditions. Table 3.15.15Difference between the rise time, ms, measured from /ba/ and the rise time, ms, measured from /da/ in all hearing aid and intensity conditions.  Table 3.16.16Difference between the fall time, ms, measured from /ba/ and the fall time, ms, measured from /da/ in all hearing aid and intensity conditions. Intensity Condition  Rise time difference between /ba/ and /da/ (ms)  Unaided Phonak Siemens Starkey   Linear Compression Linear Compression Linear Compression 65 dB SPL 49.2  -3.1  22  39.9  25.7  40.5  -14.4  45 dB SPL  49.7  12.4  37.1  51.9  58.5  37.9  8.7  Intensity Condition  Fall time difference between /ba/ and /da/ (ms)  Unaided Phonak Siemens Starkey   Linear Compression Linear Compression Linear Compression 65 dB SPL -0.9  3.5  -0.5  7  4.9  2.2  0  45 dB SPL  0.7  2.5  2.6  5.6  4.1  0.8  -1.1  80   3.3.8.1 Interpretation of rise and fall time effects on MMN Speech  Different patterns were noted for intensity and gain setting, suggesting that the gain setting can affect the rise times. Subtracting the measured rise time of /da/ from /ba/ in the unaided condition yielded a difference of 49.2 ms in the 65 dB SPL condition, and 49.7 ms in the 45 dB SPL condition. The largest difference from unaided was found for Phonak (linear), where the difference between the measured rise times of the standard and deviant stimulus was reduced to 3.1 ms. Fall time differences between the two syllables, however, were all within 7 ms, with no particular pattern observed for a specific hearing aid condition. The magnitude of changes observed that can significantly affect the MMN amplitude or latency cannot be interpreted as no studies have varied this variable alone in the deviant and standard stimulus to observe changes to the MMN.  3.4 Duration 3.4.1 SCP Tones Durations were determined for each recording and plotted as a function of the stimulus duration. See Tables 3.17 to 3.19 for the equations of the linear trend lines that were fit to the data from each frequency in each hearing aid and intensity condition. See Figures 3.14 for measured durations of 500 Hz tones in all hearing aid conditions in the 65 dB SPL condition to demonstrate the amount of overlap of data found for tones of each frequency recorded in all hearing aid and intensity conditions.       81  Table 3.17.17Trend line equations fit to recorded 500 Hz tone durations as a function of stimulus durations.  Table 3.18.18Trend line equations fit to recorded 1000 Hz tone durations as a function of stimulus durations     Hearing aid Intensity condition Gain setting Linear trend line equation Unaided 65 dB SPL N/A y = 1.0361x + 19.087 Phonak  65 dB SPL Linear y = 0.9984x + 16.043 Compression y = 0.987x + 10.331 Siemens 65 dB SPL Linear y = 1.0051x + 15.545 Compression y = 1.002x + 13.798 Starkey 65 dB SPL Linear y = 1.0095x + 29.121 Compression y = 1.0025x + 12.783 Unaided 45 dB SPL  N/A y = 0.9918x + 6.171 Phonak  45 dB SPL  Linear y = 0.9559x + 6.5074 Compression y = 0.9883x - 7.7177 Siemens 45 dB SPL  Linear y = 1.0033x + 1.4721 Compression y = 0.9962x + 7.6777 Starkey  45 dB SPL  Linear y = 0.9942x + 4.0611 Compression  y = 1.0002x + 5.3252 Hearing aid Intensity condition Gain setting Linear trend line equation Unaided 65 dB SPL N/A y = 0962x + 45.099 Phonak  65 dB SPL Linear y = 0.989x + 24.029 Compression y = 0.9976x + 5.254 Siemens  65 dB SPL Linear y = 0.9913x + 14.903 Compression y = 1.0061x + 9.2235 Starkey  65 dB SPL Linear y = 0.9968x + 8.4458 Compression y = 0.9984x + 6.0155 Unaided 45 dB SPL  N/A y = 0.9992x + 1.3981 Phonak  45 dB SPL  Linear y = 0.9856x - 6.1599 Compression y = 0.9934x + 0.313 Siemens  45 dB SPL  Linear y = 0.9965x + 0.4642 Compression y = 1.0003x + 2.047 Starkey  45 dB SPL  Linear y = 0.9954x + 0.7783 Compression  y = 0.9946x + 1.2116 82  Table 3.19.19Trend line equations fit to recorded 2000 Hz tone durations as a function of stimulus durations   Hearing aid Intensity condition Gain setting Linear trend line equation Unaided 65 dB SPL N/A y = 0.9997x + 34.993 Phonak  65 dB SPL Linear y = 0.9953x + 26.196 Compression y = 0.9917x + 17.331 Siemens  65 dB SPL Linear y = 1.0049x + 17.79 Compression y = 1.0027x + 11.765 Starkey  65 dB SPL Linear y = 0.9977x + 20.023 Compression y = 0.9896x + 12.54 Unaided 45 dB SPL  N/A  y = 0.9968x - 1.4102 Phonak  45 dB SPL  Linear y = 0.9919x + 6.3676 Compression y = 0.99x + 5.3149 Siemens  45 dB SPL  Linear y = 1.0156x - 3.8659 Compression y = 0.9942x - 2.5585 Starkey  45 dB SPL  Linear y = 0.9788x + 7.6823 Compression  y = 0.9848x + 8.3373 83   Figure 3.14.15Measured durations, ms, of 500 Hz tones plotted as a function of stimulus durations in all hearing aid conditions, with stimuli presented at 65 dB SPL.10100100010 100 1000Measured Duration (ms)Generated Duration (ms) Starkey LinearSiemens LinearPhonak LinearStarkey CompressionSiemens CompressionPhonak CompressionUnaided84  3.4.1.1 Interpretation of duration effects on SCP Tones  With hearing aid processing, duration information was expected to be preserved (e.g., Souza & Tremblay, 2006; Tremblay, Billings, et al., 2006). Such were the findings with SCP Tones. When rounded to the nearest integer, the slopes of the linear trend line equations fit to data comparing measured durations to stimulus durations all had a value of 1, suggesting that the measured durations increased at the same rate as the stimulus durations. The y-intercepts of all the trend lines were greater than 0, suggesting that the measured durations were longer than stimulus durations; this was expected from a sound-field set up as reverberation effects that were incorporated into the recording would elongate the stimuli slightly (Dillon & Walker, 1982). As seen in Figure 3.14 data from all aided and unaided conditions overlapped with one another, suggesting that hearing aid processing does not produce additional large changes to duration. Generally, more spread was seen with shorter durations, such 30 ms, most likely due to amplification of the ringing by the hearing aids. However, as seen in Figure 3.14, and from the y-intercepts, recordings in the unaided conditions were not consistently shorter than those measured from aided conditions. Thus, hearing aid processing effects on the durations of tonal stimuli are not expected to affect the morphology of the SCP (Onishi & Davis, 1968).   3.4.2 SCP Noise Durations were determined for each stimulus and plotted as a function of the generated duration. See Table 3.20 for the equation of the linear trend lines that were fit to the data from each hearing aid and intensity condition. The pattern of the data was similar to that shown in Figure 3.14.    85   Table 3.20.20Trend line equations fit to noise burst durations as a function of generated durations.  3.4.2.1 Interpretation of duration effects on SCP Noise Similar to SCP Tones, durations were expected to be preserved with hearing aid processing for noise bursts as well. When rounded to the nearest integer, the slopes of the linear trend line equations fit to data comparing measured durations to stimulus durations all had a value of 1, suggesting that the measured durations grew at the same rate as the generated durations. The y-intercepts of all the trend lines were greater than 0, suggesting that the measured durations were longer than generated durations; this was expected from a sound-field set up as reverberation effects that were incorporated into the recording would elongate the stimuli slightly (Dillon & Walker, 1982). Similar to the overlap observed in Figure 3.13, data from all aided and unaided conditions overlapped with one another, suggesting that hearing aid processing does not produce additional large changes to duration. Greater spread was generally seen with noise bursts of shorter durations, such as 30 ms, most likely due to the amplification of the ringing by the hearing aids. However, as seen by the y-intercepts of the trend line equations, recordings in the unaided conditions were not consistently shorter than those measured from aided conditions. Generally, hearing aid Hearing aid Intensity condition Gain setting Linear trend line equation Unaided 65 dB SPL N/A  y = 0.9833x + 17.584 Phonak  65 dB SPL Linear y = 0.9806x + 24.839 Compression y = 0.9891x + 16.729 Siemens  65 dB SPL Linear y = 0.9865x + 22.691 Compression y = 0.9812x + 27.147 Starkey  65 dB SPL Linear y = 0.9897x + 23.369 Compression y = 0.9925x + 18.481 Unaided 45 dB SPL  N/A  y = 0.9873x + 10.492 Phonak  45 dB SPL  Linear y = 0.9927x + 8.5099 Compression y = 0.9883x + 13.952 Siemens  45 dB SPL  Linear y = 0.9887x + 7.1727 Compression y = 0.9913x + 18.81 Starkey  45 dB SPL  Linear y = 0.9954x + 6.4753 Compression  y = 0.9933x + 6.4233 86  processing effects on the duration of noise bursts used in this study are not expected to affect the morphology of the SCP (Onishi & Davis, 1968).   3.4.3 ACC Tonal stimuli Durations were determined for ACC tonal stimuli in each hearing aid and intensity condition. See Table 3.21 for measured durations. Opposite polarity data are not displayed.  Table 3.21.21Measured durations, ms, for ACC tonal stimuli in all hearing aid and intensity conditions 3.4.3.1 Interpretation of duration effects on ACC Tonal stimuli  The stimulus duration was 600 ms and measured durations from all hearing aid conditions were within 10 ms to 22 ms of 600 ms in the 65 dB SPL condition, with the unaided recording having the longest duration, suggesting the introduction of reverberation effects with a sound-field set-up, as expected (Dillon & Walker, 1982). The observed effects were similar between polarities unless otherwise noted. Measured durations in the 45 dB SPL condition had a smaller difference range, between 1 to 7 ms of 600 ms, possibly due to the higher noise floor that may mask the true end point. We also observed from the comparison of the waveforms that the transition point remained in the 13 230 sample region, which is within the 300 ms duration generated for each frequency in both aided and unaided Hearing aid Intensity condition Gain setting Measured Duration (ms) Unaided 65 dB SPL N/A  622 Phonak  65 dB SPL Linear 612 Compression 610 Siemens  65 dB SPL Linear 617 Compression 611 Starkey  65 dB SPL Linear 613 Compression 610 Unaided 45 dB SPL  N/A  603 Phonak  45 dB SPL  Linear 596 Compression 604 Siemens  45 dB SPL  Linear 603 Compression 601 Starkey 45 dB SPL  Linear 605 Compression  607 87  conditions. See Figure 3.15 for a comparison of the stimulus waveform of the ACC tonal stimulus and three other hearing aid conditions for an illustration of this; see Appendix B to examine all waveforms. Thus, ACC morphology is not predicted to be affected by the duration of processed stimuli. 88  Figure 3.15.16Stimulus waveforms of ACC tonal stimuli in the 65 dB SPL condition. The pre-transition segment was 1680 Hz and 300 ms in duration; the post-transition segment was 1705 Hz and 300 ms in duration; thus, the total stimulus duration was 600 ms. Different panels show different hearing aid conditions: (A) Siemens (linear); (B) Phonak (linear); (C) Starkey (linear); (D) Unaided. Transition points, denoted by the arrow heads, are all in the 13 230 sample (i.e., 300 ms) region. A  B  C  D  89  3.4.4 SCP Speech and ACC Speech Durations were determined for all speech syllables and phonemes recorded in each hearing aid and intensity condition. See Figures 3.16 for measured durations. Figures for opposite polarity data are not shown.  90   Figure 3.16.17Duration, ms, of all syllables and phonemes recorded in all hearing aid and intensity conditions.  0100200300400500600700800/da//ta//sa//ʃa/ /a//i//a/ cut pre-recording/a/ cut post-recording/da//ta//sa//ʃa/ /a//i//a/ cut pre-recording/a/ cut post-recording/da//ta//sa//ʃa/ /a//i//a/ cut pre-recording/a/ cut post-recording/da//ta//sa//ʃa/ /a//i//a/ cut pre-recording/a/ cut post-recording/da//ta//sa//ʃa/ /a//i//a/ cut pre-recording/a/ cut post-recording/da//ta//sa//ʃa/ /a//i//a/ cut pre-recording/a/ cut post-recording/da//ta//sa//ʃa/ /a//i//a/ cut pre-recording/a/ cut post-recordingLINEAR COMPRESSION LINEAR COMPRESSION LINEAR COMPRESSIONUNAIDED STARKEY SIEMENS PHONAKDuration (ms)Hearing Aid Condition65 dB SPL Condition 45 dB SPL Condition91  3.4.4.1 Interpretation of duration effects on SCP and ACC Speech  As seen for most SCP stimuli discussed, durations remained stable with hearing aid processing, as also seen in other studies (e.g., Souza & Tremblay, 2006; Tremblay, Billings, et al., 2006). Observed effects were similar between polarities unless otherwise noted. Absolute differences from unaided recordings ranged from 0.1 ms to 27.3 ms in the 65 dB SPL condition and from 0.7 ms to 35.1 ms in the 45 dB SPL condition, with no particular hearing aid or gain setting showing consistently greater differences from unaided. As mentioned, reverberation effects of a sound field set-up were expected to contribute to these measurable differences (Dillon & Walker, 1982). All the measured durations were shorter in the 45 dB SPL condition as compared to the durations measured in the 65 dB SPL condition, likely due to the higher noise floor masking the final end points of the stimuli. As duration plateau effects greater than 30 ms do not affect the SCP morphology, as noted in Table 1.1, and recordings are greater than 30 ms, SCP morphology is not expected to be affected by any duration effects of hearing aid processing. 3.4.5 HEARLab™ stimuli Durations were determined for all HEARLab™ phonemes recorded in each hearing aid and intensity condition. See Figures 3.17 for measured durations.  Figures for opposite polarity data are not shown.  92   Figure 3.17.18Duration, ms, of all HEARLab™ phonemes recorded in all hearing aid and intensity conditions.     020406080100120140160180200/m/ /t/ /s/ /g/ /m/ /t/ /s/ /g/ /m/ /t/ /s/ /g/ /m/ /t/ /s/ /g/ /m/ /t/ /s/ /g/ /m/ /t/ /s/ /g/ /m/ /t/ /s/ /g/LINEAR COMPRESSION LINEAR COMPRESSION LINEAR COMPRESSIONUNAIDED STARKEY SIEMENS PHONAKDuration (ms)Hearing Aid Condition65 dB SPL Condition 45 dB SPL Condition93  3.4.5.1 Interpretation of duration effects on HEARLab™ stimuli  As seen for most SCP stimuli discussed, durations of stimuli were generally similar to those measured for unaided stimuli, with measured absolute differences below 22.3 ms for three of the four phonemes. The magnitude of difference observed did not have noticeable differences with different hearing aids and gain settings. Effects were similar between polarities unless otherwise noted. The reverberation effects of a sound-field set-up likely contributed to the longer durations measured for each stimulus (Dillon & Walker, 1982) The stimulus durations were 30 ms, 30 ms, 50 ms, and 21 ms for /m/, /t/, /s/, and /g/, respectively. Even though the /m/ and /t/ stimuli had the same stimulus durations, the durations of /m/ and /s/ were more similar after being presented in sound field in the unaided condition. This pattern was kept with hearing aid processing. Thus, hearing aid processing of duration is unlikely to cause additional differences; however, it is important to be aware of the reverberation effects in sound-field for stimuli of these durations. This is because the elongation and shortening of durations in the range of 30 to 70 ms can make a difference to the SCP morphology. For example, phonemes that were originally 21 ms, such as /g/, and elongated to 30 ms to 60 ms in duration, may elicit SCPs with larger amplitudes in sound field than if the /g/ were presented through another transducer where there may be less acoustic ringing, such as over-the-ear earphones (Gorga & Thornton, 1989; Onishi & Davis, 1968).  The phoneme /m/ seemed to have measured differences that were larger with hearing aid processing for the recordings from Siemens (compression) and Phonak (compression) in the 65 dB SPL condition only. Measured differences from unaided recordings were elongations of 69.7 ms and 70.1 ms in the /m/ recordings from Siemens and 33.5 ms and  94  33.8 ms in the /m/ recordings from Phonak, while all other aided conditions produced measured differences from unaided that ranged from 0.27 to 4.35 ms. However, as these elongations are out of the 30 ms to 70 ms absolute duration range that have been shown to increase SCP morphology, hearing aid processing is unlikely to cause any increases in SCP amplitude (Onishi & Davis, 1968). In the 45 dB SPL condition, however, hearing aid processing seemed to consistently shorten the /m/ phoneme. While Starkey (compression) shortened the stimulus by 2.1 to 2.4 ms, all other hearing aid conditions shortened the /m/ stimulus by differences ranging from 28.8 ms to 45.1 ms. As the unaided stimulus duration was measured between 82.5 ms to 82.8 ms, the shortening of the stimulus by that range may see a decrease in SCP amplitude that was unintended. However, hearing aid processing is not generally expected to affect duration; the changes noted may be a combination of room effects and measurement error, that are seen more prominently in the frequency region between 250 Hz and 500 Hz where /m/ lies, because reverberation times are generally longer for low frequencies as compared to high frequencies (Beranek, 1992).  3.4.6 ABR Tone bursts and click  As reverberation effects caused ABR stimuli to persist until the next stimulus was presented, the duration of the most intense part of the brief stimulus was determined. See section 2.5.2 for how this point was defined. Durations were determined for all tone bursts and rarefaction clicks recorded in each hearing aid and intensity condition. See Figure 3.18 for measured durations. Figures for opposite polarity data are not shown.  95   Figure 3.18.19Duration, ms, of the most intense segment of the ABR tone bursts and rarefaction click for all hearing aid and intensity conditions.   0510152025500 Hz1000 Hz2000 Hz4000 HzRarefaction Click500 Hz1000 Hz2000 Hz4000 HzRarefaction Click500 Hz1000 Hz2000 Hz4000 HzRarefaction Click500 Hz1000 Hz2000 Hz4000 HzRarefaction Click500 Hz1000 Hz2000 Hz4000 HzRarefaction Click500 Hz1000 Hz2000 Hz4000 HzRarefaction Click500 Hz1000 Hz2000 Hz4000 HzRarefaction ClickLINEAR COMPRESSION LINEAR COMPRESSION LINEAR COMPRESSIONUNAIDED STARKEY SIEMENS PHONAKDuration (ms)Hearing Aid Condition85 ppe SPL condition 65 ppe SPL condition96  3.4.6.1 Interpretation of duration effects on ABR Tone bursts and click  The lower the frequency of the tone-burst, the longer the duration measured, suggesting that the reverberation effects, as discussed throughout section 3.4, contribute to longer measured durations as compared to the 5-cycle duration of the stimulus prior to recording. The frequency dependent elongations were expected (Beranek, 1992).  The number of additional cycles of ringing was frequency dependent, with the number of cycles doubling with the doubling of frequency: 500 Hz tone bursts had approximately 6 cycles of ringing, 1000 Hz tone bursts had approximately 12 cycles of ringing, 2000 Hz tone bursts had approximately 28 cycles of ringing, and 4000 Hz tones bursts having approximately 50 cycles of ringing. Absolute differences between the durations measured from aided recordings and unaided recordings ranged from 0.1 ms to 1.5 ms; differences were consistent across hearing aids. Observed changes were similar between polarities unless otherwise noted. Funasaka & Ito (1986) found that the greatest change in latency and amplitudes of Wave V occurred between 0 to 10 ms, with a decreased rate of increase of both between 10 ms to 20 ms, and no significant changes or trends for increases in duration between 20 ms and 30 ms. Thus, the differences are not likely to change the morphology of the ABR elicited by a 500 Hz tone-burst with hearing aid processing; changes may be seen with ABR morphology when elicited by stimuli of higher frequencies. However, some researchers suggest that there are no significant amplitude effects of duration on Wave V (Hecox et al., 1976), and others suggest no duration effects at all (Davids, Valero, Papsin, Harrison, & Gordon, 2008).    97  3.4.7 MLR Tone bursts and click The durations of the most intense segment of all tone bursts and rarefaction clicks recorded in each hearing aid and intensity condition were determined. See section 2.5.2 for how this point was defined. See Figure 3.19 for the measured duration of the most intense segments of the stimuli. See Figure 3.20 for a graphical representation of the measured duration of the full stimulus.  Figures for opposite polarity data are not shown.  98   Figure 3.19.20Duration, ms, of the most intense segment of the MLR tone bursts and rarefaction click for in all hearing aid and intensity conditions.   0510152025303540500 Hz (2-1-2)1000 Hz  (2-1-2)2000 Hz (2-1-2)4000 Hz (2-1-2)500 Hz (4-2-4ms)1000 Hz (4-2-4ms)2000 Hz (4-2-4ms)500 Hz (2-1-2)1000 Hz  (2-1-2)2000 Hz (2-1-2)4000 Hz (2-1-2)500 Hz (4-2-4ms)1000 Hz (4-2-4ms)2000 Hz (4-2-4ms)500 Hz (2-1-2)1000 Hz  (2-1-2)2000 Hz (2-1-2)4000 Hz (2-1-2)500 Hz (4-2-4ms)1000 Hz (4-2-4ms)2000 Hz (4-2-4ms)500 Hz (2-1-2)1000 Hz  (2-1-2)2000 Hz (2-1-2)4000 Hz (2-1-2)500 Hz (4-2-4ms)1000 Hz (4-2-4ms)2000 Hz (4-2-4ms)500 Hz (2-1-2)1000 Hz  (2-1-2)2000 Hz (2-1-2)4000 Hz (2-1-2)500 Hz (4-2-4ms)1000 Hz (4-2-4ms)2000 Hz (4-2-4ms)500 Hz (2-1-2)1000 Hz  (2-1-2)2000 Hz (2-1-2)4000 Hz (2-1-2)500 Hz (4-2-4ms)1000 Hz (4-2-4ms)2000 Hz (4-2-4ms)500 Hz (2-1-2)1000 Hz  (2-1-2)2000 Hz (2-1-2)4000 Hz (2-1-2)500 Hz (4-2-4ms)1000 Hz (4-2-4ms)2000 Hz (4-2-4ms)LINEAR COMPRESSION LINEAR COMPRESSION LINEAR COMPRESSIONUNAIDED STARKEY SIEMENS PHONAKDuration (ms)Hearing Aid Condition85 ppe SPL Condition 65 ppe SPL Condition99   Figure 3.20.21Duration, ms, of the full MLR tone bursts and rarefaction click, including reverberation effects, in all hearing aid and intensity conditions. 0102030405060500 Hz (2-1-2)1000 Hz  (2-1-2)2000 Hz (2-1-2)4000 Hz (2-1-2)500 Hz (4-2-4ms)1000 Hz (4-2-4ms)2000 Hz (4-2-4ms)Rarefaction Click500 Hz (2-1-2)1000 Hz  (2-1-2)2000 Hz (2-1-2)4000 Hz (2-1-2)500 Hz (4-2-4ms)1000 Hz (4-2-4ms)2000 Hz (4-2-4ms)Rarefaction Click500 Hz (2-1-2)1000 Hz  (2-1-2)2000 Hz (2-1-2)4000 Hz (2-1-2)500 Hz (4-2-4ms)1000 Hz (4-2-4ms)2000 Hz (4-2-4ms)Rarefaction Click500 Hz (2-1-2)1000 Hz  (2-1-2)2000 Hz (2-1-2)4000 Hz (2-1-2)500 Hz (4-2-4ms)1000 Hz (4-2-4ms)2000 Hz (4-2-4ms)Rarefaction Click500 Hz (2-1-2)1000 Hz  (2-1-2)2000 Hz (2-1-2)4000 Hz (2-1-2)500 Hz (4-2-4ms)1000 Hz (4-2-4ms)2000 Hz (4-2-4ms)Rarefaction Click500 Hz (2-1-2)1000 Hz  (2-1-2)2000 Hz (2-1-2)4000 Hz (2-1-2)500 Hz (4-2-4ms)1000 Hz (4-2-4ms)2000 Hz (4-2-4ms)Rarefaction Click500 Hz (2-1-2)1000 Hz  (2-1-2)2000 Hz (2-1-2)4000 Hz (2-1-2)500 Hz (4-2-4ms)1000 Hz (4-2-4ms)2000 Hz (4-2-4ms)Rarefaction ClickLINEAR COMPRESSION LINEAR COMPRESSION LINEAR COMPRESSIONUNAIDED STARKEY SIEMENS PHONAKDuration (ms)Hearing Aid Condtion85 ppe SPL Condition 65 ppe SPL Condition100  3.4.7.1 Interpretation of duration effects on MLR Tone bursts and click  The full reverberation effects of using a sound-field system may have reduced the differences between the stimuli of different durations; unaided measures were at least 33 ms longer than the duration of the stimuli prior to recording, even before hearing aid processing. However, the relative differences between durations were preserved with hearing aid processing. Observed effects were similar between polarities unless otherwise noted. Processed stimuli differed by at most 6.5 ms in the 85 ppe SPL condition and 20.9 ms in the 65 ppe SPL condition. When comparing the measured duration of the most intense segment of the stimulus, similar duration effects were noted as the ABR tone bursts and click, where the frequency dependent elongations from the stimulus durations were observed for all hearing aid conditions and intensity conditions, suggesting that this may be related to the physical factors of the sound-field set-up. Nonetheless, there have been no reported effects of duration on MLR morphology (e.g., Davids et al., 2008; McGee, Kraus, & Manfredi, 1988; Vivion, Hirsh, Frye-Osier, Goldstein, 1980). Thus, changes to duration with hearing aid processing are unlikely to affect MLR morphology.  3.4.8 MMN Speech Durations were determined for /ba/ and /da/, the standard and deviant stimulus recorded in each hearing aid and intensity condition. See Table 3.22 for the difference between the measured durations of /ba/ and /da/ in all hearing aid and intensity conditions, with the difference calculated by subtracting the duration of /da/ from /ba/.   101  Table 3.22.22Duration difference, ms, between /ba/ and /da/ in all hearing aid and intensity conditions. 3.4.8.1 Interpretation of duration effects on MMN Speech  All measured durations of stimuli were within 151.5 to 162.8 ms in the 65 dB SPL condition, and between 144.4 to 157.5 ms in the 45 dB SPL condition. Duration differences increased MMN amplitude significantly only when the stimulus duration was increased to 300 ms from 4 ms or decreased from 300 ms to 4 ms (Paavilainen, Jiang, Lavikainen, & Näätänen, 1993). Thus, hearing aid processing is not predicted to affect MMN morphology.  3.4.9 ASSR 40 Hz and 80 Hz AM Tones  Durations were not determined for ASSR 40 Hz and 80 Hz AM tones as acoustic parameters were all measured from a continuous one-minute recording. From examining the spectra, the AM modulation frequency was kept. Thus, it was inappropriate to measure and report on durations for this set of stimuli. 3.5 Intensity 3.5.1 SCP Tones Intensities were measured for each stimulus in each hearing aid and intensity condition. Intensities were measured in volts and converted to dB volt with reference to 1 volt. Gains were calculated by subtracting the intensities of the unaided stimuli from the intensities of the respective aided stimuli in each hearing aid and intensity condition. See Tables 3.23 to 3.25 for gain values relative to unaided of tones of each frequency in all hearing aid and intensity conditions.  Opposite polarity data are not shown. Intensity Condition  Duration difference between /ba/ and /da/ (ms)  Unaided Phonak Siemens Starkey   Linear Compression Linear Compression Linear Compression 65 dB SPL -0.1  -0.0  0.9  -0.0  3.7  1.6  4.7  45 dB SPL  4.6  -13.1  2.1  2.2  0.8  -0.6  0.2  102  Table 3.23.23Gain relative to unaided, dB, in all hearing aid and intensity conditions for 500 Hz tones. RF = rise and fall time.  Stimulus Description Intensity Condition Gain (dB) Hearing Aid  Phonak Siemens Starkey Gain setting  Linear Compression Linear Compression Linear Compression 30 ms duration; 5 ms RF 65 dB SPL -26.4 -12.9 -17.6 -15.5 -9.1 -4.0 30 ms duration; 10 ms RF -27.0 -13.4 -17.6 -15.6 -9.1 -4.4 60 ms duration; 5 ms RF -26.8 -11.4 -15.5 -15.2 -8.8 -7.5 60 ms duration; 10 ms RF -32.7 -11.7 -15.5 -15.4 -8.9 -7.8 60 ms duration; 20 ms RF -16.1 -11.2 -15.2 -15.3 -8.5 -8.1 120 ms duration; 5 ms RF -7.3 -10.4 -12.7 -13.7 -8.6 -9.3 120 ms duration; 10 ms RF -7.4 -10.4 -12.7 -13.7 -8.6 -9.7 120 ms duration; 20 ms RF -8.0 -10.7 -13.0 -14.3 -8.8 -10.7 480 ms duration; 5 ms RF -2.7 -10.1 -9.7 -12.5 -9.0 -12.3 480 ms duration; 10 ms RF -2.7 -10.2 -9.7 -12.4 -9.1 -12.4 480 ms duration; 20 ms RF -2.6 -10.1 -9.6 -12.5 -9.1 -12.6 30 ms duration; 5 ms RF 45 dB SPL  -9.1 1.3 -2.8 3.3 6.8 14.5 30 ms duration; 10 ms RF -9.9 -0.2 -3.1 2.9 6.5 14.1 60 ms duration; 5 ms RF -1.1 10.0 -0.4 5.9 5.8 14.8 60 ms duration; 10 ms RF -2.2 9.6 -0.7 5.7 5.6 14.8 60 ms duration; 20 ms RF -3.7 8.2 -1.2 5.1 6.4 14.3 120 ms duration; 5 ms RF 5.9 14.7 2.1 8.6 6.3 14.8 120 ms duration; 10 ms RF 5.6 15.1 2.1 9.3 5.8 15.4 120 ms duration; 20 ms RF 5.1 14.0 1.7 8.3 6.5 14.4 480 ms duration; 5 ms RF 10.3 17.5 5.7 13.0 6.5 15.3 480 ms duration; 10 ms RF 10.3 17.5 5.6 13.0 6.3 15.2 480 ms duration; 20 ms RF 10.3 17.4 5.6 13.0 6.7 15.2 103          Stimulus Description       Intensity Condition        Gain (dB)     Hearing Aid  Phonak Siemens Starkey Gain setting  Linear Compression Linear Compression Linear Compression 30 ms duration; 5 ms RF 65 dB SPL  -18.5 -9.6 -12.6 -13.4 -8.0 -4.8 30 ms duration; 10 ms RF -18.9 -10.0 -12.6 -13.7 -7.9 -4.9 60 ms duration; 5 ms RF -13.4 -11.3 -11.4 -13.0 -8.1 -7.8 60 ms duration; 10 ms RF -13.5 -11.7 -11.3 -13.2 -8.0 -8.4 60 ms duration; 20 ms RF -13.6 -13.8 -11.2 -14.9 -7.9 -10.5 120 ms duration; 5 ms RF -10.1 -12.6 -9.5 -12.3 -8.1 -10.4 120 ms duration; 10 ms RF -10.1 -12.7 -9.5 -12.4 -8.1 -11.1 120 ms udration; 20 ms RF -9.6 -12.8 -9.1 -12.3 -7.7 -11.4 450 ms duration; 10 ms RF -7.7 -14.4 -7.0 -11.0 -8.0 -13.0 480 ms duration; 5 ms RF -7.5 -12.7 -6.7 -10.6 -7.9 -13.0 480 ms duration; 10 ms RF -7.4 -12.8 -6.6 -10.6 -7.8 -12.8 480 ms duration; 20 ms RF -7.4 -12.8 -6.5 -10.6 -7.8 -13.2 757 ms duration; 7.57 ms RF -7.2 -13.3 -6.4 -10.6 -7.8 -13.1 30 ms duration; 5 ms RF 45 dB SPL -5.0 12.1 0.9 3.9 6.7 14.4 30 ms duration; 10 ms RF -5.5 12.2 1.1 3.7 6.6 14.6 60 ms duration; 5 ms RF 0.7 12.2 3.3 6.0 6.0 14.1 60 ms duration; 10 ms RF 0.1 12.0 3.0 5.8 6.1 13.9 60 ms duration; 20 ms RF -0.7 11.7 2.7 5.5 7.0 13.7 120 ms duration; 5 ms RF 4.1 12.3 5.3 8.3 6.8 13.9 120 ms duration; 10 ms RF 4.1 12.2 5.3 8.2 6.4 13.7 120 ms udration; 20 ms RF -3.9 11.8 4.9 7.7 7.0 13.3 450 ms duration; 10 ms RF 5.9 11.5 7.4 10.5 6.8 13.0 480 ms duration; 5 ms RF 6.6 12.4 8.2 11.2 7.0 13.5 480 ms duration; 10 ms RF 6.6 12.4 8.2 11.2 6.7 13.5 480 ms duration; 20 ms RF 6.7 12.3 8.2 11.1 7.2 13.3 757 ms duration; 7.57 ms RF 5.8 11.3 7.3 10.1 6.5 12.6 Table 3.24.24Gain relative to unaided, dB, in all hearing aid and intensity conditions for 1000 Hz tones. RF = rise and fall time. 104  Table 3.25.25Gain relative to unaided, dB, in all hearing aid and intensity conditions for 2000 Hz tones. RF = rise and fall.  Stimulus Description Intensity Condition Gain (dB) Hearing Aid  Phonak Siemens Starkey Gain setting  Linear Compression Linear Compression Linear Compression 30 ms duration; 5 ms RF 65 dB SPL -6.4 -4.4 -13.2 -18.5 -5.3 1.5 30 ms duration; 10 ms RF -6.2 -4.3 -13.0 -18.8 -5.1 1.4 60 ms duration; 5 ms RF -3.2 -4.5 -10.6 -17.0 -4.6 -2.3 60 ms duration; 10 ms RF -3.2 -4.4 -10.7 -17.0 -4.6 -2.6 60 ms duration; 20 ms RF -3.7 -4.6 -11.2 -17.1 -4.9 -3.7 120 ms duration; 5 ms RF -1.7 -5.1 -9.2 -15.2 -5.4 -5.1 120 ms duration; 10 ms RF -1.6 -5.0 -9.1 -15.2 -5.3 -6.0 120 ms duration; 20 ms RF -1.4 -5.1 -8.9 -15.3 -5.1 -7.2 480 ms duration; 5 ms RF 0.9 -4.2 -5.3 -11.2 -4.7 -8.4 480 ms duration; 10 ms RF 0.9 -4.2 -5.3 -11.1 -4.7 -8.4 480 ms duration; 20 ms RF 1.1 -4.0 -5.2 -10.9 -4.6 -9.0 30 ms duration; 5 ms RF 45 dB SPL  6.2 16.1 1.9 0.5 7.2 22.7 30 ms duration; 10 ms RF 6.6 15.8 2.1 0.3 7.1 22.7 60 ms duration; 5 ms RF 10.0 19.5 4.9 2.9 9.8 21.8 60 ms duration; 10 ms RF 9.5 19.0 4.5 2.4 10.2 21.3 60 ms duration; 20 ms RF 8.6 18.1 3.7 1.7 11.3 20.5 120 ms duration; 5 ms RF 11.3 22.1 6.8 5.6 10.9 20.9 120 ms duration; 10 ms RF 11.2 22.4 6.6 6.0 10.2 20.9 120 ms duration; 20 ms RF 11.3 22.1 6.6 5.6 10.7 20.5 480 ms duration; 5 ms RF 13.0 23.3 10.2 8.7 10.7 19.9 480 ms duration; 10 ms RF 13.0 23.3 10.1 8.7 10.8 19.7 480 ms duration; 20 ms RF 13.1 23.3 10.2 8.8 11.0 19.8 105  3.5.1.1 Interpretation of intensity effects on SCP Tones  Unexpectedly, all unaided tones were measured at higher intensities than those recorded in the aided conditions in the 65 dB SPL condition, with the exception of the  2000 Hz tonal stimuli of 30 ms duration recorded from Starkey (compression). In the 45 dB SPL condition, positive gain was generally measured in aided conditions; negative gain was only measured from 500 Hz recordings of 30 ms to 60 ms duration and 1000 Hz recordings of 30 ms and 120 ms duration from Phonak (linear), as well as from the 500 Hz recordings of 30 ms to 60 ms duration from Siemens (linear). Maximum power output was set at maximum for all frequency bands and all hearing aids; as these stimuli were not presented at high intensities, gains were not expected to reach the maximum power output. Even if the output reached that limit, negative gain should not have been applied based on the hearing aid settings. These results were also not expected with expansion, as the negative gain was measured from the more intense of the two intensity conditions; expansion is applied to very soft stimuli inputs (Dillon, 2012).  The ear resonances of KEMAR were also ruled out as a possible factor (see Appendix C for measured ear resonances of KEMAR), because negative gains of similar amplitudes were found across frequencies, whereas, the resonances should have a larger effect on higher frequencies. The resonance effects should also occur consistently for all types of stimuli and conditions, but it does not, as seen throughout section 3.5. The coupler response for flat insertion gain (CORFIG) for KEMAR was calculated using measured real ear unaided gain, KEMAR real ear to coupler difference values measured by Killion & Revit (1993), and microphone location effects reported by Dillon (2012) (see Table 3.26). Applying the CORFIG to the insertion gain still does not explain the negative gain.  Even though all CORFIG correction values were negative, the only frequency where 106  negative gain was expected to be measured with a 65 dB SPL input was at 200 Hz and 6000 Hz, according to the input-output curves measured at each frequency for each hearing aid. Yet, at all frequencies, negative gain was observed in the 65 dB SPL condition. Other experimenter and order effects were also ruled out. Some discrepancies between prescribed gain and measured gain was expected as pure tones are fundamentally different stimuli from the speech stimuli used to evaluate gain in the test box (Souza & Tremblay, 2006).   Table 3.26.26CORFIG values for KEMAR.    The measured intensities for SCP Tones differed based on the hearing aid and the gain setting. This may be partially due to the differences in rise time; recordings with elongated rise times, such as those from Siemens or Phonak (linear), had measured intensities that were generally lower than those with rise times that were similar to unaided, such as those from Starkey (linear). Based on the data obtained from the 45 dB SPL data, it appeared that intensities measured from Siemens were the least affected by compression.  This was expected as its compression thresholds ranged from 50 dB SPL to 65 dB SPL across frequencies, while Phonak had compression thresholds ranging from below  40 dB SPL to 55 dB SPL and Starkey had compression thresholds below 40 dB SPL across all measured frequencies. As expected with compression, measured intensities were higher in both Starkey (compression) and Phonak (compression), with larger differences noted for the 2000 Hz stimuli as compared to the 500 Hz stimuli. The general pattern of more gain in the higher frequencies was seen, consistent with the frequency response set for the hearing aids. The effects of duration on the measured intensity of the stimuli were also observed. As expected with an increase in overall energy with increased duration, stimuli processed with Frequency (Hz) 250 Hz 500 Hz 1000 Hz 2000 Hz 3000 Hz 4000 Hz 6000 Hz CORFIG (dB) -4 -4 -3 -1 -1 -5 -21 107  hearing aids with linear gain generally had an increase in measured intensity with an increase in duration, although the amount of increase differed across hearing aids. Larger differences between durations were expected from Phonak and Siemens, especially with linear gain, due to their longer rise times. Although Phonak (compression) followed the duration pattern described, Starkey (compression) did not. With stimuli of longer durations, the measured intensities for Starkey were actually lower than those measured from stimuli of shorter durations. This may be due to how compression is applied by Starkey. As observed in Appendix B, stimulus envelopes of recorded SCP Tones from Starkey had attack time overshoots in intensity and shorter rise times that were consistent with the activation of the compressor. Thus, although it was normally expected for stimuli to increase in intensity with longer durations, when processed by Starkey (compression), stimuli of longer durations were long enough for the intensity overshoot to decline to a steady-state intensity that was lower than the initial intensity at the onset of the stimulus, while stimuli of shorter durations ended before the intensities reached a lower steady state intensity. Whether these intensity changes will affect the SCP morphology will depend on whether the stimuli are at or approximately around 60 to 80 dB nHL. If stimuli were presented below this saturation level, then hearing aid processing would be expected to increase the amplitude and decrease the latency of the SCP, assuming thresholds are below the presentation level. The effects on SCP morphology can also be considered using sensation level. Greater amplitude increases and latency decreases are expected for the first 10 dB above threshold than for increases of the same increment between 10 dB SL and 20 dB SL (Picton, 2011).  The amount of increase in amplitude and decrease in latency would differ based on the hearing aid used; for example, the recorded 757 ms duration 1000 Hz stimulus in the 45 dB SPL condition had a gain from 108  unaided of 3.7 dB when processed with Siemens (compression), but a gain from unaided of 14.6 dB when processed with Starkey (compression). Such increases from threshold, assuming that the unaided stimuli were presented below threshold, would cause significant increases in amplitude and decreases in latency, with larger effects seen if elicited by the tone processed by Starkey as compared to an SCP elicited by the tone processed by Siemens; if the intensity was well above threshold already, these effects would not be predicted (Picton, 2011).  Based on the stimuli from the 65 dB SPL condition, it may be predicted that the SCP amplitudes and latencies would be unexpectedly low and long, respectively, when elicited by hearing aid processed stimuli. 3.5.2 SCP Noise Intensities were measured for each stimulus in each hearing aid and intensity condition. Intensities were measured in volts and converted to dB volt with reference to 1 volt. Gains were calculated by subtracting the intensities of the unaided stimuli from the intensities of the respective aided stimuli in each hearing aid and intensity condition. See Table 3.27 for measured gains from unaided of noise bursts in all hearing aid and intensity conditions. Data measured for stimuli of opposite polarity are not shown.109  Table 3.27.27Gain relative to unaided, dB, in all hearing aid and intensity conditions for noise bursts. RF = rise and fall time.  Stimulus Description Intensity Condition Gain (dB) Hearing Aid  Phonak Siemens Starkey Gain setting  Linear Compression Linear Compression Linear Compression 30 ms duration; 5 ms RF 65 dB SPL -1.2 3.2 -6.7 -5.4 0.7 5.7 30 ms duration; 10 ms RF -1.2 3.2 1.6 -5.5 0.6 6.0 60 ms duration; 5 ms RF 0.4 1.9 -4.6 -3.6 1.2 3.2 60 ms duration; 10 ms RF 0.0 1.8 0.6 -4.1 1.0 2.4 60 ms duration; 20 ms RF 0.1 1.9 -5.3 -3.1 1.3 2.5 120 ms duration; 5 ms RF 2.5 1.6 -0.4 -0.3 1.4 0.8 120 ms duration; 10 ms RF 2.3 1.6 -0.8 -0.8 1.2 0.1 120 ms duration; 20 ms RF 1.8 1.0 -1.3 -1.2 0.7 -0.4 480 ms duration; 5 ms RF 3.7 0.9 0.2 2.5 1.2 -2.1 480 ms duration; 10 ms RF 3.9 0.9 -7.6 2.7 1.4 -2.2 480 ms duration; 20 ms RF 3.8 1.0 0.4 2.5 1.2 -2.6 30 ms duration; 5 ms RF 45 dB SPL  7.9 23.6 5.4 11.0 12.3 24.9 30 ms duration; 10 ms RF 8.0 22.7 4.9 10.3 11.7 24.6 60 ms duration; 5 ms RF 10.3 25.0 9.2 12.0 12.2 23.7 60 ms duration; 10 ms RF 10.8 25.8 9.3 12.1 12.5 24.3 60 ms duration; 20 ms RF 11.1 25.2 9.8 12.7 13.3 24.7 120 ms duration; 5 ms RF 10.7 25.4 10.9 14.2 11.6 23.2 120 ms duration; 10 ms RF 11.5 26.4 11.2 14.5 12.2 22.9 120 ms duration; 20 ms RF 12.1 26.3 11.5 14.6 12.2 23.2 480 ms duration; 5 ms RF 12.6 26.2 14.5 18.2 12.5 22.8 480 ms duration; 10 ms RF 12.4 26.2 14.5 18.2 12.4 22.8 480 ms duration; 20 ms RF 12.4 26.4 14.5 18.2 12.4 22.8 110  3.5.2.1 Interpretation of intensity effects on SCP Noise Unlike the pattern discussed for measured intensities of SCP Tones, stimuli recorded from aided conditions were not all below the measured intensities of unaided stimuli in the 65 dB SPL condition. In the 45 dB SPL condition, measured intensities of aided stimuli were all higher than unaided. Less discrepancy from the programmed gain may have been observed compared to the discrepancy seen for tones because noise bursts have greater resemblance than pure tones to speech stimuli, the type of stimulus used for programming the hearing aid (Souza & Tremblay, 2006).   However, similar to the pattern discussed for measured intensities of SCP Tones, the measured intensities for noise bursts differed based on the hearing aid and the gain setting. Differences between the intensities measured across hearing aids were likely partially related to the difference in rise times across hearing aids. For example, Siemens tended to have lower intensities measured due to long rise times in both intensity conditions, while Phonak with either gain setting in the 45 dB SPL condition had similar measured intensities as Starkey with either gain setting; rise times in recordings from the latter two hearing aids were similar to one another and to the unaided condition. As expected with compression, measured intensities were higher in both Starkey (compression) and Phonak (compression). Siemens had compression thresholds ranging between 50 dB SPL to 65 dB SPL across frequencies, while Phonak had compression thresholds ranging from below 40 dB SPL to  55 dB SPL and Starkey had compression thresholds below 40 dB SPL across all measured frequencies. Intensities measured in recordings from Siemens were generally the least affected by compression. Similar effects of duration on the measured intensity of the stimuli were observed for noise bursts as they were for tonal stimuli. As expected with an increase in 111  overall energy with increased duration, stimuli processed with hearing aids with linear gain had an increase in measured intensity with an increase in duration, although the amount of increase differed by hearing aids. Although Phonak (compression) followed the duration pattern described, Starkey (compression) did not; with stimuli of longer durations, the measured intensities were lower than those measured from stimuli of shorter durations. This may be due to how compression is applied by Starkey. As observed in Appendix B, stimulus envelopes of recorded SCP Noise from Starkey had attack time overshoots in intensity and shorter rise times that were consistent with the activation of the compressor. Thus, although it was normally expected for stimuli to increase in intensity with longer durations, when processed by Starkey (compression), stimuli of longer durations were long enough for the intensity overshoot to decline to a steady-state intensity that was lower than the initial intensity at the onset of the stimulus, while stimuli of shorter durations ended before the intensities reached a lower steady state intensity. Whether these intensity changes will affect the SCP morphology will also depend on whether the stimuli are at or approximately around 60 to 80 dB nHL. If stimuli were presented below these saturation levels and above threshold, then hearing aid processing would be expected to increase the amplitude and decrease the latency of the SCP. The effects on SCP morphology can also be considered using sensation level. Greater amplitude increases and latency decreases are expected for the first 10 dB above threshold than for increases of the same increment between 10 dB SL and 20 dB SL (Picton, 2011). The amount of increase in amplitude and decrease in latency would differ based on the hearing aid used. Increases from threshold, assuming that the unaided stimuli were presented below threshold, would cause significant increases in amplitude and latency. Based on the stimuli from the 65 dB SPL condition, however, it may be that the SCP 112  amplitudes and latencies would be unexpectedly low and long after hearing aid processing for some noise bursts.  3.5.3 ACC Tonal stimuli Intensities were measured for each stimulus in all hearing aid and intensity conditions; the 1680 Hz segment and the 1705 Hz segment were also measured separately. See Table 3.28 for gain values (aided subtracted from unaided) in all aided conditions in both intensity conditions and Table 3.29 for the difference between the intensities of the 1705 Hz segment and the 1680 Hz segment measured in all hearing aid and intensity conditions. The difference was calculated by subtracting the intensity measured from the 1705 Hz segment from the intensity measured from the 1680 Hz segment. Opposite polarity data are not shown.113   Table 3.28.28Gain relative to unaided, dB, of ACC tonal stimuli in all hearing aid and intensity conditions.   Table 3.29. 29Measured difference between intensities of the 1705 Hz segment and the 1680 Hz segment, dB, in all hearing aid and intensity conditions.    Stimulus Description Intensity Condition Gain (dB) Hearing Aid  Phonak Siemens Starkey Gain setting  Linear Compression Linear Compression Linear Compression 1680 to 1705 Hz tone 65 dB SPL -4.2  -15.2  -0.1  -15.1  -5.7  -13.1  1680 to 1705 Hz tone 45 dB SPL  3.5  16.2  9.9  8.0  9.4  18.7  Stimulus Description Intensity Condition  Difference between the 1705 Hz and 1680 Hz segment (dB) Hearing Aid  Unaided Phonak Siemens Starkey Gain setting   Linear Compression Linear Compression Linear Compression 1680 to 1705 Hz tone 65 dB SPL 0.4  2.9  1.3  2.7  2.6  1.8  -0.6  1680 to 1705 Hz tone 45 dB SPL  0.3  2.8  1.3  2.7  2.4  1.6  0.8  114  3.5.3.1 Interpretation of intensity effects on ACC Tonal Stimuli Similar to the results found with tonal stimuli presented at 65 dB SPL, aided stimuli had measured intensities that were below those measured for unaided stimuli. Again, this finding was perplexing and multiple factors have been ruled out, as discussed in section 3.5.1.1. Discrepancies between the gain for tonal stimuli as compared to broadband stimuli such as noise bursts and speech were expected, but negative gain was not (Souza & Tremblay, 2006). In the 45 dB SPL condition, all aided conditions results in positive gain. Starkey (compression) and Phonak (compression) also had even higher intensities as compared to unaided stimuli, as expected with compression with stimuli presented at lower levels. Changes observed were similar in both polarities unless otherwise noted. Compression was not expected to affect recordings from Siemens as the compression threshold was set at 50 dB SPL for that hearing aid. Thus, measured gain from Siemens (linear) were within 2 dB of the measured gain from Siemens (compression). Although general patterns emerge with different gain settings, the amount of change can differ based on the hearing aid. The intensity change of interest was more so between the 1680 Hz and 1705 Hz stimulus segments. An intensity increase between the two segments of 2 dB or greater (Martin & Boothroyd, 2000) with hearing aid processing could elicit a larger ACC than anticipated from frequency change alone.  Differences in intensities were unexpected with hearing aid processing. Differences in intensities in the unaided condition were no more than 0.4 dB; a difference of 0.4 dB is considered negligible as an ACC above noise floor is not observed until a 2 dB increase (Martin & Boothroyd, 2000). Hearing aid processing, increased the intensity difference between the two segments by more than 2 dB in recordings 115  from Phonak (linear) and Siemens with linear gain and compression in both intensity conditions. These changes are expected to increase ACC amplitudes and decrease ACC latencies. None of other hearing aid conditions are expected to affect ACC morphology due to intensity changes under the conditions investigated. 3.5.4 SCP Speech and ACC Speech Intensities were measured for each syllable and vowel recorded in all hearing aid and intensity conditions. Table 3.30 for gain values (aided subtracted from unaided) in all hearing aid conditions in both intensity conditions. Opposite polarity data are not shown.116   Table 3.30.30Gain relative to unaided, dB, for speech stimuli in all aided conditions and both intensity conditions. Stimulus  Intensity Condition Gain (dB) Hearing Aid  Phonak Siemens Starkey Gain setting  Linear Compression Linear Compression Linear Compression /da/ 65 dB SPL 7.3 -7.1 6.8 -9.0 7.0 -9.6 /ta/ 7.8 -6.7 8.0 -7.9 8.1 -8.3 /sa/ 13.5 -6.2 12.4 -2.7 11.1 -8.1 /ʃa/ 10.9 -5.3 10.9 -5.2 9.5 -9.2 /a/ 4.5 -7.3 4.4 -10.7 7.2 -8.0 /i/ 10.5 -5.7 11.5 -10.3 12.9 -3.5 /a/ cut pre-recording 8.4 -5.9 8.1 -7.6 8.7 -8.2 /a/ cut post-recording 9.1 -6.4 10.2 -6.0 8.9 -10.6 /da/ 45 dB SPL 3.5 13.8 6.6 8.9 7.0 14.6 /ta/ 3.7 13.9 7.0 9.3 7.5 15.3 /sa/ 10.3 16.8 12.0 13.5 10.2 16.7 /ʃa/ 6.2 18.7 10.5 11.2 8.8 16.8 /a/ 0.2 12.3 3.4 6.5 6.5 15.2 /i/ 8.4 20.8 11.9 11.6 12.6 20.3 /a/ cut pre-recording 4.0 18.1 7.9 9.2 8.0 15.7 /a/ cut post-recording 4.2 18.1 9.9 10.9 8.4 15.6 117  3.5.4.1 Interpretation of intensity effects on SCP and ACC Speech  Unlike the intensity patterns found for tonal stimuli and noise bursts, positive gain values were obtained in aided conditions in the 65 dB SPL condition; in fact, all hearing aids with linear gain had positive gain values. However, with compression, all processed stimuli had negative gain values. Gain values were all positive in the 45 dB SPL condition, no matter the gain setting, although the amount of gain changed with the hearing aid. Changes observed were similar in both polarities unless otherwise noted.  The findings of compression having negative gain in the 65 dB SPL condition were unexpected (Jenstad et al., 2000). Experimenter error, ear canal resonances, coupler gain differences, and expansion were all ruled out, as discussed in section 3.5.1.1. The findings of Starkey and Phonak aids yielding greater gain with compression than with linear gain in the 45 dB SPL condition were expected, as compression thresholds were either all measured to be below 40 dB SPL in Starkey, or some below 40 dB SPL in Phonak, as reported in Table 2.2.  As expected with compression thresholds ranging between 50 dB SPL to 65 dB SPL, similar intensities were measured from Siemens with both gain settings in the 45 dB SPL condition. Compared to SCP Tones and noise bursts, speech sounds were most similar to the stimuli used to adjust gain in the test box. Thus, the positive gain found in the 65 dB SPL condition, at least with the linear gain setting, was expected (Souza & Tremblay, 2006).   However, whether these intensity changes lead to changes in SCP morphology is still affected by whether the stimuli are presented near threshold. For example, with stimuli presented at 45 dB SPL, where people with a moderate hearing loss may not be able to detect unaided stimuli, gain due to hearing aid processing will likely lead to significant increases in amplitude and decreased latencies. However, as discussed, the effects of gain due to hearing 118  aid processing should also be considered with sensation levels. With hearing aids that provide a gain of greater than 10 dB above threshold with linear gain, further increases in intensity with compression may not affect SCP morphology, or only lead to smaller SCP amplitude increase and latency decreases (Picton, 2011). As intensity effects are largely determined by threshold and saturation levels, the effects of hearing aid processing can vary when interacting with subject parameters. However, as seen from the measurements, different hearing aids set to apply the same amount of gain over the programmable frequency bands still had variable gain that changed further with different gain settings.  3.5.5 HEARLab™ Stimuli Intensities were measured for each phoneme recorded in all hearing aid and intensity conditions. See Table 3.31 for measured gains from unaided of the phonemes recorded in all hearing aid and intensity conditions. Opposite polarity data are not shown. 119  Table 3.31.31Gain relative to unaided, dB, for HEARLab™ speech sounds in all aided conditions and both intensity conditions.       Stimulus  Intensity Condition Gain (dB) Hearing Aid  Phonak Siemens Starkey Gain setting  Linear Compression Linear Compression Linear Compression /m/ 65 dB SPL -17.8 -20.4 -17.5 -21.0 -7.6 -12.9 /t/ -2.2 -3.8 -2.0 -6.4 5.5 -1.0 /s/ 9.1 4.7 3.4 0.4 13.6 9.0 /g/ -14.4 -4.7 -13.2 -15.9 0.1 -0.4 /m/ 45 dB SPL -1.4 -3.4 -0.9 2.2 3.9 11.9 /t/ 10.6 23.4 12.5 11.4 15.8 22.9 /s/ 24.3 28.7 19.0 18.1 24.9 30.4 /g/ -3.0 10.8 1.2 1.3 9.9 17.3 120  3.5.5.1 Interpretation of intensity effects on HEARLab™ Stimuli  As with results obtained with SCP Speech, measured gains were not all negative in the 65 dB SPL condition: all aided conditions provided gain for the phoneme /s/ while all the other phonemes had negative gain, with the exception of /t/ and /g/ when processed with Starkey (linear). Stimuli processed by hearing aids in the 45 dB SPL condition had positive gain, with the exception of three instances in one polarity – the /m/ processed by Siemens (linear) and by Phonak with both gain settings – and four instances in the other polarity, with the addition of /g/ processed by Phonak (compression).  Changes observed were similar for both polarities unless otherwise noted. Experimental error, expansion, and ear canal resonances of KEMAR were ruled out for the unexpected negative gain results for all phonemes other than /s/ in the 65 dB SPL condition, as discussed in section 3.5.1.1.  As with the results obtained throughout this section, the amount of gain for the same stimulus and the same gain settings still differed across hearing aids. The compression threshold range of Siemens explains the small intensity difference between linear gain and compression for /s/. For the other two hearing aids, larger differences in intensity between gain settings suggest that the compressor was activated in the 45 dB SPL condition, consistent with their compression thresholds reported in Table 2.2. With the 65 dB SPL condition, the measured intensities for /s/ were also expected, as the measured intensities from the hearing aids with compression had lower measured intensities than the hearing aids with linear gain. However, although all hearing aids were set to have a compression ratio of 2:1, the difference between the measured intensities still differed based on the hearing aid, with Starkey and Phonak both attenuating the gain by 5 dB and Siemens attenuating the gain by 3 dB.  121   However, as with the discussion for other SCP eliciting stimuli, whether these intensity changes lead to changes in SCP morphology is affected by whether the stimuli are presented near threshold, with the largest changes expected within the first 10 dB above threshold (Picton, 2011).  It is important to be aware of the interactions of the spectrum of the stimulus and the hearing aid settings; as we see with intensities from /m/ in various hearing aid conditions, unexpected negative gain relative to unaided may decrease the amplitude and latency of the SCP morphology.  3.5.6 ABR Tone bursts and click  Intensities were measured for all the tone bursts and rarefaction clicks recorded in all hearing aid and intensity conditions. See Table 3.32 gain values of the tone bursts and rarefaction clicks in aided conditions and both intensity conditions. Unplanned ad hoc analyses of the intensities of the first and third repetition of the stimuli in each recording were measured to observe the intensity effects of different gain settings. The difference between the two intensities were calculated by subtracting the intensity of the third repetition from the first presentation. The comparison was between the first presentation and third repetition because the second repetition was of the opposite polarity. See Table 3.33 for intensity differences calculated between the first presentation and the third repetition in all hearing aid and intensity conditions. Another set of unplanned ad hoc analyses were performed to objectively measure the onset effect that Phonak had on the first six seconds of the full recording: the intensities of the first and second six seconds of the recording in each hearing aid condition and intensity conditions were measured (see Figure 3.21). The difference was subsequently calculated by subtracting the intensity of the second six seconds from the intensity of the first six seconds. The duration of six seconds was chosen as the 122  higher intensity onset was measured to last approximately six seconds by the experimenter. The severity of the onset effect varied with frequency for ABR stimuli. See Table 3.34 for intensity differences calculated between the first and second six seconds of the full recording in all hearing aid and intensity conditions. Opposite polarity data are not shown.123  Table 3.32.32Gain relative to unaided, dB, for 2-1-2 tone bursts of all presented frequencies and rarefaction clicks in all aided conditions and both intensity conditions.  Table 3.33. 33Intensity difference between the first presentation and third repetition, dB, of the full recording of 2-1-2 tone bursts of each frequency and rarefaction clicks in all hearing aid conditions and both intensity conditions.  Stimulus  Intensity Condition Gain (dB) Hearing Aid  Phonak Siemens Starkey Gain setting  Linear Compression Linear Compression Linear Compression 500 Hz 85 ppe SPL 10.8 0.8 6.5 -0.7 7.4 0.5 1000 Hz 4.7 2.3 8.4 1.9 8.6 2.0 2000 Hz 10.5 5.1 10.5 6.1 8.6 3.7 4000 Hz 16.0 5.7 19.5 19.3 15.9 7.5 Rarefaction Click  8.8 18.2 15.1 14.7 14.2 17.6 500 Hz 65 ppe SPL -1.7 11.5 5.8 10.5 6.6 12.5 1000 Hz -3.1 9.4 7.7 9.6 7.8 12.6 2000 Hz 3.6 8.5 10.0 8.4 9.0 16.8 4000 Hz 9.0 17.3 18.3 19.4 15.7 19.6 Rarefaction Click  2.2 16.8 11.2 10.2 14.4 22.6 Stimulus  Intensity Condition  Intensity difference between first presentation and third repetition (dB) Hearing Aid  Unaided Phonak Siemens Starkey Gain setting   Linear Compression Linear Compression Linear Compression 500 Hz 85 ppe SPL -0.1 -19.8 -11.7 -5.9 1.8 -0.1 12.0 1000 Hz 0.0 -14.8 3.6 -4.2 1.3 0.0 9.1 2000 Hz 0.0 -11.0 -5.7 -5.0 -4.4 0.0 16.2 4000 Hz 0.0 -10.1 5.3 -5.7 -5.8 0.0 15.0 Rarefaction Click  0.0 -4.8 -3.5 -4.5 -5.0 0.0 4.8 500 Hz 65 ppe SPL -0.1 -7.4 -6.9 -4.8 -5.3 -0.1 1.0 1000 Hz -0.1 -5.9 -2.5 -3.6 -3.5 0.1 2.3 2000 Hz 0.0 -6.6 -6.6 -4.1 -3.4 -0.1 3.4 4000 Hz 0.0 -8.7 -6.0 -4.4 -4.4 0.1 0.7 Rarefaction Click  0.0 -1.2 -1.2 -1.1 -1.1 0.0 0.2 124   Table 3.34.34Intensity difference between the first and six seconds, dB, of the full recording of 2-1-2 tone bursts of each frequency and rarefaction clicks in all hearing aid conditions and both intensity conditions.  Stimulus  Intensity Condition  Intensity difference between first and second six seconds of recording (dB) Hearing Aid  Unaided Phonak Siemens Starkey Gain setting   Linear Compression Linear Compression Linear Compression 500 Hz 85 ppe SPL -1.0 1.5 4.0 -1.1 -1.1 -1.0 -0.6 1000 Hz -1.0 0.0 -1.1 -1.1 -1.0 -1.0 -0.4 2000 Hz -1.0 2.0 2.8 -1.1 -1.0 -1.0 0.1 4000 Hz -1.0 2.0 1.3 -1.1 -1.1 -1.0 0.3 Rarefaction Click  -1.0 -1.0 0.5 -1.1 -1.1 -1.0 -0.8 500 Hz 65 ppe SPL -1.0 3.5 0.3 -1.1 -1.1 -0.9 -0.8 1000 Hz -1.0 0.0 -0.3 -1.1 -1.1 -0.9 -0.6 2000 Hz -1.0 0.2 0.2 -1.1 -1.1 -0.9 -0.8 4000 Hz -1.0 0.0 -0.4 -1.1 -1.1 -1.0 -0.8 Rarefaction Click  -0.9 -0.9 -0.8 -1.0 -1.1 -0.8 -0.9 125                  Figure 3.21.22Waveform of the full recording of 4000 Hz in Unaided (A), Phonak (linear) (B), and  Starkey (linear) (C) in the 85 ppe SPL condition. Onset effect, observed only in Phonak, lasts approximately 6 seconds. 0              0.5  1    1.5      2       2.5              (×106) 0.25 0.15 0.05 0 0.05 0.15 -0.25  0.4 0.15 0.05 0 0.05 -0.3 0.4 0.15 0.05 0 0.05 0.15 -0.4 Amplitude (volts) Time (samples) A B C 126  3.5.6.1 Interpretation of intensity effects on ABR Tone bursts and click  Unlike the intensity patterns seen with longer stimuli such as in SCP Tones where gain was negative in the more intense intensity condition, gain was generally positive in both intensity conditions for ABR tone bursts, with the exception of the 500 Hz tone burst recorded from Siemens (compression) in the 85 ppe SPL condition, and the 500 Hz and  1000 Hz tone bursts recorded from Phonak (linear). Effects were observed with compression, as Phonak and Starkey both had greater gain with compression than with linear gain in the  65 ppe SPL condition, but less gain with compression than with linear gain in the 85 ppe SPL condition. Siemens had similar gain with either gain setting. This was expected given the compression thresholds of each hearing aid, as reported in Table 2.2; 85 ppe SPL and 65 ppe SPL, when measured using dB SPL with A-weighting over approximately 35 seconds, ranged between approximately 60 to 70 dB SPL and 45 to 55 dB SPL, respectively. Another effect observed with compression was within the repetitions of the stimulus in the recording. The difference between the first presentation and third repetition of the tone burst was generally positive for two of the three hearing aids with compression in the 85 ppe SPL condition. In the 65 ppe SPL condition, positive differences were only measured from Starkey (compression), although the magnitude of difference was smaller than in the  85 ppe SPL condition. Changes observed were similar for both polarities unless otherwise noted. Unlike longer stimuli where an overshoot in intensity was seen before stimuli reached a steady state in intensity, brief tone-bursts and clicks used to elicit ABR were possibly too short, such that the overshoot in intensity lasted throughout the full first presentation of the stimulus, before the compressor was activated and a lower intensity was measured in subsequent repetitions.   127  With measured gains ranging from -0.7 dB to 19.5 dB in the 85 ppe SPL condition, and -3.0 dB to 22.6 dB in the 65 ppe SPL condition, hearing aid processing is expected to affect the latency and amplitude of waves I-V, although wave V is more resilient to changes in amplitude unless the intensities are near threshold, with negative gain leading to longer latencies and smaller amplitudes and positive gain leading to shorter latencies and increased amplitudes (Picton, 2011).  To objectively measure the onset effect observed in Phonak through visual inspection of the one-minute recording, intensity differences between the first and six seconds of the recording were measured across all hearing aid and intensity conditions. Phonak with either gain setting was the only hearing aid with positive differences between the two segments of the recording. This discrepancy in intensity was dependent on frequency and presentation level, as generally greater differences were noted in the lower frequencies and the  85 ppe SPL condition, than in the higher frequencies and the 65 ppe SPL condition. This onset effect was not consistently seen with clicks; the difference was negative in three of the four instances that the click was processed with Phonak hearing aids. However, the reason to this inconsistency remains unclear as we were unable to determine whether this is related to the spectral details, intensity, rate of presentation of the clicks, or gain setting, as it was mainly seen in Phonak (compression) in the 85 ppe SPL condition.  This onset effect was also observed in ASSR tonal stimuli (see section 3.5.9), where stimuli were presented at  40 to 80 Hz, but not in MLR tone bursts and clicks (see section 3.5.7), where a portion of the stimuli only differed from the ABR stimuli by the rate of presentation. As the rarefaction clicks in the ABR set of stimuli were presented at 19.1/s, a rate between 9.7/s, where the onset effect was not observed, and 39.1/s, where the onset effect was observed, the data 128  cannot provide strong support for one particular factor. Nevertheless, this effect was seen with Phonak, suggesting that approximately 235 of the repetitions are of greater intensities, and the latency and amplitude of the ABR may be slightly earlier than the rest of the stimuli over which the response is averaged. This may lead to constructive or destructive interactions with the other responses averaged, especially at lower frequencies (Picton, 2011). However, if we consider an averaged response of 2000 sweeps, this intensity change may not play a large role in the resulting wave as it can potentially only affect approximately ten percent of the sweeps; either an average containing more sweeps to compensate for this onset effect, or simply delaying the collection of data for approximately 6 seconds could minimize or prevent this onset effect from confounding the data. 3.5.7 MLR Tone bursts and click Intensities were measured for all the tone-bursts and rarefaction clicks recorded in all hearing aid and intensity conditions. See Table 3.35 for gain values of the tone bursts and rarefaction clicks in aided conditions and both intensity conditions. Unplanned ad hoc analyses of the intensities of the first and third repetition of the stimuli in each recording were measured to observe the intensity effects of different gain settings. The difference between the two intensities were calculated by subtracting the intensity of the third repetition from the first presentation. The comparison was between the first presentation and third repetition because the second repetition was of the opposite polarity. See Table 3.36 for intensity differences calculated between the first presentation and the third repetition in all hearing aid and intensity conditions. The other set of unplanned ad hoc analyses were also performed to objectively measure and compare the lack of the onset effect seen with recordings from Phonak in the set of ABR stimuli: the intensities of the first and second six 129  seconds of the recording in each hearing aid condition and intensity conditions were measured and the differences between the two segments were calculated. See section 3.5.6 for further discussion of the calculation and the choice of six seconds. See Figure 3.22 for a comparison of waveforms of a full recording of the 4000 Hz 2-1-2 tone burst presented at 10/s in the unaided condition and when processed by Phonak (linear). See Table 3.37 for intensity differences calculated between the first and second six seconds of the full recording in all hearing aid and intensity conditions. Opposite polarity data are not shown.  130  Table 3.35.35Gain relative to unaided, dB, for all 2-1-2 and 4-2-4 ms tone bursts and rarefaction clicks in all aided conditions and both intensity conditions.        Stimulus  Intensity Condition Gain (dB) Hearing Aid  Phonak Siemens Starkey Gain setting  Linear Compression Linear Compression Linear Compression 500 Hz (2-1-2) 85 ppe SPL -6.8 11.8 4.4 4.8 7.3 8.5 1000 Hz (2-1-2) -5.3 10.8 6.0 7.1 8.6 10.2 2000 Hz (2-1-2) 2.9 5.9 7.2 5.8 8.7 13.6 4000 Hz (2-1-2) 10.8 16.2 15.4 16.4 15.8 17.2 500 Hz (4-2-4 ms) -8.8 10.4 3.8 4.7 7.3 9.1 1000 Hz (4-2-4 ms)  -4.3 8.4 6.6 4.8 8.3 7.5 2000 Hz (4-2-4 ms) 4.5 5.4 5.9 2.3 9.7 10.6 Rarefaction Click  10.3 19.7 14.5 14.3 17.2 22.8 500 Hz (2-1-2) 65 ppe SPL -16.3 -8.4 -0.2 5.5 6.6 14.3 1000 Hz (2-1-2) -11.7 8.7 3.1 4.9 7.9 15.0 2000 Hz (2-1-2) 0.3 3.8 4.7 3.2 9.0 20.1 4000 Hz (2-1-2) 6.0 16.8 12.2 13.2 15.6 23.3 500 Hz (4-2-4 ms) -16.5 -9.3 -0.7 5.1 6.6 14.2 1000 Hz (4-2-4 ms)  -10.2 9.6 3.9 5.5 7.6 13.9 2000 Hz (4-2-4 ms) 1.7 6.9 3.9 1.2 10.0 19.4 Rarefaction Click  6.9 19.3 11.7 11.4 17.5 25.6 131   Table 3.36.36Intensity difference between the first presentation and third repetition, dB, of the full recording of all 2-1-2 and 4-2-4 ms tone bursts and rarefaction clicks in all hearing aid conditions and both intensity conditions.      Stimulus  Intensity Condition Intensity difference between first presentation and third repetition of recording (dB) Hearing Aid  Unaided Phonak Siemens Starkey Gain setting   Linear Compression Linear Compression Linear Compression 500 Hz (2-1-2) 85 ppe SPL 0.0 -8.3 -7.5 -8.2 -4.7 0.0 5.4 1000 Hz (2-1-2) 0.0 -7.2 -2.5 -4.9 -4.5 0.0 5.7 2000 Hz (2-1-2) 0.0 2.0 0.1 -4.7 -4.4 0.0 9.0 4000 Hz (2-1-2) 0.0 -2.8 0.0 -5.0 -5.8 0.0 9.0 500 Hz (4-2-4 ms) 0.0 -6.7 -17.5 -8.0 -4.6 0.0 6.1 1000 Hz (4-2-4 ms)  0.0 -8.0 -0.6 -5.2 -3.1 2.2 6.5 2000 Hz (4-2-4 ms) 0.0 -0.4 2.3 -4.8 -4.3 -6.7 10.1 Rarefaction Click  0.0 -1.2 -1.5 -3.1 -3.5 0.0 3.7 500 Hz (2-1-2) 65 ppe SPL 0.1 0.3 -0.3 -1.9 -3.1 0.0 -0.1 1000 Hz (2-1-2) -0.1 -0.1 -1.1 -1.8 -2.3 0.0 0.5 2000 Hz (2-1-2) -0.2 -1.2 0.6 -1.7 -2.1 0.1 0.5 4000 Hz (2-1-2) 0.0 -0.9 -1.1 -2.8 -2.3 -0.1 0.4 500 Hz (4-2-4 ms) 0.1 0.0 -0.6 -1.6 -3.5 0.1 0.2 1000 Hz (4-2-4 ms)  0.0 -0.8 1.1 -2.5 -1.4 0.0 0.8 2000 Hz (4-2-4 ms) 0.0 1.1 2.3 -3.0 -2.6 0.0 2.0 Rarefaction Click  -0.3 -0.1 -0.3 -1.0 -0.7 0.0 0.1 132  Table 3.37.37Intensity difference between the first and six seconds, dB, of the full recording of all 2-1-2 and 4-2-4 ms tone bursts and rarefaction clicks in all hearing aid conditions and both intensity conditions.       Stimulus  Intensity Condition Intensity difference between first and second six seconds of recording (dB) Hearing Aid  Unaided Phonak Siemens Starkey Gain setting   Linear Compression Linear Compression Linear Compression 500 Hz (2-1-2) 85 ppe SPL -1.0 -0.9 -0.9 -1.1 -1.1 -1.0 -0.8 1000 Hz (2-1-2) -1.0 -1.1 -0.9 -1.1 -1.0 -1.0 -0.8 2000 Hz (2-1-2) -1.0 -0.6 -0.7 -1.1 -1.1 -0.9 -0.5 4000 Hz (2-1-2) -1.0 -1.0 -0.9 -1.1 -1.1 -0.9 -0.4 500 Hz (4-2-4 ms) -0.9 -0.8 -1.0 -1.1 -1.0 -0.9 -0.6 1000 Hz (4-2-4 ms)  -0.9 -1.0 -0.9 -1.0 -1.0 -1.0 -0.6 2000 Hz (4-2-4 ms) -0.9 -0.7 -0.7 -1.1 -1.0 -0.9 0.3 Rarefaction Click  -1.1 -1.0 -1.0 -1.2 -1.2 -1.1 -1.1 500 Hz (2-1-2) 65 ppe SPL -0.9 -0.1 -0.1 -0.7 -0.9 -0.6 -0.6 1000 Hz (2-1-2) -0.9 0.0 -0.3 -0.9 -1.0 -0.6 -0.6 2000 Hz (2-1-2) -0.9 -0.5 -0.4 -1.0 -1.0 -0.7 -0.8 4000 Hz (2-1-2) -0.9 -1.0 -0.9 -1.1 -1.0 -0.9 -0.8 500 Hz (4-2-4 ms) -0.9 0.0 -0.2 -0.7 -0.9 -0.6 -0.5 1000 Hz (4-2-4 ms)  -0.9 -0.3 -0.5 -0.9 -0.9 -0.7 -0.6 2000 Hz (4-2-4 ms) -0.9 -0.6 -0.4 -1.0 -0.9 -0.8 -0.7 Rarefaction Click  -0.9 -0.7 -0.7 -0.9 -1.0 -0.9 -0.9 133                                           Figure 3.22.23Waveform of the full recording of 4000 Hz 2-1-2 tone burst in Unaided (A), and Phonak (linear) (B) in the 85 ppe SPL condition. Onset effect is not observed with Phonak in this set of stimuli.  0              0.5  1    1.5      2       2.5              (×106) 0.3 0.15 0.05 0 0.05 0.15 -0.3 0.2 0.15 0.05 0 0.05 0.15 -0.2 Amplitude (volts) Time (samples) A B 134  3.5.7.1 Interpretation of intensity effects on MLR Tone bursts and click  Unlike the intensity patterns seen with longer stimuli, such as in SCP Tones where gain was negative in the more intense intensity condition, gain was generally positive in both intensity conditions, with the exceptions of the 500 Hz and 1000 Hz tone bursts recorded from Phonak (linear) in the 85 ppe SPL and 65 ppe SPL conditions, and the 500 Hz tone bursts recorded from Siemens (linear) and Phonak (compression) in the 65 ppe SPL condition. The rate of presentation seemed to have an effect on how Phonak processed the brief 500 and 1000 Hz stimuli. The same 2-1-2 tone bursts presented at 39.1/s led to generally positive gain. However, when presented at a slower rate (i.e., 10/s), gain was negative in Phonak (linear) in both intensity conditions, and for 500 Hz stimuli processed by Phonak (compression) in the 65 ppe SPL condition. Tone bursts recorded from Starkey (compression) and Phonak (compression) had measured gains that were greater in the 65 ppe SPL condition than those measured from Starkey (linear) and Phonak (linear), likely due to the increased gain from compression. A similar pattern was noted in the 85 ppe SPL condition, albeit with smaller increases in intensity with compression as compared to with linear gain. Another effect observed with compression was within the repetitions of the stimulus in the recording. The difference between the first presentation and third repetition of the tone burst was generally positive for Starkey (compression) in the 85 ppe SPL condition. Unlike longer stimuli where an overshoot in intensity was seen before stimuli reached a steady state in intensity, brief tone-bursts and clicks used to elicit MLR were possibly too short, such that the overshoot in intensity lasted throughout the full first presentation of the stimulus, before the compressor was activated and a lower intensity was measured in 135  subsequent repetitions. Changes observed were similar for both polarities unless otherwise noted With measured gains ranging from -8.7 dB to 22.8 dB in the 85 ppe SPL condition, and -8.3 dB to 25.6 dB in the 65 ppe SPL condition, hearing aid processing is expected to affect the amplitude and latency measured from Na-Po with positive gain increasing its amplitude and decreasing its latency, up to about 60 dB above the hearing threshold (Borgmann et al., 2001). The amplitudes of the other peaks of the MLR are relatively unaffected by intensity effects (Maurizi et al., 1984). Negative gain, on the other hand, is expected to increase the latency of Po, but not affect the latencies of the other peaks in a clear manner; that is, the peaks do not seem to consistently decrease or increase in latency with intensity decrements (Maurizi et al., 1984). However, if the negative gain causes the stimuli to be near threshold, the negative peak of the MLR may be hard to identify as only the early positive peak, such as Po, is stable at threshold (Maurizi et al., 1984).   As observed from the data presented in Table 3.37 and Figure 3.22, the onset effect of Phonak observed with ABR and ASSR stimuli was not observed with any tone bursts and rarefaction clicks in any of the hearing aid and intensity conditions. This suggests a possible partial involvement of rate; from the data collected in this study, this effect is seen in stimuli presented at 39.1/s and above. 3.5.8 MMN Speech See Table 3.38 for gain values (unaided subtracted from unaided) were calculated for /ba/ and /da/ were measured in all hearing aid and intensity conditions. In addition, the intensity difference between the two syllables, calculated by subtracting the intensity 136  measured from /da/ from the intensity measured from /ba/, were also measured in all hearing aid conditions and both intensity conditions (See Table 3.39).  Table 3.38.38Gain relative to unaided, dB, of the syllables /ba/ and /da/ for all aided conditions in both intensity condition.   Table 3.39.39Intensity difference between /ba/ and /da/, dB, in all hearing aid and intensity conditions. 3.5.8.1 Interpretation of intensity effects on MMN Speech  Measured gain from unaided differed by hearing aid conditions for both intensity conditions. Unexpected based on previous speech findings, Phonak (linear) provided negative gain for both syllables in the 45 dB SPL condition, and for /da/ in the 65 dB SPL condition; recordings from hearing aids with compression in the 65 dB SPL condition had positive gain, unexpected from the negative gains measured in aided conditions in the SCP Speech set of stimuli in the same intensity condition (see section 3.5.4). However, the /ba/ and /da/ syllables were shorter than those presented in the SCP Speech set of stimuli. Although it was expected that the same amount of gain would be provided for both /ba/ and /da/ as they were relatively similar in their frequency make-up (see section 3.7.8), differences in gain were noted, even in the unaided condition, although the stimuli were calibrated on the Stimulus Intensity Condition Gain (dB) Hearing Aid Phonak Siemens Starkey Gain Setting Linear Compression Linear Compression Linear Compression /ba/ 65 dB SPL 0.9 6.3 2.4 3.3 5.9 8.1 /da/ -1.2 6.6 -0.2 1.6 4.9 7.6 /ba/ 45 dB SPL  -0.9 11.8 2.4 6.4 6.7 15.9 /da/ -1.3 8.7 0.0 4.6 4.8 14.3 Intensity Condition  Intensity difference between /ba/ and /da/ (dB) Hearing Aid Unaided Phonak Siemens Starkey Gain Setting  Linear Compression Linear Compression Linear Compression 65 dB SPL -2.4 -0.3 -2.8 0.2 0.6 -1.4 -1.9 45 dB SPL  -4.1 -2.8 -1.0 -1.7 -2.2 -2.1 -2.5 137  recording day. These small intensity changes between the standard and the deviant stimulus are unlikely to affect the latency and amplitude of the MMN morphology. However, it is important to note that if hearing aid processing brought the stimulus to an audible level from an inaudible level, one could expect the presence of an MMN response where there was not prior to aiding.   3.5.9 ASSR 40 and 80 Hz Amplitude Modulated (AM) Tones Intensities were measured for the full one-minute recording of the simultaneously presented 40 Hz and 80 Hz amplitude modulated tones in each hearing aid and intensity condition. See Table 3.40 for gain values relative to unaided in aided conditions in both intensity conditions. Unplanned ad hoc analyses of the intensities of the first seven seconds and intensity of the rest of the stimulus were also measured to observe and objectively measure the onset effects of Phonak on the modulated tonal stimuli in each hearing aid and intensity condition. See Table 3.41 for intensity difference between the two segments by subtracting the intensity of the second segment from the first seven seconds of the recording and see Figure 3.23 for a visual of the onset effect observed in the 40 Hz and 80 Hz AM waveform after Phonak (linear) processing in the 65 dB SPL condition.  138  Table 3.40.40Gain relative to unaided, dB, for all AM tones in all aided conditions and both intensity conditions.   Table 3.41.41Intensity difference between the first seven seconds and the rest of the recording, dB, for all AM tones in all hearing aid and intensity conditions.        Stimulus Intensity Condition Gain (dB)  Hearing Aid Phonak Siemens   Starkey  Gain Setting Linear Compression Linear Compression Linear Compression 40 Hz AM Tone 65 dB SPL 8.6 -9.6 18.0 2.8 -4.7 -28.0 80 Hz AM Tone 9.1 -8.7 17.5 2.1 -5.2 -29.9 40 Hz AM Tone 45 dB SPL  2.6 13.7 14.7 25.3 8.3 20.1 80 Hz AM Tone 1.9 13.9 14.1 24.7 7.9 19.6 Stimulus Intensity Condition Intensity difference between the first seven seconds and the rest of the recording (dB)  Hearing Aid Unaided Phonak  Siemens   Starkey  Gain Setting  Linear Compression Linear Compression Linear Compression 40 Hz AM Tone 65 dB SPL -0.2 0.6 1.1 -0.3 -0.3 -0.2 0.1 80 Hz AM Tone -0.2 1.8 0.1 -0.3 -0.3 -0.2 0.1 40 Hz AM Tone 45 dB SPL  -0.2 1.6 1.0 -0.3 -0.3 -0.2 -0.1 80 Hz AM Tone -0.1 1.5 1.1 -0.3 -0.3 -0.1 -0.1 139                                                    Figure 3.23.24Waveform of the 40 Hz AM tone (A), and 80 Hz AM tone (B) processed by Phonak (linear) in the 65 dB SPL condition. The black outline is the envelope of the recording.   A B 140  3.5.9.1 Interpretation of intensity effects on ASSR 40 and 80 Hz AM Tones  Interestingly, although ASSR stimuli were tonal in nature, negative gain was not measured across all hearing aid conditions in the 65 dB SPL condition. Negative gain was measured for recordings from Starkey with both gain settings and Phonak (compression), while positive gain was measured from all other hearing aid conditions. However, it is important to note that ASSR stimuli were amplitude modulated and continuous, with the tonal stimuli of different frequencies presented simultaneously. Thus, the ASSR stimuli have more speech-like qualities than the tones recorded in the SCP Tones, which may lessen the discrepancy between measured and programmed gain (Souza & Tremblay, 2006). As the ASSR amplitude increases with intensity with no saturation effects, at least up to 70 dB nHL, it is expected that negative gain can lead to decreased amplitudes while positive gain would lead to increased amplitudes (Picton, Skinner, Champagne, Kellett, & Maiste, 1987). Thus, hearing aid processing could decrease or eliminate ASSR responses in some cases. .   As discussed in sections 3.5.6.1, an onset effect was visually observed with Phonak recordings. From the data presented in Table 3.41, the measured intensity of the first seven seconds of the stimulus was also larger than the rest of the stimuli in three of the four recordings from Phonak: the one exception was the 80 Hz AM tone in the 65 dB SPL condition had a gain of 0.1 dB, which is considered negligible as the recordings from unaided had differences of 0.1 dB between the two segments. As the ASSR is also identified over time using a statistical manner, a gain of at most 1.8 dB over 6 to 7 seconds is unlikely to cause a response to reach significance as recordings are generally made over minutes (Picton, Dimitrijevic, Perez-Abalo, & van Roon, 2005).  141  3.6 SNR 3.6.1 SCP Tones  SNRs, calculated in dB, were measured for each stimulus in all hearing aid and intensity conditions. See Tables 3.42 to 3.44 for measured SNRs of tones of each frequency in all hearing aid and intensity conditions.  Data measured for stimuli of opposite polarity are not shown.142  Table 3.42.42SNR, dB, of 500 Hz tones in all hearing aid and intensity conditions. RF = rise time.   Stimulus Description Intensity Condition  SNR (dB) Hearing Aid  Unaided Phonak Siemens Starkey Gain setting   Linear Compression Linear Compression Linear Compression 30 ms duration; 5 ms RF 65 dB SPL 32.5 5.4 12.6 14.6 17.5 19.7 19.3 30 ms duration; 10 ms RF 29.9 4.8 11.1 15.5 17.0 19.8 18.8 60 ms duration; 5 ms RF 30.4 16.5 13.7 16.7 17.5 19.9 15.6 60 ms duration; 10 ms RF 31.5 16.3 13.2 16.9 17.2 19.7 15.7 60 ms duration; 20 ms RF 30.4 14.8 13.5 16.8 17.4 19.9 14.7 120 ms duration; 5 ms RF 31.0 24.4 15.1 19.6 18.8 19.9 13.5 120 ms duration; 10 ms RF 30.9 24.3 14.6 19.4 19.0 19.6 13.4 120 ms duration; 20 ms RF 30.7 24.2 14.7 19.5 18.7 19.8 12.4 480 ms duration; 5 ms RF 31.6 29.8 15.6 23.3 20.9 20.1 11.4 480 ms duration; 10 ms RF 30.8 29.8 15.6 23.1 20.8 20.0 11.1 480 ms duration; 20 ms RF 31.6 29.7 15.7 23.1 20.8 20.1 11.0 30 ms duration; 5 ms RF 45 dB SPL  -3.3 -13.8 -9.6 -7.2 -0.6 -0.4 5.2 30 ms duration; 10 ms RF -3.2 -14.1 -12.6 -7.0 -1.5 -0.9 5.1 60 ms duration; 5 ms RF -4.8 -6.3 -2.2 -5.7 0.8 -1.2 8.5 60 ms duration; 10 ms RF -4.7 -7.5 -3.2 -5.8 0.4 -1.1 6.5 60 ms duration; 20 ms RF -4.3 -8.8 -5.6 -6.0 0.4 -1.5 8.5 120 ms duration; 5 ms RF -4.6 0.2 0.6 -3.1 3.5 -1.0 4.4 120 ms duration; 10 ms RF -4.9 -0.2 1.0 -3.6 3.6 -1.4 4.2 120 ms duration; 20 ms RF -4.5 -0.2 0.5 -3.4 2.9 -1.3 4.3 480 ms duration; 5 ms RF -4.8 4.8 3.9 0.3 7.6 -1.1 6.0 480 ms duration; 10 ms RF -4.8 4.8 3.5 0.3 7.5 -1.0 5.7 480 ms duration; 20 ms RF -4.8 4.8 3.6 0.3 7.5 -1.0 5.9 143          Stimulus Description       Intensity Condition         SNR (dB)  Hearing Aid  Unaided Phonak Siemens Starkey Gain setting   Linear Compression Linear Compression Linear Compression 30 ms duration; 5 ms RF 65 dB SPL  29.2 12.8 14.3 18.6 18.4 19.8 17.9 30 ms duration; 10 ms RF 28.8 12.9 15.2 19.8 18.4 20.0 17.6 60 ms duration; 5 ms RF 30.5 17.3 12.8 20.1 19.0 19.7 14.6 60 ms duration; 10 ms RF 30.8 17.4 12.9 19.8 18.8 19.5 13.5 60 ms duration; 20 ms RF 29.2 17.7 10.6 20.1 16.9 19.9 11.8 120 ms duration; 5 ms RF 29.2 21.2 12.1 22.5 19.9 19.7 12.0 120 ms duration; 10 ms RF 30.3 21.2 12.2 22.3 19.6 19.9 11.3 120 ms duration; 20 ms RF 27.9 19.6 10.0 20.8 18.0 18.3 9.4 450 ms duration; 10 ms RF 28.7 22.3 9.1 23.9 19.9 18.6 8.1 480 ms duration; 5 ms RF 30.3 23.9 11.6 25.3 21.7 20.2 9.8 480 ms duration; 10 ms RF 30.5 24.2 11.7 25.4 21.7 20.3 9.9 480 ms duration; 20 ms RF 30.2 23.9 11.4 25.2 21.5 20.2 9.4 757 ms duration; 7.57 ms RF 26.0 19.9 7.5 20.6 17.6 16.1 5.3 30 ms duration; 5 ms RF 45 dB SPL -4.7 -9.5 -1.0 -3.3 -0.8 -0.8 -2.1 30 ms duration; 10 ms RF -3.9 -9.6 -0.2 -3.0 -0.6 -0.7 -3.8 60 ms duration; 5 ms RF -4.8 -5.5 -2.4 -2.3 0.8 -0.8 -0.4 60 ms duration; 10 ms RF -4.6 -5.0 -1.9 -2.1 0.6 -0.6 -0.8 60 ms duration; 20 ms RF -4.2 -5.8 -2.7 -2.3 0.7 -0.9 -0.5 120 ms duration; 5 ms RF -5.1 -1.9 -2.3 -0.7 2.5 -0.9 -3.2 120 ms duration; 10 ms RF -5.3 -1.7 -2.3 -0.4 2.5 -1.1 -2.3 120 ms duration; 20 ms RF -6.2 -10.9 -3.9 -1.9 0.9 -2.6 -5.7 450 ms duration; 10 ms RF -5.9 -0.6 -2.9 1.0 4.1 -2.1 -2.1 480 ms duration; 5 ms RF -5.0 0.9 -1.7 2.7 5.6 -0.6 -1.0 480 ms duration; 10 ms RF -4.9 0.9 -1.7 2.6 5.6 -0.7 -2.0 480 ms duration; 20 ms RF -4.9 1.2 -2.0 2.7 5.5 -0.7 -1.8 757 ms duration; 7.57 ms RF -7.9 -2.8 -5.6 -1.2 1.6 -4.4 -0.2 Table 3.43.43SNR, dB, of 1000 Hz tones in all hearing aid and intensity conditions. RF = rise and fall time. 144  Table 3.44.44SNR, dB, of 2000 Hz tones in all hearing aid and intensity conditions. RF = rise and fall time.  Stimulus Description Intensity Condition  SNR (dB) Hearing Aid  Unaided Phonak Siemens Starkey Gain setting   Linear Compression Linear Compression Linear Compression 30 ms duration; 5 ms RF 65 dB SPL 31.1 25.6 21.9 19.2 14.7 23.0 24.7 30 ms duration; 10 ms RF 30.1 25.7 21.4 19.3 14.0 23.4 24.6 60 ms duration; 5 ms RF 30.9 27.7 20.7 21.3 15.2 23.6 20.2 60 ms duration; 10 ms RF 30.2 28.4 20.3 21.8 15.1 23.2 19.9 60 ms duration; 20 ms RF 30.5 27.9 19.8 20.9 15.4 23.5 19.7 120 ms duration; 5 ms RF 31.7 30.8 20.1 23.7 18.1 23.7 18.4 120 ms duration; 10 ms RF 31.0 30.9 20.5 23.7 17.8 23.8 17.6 120 ms duration; 20 ms RF 31.1 30.8 20.6 23.6 17.7 23.9 15.9 480 ms duration; 5 ms RF 31.2 33.0 21.0 27.2 21.8 24.1 14.9 480 ms duration; 10 ms RF 30.4 32.9 21.2 27.1 21.9 24.0 14.9 480 ms duration; 20 ms RF 29.7 32.9 20.9 27.2 21.8 23.9 14.1 30 ms duration; 5 ms RF 45 dB SPL  -3.5 2.2 2.8 -2.7 -3.7 2.5 0.2 30 ms duration; 10 ms RF -3.6 2.0 2.4 -2.6 -4.3 3.3 0.1 60 ms duration; 5 ms RF -5.0 4.2 5.0 -0.8 -2.8 3.4 0.8 60 ms duration; 10 ms RF -4.6 4.5 4.8 -0.5 -2.9 3.7 0.4 60 ms duration; 20 ms RF -4.1 4.0 5.0 -0.9 -2.9 3.5 0.7 120 ms duration; 5 ms RF -4.6 5.8 7.8 1.6 0.1 3.2 0.0 120 ms duration; 10 ms RF -4.7 5.6 8.8 1.2 0.7 2.9 -0.1 120 ms duration; 20 ms RF -4.6 5.8 7.6 1.5 0.3 3.1 0.0 480 ms duration; 5 ms RF -4.7 7.5 9.1 4.8 3.3 3.3 0.4 480 ms duration; 10 ms RF -4.7 7.6 9.2 5.0 3.4 3.3 0.3 480 ms duration; 20 ms RF -4.8 7.5 8.5 4.8 3.3 3.0 0.3 145  3.6.1.1 Interpretation of SNR effects on SCP Tones  As expected from intensity measurements shown in section 3.5.1 for tonal stimuli, measured SNRs were higher in unaided than aided conditions in the 65 dB SPL condition. Although measured intensities were higher than unaided in the 45 dB SPL condition, patterns for measured SNRs were not as clear-cut. Although most aided conditions in the 45 dB SPL condition generally had measured SNRs that were higher, SNRs were lower than unaided recordings for some tones. However, having poorer or lower SNRs in aided conditions in the 65 dB SPL condition was not unexpected, based on findings by Billings et al. (2009). Nevertheless, there was considerable variability among hearing aid conditions. Hearing aid processing did not always lead to poorer SNRs, and gain settings were able to further introduce more variability depending on intensity condition. For example, while Siemens (compression) in the 45 dB SPL condition provided more favourable SNRs than with linear gain across frequencies, the same was not observed for Phonak. In fact, while compression generally provided more favourable SNRs for stimuli processed by Phonak in the 45 dB SPL condition, compression tended to provide less favourable SNRs in all other hearing aid conditions tested. The discrepancy is greater with longer durations. This pattern may be due to larger observed overshoot, such as those observed in recordings from Starkey, that had no time to reach a lower steady state intensity for shorter durations. Thus, lower SNRs may have been measured because of the combination of lower steady state intensities, and an increase in aided gain of low background noise. This demonstrated the variability that gain settings can add to SNRs obtained from stimuli that have been processed by hearing aids.   The effects of SNR changes to SCP morphology can depend on the rate of presentation of the stimulus (Papesh et al., 2015). With a slower rate of presentation (e.g., 146  ISIs of 1900 ms), recordings from aided conditions in the 65 dB SPL condition can decrease N1 amplitudes (Billings et al., 2009; Papesh et al., 2015). Although N1 amplitudes can be enhanced with lower SNRs between 10 dB and 30 dB SNR as compared to quiet at higher stimulation rates (e.g., ISIs of 900 ms), some background noise was measured in unaided conditions; since N1 amplitudes did not differ significantly with 10 dB SNR compared to 30 dB SNR, N1 amplitudes are not expected to be enhanced with hearing aid processing (Papesh et al., 2015). P2 amplitudes are expected to decrease with hearing aid processing in the 65 dB SPL condition and increase in the 45 dB SPL condition, using either rate of presentation (Billings et al., 2009; Papesh et al., 2015). The 1000 Hz tonal stimuli based on Billings et al (2007 & 2012), on the other hand, may see a decline in N1 amplitudes, if the ISI is kept at 1900 ms (Billings et al., 2007; Billings et al., 2009; Papesh et al., 2015). In the 45 dB SPL, hearing aid processing may enhance SCP amplitudes as compared to unaided conditions due to improved SNRs (Billings et al., 2009). However, with SNRs below 10 dB SNR, it is unknown whether the rate of presentation may further affect the amplitude of the SCP as SNRs poorer than 10 dB were not tested by Papesh et al. (2015). Latencies would be expected to be longer for lower SNRs as compared to unaided, no matter the rate of presentation (Papesh et al., 2015). However, it is important to keep in mind that Papesh et al. (2015) used a speech stimulus, and faster rates only enhanced N1 amplitudes significantly when stimuli were presented in noise binaurally relative to quiet. Interpretations are based on the assumption of binaural amplification and stimuli presented in sound field. Replicability of the results from Papesh et al. (2015) with tonal stimuli remains to be investigated.     147  3.6.2 SCP Noise SNRs, calculated in dB, were measured for each stimulus in all hearing aid and intensity conditions. See Table 3.45 for measured SNRs of noise bursts in all hearing aid and intensity conditions. Opposite polarity data are not shown.148  Table 3.45.45SNR, dB, of noise bursts in all hearing aid and intensity conditions.  Stimulus Description Intensity Condition  SNR (dB) Hearing Aid  Unaided Phonak Siemens Starkey Gain setting   Linear Compression Linear Compression Linear Compression 30 ms duration; 5 ms RF 65 dB SPL 16.7 27.9 25.0 22.6 23.7 26.7 25.1 30 ms duration; 10 ms RF 17.6 28.3 26.4 31.3 24.0 27.2 26.1 60 ms duration; 5 ms RF 16.6 28.8 23.0 24.4 25.3 27.0 22.9 60 ms duration; 10 ms RF 17.0 29.0 24.5 30.0 25.6 27.3 22.4 60 ms duration; 20 ms RF 16.8 29.0 24.1 24.2 26.5 27.3 22.3 120 ms duration; 5 ms RF 16.4 31.2 23.1 28.8 28.7 27.0 20.5 120 ms duration; 10 ms RF 16.9 31.2 23.7 28.7 28.5 27.3 19.9 120 ms duration; 20 ms RF 17.0 30.8 22.9 28.1 28.1 26.8 19.1 480 ms duration; 5 ms RF 17.3 33.1 23.0 30.0 32.3 28.0 17.9 480 ms duration; 10 ms RF 17.3 33.2 23.1 22.1 32.4 28.0 17.7 480 ms duration; 20 ms RF 17.2 33.1 23.1 30.1 32.2 27.8 17.3 30 ms duration; 5 ms RF 45 dB SPL  -1.8 5.2 9.9 2.9 7.4 6.7 10.0 30 ms duration; 10 ms RF -1.1 5.7 10.1 2.5 7.7 6.5 9.2 60 ms duration; 5 ms RF -1.7 7.5 12.4 6.8 9.2 6.7 8.9 60 ms duration; 10 ms RF -2.1 7.5 12.2 6.4 8.7 7.4 8.9 60 ms duration; 20 ms RF -2.0 7.9 11.8 7.3 9.5 8.2 10.0 120 ms duration; 5 ms RF -1.6 8.1 12.3 8.4 11.2 6.6 8.0 120 ms duration; 10 ms RF -1.8 8.7 13.1 8.7 11.3 7.1 8.0 120 ms duration; 20 ms RF -1.8 9.2 12.7 9.0 11.4 6.7 7.7 480 ms duration; 5 ms RF -1.8 9.9 13.1 12.0 15.2 7.4 7.8 480 ms duration; 10 ms RF -1.7 9.7 13.0 11.9 15.2 7.2 7.6 480 ms duration; 20 ms RF -1.7 9.8 13.2 12.1 15.2 7.4 7.8 149  3.6.2.1 Interpretation of SNR effects on SCP Noise Unlike the findings from SCP Tones, noise bursts of all durations in the aided conditions had measured SNRs that were greater than unaided in both intensity conditions. This was somewhat unexpected given that gain values of SCP noise were negative for some hearing aid conditions in the 65 dB SPL condition. Effects were observed with compression: in the 45 dB SPL condition, hearing aids with compression had consistently higher SNRs than their linear counterparts, while in the 65 dB SPL condition, hearing aids with compression generally had lower SNRs measured than their linear counterparts.  In addition, the measured SNRs also followed the programmed compression ratio, as many of the recorded stimuli differed from their respective linear counterparts by approximately 3 dB. However, as expected from the intensity data, there was variability in the SNR measured from each hearing aid condition and each intensity condition.    The effects of SNR changes to SCP morphology can depend on the rate of presentation of the stimulus, as discussed in section 3.6.1 (Papesh et al., 2015). Although N1 amplitudes can be enhanced with lower SNRs between 10 dB and 30 dB SNR as compared to quiet at higher stimulation rates (e.g., ISIs of 900 ms), background noise was measured in unaided conditions; since N1 amplitudes did not differ significantly with 10 dB SNR compared to 30 dB SNR, N1 amplitudes are not expected to be decreased with hearing aid processing (Papesh et al., 2015). P2 amplitudes are expected to increase with hearing aid processing using either rate of presentation in both intensity conditions (Billings et al., 2009; Papesh et al., 2015). In the 45 dB SPL condition, the effect of rate is unknown as some measured SNRs were below 10 dB SNR. Latencies would be expected to be longer for lower SNRs as compared to unaided, no matter the rate of presentation (Papesh et al., 2015). 150  However, it is important to keep in mind that Papesh et al. (2015) used a speech stimulus, and faster rates only increased SCP amplitudes significantly when stimuli were presented binaurally. Interpretations are based on the assumption of binaural amplification and stimuli presented in sound field. Replicability of the results from Papesh et al. (2015) with noise bursts remains to be investigated 3.6.3 ACC Tonal stimuli SNRs, calculated in dB, were measured for each stimulus in all hearing aid and intensity conditions. See Table 3.46 for measured SNRs of ACC tonal stimuli in all hearing aid and intensity conditions. Opposite polarity data are not shown.151  Table 3.46.46SNR, dB, of ACC tonal stimuli in all hearing aid and intensity conditions.   Stimulus Description Intensity Condition  SNR (dB) Hearing Aid  Unaided Phonak Siemens Starkey Gain setting   Linear Compression Linear Compression Linear Compression 1680 to 1705 Hz tone 65 dB SPL 28.4  23.7  7.8  27.8  15.4  20.0  7.6  1680 to 1705 Hz tone 45 dB SPL  -6.5  -3.8  1.8  2.7  0.9  -0.2  2.0  152  3.6.3.1 Interpretation of SNR effects on ACC Tonal Stimuli  Measured SNRs changed with hearing aid conditions, with hearing aids set to compression having lower measured SNRs than their respective linear counterparts in the  65 dB SPL condition, and for Phonak and Starkey in the 45 dB SPL condition. In fact, all measured SNRs were below 5 dB in the 45 dB SPL condition, which would not be considered favourable (Billings et al., 2011).  Lower SNRs may be a combination of the discrepancy between measured and programmed gain for tonal stimuli and gain provided to low input noise levels by the hearing aids. Changes observed were similar for both polarities. As seen with all sections in Chapter 3, there was considerable variability among hearing aids, in terms of the measured SNRs from each hearing aid condition and intensity condition. As the effects of SNR on the ACC have not been studied in a systematic manner, these results cannot be interpreted for their significance on the ACC morphology.  3.6.4 SCP Speech and ACC Speech SNRs, calculated in dB, were measured for each stimulus in all hearing aid and intensity conditions. See Table 3.47 for measured SNRs of syllables and vowels recorded in all hearing aid conditions and intensity conditions. Opposite polarity data are not shown. 153   Table 3.47.47SNR, dB, of all speech stimuli measured in all hearing aid and intensity conditions.      Stimulus  Intensity Condition  SNR (dB) Hearing Aid  Unaided Phonak Siemens Starkey Gain setting   Linear Compression Linear Compression Linear Compression /da/ 65 dB SPL 31.2 30.9 19.2 32.1 24.5 24.4 14.8 /ta/ 29.9 29.3 17.6 31.1 23.8 23.6 14.2 /sa/ 28.6 33.5 16.6 34.0 27.6 25.0 13.0 /ʃa/ 30.3 33.1 19.5 34.5 27.1 25.2 13.6 /a/ 30.4 26.2 17.7 28.0 21.4 22.6 14.9 /i/ 33.2 35.2 22.0 38.1 24.5 31.7 22.3 /a/ cut pre-recording 30.3 30.4 19.2 31.8 24.5 24.4 14.9 /a/ cut post-recording 31.1 32.3 19.2 34.8 27.1 25.5 13.1 /da/ 45 dB SPL -2.6 0.1 3.8 3.4 5.8 1.5 2.0 /ta/ -4.7 -1.7 1.7 1.8 4.1 -0.1 0.8 /sa/ -5.4 4.2 3.8 6.0 7.6 1.9 1.4 /ʃa/ -4.3 1.3 7.0 5.7 6.5 1.8 2.8 /a/ -4.7 -5.7 0.1 -2.2 1.1 -1.3 0.1 /i/ -1.7 5.7 11.5 9.5 9.1 7.7 8.4 /a/ cut pre-recording -4.0 -0.8 6.4 3.4 4.7 0.9 1.8 /a/ cut post-recording -3.7 -0.2 6.9 5.6 6.8 1.9 2.1 154  3.6.4.1 Interpretation of SNR effects on SCP and ACC Speech  While positive gain was measured in the 65 dB SPL condition for linear hearing aids, measured SNRs for five out of the six aided conditions resulted in lower SNRs as compared to their unaided counterparts. Changes observed were similar for both polarities unless otherwise noted. Only Siemens (linear) produced higher SNRs compared to the unaided SNRs, with the exception of /a/.  In the 45 dB SPL condition, most aided SNRs were more favourable than the unaided condition; the /a/ recorded from Phonak (linear) was the only stimulus to have a poorer SNR than unaided. Having poorer SNRs in the 65 dB SPL condition was expected (Billings et al., 2011). As the unaided stimulus was already well above the noise floor, hearing aids may provide gain to the noise floor that can be detrimental to the SNRs (Billings et al., 2011).  This may be why even poorer SNRs were measured from stimuli processed by hearing aids with compression as compared to hearing aids with linear gain; as compression may have provided more gain to less intense inputs. In the 65 dB SPL condition, compression may have decreased the difference between the signal and the noise by increasing the noise floor and decreasing the signal intensity (Souza, 2002). Contrarily, in the 45 dB SPL condition, the stimulus and noise floor were similar in intensity; aided conditions may have provided the stimulus with more gain to be above noise floor. For example, stimuli with higher frequency content, such as /i/, had SNRs that were more favourable than stimuli of lower frequency content, such as /a/.  Of the stimuli measured to elicit SCPs, this set of stimuli most closely resembled those used by Papesh et al. (2015). The effects of SNR changes to SCP morphology can depend on the rate of presentation of the stimulus, as discussed in section 3.6.1 (Papesh et al., 2015). With a slower rate of presentation (e.g., ISIs of 1900 ms), N1 amplitudes are expected 155  to be enhanced with linear hearing aids in the 65 dB SPL condition and all aided conditions in the 45 dB SPL condition, but decreased with compression in the 65 dB SPL condition (Billings et al., 2009; Papesh et al., 2015). Although N1 amplitudes can be enhanced with lower SNRs between 10 dB and 30 dB as compared to quiet at higher stimulation rates (e.g., ISIs of 900 ms), background noise was measured in unaided conditions; since N1 amplitudes did not differ significantly with 10 dB SNR compared to 30 dB SNR, N1 amplitudes are not expected to be enhanced with compression in the 65 dB SPL condition, or decreased with hearing aid processing in other conditions (Papesh et al., 2015). P2 amplitudes are expected to decrease with compression in the 65 dB SPL condition, but increase in hearing aid processing where SNRs were improved (Billings et al., 2009; Papesh et al., 2015). Latencies would be expected to be longer for lower SNRs as compared to unaided, regardless of the rate of presentation (Billings et al., 2009; Papesh et al., 2015). Fewer latency effects may be observed for speech stimuli with higher frequency emphasis, such as in /sa/ and /ʃa/; please note that less of a latency effect was observed with higher frequency emphasis when multi-talker babble was the noise signal in the study by Kuruvilla-Mathew et al (2015), while the noise measured in this study does not contain any speech-like stimuli.  3.6.5 HEARLab™ Stimuli SNRs, calculated in dB, were measured for each stimulus in all hearing aid and intensity conditions. See Table 3.48 for measured SNRs of HEARLab™ speech sounds recorded in all hearing aid and intensity conditions. Opposite polarity data are not shown.156  Table 3.48.48SNR, dB, of all HEARLab™ speech sounds measured in all hearing aid and intensity conditions.     Stimulus  Intensity Condition  SNR (dB) Hearing Aid  Unaided Phonak Siemens Starkey Gain setting   Linear Compression Linear Compression Linear Compression /m/ 65 dB SPL 34.1 16.7 -6.1 17.6 14.4 19.7 14.3 /t/ 35.8 34.0 -6.3 34.7 31.0 34.2 27.5 /s/ 28.5 38.5 9.6 33.0 31.0 35.9 30.6 /g/ 29.5 16.2 9.6 17.2 16.0 23.6 22.5 /m/ 45 dB SPL -3.0 -6.1 -13.9 -4.8 -1.8 -3.0 -0.9 /t/ -0.3 9.6 15.1 11.7 10.7 12.4 13.2 /s/ -5.9 17.2 15.7 12.4 11.7 15.3 15.1 /g/ -4.2 -7.5 -0.6 -3.9 -3.8 1.6 3.4 157  3.6.5.1 Interpretation of SNR effects on HEARLab™ Stimuli With the exception of /s/, SNRs in the 65 dB SPL condition were lower in all aided conditions than unaided conditions. In the 45 dB SPL condition only /m/ recorded from hearing aids with linear gain, and /g/ recorded from Phonak (linear) had poorer SNRs than the unaided condition; all other phonemes recorded in aided conditions had higher SNRs than the unaided phonemes.  Changes observed were similar for both polarities unless otherwise noted. As gain measured for /m/ was consistently negative, the lower measured SNRs were expected with a higher noise floor.  Unlike the results found with SCP Speech in section 3.6.4, setting the hearing aids with compression did not seem to make a large difference in the 45 dB SPL condition; that is, SNRs from stimuli processed by hearing aids with compression did not consistently provide higher SNRs. On the other hand, hearing aids with compression in the 65 dB SPL condition produced SNRs that were consistently lower than those measured from their counterparts recorded with the hearing aids with linear gain, with the largest difference seen from SNRs measured from recordings processed by Phonak (compression). The pattern seen with Phonak was more similar to the pattern observed in SCP Speech syllables, while the other hearing aids may not have treated the HEARLab™ stimuli in a similar way as they treated the SCP Speech syllables. Compression should have provided more gain to less intense inputs; in the 65 dB SPL condition, compression may have decreased the difference between the signal and the noise by increasing the noise floor and decreasing the signal intensity (Souza, 2002). Even though greater gain was measured in the 45 dB SPL condition with compression, similar SNRs were obtained with either gain setting.  Applying compression may not have made as large of a difference in the 45 dB SPL condition because the shorter duration of the stimulus allowed for less time for the 158  compressor to activate and stabilize before the stimulus ended (Souza, 2002).  Smaller differences were noted between gain settings in the 65 dB SPL condition for Starkey and Siemens, as compared to the differences measured for SCP Speech, possibly because of the different duration and nature of these stimuli. HEARLab™ stimuli were brief, and only contained the components of the consonant, while SCP Speech contained consonant-vowel syllables.    The effects of SNR changes to SCP morphology can depend on the rate of presentation of the stimulus as discussed in section 3.6.1 (Papesh et al., 2015). With a slower rate of presentation (e.g., ISIs of 1900 ms), N1 amplitudes are expected to be decreased with hearing aid processing except for with the phoneme /s/ in the 65 dB SPL condition, and increase with hearing aid processing in the 45 dB SPL condition for /t/, /s/, and /g/ presented for most aided recordings (Billings et al., 2009; Papesh et al., 2015). Although N1 amplitudes can be enhanced with lower SNRs between 10 dB and 30 dB SNR as compared to quiet at higher stimulation rates (e.g., ISIs of 900 ms), background noise was measured in unaided conditions; since N1 amplitudes did not differ significantly with 10 dB SNR compared to 30 dB SNR, N1 amplitudes are not expected to be enhanced (Papesh et al., 2015). P2 amplitudes are expected to decrease and increase in the same conditions as the N1 amplitudes with either rate of presentation (Billings et al., 2009; Papesh et al., 2015). There may not be any latency effects for /s/, not only because of its higher SNR, but also because of its higher frequency content; please note that less latency effects were observed with higher frequency emphasis when multi-talker babble was the noise signal in the study by Kuruvilla-Mathew et al (2015), while the noise measured in this study does not contain any speech-like stimuli. Latencies would be expected to be longer for lower SNRs as compared to unaided, 159  no matter the rate of presentation (Billings et al., 2009; Papesh et al., 2015). Similar to SCP Speech, it is expected that aided conditions have higher amplitudes and shorter latencies due to their higher measured SNRs for stimuli in the 45 dB SPL condition; however, whether the possible enhancements to N1 may be provided with the rate is unknown as results with SNRs lower than 10 dB were not measured (Billings et al., 2009; Papesh et al., 2015). 3.6.6 ABR Tone bursts and click  SNRs, calculated in dB, were measured for each stimulus in all hearing aid and intensity conditions. See Table 3.49 for measured SNRs of the tone bursts and rarefaction clicks recorded in all hearing aid and intensity conditions. Opposite polarity data are not shown.160  Table 3.49.49SNRs, dB, of 2-1-2 tone-bursts and rarefaction clicks recorded in all hearing aid and intensity conditions. Stimulus  Intensity Condition  SNR (dB) Hearing Aid  Unaided Phonak Siemens Starkey Gain setting   Linear Compression Linear Compression Linear Compression 500 Hz 85 ppe SPL 37.1 44.1 23.0 39.5 33.3 33.4 18.9 1000 Hz 35.8 36.0 21.3 40.7 34.5 34.3 19.0 2000 Hz 36.6 43.8 28.3 43.5 39.9 34.5 21.1 4000 Hz 39.4 51.4 30.6 55.0 55.1 43.9 27.4 Rarefaction click 25.5 30.4 28.6 36.5 36.3 28.9 23.1 500 Hz 65 ppe SPL 19.2 7.8 9.6 15.8 20.5 7.5 5.9 1000 Hz 17.8 6.1 6.5 16.3 18.7 8.2 4.5 2000 Hz 20.2 13.0 5.4 19.5 19.2 10.4 9.6 4000 Hz 20.3 19.8 16.0 29.0 32.6 18.6 14.0 Rarefaction click 10.9 2.6 5.1 10.9 11.1 6.6 6.3 161  3.6.6.1 Interpretation of SNR effects on ABR Tone bursts and click  Given the patterns observed with measured gain, the pattern of measured SNRs in the aided conditions were expected in the 85 ppe SPL condition and unexpected in the  65 ppe SPL condition. Changes observed were similar for both polarities unless otherwise noted. As measured gain from stimuli processed by hearing aids with compression were consistently lower than those processed by hearing aids with linear gain in the 85 ppe SPL condition, SNRs were expected to be lower in aided conditions with hearing aids set to compression, especially with the possible raising of the noise floor with compression. Contrarily, although measured gain was higher in aided conditions with hearing aids set to compression compared to their linear counterparts in the 65 ppe SPL condition, SNRs were lower in the recordings from Phonak (compression) and Starkey (compression). The overall pattern of SNRs within each hearing aid condition is consistent with the frequency response set for the hearing aids, as SNRs improved with tone bursts of higher frequencies, where more gain is expected.   Latencies of ABR clicks begin to increase at approximately 20 dB EM, and amplitudes begin to decrease at approximately 10 to 20 dB EM for a stimulus presented at 60 dB nHL (Owen & Burkard, 1991). Using a different measure of noise, Hecox et al. (1989) also showed that latency shifts were not seen between 0 dB and 10 dB SNR, but amplitudes of Wave V decreased over this range. It is expected that significant decreases in ABR amplitude would be observed from responses elicited by stimuli processed by Starkey (compression) and Phonak (compression) compared to unaided stimuli in both intensity conditions; no latency shifts are expected as there were no measured SNRs below 0 dB (Hecox, Patterson, & Birman, 1989). In the 65 ppe SPL condition, amplitudes were expected 162  to decrease with responses elicited by stimuli between 500 Hz to 1000 Hz processed by Starkey and Phonak, which have an approximately 10 dB decrease in SNR from unaided (Hecox et al., 1989). Amplitudes should increase with improving SNR, up to approximately 40 dB SNR (Hecox et al., 1989). However, whether 10 dB increases in SNR are significant for ABR latency, such as those measured in some recordings from Siemens set with either gain setting and Phonak (linear) in the 85 ppe SPL condition, will depend on whether the noise is in the 0 to 10 dB EM region. No decreases to latency are expected if noise is in the 10 to 0 dB EM region after hearing aid processing. Contrarily, if the noise is more intense than 10 dB EM, shorter latencies are expected with improved SNRs provided by these aided conditions as compared to unaided (Owen & Burkard, 1991). It is important to also take into account the processing delays of hearing aids for responses with such short latencies because processing delays ranged from 4 to 7 ms, as outlined in Table 3.1, which is in the region where some peaks of the ABR are expected.   3.6.7 MLR Tone bursts and click SNRs, calculated in dB, were measured for each stimulus in all hearing aid and intensity conditions. See Table 3.50 for measured SNRs of the tone bursts and rarefaction clicks recorded in all hearing aid conditions and intensity conditions.  Opposite polarity data are not shown.163  Table 3.50.50SNRs, dB, of 2-1-2 and 4-2-4 ms tone bursts as well as clicks recorded in all hearing aid and intensity conditions.      Stimulus  Intensity Condition  SNR (dB) Hearing Aid  Unaided Phonak Siemens Starkey Gain setting   Linear Compression Linear Compression Linear Compression 500 Hz (2-1-2) 85 ppe SPL 33.8 22.4 30.2 33.4 34.3 29.6 22.1 1000 Hz (2-1-2) 32.3 22.6 28.5 34.4 36.2 29.9 23.0 2000 Hz (2-1-2) 33.8 32.4 23.8 36.6 35.8 31.0 27.9 4000 Hz (2-1-2) 34.6 41.3 35.9 46.1 47.4 39.5 32.5 500 Hz (4-2-4 ms) 33.4 20.3 27.3 33.0 34.3 29.8 22.8 1000 Hz (4-2-4 ms) 35.2 27.0 28.4 38.0 36.7 39.6 23.6 2000 Hz (4-2-4 ms) 39.2 39.0 29.0 40.8 37.4 44.4 30.3 Rarefaction Click 27.9 33.7 32.3 38.0 38.7 34.1 31.6 500 Hz (2-1-2) 65 ppe SPL 16.8 -9.8 -14.0 6.2 12.7 4.6 4.1 1000 Hz (2-1-2) 15.9 -5.1 2.6 9.3 12.2 6.0 4.1 2000 Hz (2-1-2) 15.6 6.8 -1.4 12.0 10.6 7.9 10.2 4000 Hz (2-1-2) 18.2 14.7 12.9 20.5 22.6 15.5 15.0 500 Hz (4-2-4 ms) 17.4 -9.2 -14.3 6.0 13.0 5.5 4.2 1000 Hz (4-2-4 ms) 15.6 -1.4 5.0 13.3 15.3 8.2 5.8 2000 Hz (4-2-4 ms) 19.7 13.4 6.4 15.8 13.9 13.1 13.4 Rarefaction Click 11.8 9.4 8.6 13.4 13.4 11.5 10.2 164   3.6.7.1 Interpretation of SNR effects on MLR Tone bursts and click In the 85 ppe SPL condition, Starkey (compression) and Phonak (linear) had measured SNRs that were most different from unaided, with differences ranging from  -2.1 dB to -11.8 dB in recordings from Starkey and from -0.1 dB to -11.4 dB in recordings from Phonak. Other aided conditions had differences below 10 dB, with some stimuli, such as those with higher frequency content having more favourable SNRs as compared to unaided. This pattern is consistent with the frequency response set for each hearing aid. In the 65 ppe SPL condition, Starkey had larger differences than Siemens from unaided, no matter the gain setting, while setting the hearing aid with compression seemed to lessen the differences from unaided in Siemens. The odd intensity behaviour at 500 Hz and 1000 Hz observed in recordings from Phonak described in section 3.5.7.1 also affected the SNRs measured in recordings from Phonak, with negative measured SNRs. Changes observed were similar for both polarities unless otherwise noted. As evidenced, different hearing aids and different gain settings can lead to various results regarding measured SNRs.  As there are currently no systematic studies of the effects of SNR on MLR morphology, the significance of the differences noted in aided conditions from unaided conditions on the morphology of the MLR remains to be investigated.  Given the various effects of SNR on many different AEPs, it is expected that SNR will have some effect on MLR. However, the saturation levels, if any, or minimal changes to SNR that may affect the MLR morphology remain to be investigated.    165  3.6.8 MMN Speech SNRs, calculated in dB, were measured for each stimulus in all hearing aid and intensity conditions. See Table 3.51 for measured SNRs of /ba/ and /da/ used in an odd-ball paradigm, recorded in all hearing aid and intensity conditions.   166  Table 3.51.51SNRs, dB, of /ba/ and /da/ recorded in all hearing aid and intensity conditions. Stimulus  Intensity Condition  SNR (dB) Hearing Aid  Unaided Phonak Siemens Starkey Gain setting   Linear Compression Linear Compression Linear Compression /ba/ 65 dB SPL 12.4 12.0 10.1 13.6 14.8 14.2 9.0 /da/ 14.8 12.3 12.8 13.4 15.4 15.6 10.9 /ba/ 45 dB SPL -11.0 -12.6 -8.4 -9.0 -5.6 -7.3 -7.0 /da/ -7.0 -9.0 -7.4 -7.3 -3.3 -5.2 -4.5 167  3.6.8.1 Interpretation of SNR effects on MMN Speech  In the 65 dB SPL condition, all measured SNRs were positive while in the 45 dB SPL condition, all measured SNRs were negative. The differences among the syllables in the aided conditions were similar to those found in the unaided conditions.  Effects of SNR on the amplitudes measured in each hemisphere were previously found (Shytrov et al., 1996). It is expected that the stimuli recorded in the 45 dB SPL condition would elicit MMN with larger amplitudes in the right hemisphere and smaller amplitudes in the left hemisphere, while the stimuli recorded in the 65 dB SPL condition would elicit MMN with smaller amplitudes in the right hemisphere and larger amplitudes in the left hemisphere, but that hearing aid processing should have no additional effect (Shytrov et al., 1996).  3.6.9 ASSR 40 and 80 Hz AM Tones SNRs, calculated in dB, were measured for each stimulus in all hearing aid and intensity conditions. See Table 3.52 for measured SNRs of the 40 Hz and 80 Hz AM Tones recorded in all hearing aid and intensity conditions.  168  Table 3.52.52SNRs, dB, of 40 Hz and 80 Hz AM tones recorded in all hearing aid and intensity conditions.   Stimulus  Intensity Condition  SNR (dB) Hearing Aid  Unaided Phonak Siemens Starkey Gain setting   Linear Compression Linear Compression Linear Compression 40 Hz AM Tone 65 ppe SPL 33.9 34.4 19.0 45.3 38.8 27.4 17.4 80 Hz AM Tone 34.0 35.1 19.8 44.7 38.0 26.8 16.9 40 Hz AM Tone 45 dB SPL 2.7 4.3 3.5 16.8 24.6 8.0 7.6 80 Hz AM Tone 2.6 3.4 3.8 16.1 23.9 7.6 6.9 169  3.6.9.1 Interpretation of SNR effects on ASSR 40 and 80 Hz AM Tones  Hearing aid processing either increased or decreased SNR, depending on the hearing aid and gain setting in the 65 dB SPL condition. In the 65 dB SPL condition, differences larger than 10 dB between aided and unaided SNRs were noted in Starkey (compression) and Phonak (compression), as well as Siemens (linear). In the 45 dB SPL condition, aided conditions always provided greater SNRs, with Siemens having differences greater than  10 dB in both gain settings.    According to Galambos et al. (1992b), SNRs around 20 dB may actually see a slight but significant increase in amplitude, at least for the 40 Hz ASSR. Considering that the effects are more central to the auditory nerve, as suggested by Galambos et al. (1992b), the same may be expected of the 80 Hz ASSR. Thus, stimuli processed by Starkey (compression) and Phonak (compression) may elicit ASSRs of higher amplitudes than unaided stimuli, while other aided conditions may elicit similar amplitudes as those elicited by unaided stimuli. On the other hand, with lower SNRs, such as those measured in the 45 dB SPL condition, stimuli with improved SNRs of 5 dB or greater, such as those measured in recordings from Starkey and Siemens, will likely elicit ASSRs with higher amplitudes (Galambos et al., 1992b). Although phase was not measured in this study, higher SNRs are expected to lead to higher phase coherence while lower SNRs are expected to lead to lower phase coherence, with no enhancements with a low-level of noise (Galambos et al., 1992b). Please note that these are changes that may be observed if intensity was held constant throughout aided and unaided conditions. However, as discussed, ASSR AM tones intensities do change. Whether a change in intensity outweighs those noted in SNR, or vice versa, remains to be investigated. 170  3.7 Spectra  All spectra depicted are of the same polarity. Spectra of opposite polarity were essentially the same as the original polarity. See Appendix B for spectra of all stimuli, including the opposite polarity. 3.7.1 SCP Tones  Spectra were examined by performing FFTs on all stimuli for all hearing aid and intensity conditions. See Appendix B for FFTs of all stimuli; for examples of patterns observed, see Figures 3.24 to 3.31 for FFTs comparing 500 Hz and 1000 Hz tones of 30 ms and 480 ms duration with stimulus rise and fall times of 5 ms in each hearing aid condition and stimuli presented at 65 dB SPL. See Figures 3.32 and 3.33 for FFTs comparing 1000 Hz tones of 30 ms duration with stimulus rise and fall times of 5 ms in each hearing aid condition with stimuli presented at 45 dB SPL. See Figure 3.34 for an example of the changes to the spectrum expected with changing stimulus rise times demonstrated by 60 ms 1000 Hz tones with different rise times recorded from Starkey (linear). The same pattern of findings was seen with 2000 Hz tones, and in the 45 dB SPL intensity conditions, as discussed in section 3.7.1.1.  171    Figure 3.24.25FFT of 500 Hz tone bursts of 30 ms duration with stimulus rise and fall times of 5 ms, presented at 65 dB SPL. Different panels show different hearing aid conditions: (A) Unaided; (B) Starkey Linear; (C) Siemens Linear; (D) Phonak Linear. A  B  C  D  172                  A  B  C  D  Figure 3.25.26FFT of 500 Hz tone bursts of 30 ms duration with stimulus rise and fall times of 5 ms, presented at 65 dB SPL. Different panels show different hearing aid conditions: (A) Unaided; (B) Starkey Compression; (C) Siemens Compression; (D) Phonak Compression. 173                      A  B  C  D  Figure 3.26.27FFT of 500 Hz tone bursts of 480 ms duration with stimulus rise and fall times of 5 ms, presented at 65 dB SPL. Different panels show different hearing aid conditions: (A) Unaided; (B) Starkey Linear; (C) Siemens Linear; (D) Phonak Linear. 174                A  B  D  C  Figure 3.27.28FFT of 500 Hz tone bursts of 480 ms duration with stimulus rise and fall times of 5 ms, presented at 65 dB SPL. Different panels show different hearing aid conditions: (A) Unaided; (B) Starkey Compression; (C) Siemens Compression; (D) Phonak Compression. 175               A  B  C  D  Figure 3.28.29 FFT of 1000 Hz tone bursts of 30 ms duration with stimulus rise and fall times of 5 ms, presented at 65 dB SPL. Different panels show different hearing aid conditions: (A) Unaided; (B) Starkey Linear; (C) Siemens Linear; (D) Phonak Linear. 176   A  B  C  D  Figure 3.29.30FFT of 1000 Hz tone bursts of 30 ms duration with stimulus rise and fall times of 5 ms, presented at 65 dB SPL. Different panels show different hearing aid conditions: (A) Unaided; (B) Starkey Compression; (C) Siemens Compression; (D) Phonak Compression. 177   A  B  C  D  Figure 3.30.31FFT of 1000 Hz tone bursts of 480 ms duration with stimulus rise and fall times of 5 ms, presented at 65 dB SPL. Different panels show different hearing aid conditions: (A) Unaided; (B) Starkey Linear; (C) Siemens Linear; (D) Phonak Linear. 178                     Figure 3.31.32FFT of 1000 Hz tone bursts of 480 ms duration with stimulus rise and fall times of 5 ms, presented at 65 dB SPL. Different panels show different hearing aid conditions: (A) Unaided; (B) Starkey Compression; (C) Siemens Compression; (D) Phonak Compression. A  B  C  D  179                Figure 3.32.33FFT of 1000 Hz tone bursts of 30 ms duration with stimulus rise and fall times of 5 ms, presented at 45 dB SPL. Different panels show different hearing aid conditions: (A) Unaided; (B) Starkey Linear; (C) Siemens Linear; (D) Phonak Linear. A  B  C  D  180             A  B  C  D  Figure 3.33.34FFT of 1000 Hz tone bursts of 30 ms duration with stimulus rise and fall times of 5 ms, presented at 45 dB SPL. Different panels show different hearing aid conditions: (A) Unaided; (B) Starkey Compression; (C) Siemens Compression; (D) Phonak Compression. 181   Figure 3.34.35FFT of 1000 Hz tone bursts of 60 ms duration with 5 ms (A), 10 ms (B), and 20 ms (C) rise and fall time presented at 65 dB SPL recorded after being processed by Starkey (Linear). Notice the change in width of the largest peak with an increase in rise time.  A  B  C  182  3.7.1.1 Interpretation of spectral effects on SCP Tones  As shown in Figures 3.28 and 3.32 the fundamental peak was much harder to visualize in the 45 dB SPL condition than the 65 dB SPL condition, especially with stimuli of short durations. As duration, rise times, and frequencies increased, the amplitude spectra of the 45 dB SPL condition began to resemble those in for the 65 dB SPL condition, in that the fundamental peak was the largest peak. The fundamental peak was also observed to narrow with an increase in different acoustic parameters: rise time, frequency, and duration. Narrower fundamental peaks were visualized with increases in frequency due to the logarithmic scale of the x-axis of the FFT. Longer rise times and durations were expected to narrow the fundamental peak width because of less spectral splatter; the side lobes on either side are expected to decrease in intensity (Stapells & Picton, 1981; Vivion et al., 1980). This narrowing was observed in both aided and unaided conditions. With stimuli of longer durations, such as 480 ms, Siemens had slightly narrower fundamental peak widths as compared to Starkey and Phonak. This can potentially be attributed to the elongated rise times measured from Siemens; Siemens always had the longest measured rise times with both gain settings and the largest differences were noted with stimuli of the longest durations. With the frequency response of the hearing aid programmed for a sloping loss, it was expected that spectra would change in aided conditions. The higher frequency components  the noise were amplified in the aided conditions above the base line measured from unaided in both intensity conditions, although the width of the fundamental peak did not appear to be changed. The high frequency component seemed to increase even more in hearing aids with compression, observed in the spectra of stimuli from Starkey and Phonak in the 45 dB SPL 183  condition and all spectra from aided conditions with the hearing aid with compression in the 65 dB SPL condition. This was expected as compression was expected to provide greater amplification for lower intensities. In the 45 dB SPL condition, intensities of the fundamental peak are similar to the background noise; the higher frequency component of the noise become more similar in intensity to the fundamental peak with aided conditions, which is potentially due to the additional high frequency emphasis of the hearing aids (see Figures 3.32 and 3.33).   The hearing aid specifications suggest that the frequency range of the devices were between 100 Hz and approximately 6500 to 7000 Hz. As seen with the spectra of the  1000 Hz tone recordings, the low frequency energy in the noise was observed to be rolled-off in aided conditions. For example, in Figure 3.31, energy at approximately 20 Hz was higher in the unaided condition, peaking at levels above -60 dB volt, while all aided conditions had peaks below -80 dB volt at the same frequency. The high frequency roll-off differed between hearing aids, and although the reported range was the highest for Starkey, it was often observed to be the condition with the lowest frequency roll-off seen throughout Figures 3.24 to 3.33.   With the emphasis at higher frequencies, increases in amplitude and decreases in latency may be expected when the SCP is elicited by stimuli in the aided conditions as compared to unaided (Picton, 2011). The differences may be less noticeable for stimuli recorded in the 65 dB SPL condition than the 45 dB SPL condition. This is because the fundamental peak is noticeably higher than the noise with high frequency emphasis in the  65 dB SPL condition; contrarily, the intensities at the fundamental peak and the noise with high frequency emphasis are similar in the 45 dB SPL condition. Thus, the spread of 184  excitation caused by the additional high frequency emphasis may affect SCPs measured in the 45 dB SPL condition as the intensities at the higher frequencies were at similar intensities as the fundamental frequency. Please note that the measured spectra in this study were only examples of how spectra can change with hearing aid processing. The frequency response can change within the same hearing aid programmed for a different hearing loss, which can then lead to different results. However, it is important to note that even when hearing aids are programmed for the same hearing loss as they were in this study, differences were still noted among the hearing aids, such as the roll-off at the low and high extremes, the amount of low frequency content in the noise, and the number of peaks in the high frequency emphasis, to name several differences.  3.7.2 SCP Noise Spectra were examined by performing FFTs on all stimuli for all hearing aid and intensity conditions. See Appendix B for FFTs of all stimuli, but for examples of patterns observed, see Figures 3.35 to 3.38 for FFTs comparing noise bursts of 30 ms and 480 ms duration with generated rise and fall times of 5 ms in each hearing aid condition and stimuli presented at 65 dB SPL. See Figures 3.39 and 3.40 for FFTs comparing noise bursts of 30 ms duration with generated rise and fall times of 5 ms in each hearing aid condition with stimuli presented at 45 dB SPL.  See Figure 3.41 for FFTs comparing noise bursts of 60 ms duration with increasing generated rise and fall times recorded from Starkey (linear) with stimuli presented at 65 dB SPL. 185   Figure 3.35.36FFT of noise bursts of 30 ms duration with stimulus rise and fall times of 5 ms, presented at 65 dB SPL. Different panels show different hearing aid conditions: (A) Unaided; (B) Starkey Linear; (C) Siemens Linear; (D) Phonak Linear.  A  B  C  D  186   Figure 3.36.37FFT of noise bursts of 30 ms duration with stimulus rise and fall times of 5 ms, presented at 65 dB SPL. Different panels show different hearing aid conditions: (A) Unaided; (B) Starkey Compression; (C) Siemens Compression; (D) Phonak Compression. A  B  C  D  187   Figure 3.37.38FFT of noise bursts of 480 ms duration with stimulus rise and fall times of 5 ms, presented at 65 dB SPL. Different panels show different hearing aid conditions: (A) Unaided; (B) Starkey Linear; (C) Siemens Linear; (D) Phonak Linear. A  B  C  D  188   Figure 3.38.39FFT of noise bursts of 480 ms duration with stimulus rise and fall times of 5 ms, presented at 65 dB SPL. Different panels show different hearing aid conditions: (A) Unaided; (B) Starkey Compression; (C) Siemens Compression; (D) Phonak Compression.  A  B  C  D  189   A  B  C  D  Figure 3.39.40FFT of noise bursts of 30 ms duration with stimulus rise and fall times of 5 ms, presented at 45 dB SPL. Different panels show different hearing aid conditions: (A) Unaided; (B) Starkey Linear; (C) Siemens Linear; (D) Phonak Linear.  190   A  B  C  D  Figure 3.40.41FFT of noise bursts of 30 ms duration with stimulus rise and fall times of 5 ms, presented at 45 dB SPL. Different panels show different hearing aid conditions: (A) Unaided; (B) Starkey Compression; (C) Siemens Compression; (D) Phonak Compression. 191  A  B  C  Figure 3.41.42FFT of noise bursts of 60 ms duration with 5 ms (A), 10 ms (B), and 20 ms (C) rise and fall time presented at 65 dB SPL recorded after being processed by Starkey (linear).  192  3.7.2.1 Interpretation of spectral effects on SCP Noise  As expected from the frequency response of the hearing aids, more energy was measured for the high-frequency components of the noise in the aided conditions in both intensity conditions, a pattern also observed with SCP Tones. This would change with hearing aids programmed to a different hearing loss.   There was no change to the width of the peaks or the amount of energy in the higher frequency with rise time for the noise bursts, as illustrated by the similar spectra displayed in Figure 3.41, a noise burst of the same duration with different rise times. This was expected as the rise time was observed to only affect the width of the fundamental peak of SCP Tones and not the higher frequency components of noise. It has also been demonstrated that the noise bursts with different rise times do not differ in their spectrum (Hecox & Deegan, 1973).    The additional gain to frequencies above 2000 Hz in aided conditions could potentially lead to a decrease in N1 amplitude, although the change may not be significant (Antinoro & Skinner, 1968; Picton, 2011). It is important to note that this decrease in amplitude with shifts in frequency was recorded with changes in tonal stimuli (Antinoro & Skinner, 1968). In addition, the spectrum was broad in all conditions; as such, the latencies of the SCP peaks may not be expected to change with hearing aid processing (Picton, 2011).  3.7.3 ACC Tonal stimuli Spectra were examined by performing FFTs on all stimuli for all hearing aid and intensity conditions. Two separate peaks at 1680 Hz and 1705 Hz were observed in all conditions. See Appendix B for FFTs of all stimuli. For examples of FFTs obtained, see Figures 3.42 and 3.43 for FFTs comparing the ACC tonal stimuli between unaided and aided 193  conditions where hearing aids were set to linear gain in the 65 dB SPL and 45 dB SPL condition respectively.   194   Figure 3.42.43FFT of ACC tonal stimuli – a 1680 Hz tone transitioning to a 1705 Hz tone, each 300 ms long – presented at 65 dB SPL. Different panels show different hearing aid conditions: (A) Unaided; (B) Starkey Linear; (C) Siemens Linear; (D) Phonak Linear. A  B  C  D  195   Figure 3.43.44FFT of ACC tonal stimuli – a 1680 Hz tone transitioning to a 1705 Hz tone, each 300 ms long – presented at 45 dB SPL. Different panels show different hearing aid conditions: (A) Unaided; (B) Starkey Linear; (C) Siemens Linear; (D) Phonak Linear.  A  B  C  D  196  3.7.3.1 Interpretation of spectral effects on ACC Tonal Stimuli  Results from the spectra mirror the intensity difference measured between the two segments of the stimulus, as the comparison of the two intensities of the two peaks at  1680 Hz and 1705 Hz differ from unaided; there is a greater difference between the heights of the two peaks in the aided conditions as compared to the unaided conditions, showing the intensity increase for the 1705 Hz segment of the ACC stimulus.     As expected from the frequency response of the hearing aids, there was more high frequency emphasis in the aided than unaided conditions, with some roll off at extremes due to the limited frequency range of the receiver. That is, in the aided conditions, noise at the extreme low frequencies was flattened out while noise with high-frequency content was emphasized. As seen, the spectra obtained in both intensity conditions were also very similar; the only difference is the comparison of the height of the peaks compared to the height of the noise floor, with smaller differences in the 45 dB SPL condition, as expected from measured SNR data.   Although increased harmonics are expected to increase the amplitude and decrease the latency of the ACC, these changes should occur at the transition point between the  1680 Hz tone and the 1705 Hz tone (Jones & Perez, 2001). As the greater high frequency emphasis was present from the beginning in the aided conditions, with no additional change to the energy at the harmonics specifically at the transition point, the higher frequency emphasis in the aided conditions are not expected to affect the ACC morphology.  3.7.4 SCP Speech and ACC Speech Spectra were examined by performing FFTs on all stimuli in all hearing aid and intensity conditions. See Figure 3.44 for spectra of /a/ cut pre-recording, in unaided and 197  aided conditions where hearing aids were set to linear gain; see Figure 3.45 for spectra of /i/ in unaided and aided conditions where hearing aids were set to linear gain; see Figures 3.46 and 3.47 for a comparison of the spectra measured from /sa/ recorded in all hearing aid conditions with stimuli presented at 65 dB SPL.  See Figures 3.48 and 3.49 for a comparison of the spectra measured from /sa/ recorded in all hearing aid conditions with stimuli presented at 45 dB SPL.  See Figure 3.50 for a comparison of the /a/ spectra between /a/ cut pre-recording, and /a/ cut post-recording. See Appendix B for FFTs of all stimuli. 198   A  B  C  D  Figure 3.44.45FFT of /a/ cut pre-recording in unaided and aided conditions where hearing aids were set to linear gain, with stimuli presented at 65 dB SPL. Different panels show different hearing aid conditions: (A) Unaided; (B) Starkey Linear; (C) Siemens Linear; (D) Phonak Linear.  199               Figure 3.45.46FFT of /i/ in unaided and aided conditions where hearing aids were set to linear gain, with stimuli presented at 65 dB SPL. Different panels show different hearing aid conditions: A) Unaided; (B) Starkey Linear; (C) Siemens Linear; (D) Phonak Linear. A  B  C  D  200               Figure 3.46.47FFT of /sa/ presented at 65 dB SPL. Different panels show different hearing aid conditions: (A) Unaided; (B) Starkey Linear; (C) Siemens Linear; (D) Phonak Linear.  A  B  C  D  201               Figure 3.47.48FFT of /sa/ presented at 65 dB SPL. Different panels show different hearing aid conditions: (A) Unaided; (B) Starkey Compression;  (C) Siemens Compression; (D) Phonak Compression. A  B  C  D  202               A  B  C  D  Figure 3.48.49FFT of /da/ presented at 45 dB SPL. Different panels show different hearing aid conditions: (A) Unaided; (B) Starkey Linear;  (C) Siemens Linear; (D) Phonak Linear.  203                   Figure 3.49.50 FFT of /da/ presented at 45 dB SPL. Different panels show different hearing aid conditions: (A) Unaided; (B) Starkey Compression;  (C) Siemens Compression; (D) Phonak Compression. A  B  C  D  204    A  B  Figure 3.50.51FFT of /a/ cut pre-recording (A), and /a/ cut post-recording (B), recorded from Phonak (compression) with stimuli presented at 65 dB SPL condition. 205  3.7.4.1 Interpretation of spectral effects on SCP and ACC Speech  Comparing Figure 3.44 to the spectra in Figures 3.46 to 3.49, the spectra of /a/ was apparent in all of the syllables containing /a/, with the spectra of /sa/ and /da/ shown as examples. Energy at the frequency regions where the syllables are concentrated are observed, such as in the higher frequency region for /sa/ as observed in Figures 3.46 and 3.47, and in the mid regions for the /da/ as observed in Figures 3.48 and 3.49. Seen consistently throughout section 3.7, there was a high frequency emphasis due to the frequency response of the hearing aids. With syllables such as /sa/ and /∫a/, the emphasis on the high frequency component of noise was overshadowed by the energy of the syllables in the aided conditions, but it can be observed in syllables such as /da/ and /ta/, as well as in the /i/ illustrated in Figure 3.45. However, there was variability in the amount of high frequency emphasis that each hearing aid provided to the noise, and the height of the large peaks and valleys of the spectral envelope also changed based on the hearing aid. For example, the spectral envelope of Phonak and Siemens tended to have larger peaks and valleys than Starkey, especially in the high frequencies.  Generally, the patterns observed in the FFT of the stimuli recorded in the 65 dB SPL condition were very similar to those of the stimuli recorded in the 45 dB SPL condition; the obvious difference was the smaller difference between the height of the peaks of the noise floor and the peaks of the stimulus, as expected from the measured SNRs discussed in section 3.6. Comparing the spectrum of /a/ cut pre-recording and /a/ cut post-recording, one can also see that the spectrum did not change; therefore, in terms of spectral effects, the ACC paradigm and the SCP paradigm of eliciting the stimulus is not expected to result in a different morphology with hearing aid processing.  206   Agung et al. (2006) investigated the amplitude and latency differences of speech sounds of not only different frequency content, but also speech sounds of different durations. Even with stimuli that clearly differ in spectrum, measured latencies and amplitudes of the peaks in the SCP were not always significantly different across phonemes (Agung et al., 2006). For example, the measured amplitudes and latencies of N1 were not significant between the /∫/ and /s/ used in their study, but were significantly different from /m/, /u/, /a/, and /i/ N1 amplitudes and /m/ and /i/ N1 latencies for stimuli that were 500 ms in duration (Agung et al., 2006). Thus, significant changes to SCP morphology required large differences in spectrum before differences observed in morphology were significant, especially for latencies (Agung et al., 2006). Although the frequency response of the hearing aids did add greater high frequency emphasis in all aided conditions, it is not expected to cause N1 amplitudes to decrease significantly as the changes to the spectrum were not larger than those that led to significant differences in the study conducted by Agung et al. (2006).   3.7.5 HEARLab™ Stimuli Spectra were examined by performing FFTs on all stimuli for all hearing aid and intensity conditions.  See Figure 3.51 for spectra of all four phonemes in the unaided condition with stimuli presented at the 65 dB SPL condition. See Figures 3.52 to 3.57 for the spectra of all four phonemes in the aided conditions in the same intensity condition for comparison. See Appendix B for FFTs of all stimuli. 207                 A  B  C  D  Figure 3.51.52FFT of /m/ (A), /s/ (B), /g/ (C), /t/ (D) presented at 65 dB SPL and recorded in the unaided condition.   208                Figure 3.52.53 FFT of /m/ (A), /s/ (B), /g/ (C), /t/ (D) presented at 65 dB SPL and recorded from Starkey (Linear).  A  B  C  D  209               Figure 3.53.54FFT of /m/ (A), /s/ (B), /g/ (C), /t/ (D) presented at 65 dB SPL and recorded from Starkey (Compression). A  B  C  D  210               Figure 3.54.55 FFT of /m/ (A), /s/ (B), /g/ (C), /t/ (D) presented at 65 dB SPL and recorded from Siemens (Linear). A  B  C  D  211               Figure 3.55.56 FFT of /m/ (A), /s/ (B), /g/ (C), /t/ (D) presented at 65 dB SPL and recorded from Siemens (Compression). A  B  C  D  212               Figure 3.56.57 FFT of /m/ (A), /s/ (B), /g/ (C), /t/ (D) presented at 65 dB SPL and recorded from Phonak (Linear).  A  B  C  D  213   A  B  C  D  Figure 3.57.58 FFT of /m/ (A), /s/ (B), /g/ (C), /t/ (D) presented at 65 dB SPL and recorded from Phonak (Compression). 214  3.7.5.1 Interpretation of spectral effects on HEARLab™ Stimuli  As seen from the figures, the four HEARLab™ stimuli were distinguishable by spectrum before and after processing. This was expected as they all have differing frequency of emphasis (Chang et al., 2012). There was a greater high frequency emphasis in all stimuli recorded in the aided conditions, as expected from the frequency response set for all hearing aids. Although these figures only depicted the spectra obtained in the 65 dB SPL condition, the spectra of the stimuli recorded in the 45 dB SPL condition were essentially the same; the intensity of the spectral peaks as compared to the intensity of the noise floor was more similar in the 45 dB SPL condition, as expected with the SNRs measured and reported in section 3.6. While the same patterns were observed for all aided conditions, variability in the spectral envelopes was observed among hearing aids. Similar to the spectra observed for SCP and ACC Speech, the spectral envelope of Phonak and Siemens also tended to have larger peaks and valleys than Starkey, especially in the high frequencies. Thus, it was important to be aware of the variability introduced by the hearing aids.   Since the higher frequency emphasis was the main difference seen, significant differences in SCP morphology with hearing aid processing are unlikely based on spectral changes in isolation. Significant differences in amplitude and latencies require large changes in spectrum (Agung et al., 2006). With greater high frequency emphasis, aided conditions may only be expected to lessen the N1 amplitudes by non-significant but trending levels, at most, considering this frequency response. Hearing aids programmed with more irregular configurations and activated advanced features may lead to larger changes to spectra.    215  3.7.6 ABR Tone bursts and click  Spectra were examined by performing FFTs on all stimuli all hearing aid and intensity conditions.  See Figure 3.58 and 3.59 for spectra of the 4000 Hz 2-1-2 tone burst with stimuli presented at 85 ppe SPL in all hearing aid conditions, as the largest but also clearer differences are expected to be observed at this frequency with the frequency response of the hearing aids. See Appendix B for FFTs of all stimuli.216                 A  B  C  D  Figure 3.58.59 FFT of 4000 Hz 2-1-2 tone bursts presented at 85 ppe SPL. Different panels show different hearing aid conditions: (A) Unaided; (B) Starkey Linear; (C) Siemens Linear; (D) Phonak Linear.  217    Figure 3.59.60 FFT 4000 Hz 2-1-2 tone bursts presented at 85 ppe SPL. Different panels show different hearing aid conditions: (A) Unaided;  (B) Starkey Compression; (C) Siemens Compression; (D) Phonak Compression. A  B  C  D  218  3.7.6.1 Interpretation of spectral effects on ABR Tone bursts and click  Changes to rise time can change the spectral pattern of stimuli; with longer rise times, less spectral splatter is expected (Stapells & Picton, 1981). However, as rise times of stimuli remained stable across aided and unaided conditions in both intensity conditions, the peak width of the fundamental frequency was not expected to change (Stapells & Picton, 1981). The spectra measured in the 65 ppe SPL condition were also essentially the same as those measured from the stimuli recorded in the 85 ppe SPL. There was an emphasis in the higher frequency content of the noise floor due to the hearing aid frequency response. Therefore, hearing aid processing resulted in a slight widening of the spectrum with some spectral splatter towards the high frequencies. However, whether this high frequency splatter would reach threshold to increase synchronicity of firing will depend on the configuration of the hearing loss; if it does increase the synchronicity of firing, then the high frequency emphasis could potentially lead to decreased latencies and amplitudes of Wave V (Stapells & Picton, 1981).  3.7.7 MLR Tone bursts and click Spectra were examined by performing FFTs on all stimuli for all hearing aid and intensity conditions.  See Figure 3.60 and 3.61 for spectra of the 2000 Hz 2-1-2 tone burst and Figures 3.62 and 3.63 for spectra of the 2000 Hz 4-2-4 ms tone burst, with stimuli presented at 85 ppe SPL.  See Appendix B for FFTs of all stimuli.219                Figure 3.60.61 FFT of 2000 Hz 2-1-2 tone bursts presented at 85 ppe SPL Different panels show different hearing aid conditions: (A) Unaided; (B) Starkey Linear; (C) Siemens Linear; (D) Phonak Linear.  A  B  C  D  220               Figure 3.61.62FFT of 2000 Hz 2-1-2 tone bursts presented at 85 ppe SPL. Different panels show different hearing aid conditions: (A) Unaided; (B) Starkey Compression; (C) Siemens Compression; (D) Phonak Compression.  A  B  C  D  221                Figure 3.62.63FFT of 2000 Hz 4-2-4 ms tone bursts presented at 85 ppe SPL. Different panels show different hearing aid conditions:  (A) Unaided;  (B) Starkey Linear; (C) Siemens Linear; (D) Phonak Linear.  A  B  C  D  222     Figure 3.63.64 FFT of 2000 Hz 4-2-4 ms tone bursts presented at 85 ppe SPL. Different panels show different hearing aid conditions: (A) Unaided;  (B) Starkey Compression; (C) Siemens Compression; (D) Phonak Compression.  A  B  C  D  223  3.7.7.1 Interpretation of spectral effects on MLR Tone bursts and click Changes to rise time can change the spectral pattern of stimuli; with longer rise times, less spectral splatter is expected (Stapells & Picton, 1981). However, as rise times of stimuli remained stable across aided and unaided conditions in both intensity conditions, the peak width of the fundamental frequency was not expected to change (Stapells & Picton, 1981). The spectra measured in the 65 ppe SPL condition were also essentially the same as those measured from the stimuli recorded in the 85 ppe SPL. With longer stimulus rise times, the fundamental frequency peak in the spectrum was narrower, as expected (Stapells & Picton, 1981). High-frequency emphasis in the content of the noise floor due to the hearing aid frequency response was observed in all stimuli. Therefore, hearing aid processing resulted in a slight widening of the spectrum with some spectral splatter towards the higher frequencies. The spectral splatter may increase synchronicity of firing, and in so doing, increase MLR amplitudes and decrease latencies (Borgmann et al., 2001; Maurizi et al., 1984; Xu et al., 1995). The increase in synchronicity of firing may depend on the hearing loss of the individual and whether the spectral splatter would be processed by the auditory system of the individual.  3.7.8 MMN Speech Spectra were examined by performing FFTs on all stimuli for all hearing aid and intensity conditions.  See Figure 3.64 and 3.65 for spectra of /ba/ and /da/ to compare the spectra obtained in unaided conditions and aided conditions where hearing aids were set to compression to observe the greatest differences expected with aided conditions, as hearing aids set to linear gain have less high frequency emphasis. See Appendix B for FFTs of all stimuli. 224   Figure 3.64.65FFT of /ba/ (A) and /da/ (B) from the unaided condition, and /ba/ (C) and /da/ (D) from the Starkey (compression). All stimuli were presented at 65 dB SPL.   225                Figure 3.65.66 FFT of /ba/ (A) and /da/ (B) from Siemens (compression), and /ba/ (C) and /da/ (D) from Phonak (compression). All stimuli were presented at 65 dB SPL.   226  3.7.8.1 Interpretation of spectral effects on MMN Speech  There was a high frequency emphasis in the spectrum of processed stimuli above approximately 2500 Hz. The relative difference between the spectra of /ba/ and /da/ was essentially unaffected with hearing aid processing.    MMN amplitudes would be expected to increase with spectral complexity (Tervaniemi, Schröger, Saher, & Näätänen, 2000), but the additional high frequency emphasis would not be expected to add any additional spectral complexity to the already spectrally complex speech sounds used in this odd-ball paradigm. MMN amplitudes are not expected to change with these spectral changes caused by hearing aid processing (Näätänen et al., 1997).  3.7.9 ASSR 40 and 80 Hz AM Tones Spectra were examined by performing FFTs on all stimuli for all hearing aid and intensity conditions.  See Figures 3.66 and 3.67 for spectra of the 80 Hz AM tone with stimuli presented at 65 dB SPL; greater peak separation from the central carrier frequencies are expected with modulations at higher frequencies in the stimulus all hearing aid conditions, allowing for easier visualization of the side peaks. See Appendix B for FFTs of all stimuli. 227   Figure 3.66.67 FFT of 80 Hz AM tones presented simultaneously at 65 dB SPL. Different panels show different hearing aid conditions: (A) Unaided;  (B) Starkey Linear; (C) Siemens Linear; (D) Phonak Linear.  A  B  C  D  228    Figure 3.67.68 FFT of 80 Hz AM tones presented simultaneously at 65 dB SPL. Different panels show different hearing aid conditions: (A) Unaided;  (B) Starkey Compression; (C) Siemens Compression; (D) Phonak Compression. A  B  C  D  229  3.7.9.1 Interpretation of spectral effects on ASSR 40 and 80 Hz AM Tones As observed throughout section 3.7, high frequency emphasis was present in the spectra of the 40 and 80 Hz AM tones. Due to the hearing aid frequency range, we also see a roll off of low frequency noise in the aided conditions. As there were energy peaks noted in the low frequencies where the modulation frequencies were situated, F-ratios to compare the variance of those peaks with the variance of the noise floor was calculated and discussed in section 3.12. The expected peaks at the central frequencies at 500 Hz, 1000 Hz, 2000 Hz and 4000 Hz, of the tones presented simultaneously with their sidebands, at 40 Hz and 80 Hz above and below each central frequency was observed for the 40 Hz and 80 Hz AM tone, respectively. (Dmitrijevic et al., 2002). As no unexpected central frequencies with sidebands were seen in the high frequency component of the noise, it is unlikely that the high frequency emphasis would affect the amplitude of the ASSR responses.  3.8 Difference between Formant 2 and Formant 1 frequencies 3.8.1 SCP Speech and ACC Speech - /a/ and /i/  Formant 1 and Formant 2 frequencies were measured for all vowels in all hearing aid and intensity conditions. The difference was subsequently calculated by subtracting the frequency of Formant 1 from the frequency of Formant 2. See Figure 3.68 for the differences in all hearing aid and intensity conditions. Figures for opposite polarity data are not shown.230   Figure 3.68.69Differences between Formant 1 and Formant 2 in each vowel recorded in all hearing aid and intensity conditions.  .02505001500175020002250/a//i//a/ cut pre-recording/a/ cut post-recording/a//i//a/ cut pre-recording/a/ cut post-recording/a//i//a/ cut pre-recording/a/ cut post-recording/a//i//a/ cut pre-recording/a/ cut post-recording/a//i//a/ cut pre-recording/a/ cut post-recording/a//i//a/ cut pre-recording/a/ cut post-recording/a//i//a/ cut pre-recording/a/ cut post-recordingLinear Compression Linear Compression Linear CompressionUNAIDED STARKEY SIEMENS PHONAKF2 -F1 difference  (Hz)Hearing Aid Condition65 dB SPL condition 45 dB SPL condition231  3.8.1.1 Interpretation of the effects of the difference between Formant 2 and Formant 1 on SCP and ACC vowels   Although hearing aid processing was expected to change the spectral emphasis, it was not expected to cause large changes to measured formant frequencies. Thus, the difference between the first and second formant, hereafter referred to as the F2-F1 difference, was expected to remain stable after hearing aid processing (Dillon, 2012).  Changes observed were similar for both polarities unless otherwise noted. The largest difference was found for Phonak (compression) in the 45 dB SPL condition, with a 215 Hz difference from unaided found for /i/.  Several studies have looked at the F2-F1 differences and found that larger differences led to larger SCPs, as noted in section 1.4 (Agung et al., 2006; Johnson, 2011; Makela, Aku & Tiitinen, 2003; Obleser, Eulitz, & Lahiri, 2004). While the systematic study of the F2-F1 difference did not look at SCPs directly, 500 Hz step sizes were used, and the pattern of change did not change substantially until between 2500 Hz to 3000 Hz (Ohl & Schleich, 1997). Additionally, Obleser et al. (2004), found no significant differences in latency, and a small significant difference in amplitude between /e/ and /i/, which differ in their F2-F1 difference by 570 Hz. Considering that the hearing aid processing produced a maximum of 215 Hz difference from unaided and /i/ peaked below 2500 Hz, it is not likely that hearing aid processing would cause changes to the F2-F1 difference large enough to elicit a morphologically different SCP.  3.8.2 MMN Speech Formant 1 and Formant 2 frequencies were measured for /a/ in both /ba/ and /da/ in all hearing aid and intensity conditions. The difference was subsequently calculated by 232  subtracting the frequency of Formant 1 from the frequency of Formant 2. See Table 3.53 for F2-F1 differences measured in all hearing aid and intensity conditions.233  Table 3.53.53Differences between Formant 1 and Formant 2 in /a/ in /ba/ and /da/ recorded in all hearing aid and intensity conditions. Stimulus  Intensity Condition  F2-F1 difference (Hz) Hearing Aid  Unaided Phonak Siemens Starkey Gain setting   Linear Compression Linear Compression Linear Compression /a/ in /ba/ 65 ppe SPL 287 316 332 278 321 295 353 /a/ in /da/ 502 598 555 487 537 531 550 /a/ in /ba/ 45 dB SPL 293 269 288 265 289 292 328 /a/ in /da/ 548 612 483 554 556 559 563 234  3.8.2.1 Interpretation of the effects of F2-F1 difference on MMN Speech  Similar to the variability in F2-F1 differences measured in the vowels in section 3.8.1, there were differences between F2-F1 differences measured in aided conditions compared to unaided. However, also similar to the differences measured in the vowels in section 3.8.1, the differences are unlikely to cause significant changes to the MMN alone, as differences ranged from 1 Hz to 96 Hz. Latency changes are unlikely to occur because a difference of 1116 Hz in the F2-F1 difference, found between a /bi/ and a /bu/, did not lead to significant changes to the latency, so any difference smaller than that are unlikely to change the latency of the MMN (Korczak & Stapells, 2010). A small but significant change to MMN amplitude, however, was found with an 1116 Hz difference in the F2-F1 difference, with a larger difference eliciting an MMN that was 1.78 µV smaller in amplitude (Korczak & Stapells, 2010). Considering that the largest difference noted was 96 Hz, and the vowel presented was within the same category, it would be predicted that a significant change to the amplitude would not occur, although this remains to be investigated.  3.9 VOT 3.9.1 SCP Speech - /da/ and /ta/  VOTs were measured in Praat for the consonant-vowel syllables containing a stop consonant recorded in all hearing aid and intensity conditions. See Table 3.54 for VOTs in all hearing aid and intensity conditions. Opposite polarity data are not shown.   235   Table 3.54.54VOTs measured for all consonant-vowel syllables containing a stop consonant for all hearing aid and intensity conditions Stimulus  Intensity Condition  VOT (ms) Hearing Aid  Unaided Phonak Siemens Starkey Gain setting   Linear Compression Linear Compression Linear Compression /da/ 65 ppe SPL 16.4 16.7 18.8 14.7 14.8 16.8 15.0 /ta/ 103.5 104.1 104.2 104.9 104.0 104.5 103.7 /da/ 45 dB SPL 14.8 14.0 15.6 13.3 14.0 15.4 13.4 /ta/ 103.9 103.6 115.5 104.9 105.3 104.7 104.3 236  3.9.1.1 Interpretation of VOT effects on SCP stop consonants From the duration data shown in section 3.4, duration is one of the acoustic parameters that stays relatively stable with hearing aid processing. Thus, we expected measured VOTs to be stable with hearing aid processing. Changes observed for both polarities were similar unless otherwise noted. Although larger differences were found in /ta/ in the 45 dB SPL condition, they are not expected to cause any significant changes to the latency or amplitude of the resulting ACC. VOTs of /ta/ processed by hearing aids are still within the 100 ms to 150 ms range where latencies and amplitudes are expected to stay stable (Ganapathy et al., 2013). The VOT of /da/ was measured between 14.8 ms to 16.4 ms in the unaided conditions in both intensity conditions; this syllable would not be expected to elicit an ACC as the VOTs are below 30 ms (Sharma & Dorman, 1999).  As the range of differences measured from unaided were within 2.4 ms, it is also unlikely that the changes in VOT within the same syllable would elicit an SCP with significant differences in morphology (Sharma & Dorman, 1999).  3.9.2 MMN Speech - /ba/ and /da/  VOTs were measured for the syllables used in the odd-ball paradigm to elicit the MMN. See Table 3.55 for VOTs in all hearing aid and intensity conditions. 237  Table 3.55.55VOTs measured for all syllables used in the odd-ball paradigm for all hearing aid and intensity conditions. Stimulus  Intensity Condition  VOT (ms) Hearing Aid  Unaided Phonak Siemens Starkey Gain setting   Linear Compression Linear Compression Linear Compression /ba/ 65 ppe SPL 6.9 7.5 6.8 7.3 8.2 6.7 6.9 /da/ 14.5 14.0 20.0 11.4 12.2 13.9 14.7 /ba/ 45 dB SPL 7.5 7.1 9.3 7.7 9.1 6.7 8.7 /da/ 13.4 15.1 14.3 14.2 13.4 15.5 17.5 238  3.9.2.1 Interpretation of VOT effects on MMN stop consonants   Measured VOTs of the two syllables in the unaided condition differed by 7.6 ms in the 65 dB SPL condition and 5.9 ms in the 45 dB SPL condition. The shorter VOT measured from the 45 dB SPL condition may be partly due to higher noise floor that can lead to measurement error. The largest difference from unaided was a 5.6 ms increase in VOT. A difference of 5.6 ms in VOT between the two syllables is unlikely to cause any difference in the MMN amplitude or latency; this is because within category changes of even 20 ms do not elicit a significant change in the MMN (Sharma & Dorman, 1999). As these two syllables remain in separate phoneme categories with hearing aid processing it is unlikely that changes with hearing aid processing will change the MMN morphology in a significant manner (Sharma & Dorman, 1999; Wood, 1976).  These two syllables also do not only differ by VOT, but also place of articulation, where they differ in their spectral pattern of their release burst, with bilabials, such as /ba/, having a falling pattern, while alveolars, such as /da/, have a flat to rising pattern (Kent, 1992). Previous studies have shown that the place of articulation of stop consonants can have a difference in the MMN latency (Korczak & Stapells, 2010). As the difference in place of articulation is kept, as observed with the spectral patterns of the release bursts of the two syllables, small VOT changes with hearing aid processing are unlikely to lead to unaccounted changes to the MMN morphology. 3.10 Fricative Duration 3.10.1 SCP Speech - /sa/ and /ʃa/  Fricative durations were measured in Praat for the fricative consonants in all hearing aid and intensity conditions. See Table 3.56 for measured fricative durations for all fricatives in all hearing aid and intensity conditions. Opposite polarity data are not shown. 239   Table 3.56.56Fricative durations measured for all consonant-vowel syllables containing a fricative for all hearing aid and intensity conditions  Stimulus  Intensity Condition  Fricative Duration (ms) Hearing Aid  Unaided Phonak Siemens Starkey Gain setting   Linear Compression Linear Compression Linear Compression /sa/ 65 ppe SPL 248.7 253.4 249.7 249.2 249.7 249.7 247.1 /ʃa/ 228.3 226.9 226.5 227.0 226.7 227.9 226.9 /sa/ 45 dB SPL 255.4 233.8 237.2 237.6 246.5 243.5 241.9 /ʃa/ 201.3 188.4 210.6 199.8 209.4 211.5 223.7 240  3.10.1.1 Interpretation of fricative duration effects on SCP fricatives   Fricative durations were stable with hearing aid processing, as expected, across gain settings and hearing aids. This also confirms the duration and vowel onset information found by Tremblay, Billings, et al. (2006): duration differences between /si/ and /∫i/ were preserved with hearing aid processing. The differences measured in the 45 dB SPL condition in the aided conditions as compared to the unaided condition were slightly larger. Changes observed were similar for both polarities unless otherwise noted. As latencies of N1 and P2 are stable with pre-transition stimulus of 120 ms or greater, differences between measured fricative durations of 248.7 ms to 255.4 ms for the syllable /sa/ and 228.3 ms for both polarities of the syllable /∫a/ are not expected to cause any morphological changes to the ACC, even with differences measured in the 45 dB SPL condition (Ganapathy et al., 2013).  Thus, we do not expect processed fricative durations to affect the latency or the amplitude of the ACC (Ganapathy et al., 2013).   3.11 Modulation Depth  3.11.1 ASSR 40 and 80 Hz AM Tones  Modulation depths were measured and calculated from the envelopes of the AM modulated tones in all hearing aid and intensity conditions. See Table 3.57 for modulation depths of 40 Hz and 80 Hz AM tones in all hearing aid and intensity conditions.  241  Table 3.57.57Modulation depths measured and calculated from the envelopes of the AM modulated tones in all hearing aid and intensity conditions.Stimulus  Intensity Condition  Modulation depth (%) Hearing Aid  Unaided Phonak Siemens Starkey Gain setting   Linear Compression Linear Compression Linear Compression 40 Hz AM Tone 65 ppe SPL 81.4 78.7 72.0 89.6 86.6 75.7 76.9 80 Hz AM Tone 66.1 68.6 71.3 81.8 78.0 61.9 73.2 40 Hz AM Tone 45 dB SPL 58.7 68.3 66.3 87.8 93.1 74.7 79.2 80 Hz AM Tone 54.5 65.1 65.6 81.9 85.6 70.2 73.5 242  3.11.1.1 Interpretation of the effects of modulation depth on ASSR AM Tones    From Table 1.1, it has been shown that larger modulation depths led to higher ASSR amplitudes. In the 65 dB SPL condition, Siemens with either gain setting and 80 Hz AM tones processed by Starkey (compression) and Phonak with either gain setting yielded larger modulation depths than unaided; all other hearing aid conditions resulted in shallower modulation depths in this intensity condition. In the 45 dB SPL condition, all aided conditions yielded deeper modulation depths than unaided for 40 and 80 Hz stimuli. These findings may largely be influenced by the SNR, as the pattern of findings seem to follow the measured SNR findings; that is, hearing aid conditions that had lower SNRs also had shallower modulation depths. A higher noise floor would decrease the modulation depth by introducing a higher minimum amplitude value (see Equation 1 in section 2.5.2).   Picton et al. (1987) showed an increase in amplitude with increases of depth of modulation, but the rate of increase slowed down at modulation depths of greater than 50%, at least with a single carrier frequency presented at 70 dB HL. Despite the slowing of the rate of increase, the measured amplitudes were significantly different at least with steps of 20% from 10% to 90%. Dmitrijevic et al. (2001) also showed a significant difference between amplitudes of a response to a stimulus with a modulation depth of 100% as compared to a modulation depth of 50%.  As measured modulation depths were all greater than 50%, even in the 45 dB SPL condition, the variations, most of which were under 20%, with the exception of recordings from Siemens in the 45 dB SPL condition, are unlikely to cause any large changes in amplitude. In the 65 dB SPL condition, modulation depths were measured to differ by 16% or less from unaided; thus, changes to modulation depth by hearing aid processing are not expected to affect the ASSR amplitude significantly. In the 45 dB SPL 243  condition, modulation depths increased from unaided by 27% to 34% for Siemens. These changes could lead to small but significant increases to ASSR amplitude (Picton et al. (1987).  3.12 F-ratio 3.12.1 ASSR 40 and 80 Hz AM Tones F-ratios were calculated at the modulation rates to examine whether there was a significant difference between the variance of the noise floor and the variance at the modulation rates. See Table 3.58 for measured F-ratios for all AM tones in all hearing aid and intensity conditions.  244  Central Modulation Frequency 40 Hz 80 Hz Intensity Condition 65 dB SPL 45 dB SPL 65 dB SPL 45 dB SPL Unaided  1.25 0.97 0.87 1.00 Starkey Linear 1.09 0.96 0.81 1.01  Compression 1.14 1.13 0.98 1.04 Siemens Linear 1.00 0.91 1.19 0.97  Compression 1.16 0.90 1.15 1.03 Phonak Linear 1.31 0.99 0.86 0.98  Compression 0.95 1.03 1.14 0.94 Table 3.58.58F-ratios measured for all AM tones in all hearing aid and intensity conditions. 245  3.12.1.1 Interpretation of the effects on the F-ratio measured in ASSR AM Tones  Measured F-ratio values were all approximately one, with measured  F-ratios ranging from 0.86 to 1.31. The F-ratio critical value was 2.99 for an α level of 0.05 for the degrees of freedom calculated for each AM tone; thus, none of the measured F-ratios were significant (Gardner, 2001; John & Picton, 2000). Although the non-linear components of hearing aid processing were predicted to lead to artifactual responses at the modulation frequencies, based on findings that non-linearity in bone-conduction transducers and high intensities can cause artifactual responses (e.g., Small & Stapells, 2004), the data from this study suggested otherwise.               246  Chapter 4: General Discussion and Conclusions  As evidenced in Chapter 3, hearing aid processing can change several acoustic parameters of the input signal to the auditory system simultaneously, some dependent on the type of stimuli presented, and many others dependent on the hearing aid. That is, most aspects of the stimulus other than the intensity may remain the same for one hearing aid in one gain setting, as tones do when presented to Starkey hearing aids in linear settings, while the same stimulus presented to a Siemens hearing aid with the same gain setting yields a stimulus with an elongated rise time but an unchanged fall time. Some hearing aids may also handle stimuli in a more stable manner for some stimulus types as compared to others. For example, in Siemens, although various changes occur when processing stimuli of longer duration, such as in the SCP Noise or SCP Tones set of stimuli, the hearing aid seems to handle brief stimuli, such as the tone bursts from ABR stimuli, in a more stable manner as less onset effects were seen when compared to other hearing aid conditions or from the unaided condition. Therefore, assumptions that the chosen stimulus is the input signal cannot be made without recording what actual changes occur with the particular hearing aid in question. Stimuli presented in sound field to the hearing aids are not the same sounds presented to the ear by the hearing aid after processing.  Other acoustic parameters may be expected to remain stable: including the stability of effects across opposite polarities, as well as the stability of parameters of speech that do not depend heavily on onset effects, such as F2-F1 differences and VOTs, as illustrated in chapter 3. Thus, the details and diversity of the results obtained from this study illustrate two important points:  (i) based on data from previous studies investigating the effects of stimulus change on AEP morphology, hearing aid processing can lead to significant differences to 247  AEP morphology due to changes to acoustic parameters that may be unaccounted for, such as larger and variable changes noted in rise times or signal-to-noise ratios, and (ii) the knowledge of the input to the auditory system can help to correctly interpret the morphology of the auditory evoked potential.   While some acoustic parameters, discussed as a measured change on its own, are not expected to change the AEP morphology by a significant value, such as with unintended intensity changes with ACC tonal stimuli after hearing aid processing discussed in section 3.4.3, it is important to be aware that the changes to the acoustic parameters do not occur in isolation. Many systematic changes in AEP morphology produce trends of change before reaching significance, as seen in the results of previous research discussed in Table 1.1. Thus, simultaneous changes to different acoustic parameters, such as those noted in Chapter 3, may combine to change AEP morphology significantly even though an isolated change to one acoustic parameter may not reach significance. The results in this study were interpreted as isolated changes in each section for clarity. However, systematic investigations with these combined changes to human AEPs would need to be done in order to test this hypothesis. It also important to be aware that some changes to acoustic parameters can affect AEP morphology differently depending on stimulus presentation parameters. For example, this was illustrated by Papesh et al. (2015) in the case of changes to SNR. They demonstrated that decreases in SNR affected N1 peak amplitudes differently depending on the presentation rate, as discussed in section 3.5 (Papesh et al., 2015). Thus, all parameters must be taken into account for a correct and comprehensive interpretation of the AEP morphology.  Throughout chapter 3, we also observed that some groups of stimuli, such as brief and speech or speech-like stimuli, tended to be more resilient to changes with hearing aid 248  processing compared to other stimulus types examined in this study, such as longer tones and noise bursts. For brief stimuli, it is likely that the duration of the brief stimulus is too short for the hearing aid to be able to apply larger changes before the next stimulus begins. For speech or speech-like stimuli, the resilience may be due to hearing aids being programmed to optimize processing of speech signals, leading to the more consistent and stable results after processing with various hearing aids. The clinical question should always be considered if clinicians choose stimuli based on their resilience to change with differences in processing. For example, although ABR stimuli tend to remain stable with hearing aid processing, it is not useful for determining whether the client is able to tell a difference between phonemes of interest; it is, however, a useful AEP to determine whether the hearing aid has provided an input that can be recognized and used by the auditory system, at least at the brainstem level.  As another example, if the effects of compression were to be examined, then brief stimuli may also not be the most suitable as each stimulus may be too short for compression to be applied cleanly to each stimulus.  It is important to note that the results from this study do not evaluate the quality of one hearing aid over another. The data cannot be used to draw conclusions on which will provide the best results for speech intelligibility, be translated to how much benefit and real-world success can be provided, or be used for recommendations regarding choosing one aid over the other. However, the results can be used for clinicians to gain better understanding of acoustic changes to input stimuli that can happen with hearing aid processing.  One of the major limitations of this study is the setting of the hearing aids. Although various changes to measured acoustic parameters were observed with two different gain settings, all advanced features were deactivated to prevent confounding factors in this study. 249  However, such advanced features could possibly lead to greater changes in the acoustic parameters measured. For example, if frequency lowering was activated, this feature would be expected to cause larger changes to occur with the spectrum than those illustrated in this study. Moreover, many of these advanced features are often activated in many current hearing aid fittings, such as feedback cancellation, directional microphones, and noise reduction. Thus, the changes seen in the inventory in this study are unlikely to be the largest changes expected with hearing aid processing. In addition, the hearing aid was programmed for a common hearing loss, but other changes may be seen with the hearing aids programmed to another hearing loss. This study is by no means a full inventory of possible changes. However, it does provide some data as well as ideas of which acoustic parameters and types of stimuli are more vulnerable to unexpected changes with hearing aid processing as compared to others.  Future directions of this study can include the use of the recorded stimuli from this study to collect the AEPs in question; comparisons of AEPs collected from unaided and aided stimuli can be used to investigate the hypotheses noted in Chapter 3. These studies can provide a better understanding of the effects that these changes to the acoustic parameters can have on human auditory evoked potentials; this knowledge can equip clinicians with more knowledge on how to interpret aided AEPs to form evidence-based clinical decisions. If future studies replicating this study with other hearing aids continue to show the diverse effects hearing aid processing can have on these AEP stimuli, it also highlights the importance for clinicians to have specifics about the different acoustic parameters of the input stimulus in real time. Other future directions may include the development of clinical tools to record the input to the auditory stimulus in the real ear to measure the relevant 250  acoustic parameters in real time so that such changes can be taken into account during the interpretation of the aided waveforms. Further systematic studies of the effects of changes to acoustic parameters that have been affected by hearing aid processing on AEP morphology, such as the rise time effects for MMN stimuli, can also be performed.   In conclusion, this study demonstrated that various changes to acoustic parameters can occur with hearing aid processing, with additional variability introduced by differences in hearing aids as well as gain settings. Hearing aid processing with some hearing aids can cause unaccounted changes to acoustic parameters that can alter AEP morphology significantly. Thus, assumptions of input stimulus morphology after hearing aid processing cannot be made as changes to the stimulus are not the same across hearing aids.  For diagnostic assessments, conclusions are made after analyzing and correlating all the results obtained from the test battery. Similarly, clinical decisions should only be made from aided evoked potentials after a comprehensive understanding of various factors that can affect the evoked potentials, including both subject and stimulus factors.            251  References Agung-King, K., Campbell, J., Sharma, A., Martin, K., Dorman, M., & Langran, J. (2008). The representation of voice onset time in the cortical auditory evoked potentials of young children. Clinical Neurophysiology, 119, 2855-2861.  Agung, K., Purdy, S., McMahon, C.M, & Newall, P. (2006). The use of cortical auditory Evoked potentials to evaluate neural encoding of speech sounds in adults. Journal of American Academy of Audiology, 17, 559-572.  Alain, C., Woods, D., & Coarrubias, D. (1997). Acitivation of duration-sensitive auditory cortical fields in humans. Electroencephalography and Clinical Neurophysiology, 104, 531-539.  ANSI (1999). “Maximum Permissible Ambient Noise Levels for Audiometric Test Rooms”, American National Standards Institute, S3.1 1999, New York, NY. Antinoro, F., & Skinner, P. (1968). The effects of frequency on the auditory evoked response. Journal of Auditory Research, 8, 119-123.  Appleby, S. (1964). The slow vertex maximal sound evoked response in infants. Acta OtoLaryngologica (supplement), 206, 146-152. Audacity Team (2011). Audacity®, Version 1.3.14-beta. Audio editor and recorder.  Audacity Team (2014).  Audacity®, Version 2.0.6. Audio editor and recorder. Available from: http://audacityteam.org/ (Accessed September 30, 2014).  Audioscan. Verifit Online User Manual. Dorchester, Ontario: Audioscan. Available at: www.audioscan.com. Accessed December 12, 2014. 252  Ballachanda, B.B., Moushegian, G., & Stillman, R.D. (1992). Adaptation of the Auditory Brainstem Response: Effects of Click Intensity, Polarity, and Position. Journal of American Academy of Audiology, 3, 275-282.  Beattie, R.C., and Torre, P. (1997). Effects of rise-fall time and repetition rate on the auditory response to 0.5 and 1 kHz tone bursts using normal-hearin gand hearing-impaired subjects. Scandinavian audiology, 26(1), 23-32.  Bellier, L., Veuillet, E., Vesson J., Bouchet, P., Caclin, A., & Thai-Van H. (2015). Speech Auditory Brainstem Response through hearing aid stimulation. Hearing Research, 325(2015), 49-54.  Beranek, L.L. (1992). Music, Acoustics, and Architecture. Bulletin of the American Academy of Arts and Sciences, 45(8), 25-46.  British Columbia Early Hearing Program (BCEHP) (2012). BCEHP Audiology Assessment Protocol. Retrieved January 30, 2014 from http://www.phsa.ca/AgenciesAndServices/Services/BCEarlyHearing/ForProfessionals/Resources/Protocols-Standards.htm. Billings, C.J., Papesh, M.A., Penman, T.M., Batzell, L.S., & Gallun, F.J. (2012). Clinical use of aided cortical auditory evoked potentials as a measure of physiological detection or physiological discrimination. International Journal of Otolaryngology, 2012, 1-14.  Billings, C.J., Tremblay, K.L., & Souza, P.E., & Binns M.A. (2007). Effects of hearing aid amplification and stimulus intensity on cortical auditory evoked potentials. Audiology & neuro-otology, 12, 234–246. 253  Billings, C.J., Tremblay, K.L. & Miller, C.W. (2011). Aided cortical auditory evoked potentials in response to changes in hearing aid gain. International Journal of Audiology, 50, 459-467.  Billings, C.J., Tremblay, K.L., Stecker, G.C., & Tolin, W.M. (2009). Human evoked cortical activity to signal-to-noise ratio and absolute signal level. Hearing Research, 254, 15-24. Bisgaard, N., Vlaming, M.S.M.G., & Dahlquist, M. (2010). Standard Audiograms for the IEC 60118-15 Measurement Procedure. Trends in Amplification, 14, 113-120. Boersma, P. & Weenink, D. (2013). Praat: doing phonetics by computer [Computer program]. Version 5.3.56, retrieved 23 January 2015 from http://www.praat.org/ Borgmann, C., Roß, B., Dragnaova, R., & Pantev, C. (2001). Human auditory middle latency responses: influence of stimulus type and intensity. Hearing Research, 158, 57-64. Campbell, K.B., Picton, T.W., Wolfe, R.G., Maru, J., Baribeau-Braun, J., & Braun, C. (1981). Auditory potentials. Sensus, 1, 21-31.  Campbell, V.A, Crews, J.E., Moriarty, D.G., Zack, M.M, & Blackman, D.K. (1999). Surveillance for sensory impairment, activity limitation, health-related quality of life among older adults – United States, 1993-1997. MMWR CDC, Surveillance Summaries, 48 131-156.  Carter, L., Golding, M., Dillon, & H., Seymour, J. (2010). The Detection of Infant Cortical Auditory Evoked Potentials (CAEPs) Using Statistical and Visual Detection Techniques. Journal of the American Academy of Audiology, 21, 347-356.  Chang, H., Dillon, H., Carter, L., van Dun, B., & Young, S. (2012). The relationship between cortical auditory evoked potential (CAEP) detection and estimated audibility in 254  infants with sensorineural hearing loss. International Journal of Audiology, 51, 663-670. Chasin, M., & Russo, F. (2004). Hearing aids and music. Trends in Hearing, 8, 35-47.  Chen, D.S., Betz, J., Yaffe, K., Ayonayon, H.N., Kritchevsky, S, Martin, K.R.,…Lin, F.R. (2014). Association of hearing impairment with declines in physical functioning and the risk of disability in older adults. Journals of Gerontology: Medical Sciences, 2014, 1-8. doi:10.1093/Gerona/glu207  Chia, E., Jie Jin, W., Rochtchina, E., Cumming, R.R., Newall, P., & Mithcell P. (2006). Hearing Impairment and Health-Related Quality of Life: The Blue Mountains Study. Ear & Hearing, 28, 187-195.  Couch, L.W. (1990). Digital and Analogue Communication Systems. New York: Macmillan.  Davids, T., Valero, J., Papsin, B.C., Harrison, R.V., & Gordon, K.A. (2008). Effect of increasing duration of stimulation on the electrically evoked auditory brainstem and middle latency responses in pediatric cochlear implant users. Journal of Otolaryngology – Head & Neck Surgery, 37, 559-564.  Davis, H., & Zerlin, S. (1966). Acoustic relations of the human vertex potential. Journal of the Acoustical Society of America, 39(1), 109-116. Davies, F.W., Mantzaridis, H., Kenny, G.N.C., & Fisher, A.C (1996). Middle latency auditory evoked potentials during repeated transitions from consciousness to unconsciousness. Anaesthesia, 51, 107-113.  Davis, H., Mast, T., Yoshi, N., & Zerlin, S. (1966). The slow response of the human cortex to auditory stimuli: recovery process. Electroencephalography and Clinical Neurophysiology, 21, 105-113.  255  Dillon, H. (2012). Hearing Aids. Turramura, Australia: Boomerang Press.  Dillon, H., and Walker, G. (1982). Comparison of stimuli used in sound field audiometric testing. Journal of Acoustical Society of America, 71(1), 161-172. Dimitrijevic A, John, M.S., van Roon, P, Picton, T.W. (2001). Human auditory steady-state responses to tones independently modulated in both frequency and amplitude. Ear & Hearing, 22, 100–111. Dmitrijevic, A., John, M., van Roon, P., Purcel, D., Adamonis, J., Ostroff, J., et al. (2002). Estimating the audiogram using multiple auditory steady-state responses. Journal of the American Academy of Audiology, 13, 205-224.  Dimitrijevic, A., Michalewski, H.J., Zeng, F.G., Pratt, H., & Starr, A. (2008). Frequency changes in a continuous tone: auditory cortical potentials. Clinical Neurophysiology, 119, 2111-2124.  Dubno, J.R., & Schaefer, A.B. (1992). Comparison of frequency selectivity and consonant recognition among hearing-impaired and masked normal-hearing listeners. Journal of the Acoustical Society of America, 91, 2110-2121.  Durante, A.S., Wieselberg, M.B., Carvalho, S., Costa, N., Pucci, B. Gudayol, N. & Almeida, K.D. (2014). Cortical Auditory Evoked Potential: evaluation of speech detection in adult hearing aid users. CoDAS, 26, 367 – 373.  Easwar, V., Glista, D., Purcell, D.W., & Scollie. D. (2012). The effect of stimulus choice on cortical evoked potentials (CAEP): Consideration of speech segment positioning within naturally produced speech.  International Journal of Audiology, 51, 926-931 Easwar, V., Purcell, D.W., & Scollie, D. (2006). Electroacoustic Comparison of Hearing Aid Output of Phonemes in Running Speech versus Isolation: Implications for Aided 256  Cortical Auditory Evoked Potentials Testing. International Journal of Otolaryngology, 2012, 1-10.  Folsom, R.C., & Aurich, C.D. (1987). Auditory brainstem responses from human adults and infants: influence of stimulus onset. Audiology, 26, 117-122.   Fortune, T.W., Woodruff, B.D., & Preves, D.A. (1994). A new technique for quantifying temporal envelope contrasts. Ear & Hearing, 15, 93-99.  Frenzel Jr, L. E. (2008). Amplitude Modulation Fundamentals. In T. Casson & J. Plant. (Eds.), Principles of Electronic Communication Systems, 3E (pp. 93-117). New York, NY: McGraw-Hill.  Frye Electronics. (2014). HEARLab™ System Operator’s Manual. Retrieved from February 8, 2015, from http://www.frye.com/wp/wp-content/uploads/2013b/manuals/hearlab.pdf.  Frye Electronics, Inc. (2010). Fonix 7000 Hearing Aid Test System Operator’s Manual. Retrieved from March 16, 2015, from http://www.frye.com/wp/wp-content/uploads/2013b/manuals/7000.pdf.  Funasaka, S. & Ito, S. (1986). Stimulus duration and waves of auditory brainstem response. Audiology, 25, 176-183.  Gage, N. & Roberts, T. (2000). Temporal integration: Reflection in the M100 of the auditory evoked field. Neuroreport, 11, 2723-2726.  Gardner, R.C. (2001). Appendix. Tables-Sampling Distributions. In L. Pearson, J. Heffler, A. Westlake & M. Rottino (Eds.), Psychological Statistics Using SPSS for Window (pp. 291). Upper Saddle River, NJ: Prentice-Hall, Inc.  257  Galambos, R. & Makeig, S. (1992a). Physiological studies of central masking in man. I: The effects of noise on the 40-Hz steady-state response. Journal of Acoustical Society of America, 92, 2683–2690.   Galambos, R. & Makeig, S. (1992b). Physiological studies of central masking in man. II: Tonepip SSRs and the masking level difference. Journal of Acoustical Society of America, 92, 2691–2697. Ganapathy, M.K., Narne, V.J., Kalaiah, M.K., & Manjula, P. (2013). Effect of pre-transition stimulus duration on acoustic change complex. International Journal of Audiology, 52, 350-359.  Gates, G.A., Cobb, J.L., D’Agostino, R.B., & Wolf, P.A. (1993). The relation of hearing in the elderly to the presence of cardiovascular disease and cardiovascular risk factors. Archives of Otolaryngology – Head & Neck Surgery, 119, 156-161.  Golding, M., Dillon, H., Seymour, J., & Carter, .L. (2009). The detection of adult cortical auditory evoked potentials (CAEPs) using an automated statistic and visual detection. International Journal of Audiology, 48, 833-842 Golding, M., Pearce, W. , Seymour, J. , Cooper, A., Ching, T., & Dillon, H. (2007). The relationship between obligatory cortical auditory evoked potentials (CAEPs) and functional measures in young infants. Journal of American Academy of Audiology, 18, 117– 125. Gorga, M.P., & Thornton, A.R. (1989). The choice of stimuli for ABR measurements. Ear & Hearing, 10(4), 217-230.  Hall, J., & Ruth, R. (1985). Acoustic Reflexes and Auditory Evoked Responses in Hearing Aid Evaluation. Seminars in Hearing, 6, 251-277.  258  Han, J.H., & Dimitrijevic. (2015). Acoustic change responses to amplitude modulation: a method to quantify cortical temporal processing and hemispheric asymmetry. Frontiers in Neuroscience, 9, 1-15. doi: 10.3389/fnins.2015.00038 Harrison Bush, A.L., Lister J.J., Lin, F.R., Betz, J., & Edwards, J.D. (2015). Peripheral hearing and cognition: evidence from the Staying Keen in Later Life (SKILL) study. Ear & Hearing. doi:10.1097/AUD.0000000000000142 Hari, R., Kaila, K., Katila, T., Tuomisto, T., & Varpula, T. (1982). Interstimulus interval dependence of the auditory vertex response and its magnetic counterpart: Implications for their neural generation. Electroenecephalography and Clinical Neurophysiology, 54,561-569.  Hawkins, D.B., & Naidoo, S.V. (1993). Comparison of sound quality and clarity with asymmetrical peak clipping and output limiting compression. Journal of the American Academy of Audiology, 4, 221-228.  He, S., Grose, J.H., & Buchman, C.A. (2012). Auditory discrimination: the relationship between psychophysical and electrophysiological measures. International Journal of Audiology, 51, 771-782.  Hecox, K.. (1983). Role of auditory brainstem response in the selection of hearing aids. Ear & Hearing, 4, 51 – 55. Hecox, K., Patterson, J., & Birman, M. (1989). Effect of broadband noise on the human brain stem auditory evoked response. Ear & Hearing, 10(6), 346-353.  Hecox, K., Squires, N., & Galambos, R. (1976). Brainstem auditory evoked responses in man. I. Effect of stimulus rise-fall time and duration. Journal of Acoustical Society of America, 60, 1187-1192.  259  Horváth, J., Czigler, I., Jacobsen, T., Maess, B., Schroger, E., Winkler, I. (2008). MMN or no MMN: no magnitude of deviance effect on the MMN amplitude. Psychophysiology, 45, 60–69.  Hyde, M.L. (1994). The slow vertex potential: Properties and clinical applications. In J. T. Jacobson (ed.), Principles and Applications in Auditory Evoked Potentials. Boston: Allyn and Bacon, pp. 179 -218. Hyde, M. (1997). The N1 response and its applications. Audiology & Neuro-otology, 2, 281-307.  International Electrotechnial Commission (IEC). (1994). Audiometers. Auditory test signals of short duration for audiometric and neuro-otological purposes. IEC 645-3. Geneva.  Jenstad, L.M. & Lister, A. Revised Envelope Difference Index (rEDI) [Computer software].  Jenstad, L.M., Marynewich, S., & Stapells, D.R. (2012). Slow cortical potentials and amplification – Part II: Acoustic Measures. International Journal of Otolaryngology, 2012, 1-14. Jenstad, L.M., Pumford, J., Seewald, R.C., & Cornelisse, L.E. (2000). Comparison of linear gain and wide dynamic range compression hearing aid circuits II: aided loudness measures. Ear & Hearing, 21, 32-44.  Jenstad, L.M., and Souza, P. (2005). Quantifying the effect of compression hearing aid release time on speech acoustics and intelligibility. Journal of Speech, Language and Hearing Research, 48, 651-667.  John, M.S., Lins, Oo. Boucher, B. & Picton, T.W. (1998). Multiple auditory steady-state responses (MASTER): Stimulus and recording parameters. Audiology, 37, 59-82.  John, M.S., Dimitrijevic, A., and Picton, T.W. (2002). Auditory Steady-State Responses to  260  Exponential Modulation Envelopes. Ear & Hearing, 23, 106-117.  John, M.S. & Picton, T.W. (2000). MASTER: a Windows program for recording multiple auditory steady-state responses. Computer Methods and Programs in Biomedicine, 61(2), 125-150. Johnson, K. (2012). Acoustic & Auditory Phonetics. West Sussex, United Kingdom: John Wiley & Sons.  Jones, S.J., & Perez, N. (2001). The auditory ‘C-process’: analyzing the spectral envelope of complex sounds. Clinical Neurophysiology, 112, 965-975.  Joutsiniemi, S, Hari, R., & Vilkman, V. (1989). Cerebral magnetic responses to noise bursts and pauses of different durations. Audiology, 28, 325-333.  Kent, R.D., & Read C. (1992). The acoustical analysis of speech. San Diego, CA: Singular Publishing Group, Inc.  Kileny, P.R. (1991). Use of electrophysiologic measures in the management of children cochlear implants: brainstem, middle latency, and cognitive (P300) responses. American Journal of Audiology, 12, 37-47 Kileny P.R . (1982). Auditory brainstem responses as indications of hearing aid performance. Annals of Otolaryngology, 91, 61 – 64.  Kileny, P.R., Paccioretti, P.R., & Wilson, A.F. (1987). Effects of cortical lesions on middle-latency auditory evoked responses (MLR). Electroencephalography and Clinical Neurophysiology, 66(2), 108-120.  Killion, M.C., & Revit, L.J. (1993). CORFIG and GIFROC: Real ear to coupler and back. In G. Studebaker & I. Hochberg (Eds.), Acoustical Factors Affecting Hearing Aid Performance, 2nd ed. (pp. 65–85). Boston: Allyn and Bacon.   261  Kodera, K., Hink, R., Yamada, O., & Suzuki, J. (1979). Effects of rise time on simultaneously recorded auditory-evoked potentials from the early, middle, and late ranges, Audiology, 18, 395-402.  Kohn, M. Lifeshitz, K, & Litchfield, D. (1978). Averaged evoked potentials and frequency modulation. Electroencephalography and Clinical Neurophysiology, 45, 236-243.  Kohn, M. Lifeshitz, K, & Litchfield, D. (1980). Averaged evoked potentials and amplitude modulation. Electroencephalography and Clinical Neurophysiology, 50, 134-140.  Korczak, P.A., Kurtzberg, D., & Stapells, D.R. (2005). Effects of sensorineural hearing loss and personal hearing aids on cortical event-related potential and behavioural measures of speech-sound processing. Ear & Hearing, 26, 165-185. Korczak, P.A., & Stapells, D.R. (2010). Effects of various articulatory features of speech on cortical event-related potentials and behavioural measures of speech-sound processing. Ear & Hearing, 31(4), 491-504.  Kraus, N, Smith, D.I., & Mcgee, T. (1987). Rate and filter effects on the developing middle-latency responses. Audiology, 26, 257-268.  Kurnaz, M., Satar, B., & Yetiser, S. (2009). Evaluation of cochlear implant users’ performance using middle and late latency responses. European archives of oto-rhino-laryngology, 266, 343-350.  Kuruvilla-Mathew, A., Purdy, S.C., & Welch, D. (2015). Cortical encoding of speech acoustics: Effects of noise and amplification. International Journal of Audiology. Advance online publication. Retrieved from http://informahealthcare.com.ezproxy.library.ubc.ca/doi/abs/10.3109/14992027.2015.1055838 262  Levitt H., & Resnick, S.B. (1978). Speech reception by the hearing-impaired: Methods of testing and the development of new tests. Scandanavian Audiology. Supplementum. 6, 105-130. Lin, F.R., Yaffe, K., Xia, J., Xue, Q. Harris, T.B., Purhcase-Helzner, E. …Simonsick, E.M. (2013). Hearing loss and cognitive decline in older adults. JAMA Internal Medicine, 173, 293-299.  Macrae, J., & Dillon, H. (1996). An equivalent input noise level criterion for hearing aids. Journal of Rehabilitation Research and Development, 33(4), 355-362.  Mahoney T.M .(1985). Auditory brainstem response hearing aid application. In: Jacobson (ed.), The Auditory Brainstem Response. San Diego: College Hill Press, pp. 349 – 370. Makela, A.M., Aku, P., & Tiitinen, H. (2003). The auditory N1m reveals the left-hemispheric representation of vowel identity in humans. Neuroscience Letters, 353(2), 111-114.  Martin, B.A., & Boothroyd, A. (1999). Cortical, auditory, event-related potentials in response to periodic and aperiodic stimuli with the same spectral envelope. Ear & Hearing, 20(1), 33-44.  Martin, B.A., & Boothroyd, A. (2000). Cortical, auditory, evoked potentials in response to changes of spectrum and amplitude. Journal of Acoustical Society of America, 107, 2155-2161. Martin, B.A., Boothroyd, A., Ali, S., & Leach-Berth, T. (2010). Stimulus presentation strategies for eliciting the acoustic change complex: increasing efficiency. Ear & Hearing, 31, 356-366.  263  Martin, B.A., Sigal, A., Kurtzberg, D., & Stapells, D.R. (1997). The effects of decreased audibility produced by high-pass noise masking on cortical event-related potentials to speech sounds /ba/ and /da/. Journal of the Acoustical Society of America, 101, 1585- 1599. Martin, B.A., Tremblay, K.L. & Korczak, P. (2008). Speech evoked potentials: from the laboratory to the clinic. Ear & Hearing, 29, 285-313.  Marynewich, S., Jenstad, L.M., & Stapells, D.R. (2012). Slow cortical potentials and amplification – Part I: N1-P2 Measures. International Journal of Otolaryngology, 2012, 1-11. Maurizi, M., Ottaviani, F., Paludetti, G., et al. (1984). Middle latency auditory components in response to clicks and low- and middle-frequency tone pips (0 .5-1 kHz). Audiology, 23, 569-580. McGee, T., Kraus, N., & Manfredi, C. (1988). Toward a strategy for analyzing the auditory middle-latency response waveform. Audiology, 27, 119-130.  Mick, P., Kawachi, I, & Lin, F.R. (2014). The association between hearing loss and social isolation in older adults. Otolaryngol Head Neck Surg, 150, 378-384.  Miller, S., & Zhang, Y. (2014). Neural coding of phonemic fricative contrast with and without hearing aid. Ear & Hearing, 35, 122-133.  Näätänen, R., & Picton, T. (1987). The N1 wave of the human electric and magnetic response to sound: a review and an analysis of the component structure Psychophysiology, 24, 375-425.  264  Näätänen, R. Lehtokoski, A., Lennes, M., Cheour, M., Huotilainen, M., Iivonen, Antti, … Alho, K. (1997). Language-specific phoneme representations revealed bye electric and magnetic brain responses. Nature, 385, 432-434. Nachtegaal, J., Smit, J.H., Smits, C., Bezemer, P.D., van Beek, J.H., Festen, J.M., & Kramer, S.E. (2009). The association between hearing status and psychosocial health before the age of 70 years; results from an internet-based national survey on hearing. Ear & Hearing, 30, 302–312.   Oates, P.A. & Purdy, S.C. (2001). Frequency specificity of the human auditory brainstem and middle latency responses using notched noise masking. Journal of Acoustical Society of America, 110, 995-1009.  Obleser J, Eulitz C, Lahiri A. (2004). Magnetic brain response mirrors extraction of phonological features from spoken vowels. Journal of Cognitive Neuroscience, 16, 31–39. Ohl, F.W., & Scheich, H. (1997). Orderly cortical representation of vowels based on formant interaction. Proceedings of the national Academy of Sciences of the United States of America, 94(17), 9440-9444.  Onishi, S. & Davis, H. (1968). Effects of duration and rise time of tone bursts on evoked V potentials. Journal of the Acoustical Society of America, 44, 582-591.  Ostroff, J., Martin, B. A., & Boothroyd, A. (1998). Cortical evoked responses to spectral change with a syllable. Ear & Hearing, 19, 290‐297. Owen, G.A., & Burkard, R. (1991). Ipsilateral, contralateral, and binaural masking effects on the human brain-stem auditory-evoked responses to click stimuli. Journal of the Acoustical Society of America, 89, 1760-1769.   265  Özdamar, Ö, & Kraus, N. (1983). Auditory middle-latency responses in humans. Audiology, 22, 34-49.  Paavilainen, P., Jiang, D., Lavikainen, J., Näätänen, R.. (1993). Stimulus duration and the sensory memory trace: an event-related potential study. Biological Psychology, 35, 139 – 152.  Papesh, M.A., Billings, C.J., & Baltzell, L.S. (2015). Background noise can enhance cortical auditory evoked potentials under certain conditions. Clinical Neurophysiology, 126(7), 1319-1330.  Picton, T. (2011). Human Auditory Evoked Potentials. B.A. Stach (Ed.). San Diego, CA: Plural Publishing, Inc. Picton, T.W., Dmitrijevic, A., & John, M.S. (2002). Multiple auditory steady-state responses.  Annals of otology rhinology and laryngology, 111(5), 16-21. Picton, T.W., Dimitrijevic, A., Perez-Abalo, M.C., & van Roon, P. (2005). Estimating auditory steady state responses. Journal of the American Academy of Audiology, 16, 143-156.  Picton, T.W., Durieux-Smith, A. Champagnee, S.C., Whittingham, J., Moran, L.M., Giguère, C., & Beauregard, Y. (1998). Objective evaluation of aided thresholds using auditory steady-state responses. Journal of the American Academy of Audiology, 9 315-331.  Picton, T.W., Skinner, C.R., Champagne, S.C., Kellett, A.J.C., & Maiste, A.C. (1987). Potentials evoked by the sinusoidal modulation of the amplitude or frequency of a tone. The Journal of the Acoustical Society of America, 82(1), 165-175.  266  Purdy, S.C. & Abbas, P.J. (2002). ABR thresholds to tonebursts gated with Blackman and linear windows in adult with high frequency sensorineural hearing loss. Ear & Hearing, 23, 358-368.  Rance G., Cone-Wesson B., Wunderlich J. & Dowell R (2002). Speech perception and cortical event related potentials in children with auditory neuropathy. Ear & Hearing, 23, 239 – 253.  Rinne, T., Särkka, A., Degerman, A., Schröger, E. & Alho, K. (2006). Two separate mechanisms underlie auditiory change detection and involuntary control of attention. Brain Resesarch, 1077(1), 135-143. Rodriguez, R., Picton, T.W., Linden, D., Hamel, G., & Laframboise, G. (1986). Human auditory steady state responses: effects of intensity and frequency. Ear & Hearing, 7, 300–313. Schochat, E., Musiek, F.E., Alonso, R., & Ogata, J. (2010). Effect of auditory training on the middle latency response in children with (central) auditory processing disorder. Brazilian Journal of Medical and Biological Research, 43, 777-785.  Schwartz, D.M., & Larson, V.D. (1977). A comparison of three hearing aid evaluation procedures for young children. Archives of Otolaryngology – Head & Neck Surgery, 103(7), 401-406.  Scollie, S., Seewald, R. Cornelisse, L., Moodie, S. Bagatto, M. Laurnagaray, D…Pumford, J. (2005). The desired sensation level multistage input/output algorithm. Trends in Amplification, 9(4), 159-197.   267  Seewald, R.C, Hudson, S.P., Gagné, J.P., & Zelisko, D.L. (1992). Comparison of two methods for estimating the sensation level of amplified speech. Ear & Hearing, 13(3), 142-149.  Selesnick, I.W., & Burrus, C.S. (1998). Generalized digital Butterworth filter design. IEEE Transactions on Signal Processing, 46(6), 1688-1694.   Sharma, A., Cardon, G., Henion, K. & Roland, P. (2011). Cortical maturation and behavioral outcomes in children with auditory neuropathy spectrum disorder. International Journal of Audiology, 50, 98 – 106 . Sharma, A., & Dorman, M.F. (1999). Cortical auditory evoked potential correlates of categorical perception of voice-onset time. Journal of Acoustic Society of America, 106, 1078-1083.  Sharma, A. Marsh, C.M., & Dorman, M.F. (2000). Relationship between N1 evoked potential morphology and the perception of voicing. Journal of the Acoustical Society of America, 108, 3030-3035 Sharma, M., Purdy, S.C., & Bonnici, L. (2003). Behavioural and electroacoustic calibration of click and toneburst auditory brainstem response stimuli. The Australian and New Zealand Journal of Audiology, 25, 54-60. Sharma, A., Martin, K., Roland, P. , Bauer, P. , Sweeney, M.H, Gilley, P., & Dorman, M. (2005). P1 latency as a biomarker for central auditory development in children with hearing impairment. Journal of American Academy of Audiology, 16, 564 – 573.  Shemesh, R., Attias, J., Magdoub, H., Nageris B.I. (2012). Prediction of aided and unaided audiograms using sound-field auditory steady-state evoked responses. International Journal of Audiology, 51, 746-53.  268  Shytrov, Y., Kujala, T., Ahveninen, J., Tervaniemi, M., Alku, P., Ilmoniemi, R.J., & Näätänen, R. (1996). Background acoustic noise and the hemispheric lateralization of speech processing in the human brain: magnetic mismatch negativity study. Neuroscience Letters, 251, 141-144.  Small, S.A. & Stapells, D.R. (2004). Artifactual responses when recording Auditory Steady-State Responses. Ear & Hearing, 25(6), 611-623.   Small, S.A. & Stapells, D.R. (2008). Maturation of bone-conduction multiple auditory steady-state responses. International Journal of Audiology, 47:  476-488 Souza, P. (2002). Effects of compression on speech acoustics, intelligibility, and sound quality. Trends in amplification, 6(4), 131-165.  Souza, P. and Tremblay, K. (2006). New perspectives on assessing amplification effects. Trends in amplification, 10, 119-143.   Stapells, D.R. (2009). Cortical Event-Related Potentials to Auditory Stimuli. In Katz, J., Medwetsky, L., Burkard, R. & Hood, L. (Eds.) Handbook of Clinical Audiology, 6th Edition, Baltimore: Lippingcott, Williams & Wilkins. pp. 395-430. Stapells, D.R. (2000). Frequency-specific evoked potential audiometry in infants. In R. C. Seewald (Ed.): A Sound Foundation Through Early Amplification. Basel: Phonak AG. pp. 13-31. Stapells, D.R., & Picton, T.W. (1981). Technical aspects of brainstem evoked potential audiometry using tones. Ear & Hearing, 2(1), 20-29.    Stapells, D.R., Picton, T.W., & Smith, A. (1982). Normal hearing thresholds for clicks. Journal of Acoustical Society of America, 72(1), 74-79.  269  Stone, M.A., & Moore, B.C. J. (2003). Tolerable hearing aid delays. III. Effects on speech production and perception of across-frequency variation in delay. Ear & Hearing, 24, 175-183.  Stone, M.A., Moore, B.C.J., Alcántra, J.I., & Glasberg, B.R. (1999). Comparison of different forms of compression using wearable digital hearing aids. Journal of Acoustical Society of America, 106(6), 3603 - 3619.  Stone, M.A., Moore, B.C.J., Meisenbacher, K., & Derleth, R.P. (2008). Tolerable hearing aid delays. V. Estimation of limits for open canal fittings. Ear & Hearing, 28(4), 601-617.  Strawbridge, W.J., Walhagen, M.I., Shema, S.J., & Kaplan, G.A. (2000). Negative Consequences of Hearing Impairment in Old Age. The Gerontologist, 40, 320-326.  Stroebel, D., Swanepoel, D., & Groenewald, E. (2007). Aided auditory steady-state responses in infants: Respuestas auditivas de estado estable en niños con auxiliares auditivos. International journal of audiology, 46(6), 287-292. Tan, M.T., Chang, J.S., Chua, H.C., & Gwee, B.H. (2003). An investigation into the parameters affecting total harmonic distortion in low-voltage low-power Class-D amplifiers. IEEE Transactions on Circuits and Systems – I: Fundamental Theory and Applications, 50(10), 1304-1315.  Tervaniemi, M., Schröger, Saher, M., & Näätänen. (2000). Effects of spectral complexity and sound duration on automatic complex-sound pitch processing in humans – a mismatch negativity study. Neuroscience Letters, 290 (1), 66-70.  The MathWorks, Inc. (2013). MATLAB and Statistics Toolbox Release 2013b. Natick, MA.  [Computer software]. 270  The MathWorks, Inc. (2015a). Fall time of negative-going bilevel waveform transitions – MATLAB falltime. Retrieved July 16, 2016, from http://www.mathworks.com/help/signal/ref/falltime.html The MathWorks, Inc.(2015b).  Convert magnitude to decibels (dB) – MATLAB mag2db.  Retrieved July 16, 2016, from http://www.mathworks.com/help/control/ref/mag2db.html  The MathWorks, Inc. (2015c). Rise time of positive-going bilevel waveform transitions – MATLAB risetime. Retrieved July 16, 2016, from http://www.mathworks.com/help/signal/ref/risetime.html The MathWorks, Inc. (2015d). Root-mean-square level – MATLAB rms. Retrieved  July 16, 2016, from http://www.mathworks.com/help/signal/ref/rms.html   The MathWorks, Inc. (2015e). Total harmonic distortion – MATLAB thd. Retrieved April 2, 2015, from http://www.mathworks.com/brandguide/editorial/products.html  Tremblay K.L., Billings C.J., Friesen L.M., & Souza P.E. (2006). Neural representation of amplified speech-sounds. Ear & Hearing, 27, 93-103. Tremblay, K.L., Kalstein, L., Billings, C., & Souza, P.E. (2006). The neural representation of consonant-vowel transitions in adults who wear hearing aids. Trends in Amplification, 10, 155-162.  Van Campen, L.E.V., Hall III, J.W., & Grantham, D.W. (1997). Human offset auditory brainstem response: effects of stimulus acoustic ringing and rise-fall time. Hearing Research, 103, 35-46.  271  Van Maanen, A., & Stapells, D.R. (2005). Comparison of multiple auditory steady-state responses (80 versus 40 Hz) and slow cortical potentials for threshold estimation in hearing-impaired adults. International Journal of Audiology, 44, 613-624.  Vivion, M.C., Hirsch, J.E., Frye-Osier, J.L., & Goldstein, R. (1980). Effects of stimulus rise-fall time and equivalent duration on middle components of AER. Scandanavian Audiology, 9, 223-232 Walker, G., Dillon, H. & Byme, D. (1984). Sound field audiometry: recommended stimuli and procedures. Ear &Hearing, 5, 13-21.  Weihing, J., & Musiek, F.E.  (2008). An electrophysiological measure of binaural hearing in noise. Journal of American Academy of Audiology, 19, 481-495.  World Health Organization. (2015). Deafness and Hearing loss. Retrieved Nov. 30, 2015, from http://www.who.int/mediacentre/factsheets/fs300/en/ Wood, C.C. (1976). Discriminability, response bias, and phoneme categories in discrimination of voice onset time. Journal of Acoustical Society of America, 60(6), 1381-1389. Xu, Z.M., De Vel, E., Vinck, B., & Van Cauwenberge, P. (1995). Selecting the best tone-pip stimulus-envelope time for estimating an objective threshold for low-and middle-tone sensorineural hearing losses. European archives of oto-rhino-laryngology, 252, 275-279.  Yoshiniga-Itano, C., Coulter, D., & Thomson, V. (2000). The Colorado Newborn Hearing Screening Project: effects on speech and language development for children with hearing loss. Journal of Perinatology, 20(suppl 8), S132-S137.  272  Yoshinaga-Itano, C., Sedley, A.L., Coulter, D.K., & Mehl, A.L. (1998). Language of early- and later-identified children with hearing loss. Pediatrics, 102, 1161-1171. Zhang, V.W., Ching, T.Y.C., Buynder, P.V., Hou, S, Flynn, C., Burns, L., …Wong, A.O.C. (2014). Aided cortical response, speech intelligibility, consonant perception and functional performance of young children using conventional amplification or nonlinear frequency compression. International Journal of Pediatric Otorhinolaryngology, 78, 1602-1700.  Zhiromirov, H. (2012). Sound Analysis with Matlab Implementation – MATLAB Central. Retrieved July 15, 2015, from https://www.mathworks.com/matlabcentral/fileexchange/38837-sound-analysis-with-matlab-implementation/content/Sound_Analysis.m             273  Appendices  Appendix A: Hearing Aid Settings  Settings on NOAH were printed; simulated real ear measures were also collected. The information from NOAH and test box measurements of each hearing aid collected on the Audioscan Verifit 2 system were included, with test box measurements following the NOAH report print-out. The y-axis of each graph is the output intensity, in dB SPL, measured in the coupler; the x-axis of each graph is the frequency, in Hz. The red line in each NOAH report print out represents the hearing thresholds the hearing aids were programmed to. The green line represents the long-term average speech spectrum (LTASS) of speech presented at 65 dB SPL; the magenta line represents the LTASS of speech presented at 75 dB SPL; the cyan line represents the LTASS of speech presented at 55 dB SPL; the yellow line represents the maximum power output at the frequencies measured. Each program name and their gain setting was described.  Starkey Z series mini BTE i70 Memory 2 was set to linear gain and Memory 3 was set to compression. Memory 1 was not used. 274   Figure A.1.70Program settings printed from NOAH for Starkey  275   Figure A.2.71Coupler measurements of hearing aid output with varying input intensities when in Memory 2.  276   Figure A.3.72Coupler measurements of hearing aid output with varying input intensities when in Memory 3.  Siemens Motion SX Micon 5mi  Program 1 was set to linear gain and Program 2 was set to compression.     277      Figure A.4.73Program settings printed from NOAH for Program 1 of Siemens 278                       Figure A.5.74Program settings printed from NOAH for Program 2 of Siemens. 279   Figure A.6.75Coupler measurements of hearing aid output with varying input intensities when in Program 1.        280   Figure A.7.76Coupler measurements of hearing aid output with varying input intensities when in Program 2.  Phonak Bolero Q90 Calm Situation 1 was set to linear gain and Calm Situation 2 was set to compression.    281   Figure A.8.77Program settings printed from NOAH for Phonak       282   Figure A.9.78Coupler measurements of hearing aid output with varying input intensities when in Calm Situation 1.        283   Figure A.10.79Coupler measurements of hearing aid output with varying input intensities when in Calm Situation 2.            284  Appendix B: Averaged waveform and spectra of all recorded stimuli See link for all files: http://hdl.handle.net/2429/57501 Find the waveform or spectrum of the corresponding stimulus by first choosing the intensity condition, where loud corresponds to 65 dB SPL for most stimuli and 85 ppe SPL for brief stimuli. Then choose the stimulus type (e.g., SCP Tones). Then choose the hearing aid condition, by first choosing the gain setting (i.e., unaided, linear, compression), and then by choosing the hearing aid to view the stimulus recorded in a specific aided condition. Waveform images are found in the folder named “WAVEFORM” and spectral images are found in the folder named “SPECTRA.” Finally to choose the correct stimulus, read the file name using Template 1.   Hearing aid condition_gain setting (if aided)_stimulus description_intensity condition (denoted by numeric presentation intensity)“filtered”_polarity_endingname (waveformFIG for waveforms; and FIG for spectra) Template 1. Template to read stimulus file name.   For example, the waveform file name for the 1000 Hz tone burst of 30 ms duration with generated rise and fall times of 5 ms of the opposite polarity and recorded from Phonak (compression) in the 65 dB SPL condition is as follows: PHONAK_COMPRESSION_SCP_1000Hz_30msdur_5msrise_fall_65filteredopppolarity_waveformFIG. The file name for the spectrum of the same stimulus is as follows: PHONAK_COMPRESSION_SCP_1000Hz_30msdur_5msrise_fall_65filteredopppolarity_FIG.  The waveform file contains both the envelope and the acoustic waveform. The black outline on each waveform image is the envelope of the stimulus, and the wave in blue is the acoustic waveform.  Each stimulus had a 7000-sample tail of the silence in between stimulus 285  repetitions, other than the stimuli in the set of stimuli used to elicit ABRs and MLRs, which had much shorter SOAs. The SOA was the length of the waveform file for the stimuli used to elicit ABRs, and the length of the waveform file for the stimuli used to elicit MLRs was 2500 samples.                     286  Appendix C: Ear canal resonances of KEMAR  Ear canal resonances were obtained by measuring the real ear unaided gain using both a swept tonal stimulus stimuli and pink noise  presented at 45 dB SPL, 55 dB SPL and 65 dB SPL. Subsequent measurements were averaged to lead to the final measurements of the ear canal resonances.  Table C.1.59Ear canal resonances of KEMAR      Freqeuncy (Hz)  250 Hz 500 Hz 750 Hz 1000 Hz 1500 Hz 2000 Hz 3000 Hz 4000 Hz 6000 Hz Gain (dB) 0 0 -1 2 0.3 10 11 8.3 -5 

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
http://iiif.library.ubc.ca/presentation/dsp.24.1-0229467/manifest

Comment

Related Items