You may notice some images loading slow across the Open Collections website. Thank you for your patience as we rebuild the cache to make images load faster.

Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Comparisons of auditory steady-state response and behavioural air- and bone-conduction thresholds in… Casey, Kelly-Ann 2012

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

Notice for Google Chrome users:
If you are having trouble viewing or searching the PDF with Google Chrome, please download it here instead.

Item Metadata

Download

Media
24-ubc_2012_fall_casey_kellyann.pdf [ 2.14MB ]
Metadata
JSON: 24-1.0072812.json
JSON-LD: 24-1.0072812-ld.json
RDF/XML (Pretty): 24-1.0072812-rdf.xml
RDF/JSON: 24-1.0072812-rdf.json
Turtle: 24-1.0072812-turtle.txt
N-Triples: 24-1.0072812-rdf-ntriples.txt
Original Record: 24-1.0072812-source.json
Full Text
24-1.0072812-fulltext.txt
Citation
24-1.0072812.ris

Full Text

COMPARISONS OF AUDITORY STEADY-STATE RESPONSE AND BEHAVIOURAL AIR- AND BONE-CONDUCTION THRESHOLDS IN INFANTS AND ADULTS WITH NORMAL HEARING by  KELLY-ANN CASEY B.Comm./B.A., The University of Calgary, 2009 A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF  MASTER OF SCIENCE  in  The Faculty of Graduate Studies (Audiology and Speech Sciences)  THE UNIVERSITY OF BRITISH COLUMBIA (Vancouver) May 2012 © Kelly-Ann Casey, 2012  Abstract To improve our understanding of normal responses in infants, the present study compares air-conduction (AC) and bone-conduction (BC) auditory thresholds using both the auditory steady-state response (ASSR) and behavioural testing methods in normal-hearing infants (6-18 months of age) and adults. There are no correction factors available for estimating BC behavioural thresholds; this is a limiting factor to clinical implementation of the ASSR. Additionally, previous studies have reported that ASSR and visual reinforcement audiometry (VRA) thresholds (in dB HL) to air- and bone-conducted stimuli have different frequencydependent trends and suggest that infants present with an air-bone gap that is not attributable to a conductive pathology; however, this relationship has not been assessed directly. The objectives of the present study are: (i) to compare BC thresholds between methods and provide the initial step towards positing correction factors to predict BC behavioural thresholds and; (ii) to directly compare AC and BC thresholds to provide a more accurate estimation of the maturational ABG. Thresholds were estimated at 500–4000 Hz using AM2 stimuli for ASSRs and warbledtone stimuli for behavioural testing. The results indicated that BC thresholds were, on average, 7–16 dB poorer for ASSR compared to VRA, but varied largely across infants. As expected for the ASSR, frequency-dependent differences in BC sensitivity were found— the 500- and 1000Hz thresholds were better than the 2000-Hz threshold. For AC ASSR, the 500-Hz thresholds were higher than the other frequencies. There was a tendency for infant and adult ASSR thresholds to differ for BC, but not for AC. Behavioural thresholds for AC and BC were similar between infants and adults and across frequency. The results support the presence of a clinically significant maturational ABG (14 and 17 dB) in the low frequencies for infant ASSRs. The infant behavioural ABG also appeared at 500 Hz, as was posited by Hulecki and Small (2011), but was too small to be practically significant. Clinical consideration of the maturational ABG seems warranted when using ASSRs, but not for VRA. The results also provided preliminary normal levels for AC and BC ASSRs to AM2 stimuli.  ii  Preface This study was reviewed and approved by the Clinical Research Ethics Board of the University of British Columbia. The certificate number of the ethics certificate obtained is H11-00177.  iii  Table of Contents Abstract ........................................................................................................................................... ii Preface ........................................................................................................................................... iii Table of Contents ........................................................................................................................... iv List of Tables ................................................................................................................................. ix List of Figures ................................................................................................................................. x List of Abbreviations ..................................................................................................................... xi Acknowledgements...................................................................................................................... xiii  CHAPTER 1: Literature Review: Maturation of Air- and Bone-Conduction Hearing ..................1 1.1  Introduction .................................................................................................................. 2  1.2  Bone-conduction Hearing............................................................................................. 6  1.2.1  Mechanisms of Bone-Conduction......................................................................... 6  1.2.1.1 Compressional .................................................................................................. 7 1.2.1.2 Middle-Ear Inertia ............................................................................................ 7 1.2.1.3 Inertia of Cochlear Fluids ................................................................................ 8 1.2.1.4 Osseotympanic ................................................................................................. 8 1.2.1.5 Non-Osseous .................................................................................................... 9 1.2.2  Maturation of BC Thresholds ............................................................................. 10  1.2.2.1 Behavioural .................................................................................................... 10 1.2.2.2 Physiological .................................................................................................. 11 1.2.3  Other Maturational Effects on Physiological Measures with Bone Conduction 14  1.2.3.1 Wave V of the ABR ....................................................................................... 14 1.2.3.2 Two-channel Asymmetries ............................................................................ 14 1.2.4  Procedural Variables of BC Testing ................................................................... 15 iv  1.3  Air-Conduction Hearing ............................................................................................. 16  1.3.1  Maturation of Air-Conduction Thresholds ......................................................... 16  1.3.1.1 Behavioural .................................................................................................... 16 1.3.1.2 Physiological .................................................................................................. 18 1.3.2  Other Maturational Effects on Physiological Measures with Air Conduction ... 20  1.3.2.1 ABR Latency Differences .............................................................................. 20 1.3.2.2 Ipsi/Contra Asymmetries ............................................................................... 20 1.4  The Maturational Air-Bone Gap ................................................................................ 21  1.5  Accounting for Infant-Adult Differences ................................................................... 23  1.5.1  Outer/Middle Ear ................................................................................................ 24  1.5.2  Cochlea ............................................................................................................... 25  1.5.3  Brainstem ............................................................................................................ 26  1.5.4  Skull .................................................................................................................... 27  1.5.5  Non-Sensory Factors ........................................................................................... 27  1.6  Comparing Threshold Estimates across Test Methods .............................................. 28  1.7  The Auditory Steady-State Response ......................................................................... 30  1.7.1  Rate/EEG ............................................................................................................ 31  1.7.2  ASSR Generators ................................................................................................ 31  1.7.3  Factors of Age and Sleep .................................................................................... 32  1.7.4  Stimuli ................................................................................................................. 33  1.7.4.1 Single vs. Multiple Stimuli ............................................................................ 35 1.7.4.2 Stimuli and ASSR Systems ............................................................................ 37 1.7.5  Reducing EEG Noise .......................................................................................... 38  1.7.6  Detecting a Response .......................................................................................... 39 v  1.7.6.1 Amplitude....................................................................................................... 39 1.7.6.2 Phase .............................................................................................................. 40 1.7.7  Stopping Criteria ................................................................................................. 40  1.7.8  Electrode Montage .............................................................................................. 41  1.7.8.1 Two-Channel Recordings .............................................................................. 42 1.8  Visual Reinforcement Audiometry ............................................................................ 43  1.8.1  Behaviour ............................................................................................................ 44  1.8.2  Habituation .......................................................................................................... 45  1.8.3  Reinforcement Type ............................................................................................ 45  1.8.4  Stimulus Type ..................................................................................................... 46  1.9  Rationale for Thesis.................................................................................................... 47  CHAPTER 2: Comparisons of Auditory Steady-State Response and Behavioural Air- and BoneConduction Thresholds in Infants and Adults with Normal Hearing ............................................49 2.1  Introduction ................................................................................................................ 50  2.2  Methods and Materials ............................................................................................... 52  2.2.1  Participants .......................................................................................................... 52  2.2.2  Stimuli ................................................................................................................. 53  2.2.2.1 Bone Conduction (BC)................................................................................... 53 2.2.2.2 Air Conduction (AC) ..................................................................................... 55 2.2.3  Calibration........................................................................................................... 55  2.2.4  Recording ............................................................................................................ 55  2.2.5  Procedure ............................................................................................................ 57  2.2.5.1 VRA ............................................................................................................... 58 2.2.5.2 ASSR.............................................................................................................. 61 vi  2.2.6 2.3  Data and Statistical Analyses .............................................................................. 62  Results ........................................................................................................................ 63  2.3.1  The 90th Percentile .............................................................................................. 63  2.3.2  ASSR Thresholds ................................................................................................ 64  2.3.2.1 Infants............................................................................................................. 65 2.3.2.2 Adults ............................................................................................................. 67 2.3.2.3 Comparing Adult and Infant ASSRs .............................................................. 67 2.3.3  VRA Thresholds ................................................................................................. 68  2.3.4  The Relationship between ASSR and VRA Thresholds ..................................... 68  2.3.5  The Air-Bone Gap (ABG)................................................................................... 70  2.3.5.1 AC vs. BC ...................................................................................................... 70 2.3.5.2 The Presence of ABGs ................................................................................... 70 2.3.6  ASSR Amplitudes ............................................................................................... 73  2.3.7  Phase ................................................................................................................... 74  CHAPTER 3: Discussion and Conclusion ...................................................................................76 3.1  Discussion .................................................................................................................. 77  3.1.1  ASSR................................................................................................................... 77  3.1.1.1 Infant ASSR Thresholds ................................................................................ 77 3.1.1.2 Infant ASSR Amplitudes ............................................................................... 80 3.1.1.3 Adult ASSR Thresholds ................................................................................. 81 3.1.1.4 Adult ASSR Amplitudes ................................................................................ 83 3.1.2  Behavioural ......................................................................................................... 83  3.1.2.1 Infant VRA ..................................................................................................... 83 vii  3.1.2.2 Adult Behavioural .......................................................................................... 86 3.1.3  Comparing Infants to Adults ............................................................................... 87  3.1.4  The Relationship between ASSR and VRA........................................................ 89  3.1.5  The Air-bone Gap ............................................................................................... 91  3.1.6  Infant-Adult Differences ..................................................................................... 94  3.1.6.1 Sensory Factors .............................................................................................. 94 3.1.6.2 Non-sensory Factors ...................................................................................... 96 3.2  Conclusion .................................................................................................................. 97  3.2.1  Clinical Implications ........................................................................................... 99  3.2.2  Future Research ................................................................................................ 100  References................................................................................................................................... 101 Appendix A: Individual Infant Threshold Data .......................................................................... 126 Appendix B: Individual Adult Threshold Data .......................................................................... 129 Appendix C: Individual Infant Amplitude Data ......................................................................... 131 Appendix D: Individual Adult Amplitude Data ......................................................................... 136 Appendix E: Individual Infant Phase Data ................................................................................. 139 Appendix F: Individual Adult Phase Data .................................................................................. 146  viii  List of Tables Table 1.1 Threshold Correction Factors (in dB) for Infant Tone-evoked ABR and ASSR. Adapted from Stapells, 2010a....................................................................................................... 29 Table 2.1 Count and Percentage of infants who provided air- or bone-conduction behavioural thresholds for a given number of frequencies............................................................................... 52 Table 2.2 Infants sample size of VRA thresholds obtained by frequency and mode. ................. 61 Table 2.3 Infants sample size of ASSR thresholds obtained by frequency and mode. ................ 62 Table 2.4 Bone- and Air-conduction mean thresholds (in dB HL) and standard deviations by age, method and frequency. .......................................................................................................... 66 Table 2.5 Descriptive statistics of difference scores (in dB) between individual infant ASSR and VRA thresholds. ........................................................................................................................... 69 Table 2.6 Air-Bone Gap (ABG; ≥ 5 dB) for infants and adults by method: reported as the percentage of participants who presented with an ABG of a given size (percentages shown are not cumulative). Median ABGs are shown for each condition. ................................................... 71 Table 3.1 Comparisons of mean ASSR thresholds (dB HL) for both infants and adults across frequency from the present and previously published studies. ..................................................... 78 Table 3.2 Comparisons of mean behavioural thresholds (dB HL) for both infants and adults across frequency from the present and previously published studies. .......................................... 84 Table 3.3 Within study comparisons of infant-adult differences in ASSR thresholds (dB; infant minus adult) from the present and previously published studies. ................................................. 88  ix  List of Figures Figure 1.1 Comparisons between air- and bone-conduction ASSR thresholds for both infants and adults between different studies.The figure shows mean data from the following four studies: (1) infant BC thresholds were reported by Small and Stapells (2006) for infants aged 0–8 months; (2) infant AC thresholds were published by Van Maanen & Stapells (2009) for infants aged 0.5–66.2 months; (3) adult AC thresholds (converted from dB SPL to dB HL) are from Herdman & Stapells (2001) and; (4) adult BC thresholds from Ishida et al. (2011) study B. ..... 22 Figure 1.2 Acoustic spectra of AM/FM and AM2 stimuli for the carrier frequencies of 500, 1000, 2000, and 4000 Hz. Y-axis ticks represent 20 dB intervals (Adapted with permission from Wood, 2010). ................................................................................................................................ 35 Figure 2.1 Acoustic spectra of air-conducted AM2 stimuli for the carrier frequencies of 500, 1000, 2000 and 4000 Hz. Y-axis ticks represent 10 dB intervals................................................. 54 Figure 2.2 The cumulative percent of responses present for infant and adult BC and AC ASSRs across intensity at each frequency. The dotted lines represent the intensities at which 90% of the subjects for each condition have present responses. ..................................................................... 64 Figure 2.3 BC ASSR and BC VRA threshold comparisons in infants at 500, 2000 and 4000 Hz. ...................................................................................................................................................... 69 Figure 2.4 Comparisons of the mean air-bone gap (AC minus BC thresholds; ±1 SD) at each frequency for adult ASSR and infant ASSR and VRA thresholds. .............................................. 71 Figure 2.5 Mean ASSR amplitudes (±1 SD) at each carrier frequency for infants and adults. ... 73 Figure 2.6 Mean ASSR phase delays (±1 SD) at each carrier frequency for infants and adults. 75  x  List of Abbreviations Abbreviation  Definition  ABG  Air-bone gap; AC minus BC threshold  ABR  Auditory brainstem response  AC  Air conduction  AM  Amplitude modulation  AM2  Exponential (squared) amplitude modulation  ASSR  Auditory steady-state response  BC  Bone conduction  BOA  Behavioural observation audiometry  dB  Decibels  dB HL  Decibels hearing level  dB nHL  Decibels normal hearing level  dB peSPL  Decibels peak-equivalent sound pressure level  dB SPL  Decibels sound pressure level  EEG  Electroencephalogram  FFT  Fast Fourier transform  FM  Frequency modulation  Hz  Hertz  mASSR  Multiple Auditory steady-state response  MM  Mixed modulation  MRL  Minimal response level  NBN  Narrowband noise  xi  Abbreviation  Definition  nV  Nanovolt  p  Probability  SNR  Signal-to-noise ratio  STC  Soft tissue conduction  VRA  Visual reinforcement audiometry  xii  Acknowledgements I would like to extend my appreciation to my supervisor, Dr. Susan Small. I thank her for her time, insight and effort in guiding me through the thesis process. I am grateful for our many thoughtful discussions that helped shape my thoughts and direct my writing. I would also like to thank the other members of my thesis committee, Dr. David Stapells and Dr. Anna Van Maanen. I am grateful for the support and feedback they provided throughout the thesis process. I would also like to acknowledge the team of P.A.L. volunteers and Audiology students who assisted me with recruiting participants and collecting data (Griselle Leon, Aisling Smyth, Ricky Lau, Estephanie Sta. Maria, Halen Panchyk and Foong Yen Chong). Thank you so much for being flexible, reliable and interested! Finally, I want to express my gratitude to my loving Mom, who has always provided support and encouragement in any way needed. Additionally, I thank my sister and friends, who have provided unconditional support throughout my post-secondary education.  xiii  CHAPTER 1: Literature Review: Maturation of Air- and BoneConduction Hearing  1  1.1 Introduction It is recommended that hearing screening and intervention programs detect hearing loss before three months of age and provide appropriate treatment by six months of age (Joint Committee on Infant Hearing, 2007). The major motivation to meet these goals is that early identification and intervention have been shown to provide improved long-term speech and language outcomes for children identified with hearing loss (Kennedy et al., 2006; Yoshinaga-Itano, Sedey, Coulter, & Mehl, 1998). Accurate and early assessment of hearing sensitivity in children is critical in these programs to best serve a child‘s auditory needs, such as the provision of amplification. As suggested by the JCIH guidelines (2007), these assessments are most informative when they include a measure providing frequency-specific threshold information and indicate the type of hearing loss, by assessing hearing through both airconduction (AC) and bone-conduction (BC). Collecting Frequency-Specific Threshold Information As children mature, the technique that provides the most reliable information about their auditory sensitivity changes. In acquiring frequency-specific threshold information, it is prudent to acknowledge that maturation influences a child‘s responses and thus what is considered a ―normal‖ auditory response varies with age and assessment method. As is discussed in a later section, there are trends in a child‘s thresholds or minimal response levels (MRLs)1 that differ by frequency and conduction pathway and become more adult-like as the child develops. Visual reinforcement audiometry (VRA) is the recommended behavioural method for assessing hearing in infants and toddlers who are of a developmentally appropriate age to provide reliable responses (JCIH, 2007). Infants under approximately five to six months of age are often not able to provide reliable behavioural thresholds (Gravel & Traquina, 1992; Moore, Wilson, & Thompson, 1977; Northern & Downs, 1991; Widen, 1990; Widen et al., 2000), so physiological threshold assessment methods are substituted. Although physiological and behavioural thresholds are not equivalent, behavioural thresholds can be estimated from physiological measures if the relationship between the responses obtained by each measure is well understood. The tone-evoked auditory brainstem response 1  Responses obtained using VRA are often referred to as MRLs because behavioural responses are thought to be elevated due to maturation and may not reflect true auditory sensitivity (Matkin, 1977). Additionally, it is an appropriate term, if the protocol used does not involve searching for responses below a certain level.  2  (ABR) is considered to be the best physiological measure available for approximating hearing thresholds in infants and children who cannot provide reliable behavioural results (JCIH, 2007). The auditory steady-state response (ASSR) is another physiological diagnostic tool for hearing evaluation that has increasingly gained research and clinical interest for its potential to assess this population. The tone-evoked ABR assesses responses of one ear to one frequency at a time, but the ASSR has the advantage that it can determine responses to both ears at multiple frequencies, simultaneously. In this way, implementation of ASSRs has the potential to decrease clinical testing time (Picton et al, 2003; Stapells et al, 2005; Cone and Dimitrijevic, 2009). Additionally, the ASSR is a more objective measure of response presence or absence than the tone-evoked ABR, as it evaluates responses on the basis of statistical measures, while ABRs are determined by the more subjective visual identification of waveforms (specifically wave V). An existing barrier to full clinical implementation of the ASSR is insufficient research. Research involving normal-hearing infants and young children can provide expected normal levels. Van Maanen and Stapells (2009) posited normal AC multiple-ASSR levels for four frequencies—500, 1000, 2000, and 4000 Hz. Small and Stapells (2008a; 2006) have suggested normal levels for BC ASSR for the same four frequencies. However, research must also involve subjects with hearing loss to evaluate how well ASSR differentiates between varying types and degrees of hearing loss. Currently, a number of studies on infants with hearing loss exist for AC ASSR, but no equivalent studies, with appropriate threshold confirmation, have been published for BC ASSR (Stapells, 2010a). To assess the performance of the ASSR, it should be compared to the current gold standards in the field, such as VRA or threshold estimation using the tone-evoked ABR for the population of interest (Stapells, 2010a). High correlations between AC ASSR thresholds and thresholds provided by goldstandard measures, indicate that the ASSR can provide accurate estimates of hearing threshold in young children with hearing loss (Chou, Chen, Yu, Wen, & Wu, 2012; Han, Mo, Liu, Chen, & Huang, 2006; Luts, Desloovere, & Wouters, 2006; Rance & Briggs, 2002; Rance et al., 2005; Rodrigues & Lewis, 2010; Van Maanen & Stapells, 2010). However, Van Maanen and Stapells (2010) point out that the current ASSR studies that include appropriate threshold confirmation differ substantially, and as such only provide data applicable to their own specific technique or stimulus used. Overall, there are limited data (i.e., a small sample size) for AC ASSRs in young children with hearing loss, especially for multiple stimuli presented simultaneously.  3  Research on BC ASSRs is even more limited. There is currently only one published study of BC ASSRs in children with hearing loss (Swanepoel, Ebrahim, Friedland, Swanepoel, & Pottas, 2008). However, it examined thresholds for a wide age range (3 months to 11.5 years) of children and the accuracy of the ASSR results cannot be verified as the hearing losses were not confirmed using a goldstandard tool. There is no study that directly assesses the association between BC ASSRs and either tone-evoked ABR or VRA using bone-conducted stimuli in infants with normal or impaired hearing. Hulecki and Small (2011) recently investigated behavioural BC MRLs in infants, using VRA, and found that behavioural thresholds exhibit the same frequency-dependent trend that has been observed using ASSRs (Small & Stapells, 2006, 2008a). Specifically, infant BC MRLs are better at low frequencies than high frequencies, whereas adults show the opposite pattern—better thresholds for high frequencies than low frequencies. A limitation of their study was that they did not estimate BC ASSR thresholds and were not able to make direct comparisons between physiological and behavioural responses to boneconducted stimuli. For any assessment of hearing, the aim is to ultimately estimate behavioural thresholds. In the protocols of Universal Newborn Hearing Screening and Early Hearing Detection and Intervention programs, this requires the application of a correction factor to the physiological estimate (British Columbia Early Hearing Program, 2008; Ontario Infant Hearing Program, 2008; for review of estimation techniques see Picton, John, Dimitrijevic, & Purcell, 2003; Stapells, 2010a). Application of the correction factor for ASSR and tone-evoked ABR thresholds results in an estimated behavioural hearing level, eHL, which is the closest approximation of the infant‘s hearing thresholds. An eHL value is valuable because it provides a reasonable guide for fitting amplification. Correction factors are available for AC tone-evoked ABR thresholds and conservative preliminary values have been proposed for AC ASSR thresholds. Hatton, Janssen and Stapells (submitted) recently posited a correction factor for 2000-Hz BC ABR; however, there are no correction factors at other frequencies for BC ABR or at any frequencies for BC ASSR. Until further data are obtained to determine appropriate correction factors, ASSR can only be used as a screening tool or a supplement to the ABR (Stapells, Herdman, Small, Dimitrijevic, & Hatton, 2005; Van Maanen & Stapells, 2010). Establishing correction factors requires estimation of actual versus predicted thresholds for infants, where a gold-standard measure (such as tone-evoked ABR or VRA) provides actual thresholds that are compared to the predicted thresholds of ASSR (Stapells, 2010a). As a precursor to establishing BC ASSR eHLs, it is necessary to acquire BC ASSR and behavioural thresholds in normal-hearing infants. Estimating the mean difference  4  of VRA and ASSR thresholds will contribute to the initial step in establishing correction factors for BC ASSR. To posit eHL correction factors, it would then be necessary to obtain the same information in infants who have hearing loss. Determining the Type of Hearing Loss Acquiring threshold information for both air- and bone-conducted stimuli allows the differentiation of conductive, mixed and sensorineural hearing losses. In adults, a gap between the AC and BC thresholds is indicative of a conductive component and is commonly called an air-bone gap (ABG) (Lierle & Reger, 1946; Carhart, 1950). The reported frequency-dependent trends for infant AC and BC ASSR thresholds in dB HL suggest that normal-hearing infants present with an ABG that cannot be explained by a conductive pathology. AC and BC thresholds estimated in dB HL account for adult sensitivities across frequency; therefore, the ABG in infants with normal middle ear function are likely explained by infant-adult differences in sensitivity to AC and BC stimuli. In an infant, to accurately diagnose conductive involvement in a hearing loss, it is necessary to recognize that an ABG consists of a maturational component. By determining the size of the maturational ABG, diagnosis of a conductive component can become more accurate. Recent research suggests that this maturational ABG may also be apparent in the behavioural testing of infants. Hulecki and Small (2011) established that the maturational differences found for physiological BC thresholds also exist for behavioural thresholds. By comparing their results to AC MRLs published by Parry, Hacking, Bamford & Day (2003), they found that a behavioural maturational ABG also exists in infants, but to a smaller extent than the implied physiological ABG. However, these behavioural and physiological estimates of the ABG are based on group data comparisons of AC and BC between different studies (i.e., AC and BC testing were not conducted in the same children). Comparison of AC and BC thresholds in the same infant will provide a more accurate estimation of the maturational gap, by controlling for variability due to age and differences in test procedures (Hulecki & Small, 2011). Comparison of the maturational ABG between methods will indicate how well the ASSR predicts the behavioural ABG. For completeness, future research should obtain the same measures for both infants with normal hearing and those with hearing loss. The present study compares AC and BC thresholds using both VRA and ASSR methods in normal-hearing infants to AC and BC ASSR thresholds and behavioural thresholds in adults. Comparing  5  BC thresholds across different testing techniques, and in relation to AC thresholds, can increase our understanding of what constitutes a normal response for an infant. The study has the following two objectives: (1) to compare BC responses between VRA and ASSR testing methods to further define their relationship and; (2) to compare BC and AC sensitivity for both methods to better quantify the maturational ABG. Adult subjects will serve as a control group for making maturational comparisons. The following sections will present background information applicable to the current study: a. Bone-conduction hearing mechanisms, threshold maturation and procedural variables; b. Air-conduction hearing threshold maturation; c. The maturational air-bone gap; d. Accounting for differences between adults and infants; e. Comparing threshold estimates across test methods; f. Overview of test methods—The Auditory Steady State Response and Visual Reinforcement Audiometry.  1.2 Bone-conduction Hearing Bone-conduction testing is an important clinical tool in the evaluation of hearing sensitivity. It is used as an indicator of the area of the auditory system responsible for a hearing deficit. It differs from AC transmission by theoretically bypassing the conductive components of the auditory system (i.e., the outer and middle ear) to reach the cochlea. In this way, its elevation suggests sensorineural involvement. The following sections address the mechanisms and maturation of BC hearing as well as procedural considerations when testing infants.  1.2.1 Mechanisms of Bone-Conduction Past research has shown that the basilar membrane is stimulated similarly by air- and boneconducted sounds (for review see Stenfelt & Goode, 2005). In contrast to air-conducted sounds, which have only one pathway for sound to travel to the cochlea by, there are a number of pathways that have been theorized to contribute to BC hearing. Although early research by Allen and Fernandez (1960) posited a single mechanism of skull vibrations inducing cochlear fluid movement, continued research supports multiple BC pathways that vary in influence. Three osseous pathways have widely been considered to contribute to BC hearing: compressional, inertial and osseotympanic (Barany, 1938;  6  Kirikae, 1959; Vento & Durrant, 2009; Wever & Lawrence, 1952; Tonndorf, 1966). Stenfelt and Goode (2005) re-categorized the pathways and added an additional non-osseous pathway to describe the five most relevant mechanisms of BC hearing: compressional, ossicular inertia, cochlear fluid inertia, osseotympanic, and the cerebrospinal fluid. 1.2.1.1  Compressional The compressional mechanism was first described by Herzog and Krainz in 1926, and was  further defined by von Bekesy (1932, as cited in Hood, 1962) and Tonndorf (1966). This mechanism describes compression and dilation of the cochlea‘s shell that correspond to the BC sound waves, resulting in distortion of the cochlea‘s shape (Tonndorf, 1966). This mechanism relies on two asymmetries to create a pressure gradient: the scala vestibuli having a greater volume of perilymph than the scala tympani and the round window being more compliant than the oval window (Hood, 1962). In theory, when the cochlear walls are distorted, these asymmetries allow for fluid displacement, triggering a wave along the basilar membrane. In otosclerotic ears, the oval window is unyielding; compression of the walls would be expected to produce a larger fluid displacement towards the round window and thus, improved BC sensitivity (Stenfelt & Goode, 2005). However, BC sensitivity does not improve in otosclerotic ears (Carhart, 1962), suggesting compression is not a dominant factor in BC hearing. 1.2.1.2  Middle-Ear Inertia Middle-ear inertia was first described by Barany (1938). Bone vibrations are transferred from the  bony walls of the middle ear cavity to the ossicles by connective tissues, but this transfer introduces an inertial lag such that the vibrations of the ossicles and skull become asynchronous (Barany, 1938; Humes, 1979; Stenfelt & Goode, 2005). This difference causes relative movement between the stapes and the cochlea, which triggers displacement of the cochlear fluid (Barany, 1938; Kirikae, 1959; Stenfelt & Goode, 2005). There are frequency-dependent effects of the inertial mechanism, such that it potentially plays a role in BC hearing mainly at the frequencies between 1000 and 3000 Hz (Stenfelt & Goode, 2005). However, studies have shown that in conditions where the ossicles cannot function normally, there is only a small impact on BC sensitivity (Stenfelt & Goode, 2005). For example, Everberg (1968) found that despite the lack of an oval window, which eliminates the possible contributions of ossicular inertia, subjects exhibited near-normal BC thresholds. Møller (2000) also reported near-normal BC thresholds in the presence of ossicular discontinuity. These findings suggest  7  that other mechanisms play a much greater role in bone-conducted hearing sensitivity than middle-ear inertia. 1.2.1.3  Inertia of Cochlear Fluids Similar to the ossicular inertia mechanism, Stenfelt and Goode (2005) describe that inertia of the  cochlear fluids involves bone vibrations directly acting on the cochlea. Although the fluids are incompressible, the membranous oval and round windows act together, by way of a pressure gradient between them, to allow the fluid to move and create a traveling wave on the basilar membrane. Relevant for BC testing, any additional opening into the cochlea that may provide fluid motion, a ―third window‖, may include aqueducts and nerve fibers (Küçük et al., 1991; Stenfelt & Goode, 2005; Tonndorf, 1966). Stenfelt and Goode (2005) found that the fluid displacement differs at the oval and round window, which can be accounted for by the existence of a third window. This theory also explains why nearnormal BC thresholds can accompany certain conditions where the oval window is no longer acting on the pressure gradient, such as is the case when there is no oval window or in an otosclerotic ear. The authors suggest that fluid inertia is probably the major mechanism of BC hearing below 1000 Hz. 1.2.1.4  Osseotympanic The osseotympanic mechanism is involved when a bone-conducted sound‘s vibrations transfer to  the cartilaginous walls of the external auditory canal (von Bekesy, 1932 as cited in Hood, 1962; Stenfelt, Wild, Hato, & Goode, 2003). This movement of the walls creates pressure changes in the canal. Just as with an air-conducted sound, the pressure change in the canal is passed on to the tympanic membrane and through the middle ear to stimulate the cochlea (von Bekesy, 1932 as cited in Hood, 1962). An important consequence of this pathway is the occlusion effect, whereby hearing sensitivity to low-frequency (≤ 1000 Hz) bone-conducted sound is greater when the external auditory canal is blocked. Huizing (1960) originally speculated that the occlusion affect was a result of resonance changes with the canal. This theory is thought to be relevant only for the slight changes in ear-canal sound pressure that occur in the higher frequencies (equal to and above 2000 Hz) with occlusion (Stenfelt & Goode, 2005). As described by Tonndorf (1966), the much larger occlusion effect is at the lower frequencies and is attributable to the open ear canal acting as a high-pass filter, an effect which is  8  absent in the occluded canal. Without the high-pass filter, there is an increase in the intensity of lowfrequency sounds in the occluded canal relative to the open ear canal. 1.2.1.5  Non-Osseous Yoshida and Uemura (1991) studied the role of cerebrospinal fluid (CSF) on cochlear activity  and provided evidence that sound vibrations can be transmitted through fluid instead of bone. This type of transmission is the non-osseous fourth pathway, which has also been supported in recent literature (Freeman, Sichel, & Sohmer, 2000; Sohmer, Freeman, Geal-Dor, Adelman, & Savion, 2000; Sohmer & Freeman, 2004). This mode is supported by recent studies in rodents (Freeman et al., 2000; Sohmer et al., 2000) and humans (Sohmer et al., 2000) which demonstrated that presenting the bone-conducted sound to the skull contents stimulates the cochlea. The authors concluded that this was likely attributable to fluid communications between the CSF and cochlear fluids through the cochlear and vestibular aqueducts and perineural and perivascular spaces. This theory is supported by comparable auditory responses (through ABR and pure-tone audiometry) obtained through non-osseous and osseous stimulation, which suggests that ‗bone‘ is not necessary to achieve BC hearing and that fluid is actually the major factor (Sohmer et al., 2000). Furthermore, reducing the volume of CSF produces BC threshold elevations in rats (Freeman et al., 2000). De Jong and colleagues (2011) propose that the non-osseous mechanism is actually a new mode of conduction, soft-tissue conduction, in addition to air- and boneconduction. The authors highlight the fact that soft-tissue conduction only exhibits a minimal occlusion effect compared to BC; based on this they conclude that soft-tissue conduction must be a separate phenomenon that may be useful clinically as a measure of cochlear sensitivity. Further research is needed to expand our understanding of the non-osseous mechanism. It is clear that there are many mechanisms that contribute to BC hearing to varying degrees. However, much of what we know about these mechanisms is based on research in either adult human or animal subjects and is not necessarily applicable to infants. Infants exhibit structural and neural immaturities that may alter the transmission of bone-conducted sound in comparison to adults. Small, Hatton, and Stapells (2007) provided evidence for age-related differences in BC mechanisms with the discovery that young infants (approximately 0-7 months) do not have an occlusion effect. A follow-up study replicated these findings for young infants (0-7 months), and found that the occlusion effect was emerging in older infants (10-22 months) (Small & Hu, 2011). This suggests the osseotympanic pathway  9  is not yet fully developed in infants. It is currently unclear to what extent other identified adult BC mechanisms contribute to infant BC hearing. Specific origins of these and other infant-adult differences are discussed in a later section.  1.2.2 Maturation of BC Thresholds Prior to discussing the maturation of BC thresholds, it is necessary to recognize that the auditory system is not fully developed at birth, but matures into the late teenage years. There are many anatomical changes that occur during the transition from infancy to adulthood that also influence responses to physiological and behavioural tests of auditory sensitivity (see review in Moore & Linthicum, 2007). The decibel scale commonly used to express both AC and BC hearing sensitivity across frequencies is dB hearing level, which is referenced to normal young adult thresholds (0 dB HL). Thresholds that differ significantly from 0 dB HL are considered elevated and indicate a hearing loss. Because the infant auditory system and associated auditory behaviours are still in development, it is not surprising that the definition of normal for infants differs from adults. To accurately interpret infant hearing assessments, it is critical to understand normal development of air- and bone-conduction thresholds for both behavioural and physiological test methods. Despite the fact that BC testing is critical for establishing cochlear integrity across the lifespan, there is limited research on the maturation of BC hearing sensitivity. As previously discussed, there are multiple routes by which bone-conducted stimuli may reach the cochlea. Although these routes have not been well investigated in infants, it is reasonable to expect that immaturities in any component of these pathways have the potential to alter the transmission of a bone-conducted signal and thus may influence BC hearing sensitivity (Nousak & Stapells, 1992). This includes developmental changes affecting any component of the auditory system or skull. It is believed that age- and frequency-dependent differences in the BC thresholds of normal-hearing infants do exist and thus are important to understand for accurate assessment of hearing sensitivity. 1.2.2.1  Behavioural Research indicates that BC MRLs obtained using VRA are reliable (Gravel & Traquina, 1992),  and thus for their diagnostic value they can, and should, be included in the infant audiological assessment. Despite being of relevance for determining type of hearing loss, the maturation of BC 10  hearing thresholds is not well understood. As is the case with normal hearing or conductive hearing loss, BC MRLs are expected to be within the same normal hearing limits accepted for AC responses (≤ 20 to 25 dB HL). There is very little published evidence on the mean BC thresholds of infants; although, it is likely that there is much unpublished clinical and research data that exist. This may be because of the lack of perceived diagnostic value that actual normal BC thresholds provide in addition to simply knowing that BC hearing sensitivity is within the normal range. Hulecki and Small (2011) argue that it is critical to understand the developmental trends of normal-hearing infant behavioural BC MRLs to accurately interpret any disparities between air- and bone-conduction responses. The importance of this is further discussed in a later section (see The Maturational Air-Bone Gap section). Gravel (1989) used VRA in a clinical setting and collected estimates of both AC and BC hearing sensitivity in normal hearing infants and infants with middle-ear pathology. Published figures included audiograms with both mean AC and BC VRA thresholds ( 1 SD) for normal hearing infants age 6–12 months. Based on visual assessment, the data do not appear to suggest any significant frequencydependent threshold differences for either AC or BC thresholds, nor do they exhibit an air-bone gap. Inspired by the frequency-dependent differences observed in the physiological thresholds of infants, Hulecki and Small (2011) published behavioural BC MRLs for young (7–15 months) and older (18–30 months) infants. They confirmed that infant responses were significantly better in the lower frequencies (500 and 1000 Hz) compared to the higher frequencies (2000 and 4000 Hz) with no age-dependent differences. The mean MRLs were between 5 and 8 dB worse for the high frequencies compared to the lower frequencies. This trend is in agreement with physiological findings in infants (Stapells & Ruben, 1989; Foxe & Stapells, 1993; Cone-Wesson & Ramirez, 1997; Small & Stapells, 2006, 2008a), as discussed in the next section. Given the sparcity of published data and contradictory findings between the two studies mentioned, further research is necessary to clarify our understanding of BC behavioural thresholds. 1.2.2.2  Physiological Many studies of ABR and ASSR BC thresholds in normal-hearing infants have shown that their  thresholds differ from adults. In some research on click-ABR thresholds, authors found that adults had poorer thresholds than newborn infants (Cone-Wesson & Ramirez, 1997; Muchnik, Neeman, &  11  Hildesheimer, 1995; Stuart, Yang, & Green, 1994). However, Cornacchia, Martinia, & Morra (1983) found no differences between older infants (16–20 months) and adults. Frequency-specific stimuli are the preferred stimuli for thorough assessment of hearing sensitivity. Research on ABR and ASSR with frequency-specific bone-conducted stimuli also supports age-dependent differences in thresholds across frequencies (Foxe & Stapells, 1993; Nousak & Stapells, 1992; Small & Stapells, 2006, 2008a; Stapells & Ruben, 1989; Swanepoel et al., 2008; Vander Werff, Prieve, & Georgantas, 2009). Small and Stapells (2008a) assessed BC ASSR thresholds in young infants (0.5–44 weeks), older infants (12–24 months) and adults (19–48 years) and found that infant thresholds continue to mature up to at least two years of age. Compared to adults, young infants have better BC thresholds at 500 Hz and slightly worse thresholds at 2000 Hz (Foxe & Stapells, 1993; Nousak & Stapells, 1992; Small & Stapells, 2006, 2008a; Stapells & Ruben, 1989; Vander Werff et al., 2009). This is opposite to the trend seen for infant AC thresholds. Less research has been conducted on BC thresholds at 1000 Hz and 4000 Hz. Thresholds to 500-Hz bone-conducted stimuli are reported to be lower (i.e., better) in infants compared to adults (Cone-Wesson & Ramirez, 1997; Foxe & Stapells, 1993; Small & Stapells, 2006, 2008a; Stapells & Ruben, 1989). Comparisons across studies reveal that the 500-Hz ABR threshold worsens by approximately 16-27 dB from birth to approximately 5- to 6-months of age (Cone-Wesson & Ramirez, 1997; Foxe & Stapells, 1993; Stapells & Ruben, 1989). Adults have 500-Hz thresholds only about 2–7 dB worse than those of 5- to 6-month old infants (Foxe & Stapells, 1993; Stapells & Ruben, 1989). The maturation of the BC hearing sensitivity at 1000 Hz is similar to 500 Hz, as evidenced by ASSR studies comparing infant and adult thresholds. Small and Stapells (2008a) found that in young infants (mean age: 16 weeks) the 1000-Hz threshold was significantly better than it was at the other frequencies (500, 2000 and 4000 Hz), but like the 500-Hz threshold became worse with age (Small & Stapells, 2008a). The 1000-Hz thresholds of older infants (mean age: 18 months) were only significantly better than the 2000-Hz thresholds, and no different than the 500- and 4000-Hz thresholds. At 2000 Hz, studies indicate that BC thresholds exhibit the opposite trend that is observed in the lower frequencies (Foxe & Stapells, 1993; Small & Stapells, 2006, 2008a; Stapells & Ruben, 1989). For the tone-evoked ABR, hearing sensitivity at 2000 Hz improves slightly (2–4 dB) from 5- to 6-months of age into adulthood (Foxe & Stapells, 1993; Stapells & Ruben, 1989). Small and Stapells (2006) reported  12  this trend to a greater degree in pre- and post-term infants, who had 2000-Hz thresholds that were 19 and 8 dB worse than adult thresholds, respectively. Small and Stapells (2006) pooled the high-frequency thresholds (2000 and 4000 Hz) and found that pre-term infants (mean age: 34.5 weeks postconceptional age) had significantly higher (i.e., worse) thresholds in the higher frequencies compared to adults. The difference between the high-frequency thresholds of post-term infants (mean age: 17 weeks) and adults did not reach significance. However, the authors acknowledged that these maturational differences may be overstated due to high ambient noise when testing the pre-term newborns. Subsequently, Small and Stapells (2008a) concluded that there is no significant difference between the 2000-Hz thresholds of young (mean age: 16 weeks) and older infants (mean age: 18 months). Consistent with previous ABR findings, infant thresholds are poorest at 2000 Hz, compared to the other frequencies, and approximately 6 dB worse than adult thresholds (Small & Stapells, 2008a). There are more limited data to draw from regarding maturational effects at 4000 Hz. For the tone-evoked ABR, Cone-Wesson and Ramirez (1997) reported newborn infant thresholds (N=60) that were 5 dB better than adult thresholds (N=3). As already discussed for the ASSR, Small & Stapells (2006) pooled the thresholds for 2000 and 4000 Hz and found that only the thresholds of the pre-term infants were significantly higher than those of adults. In a later and larger study, the authors again found no significant difference in the 4000-Hz thresholds between older infants and adults; yet young infants had significantly higher thresholds than the older infant subjects (Small & Stapells, 2008a). In contrast to their previous study, they also found that the 4000-Hz threshold was significantly better than the 2000-Hz threshold. The authors considered the findings to be a more reliable representation of the population than the previous study. It is clear that infants have frequency-dependent BC threshold differences, with better physiological thresholds in the low frequencies, poorer thresholds at 2000 Hz and similar thresholds at 4000 Hz compared to adults. Currently, infant BC ASSRs are considered to be normal in infants 0–11 months of age when present at the levels of 30, 20, 40 and 30 dB HL and in infants 12–24 months of age when present at the levels of 40, 20, 40 and 20 dB HL at 500, 1000, 2000 and 4000 Hz, respectively (Small & Stapells, 2008a). In contrast to the infant threshold pattern, adult BC ASSR thresholds do not differ between 1000, 2000 and 4000 Hz, and are actually poorer at 500 Hz compared to the higher frequencies (Small & Stapells, 2008a). These relatively recent findings are supported by previous studies on BC ASSR (Dimitrijevic et al., 2002; Jeng, Brown, Johnson, & Vander Werff, 2004; Lins et  13  al., 1996). However, Ishida, Cuthbert, and Stapells (2011) also investigated ASSR thresholds to boneconducted stimuli in adults and for one of their two studies, did not find the 500-Hz threshold to be significantly different than the thresholds at higher frequencies. Ishida et al. (2011) reported that the 500-Hz threshold is actually more variable across individuals. This finding may also account for the inconsistency across studies regarding the threshold at 500 Hz relative to thresholds at other frequencies.  1.2.3 Other Maturational Effects on Physiological Measures with Bone Conduction 1.2.3.1  Wave V of the ABR The ABR can be composed of a number of waves, each attributable to different  generator/generators, but wave V is the most easily identified in threshold assessments and is valuable because it indicates auditory responsiveness at the level of the brainstem. Maturational effects are also seen in wave V of the BC ABR. Infants have shorter latencies compared to adults for 500-Hz tones, and slightly longer latencies for 2000 Hz tones (Foxe & Stapells, 1993; Nousak & Stapells, 1992). The authors of these studies state that these shorter latencies indicate the low-frequency bone-conducted stimuli are more intense (by up to 17 dB) when presented to an infant versus an adult skull and thus elicit an earlier response from the cochlea. 1.2.3.2  Two-Channel Asymmetries Another maturational difference between infant and adult ABRs and ASSRs relates to two-  channel recordings, where the EEG channels both ipsilateral and contralateral to the stimulus ear are recorded. In adults, bone-conducted stimuli stimulate both cochleae (e.g., Studebaker, 1967); however, in infants, there are differences in the stimulation of each cochlea and these differences offer insight not available in the adult population (Foxe & Stapells, 1993; Stapells & Ruben, 1989; Stuart, et al., 1996; Small & Stapells, 2008b). These asymmetries can be used to identify which cochlea is responding to the stimulus (Foxe & Stapells, 1993; Stapells, 2000a; Stapells & Ruben, 1989). Adult BC ABRs have comparable wave V amplitudes and latencies between channels (Foxe & Stapells, 1993; Stapells, 2000a). In contrast, infant BC ABRs have shown that compared to the ipsilateral EEG channel, wave V is smaller in amplitude and has a longer latency in the contralateral EEG channel (Foxe & Stapells, 1993; Picton, Durieux-Smith, & Moran, 1994; Stapells & Ruben, 1989; Stapells & Oates, 1997; Stuart, Yang, & Botea, 1996). These infant asymmetries between the channels 14  have been shown to be useful in determining which cochlea is primarily responding to the stimulus (Foxe & Stapells, 1993; Stapells & Ruben, 1989; Stapells, 2000a). Small and Stapells (2008b) also identified asymmetries when recording two-channel ASSRs in infants and adults. The authors found that for bone-conducted stimuli, infants have significantly larger amplitude responses in the ipsilateral EEG channel compared to the contralateral channel across frequencies (500–4000 Hz). This amplitude difference between channels is considerably larger in infants compared to adults. It is suggested that these channel differences reflect the stimulation of each cochlea. In adults, there is less of a difference between the channels because both cochleae respond to a similar degree. In contrast, for infants, the cochlea ipsilateral to the stimulus is a much more dominant contributor to the response than the contralateral cochlea. The authors attribute these differences to immature neural pathways and skull structure. Skull structure and size specifically affects how efficiently bone-conducted sound can be transferred to the cochleae; as is the case with infants, the skull‘s small size, membranous structure and unfused sutures results in a larger interaural attenuation compared to adults (Anson & Donaldson, 1981; Eby & Nadol, 1986; Small & Stapells, 2008b). Small and Stapells (2008b) estimated the interaural attenuation in infants to be at least 10–30 dB, a value that decreases with maturation (Yang, Rupert, & Moushegian, 1987). In adults, BC interaural attenuation is commonly thought to be negligible and thus conservatively valued at 0 dB across frequencies, but may be as much as 15 dB at higher frequencies (Studebaker, 1967; Yacullo, 2009). Additionally, immaturities in the structures of the brainstem affect dipole orientation, which is thought to contribute to asymmetries between the channels (Edwards, Durieux-Smith, & Picton, 1985).  1.2.4 Procedural Variables of BC Testing Research conducted by Small, Hatton and Stapells (2007) addressed how the procedural variables that affect adult BC threshold estimation apply to infants, including coupling method and placement location. The bone oscillator must be applied to the skull with a certain and continuous amount of force to maintain a calibrated output. In adults and children, the most commonly accepted method of coupling the bone oscillator to the head is by using a steel headband. However, testing with a steel headband may be difficult due to factors such as head size, practicality and the infant‘s willingness to wear the headband. Small and colleagues (2007) have shown that in fact there are two other options for coupling the bone-oscillator to the infant skull that provide equivalent threshold estimations; these  15  options include hand-holding of the oscillator by a trained assistant or coupling of the oscillator using an elastic headband. In the current study, the coupling method for ASSR will invariably be hand-held and for VRA the method best tolerated by each child will be used. The location where the bone-oscillator is positioned also affects the intensity of the signal. For example, Durrant and Hyre (1993) provided evidence that the most efficient placement in adults is the mastoid of the test ear; whereas, the forehead was found to give higher thresholds. Studies with infants have also recommended temporal bone placement over the forehead (Small et al., 2007; Stuart, Yang, & Stenstrom, 1990; Yang et al., 1987). For ABR, Stuart et al. (1990) found that wave V response latencies varied across three bone-oscillator positions they tested, and as such, suggested that bone-oscillator placement remain consistent throughout testing. For ASSR, Small et al. (2007) investigated boneoscillator placement in pre-term infants and found no significant differences between the ASSRs with superior-posterior or posterior (i.e., mastoid bone) placement on the temporal bone relative to the ear canal. Thus, positions along the temporal bone provide equivalent and acceptable estimations of BC ASSR hearing thresholds in infants. For the current study, superior-posterior placement will be used for ASSR and either superior-posterior or posterior placement will be used for VRA, depending on the coupling method.  1.3 Air-Conduction Hearing AC thresholds indicate whether a hearing loss is present or not, and if so, to what degree. The following sections address the maturation of AC hearing when assessed using both physiological and behavioural measures.  1.3.1 Maturation of Air-Conduction Thresholds 1.3.1.1  Behavioural Appropriate behavioural methods differ with age and can have multiple procedural variations  that affect MRLs, which makes direct comparisons of auditory sensitivity across studies impractical. It is evident that auditory responses tend to improve with maturation becoming adult-like by 10 to 12 years of age (Northern & Downs, 1991; Werner, 1996). The remainder of this discussion will focus on the behavioural procedure used in the current study, VRA.  16  For VRA specifically, the literature asserts that air-conduction thresholds are substantially worse in infants than they are for pure-tone audiometry in adults; estimates of the infant-adult differences in behavioural thresholds range from 10 to 15 dB (Berg & Smith, 1983; Olsho, Koch, & Carter, 1988; Parry et al., 2003). These findings are consistent regardless of transducer: in the sound field (McDermott & Hodgson, 1982; Primus, 1988; Sabo, Paradise, Kurs-Lasky, & Smith, 2003; Trehub, Schneider, & Endman, 1980), with supra-aural headphones (Wilson & Moore, 1978 as cited in Parry et al., 2003; Berg & Smith, 1983; Nozza & Wilson, 1984; Nozza, 1995) or with insert earphones (Nozza, 1995; Nozza & Henson, 1999; Parry et al., 2003; Vander Werff et al., 2009). These studies also support that these infant-adult differences are greatest at the lower frequencies (500 and/or 1000 Hz). Additionally, in many of the ear-specific studies of VRA, researchers have found that infants exhibit frequency-dependent threshold differences, such that they have poorer thresholds in the lower frequencies (500 and/or 1000 Hz) compared to the higher frequencies (2000 and/or 4000 Hz). Parry et al. (2003) found considerable differences between the high and low frequencies with MRLs of 16, 13, 7 and 6 for 500, 1000, 2000 and 4000 Hz, respectively. Berg and Smith (1983) provide conflicting evidence, suggesting there is no significant difference between the 500 and 2000 Hz thresholds of infants (6 and 10 months of age). A recent study by Vander Werff et al. (2009) also did not find frequency-dependent threshold differences in infants who were approximately 10 months of age; however, the examiners of the study did not test below 10 dB HL. The MRLs/threshold differences between frequencies may be underestimated in the Vander Werff et al. (2009) study as Parry and colleagues (2003) found infant MRLs that ranged from -5 to 25 dB HL. Supporting the existence of frequency-dependent behavioural thresholds in infants, Franklin, Johnson, Smith-Olinde and Nicholson (2009) provide adult behavioural data that is comparable to Parry and colleagues (2003). The study presents mean threshold data for warbled tones from 250–8000 Hz and a general trend of decreasing threshold with increasing frequency is evident. The adult thresholds from 500–4000 Hz show that only the 500-Hz threshold is significantly different, and is in fact elevated compared to the thresholds at the other three frequencies. From the data provided by Franklin et al. (2009) and Parry et al. (2003), it appears that compared to adults, infants have better thresholds in the high frequencies and worse thresholds in the low frequencies. This apparent pattern of relatively poorer low-frequency and better high-frequency AC behavioural MRLs is the reverse of the pattern seen with BC behavioural MRLs (Hulecki & Small,  17  2011). However, it is consistent with the trend for physiological thresholds in infants, as discussed in the next section. 1.3.1.2  Physiological Although there is conflicting research, the majority of studies, particularly ones of the last  decade, suggest that infants show frequency-dependent differences in AC ABR thresholds in dB nHL (Balfour, Pillion, & Gaskin, 1998; Lee, Hsieh, Pan, & Hsu, 2007; Marcoux, 2011; Qian et al., 2010; Rance, Tomlin, & Rickards, 2006; Stapells et al., 1995; Stapells, 2000b; Vander Werff et al., 2009). The trend is that lower frequency thresholds (500 and 1000 Hz) tend to be significantly higher (i.e., poorer) than higher frequency thresholds (2000 and 4000 Hz). This trend is also seen in comparable adult data, as infant thresholds are actually very similar to adult thresholds when calibrated in dB nHL (Marcoux, 2011; Sininger, Abdala, & Cone-Wesson, 1997; Stapells, 2000b; Vander Werff et al., 2009; Werner, Folsom, & Mancl, 1993). This infant-adult similarity is surprising as infants undergo maturational changes that are known to affect the ABR. Infants actually require a greater sound pressure to elicit a threshold (Sininger et al., 1997). Additionally, the infant ear canal is smaller and has different acoustic properties than an adult canal. Measurements in the ear canal indicate that the same stimulus presented to an infant and an adult is more intense in the infant ear canal (Sininger et al., 1997). It may be that there is a trade-off between the maturational influences (i.e., ear canal volume and neural synchronization) on the ABR, such that the ABR thresholds remain relatively similar from infancy to adulthood (Sininger & Hyde, 2009). For example, when immature neural synchrony causes threshold elevation, greater sound pressure in the ear canal may compensate and vice versa. This is in agreement with the findings of Rance et al. (2006), that ABR thresholds (calibrated in dB nHL) do not change significantly in the first six weeks of life. However, when the ABR thresholds are calibrated in the ear canal (in dB peSPL, also called in situ SPL), Sininger et al. (1997) found that infant thresholds were elevated compared to adults at 1500, 4000 and 8000 Hz. Similarly, Marcoux et al. (2011) concluded that the trend of higher (i.e., worse) infant thresholds in dB peSPL relative to adults persists until approximately 4 months of age for 500 and 2000 Hz, and 6 months of age for 4000 Hz. It follows that when measured in the ear canal, infant thresholds improve with age.  18  Similar to the maturation of ABR thresholds, ASSR thresholds in normal-hearing infants improve with age, especially in the first few months of life (John, Brown, Muir, & Picton, 2004; Lins et al., 1996; Rance & Rickards, 2002; Rance et al., 2006). Rance and Rickards (2002) found that infant (mean age: 3 months) thresholds to single-stimuli ASSR were poorer compared to the thresholds of adults and older children published by Rance, Rickards, Cohen, De Vidi and Clark (1995). While Rance et al. (1995) and Rance and Rickards (2002) evoked ASSRs using a single stimulus presented in isolation, ASSRs can also be obtained by presenting multiple stimuli simultaneously (mASSR). Infant mean AC ASSR thresholds are also often elevated, although this can be frequency-dependent, in studies that used multiple simultaneous stimuli (Lins et al., 1996; Luts et al., 2006; Swanepoel & Steyn, 2005; Van Maanen and Stapells, 2009) compared to adult thresholds (e.g., Lins et al., 1996; Luts et al., 2006; Herdman & Stapells, 2001; Alaerts et al., 2010; D‘haenens et al., 2008). For example, Luts and colleagues (2006) compared infant (age < 3 months) and adult AC mASSR and revealed thresholds that were on average 11 dB worse across all frequencies in infants relative to adults. Threshold data of newborns also suggest higher (i.e., poorer) ASSR thresholds compared to older infants (Cone-Wesson, Parker, Siderski, & Rickards, 2002a; John et al., 2004; Luts et al., 2006; Rance & Tomlin, 2006). Rance and Tomlin (2006) suggest neural development is the most likely reason for differences between very young and older infants. Overall, it has been suggested that the differences between infants and adults may reflect multiple maturational changes, including: neural development, ear canal growth, and/or vernix in the ear canal (John et al., 2004; Luts et al., 2006; Rance et al., 2006). With these maturational changes, thresholds improve, amplitudes increase and the infant AC ASSRs gradually becomes more adult-like (John et al., 2004; Rance & Tomlin, 2006; Savio, Cardenas, Abalo, Gonzalez, & Valdes, 2001). Currently, infant AC ASSRs are considered within the normal range, in a wide age range of infants (mean age: 18.3 months, range: 0.7–66.2 months), when present at the levels of 50, 45, 40 and 40 dB HL at 500, 1000, 2000 and 4000 Hz, respectively (Van Maanen & Stapells, 2009, 2010). Van Maanen and Stapells (2009) investigated ASSR thresholds for 500–4000 Hz and found better infant thresholds with increasing carrier frequency. The worst thresholds were at 500 Hz and the best were at 4000 Hz. Ribeiro, Carvallo, and Marcoux (2010) assessed the same frequencies and also found that the AC ASSR thresholds of full-term infants were the worst at 500 Hz. In contrast to Van Maanen and Stapells (2009), Ribeiro et al. (2010) found the mean 4000-Hz threshold to be the second worst, with the  19  best thresholds being for 1000 and 2000 Hz. Other studies have reported no significant difference between thresholds at the 4000-Hz carrier frequency compared to 1000 and 2000 Hz (Lins et al., 1996; Luts et al., 2006; Rance et al., 2005). Despite disagreement at 4000 Hz, the literature does agree that AC mASSR thresholds for 500-Hz stimuli are much poorer compared to other frequencies (Lins et al., 1996; Luts et al., 2006; Ribeiro et al., 2010; Swanepoel & Steyn, 2005; Van Maanen & Stapells, 2009). In some studies, this elevation is also evident in the 500-Hz ASSRs of adults and older children (Dimitrijevic et al., 2002; Luts & Wouters, 2004; Perez-Abalo et al., 2001) and is consistent with ABR findings discussed above. As ASSR is calibrated in dB HL, which is referenced to normal hearing young adults, frequency-dependent differences are unexpected and thus may be attributable to non-sensory factors.  1.3.2 Other Maturational Effects on Physiological Measures with Air Conduction 1.3.2.1  ABR Latency Differences Numerous studies indicate that with maturation, two of the major peaks present in newborn  ABRs (waves III & V) exhibit a decrease in latency (e.g., Hecox & Galambos, 1974; Krumholz, Felix, Goldstein, & McKenzie, 1985; Mochizuki, Go, Ohkubo, & Motomura, 1983; Salamy, McKean, & Buda, 1975; Salamy & McKean, 1976; Starr, Amlie, Martin, & Sanders, 1977). These latencies decrease most rapidly in the first year of life, but continue to decrease at a slower rate until approximately three years of age (Fria & Doyle, 1984; Jiang, Zheng, Sun, & Liu, 1991; Moore, Ponton, Eggermont, Wu, & Huang, 1996; Salamy, 1984). This maturational effect, which is greatest for waves III and V, is a result of increased myelination and growth of the brainstem, structural changes which in turn improve conduction velocity and synaptic efficiency (Moore et al., 1996; Ponton, Moore, Eggermont, Wu, & Huang, 1994 as cited in Moore et al.1996). 1.3.2.2  Ipsi/Contra Asymmetries In adult AC ABRs, there are differences between the ipsilateral and contralateral EEG channels.  Compared to recordings of wave V in the ipsilateral channel, wave V of the contralateral channel has a slightly longer latency, occurring 0.1 ms later, and is approximately 70% of the amplitude (Picton, Stapells, & Campbell, 1981; Stapells & Mosseri, 1991). In infants, these differences are even more pronounced; responses in the ipsilateral channel have greater amplitudes and much shorter latencies than  20  in the contralateral channel (Edwards et al., 1985; Salamy, Eldredge, & Wakeley, 1985; Stapells & Mosseri, 1991). These infant-adult differences have been attributed to the dipole orientation that relates to the immaturity of infant brainstem structures and possibly to asymmetrical myelination of the ipsilateral and contralateral auditory pathways (Picton, 2011 p. 468; Edwards et al., 1985). These asymmetries between channels and between infants and adults also exist for two-channel AC ASSRs. In 2005, van der Reijden, Mens, and Snik reported that the infant single-stimulus ASSRs in the contralateral EEG channel were either smaller than in the ipsilateral channel or completely absent. Small and Stapells (2008b) used monotic ASSRs and also found that the responses in the contralateral EEG channel were approximately 33.3% and 66.7% the amplitude of the responses in the ipsilateral channel for infants and adults, respectively. Van Maanen and Stapells (2009) investigated dichotic mASSRs in infants and found that if ipsilateral responses are present, most often the contralateral responses are absent, which is consistent with the findings of van der Reijden et al. (2005). Because of the inconsistency in responses from the contralateral channel, it is currently recommended to not consider it in judging the presence of a response in the ipsilateral channel (Van Maanen & Stapells, 2009).  1.4 The Maturational Air-Bone Gap Before discussing the implications of comparing air- and bone-conduction thresholds in infants, it is helpful to recall what their relationship suggests in adults. As addressed previously, adult thresholds are referenced to normal young adult thresholds and expressed in dB HL. In adults, there is a well known relationship between the air- and bone-conduction responses. In the absence of a hearing loss, air- and bone-conduction thresholds will be within the normal range of hearing and differ by no more than 10 dB at any frequency. When AC thresholds are elevated, a hearing loss exists. It is sensorineural in nature if AC thresholds are near equivalent with that of BC. In certain circumstances, AC thresholds are elevated compared to BC creating and air-bone gap. In adults, an ABG is the difference between airand bone-conduction thresholds (dB HL) that quantifies the conductive component of a hearing loss (Lierle & Reger, 1946; Carhart, 1950). Due to maturational differences, the extent of the ABG‘s applicability to infants requires further investigation both physiologically and behaviourally.  21  Strong correlations between AC and BC ASSR and AC and BC behavioural thresholds suggest that the ASSR can likely accurately predict the ABG (see Comparing Threshold Estimates across Test Methods section). Jeng et al. (2004) and King (2007) directly assessed the ABG and they both found that, in adult subjects with simulated conductive hearing loss, the ASSR ABG was strongly correlated with the behavioural ABG. Despite this adult evidence, further research is required to directly assess how well the ABG is predicted in infants. Figure 1.1 compares four individual studies on ASSR (Herdman & Stapells, 2001; Ishida et al., 2011; Small & Stapells, 2006; Van Maanen & Stapells, 2009). As discussed previously, and illustrated in Figure 1.1, it is clear that normal-hearing infants present with different air- and bone-conduction threshold patterns than adults do when assessed using ASSR calibrated in adult dB HL. Specifically, BC thresholds are better in the low-frequencies compared to the higher frequencies; conversely, AC thresholds tend to be worse in the lower frequencies relative to the higher frequencies. Therefore, the relationship between air- and bone-conduction thresholds must be different between infants and adults.  Figure 1.1 Comparisons between air- and bone-conduction ASSR thresholds for both infants and adults between different studies. The figure shows mean data from the following four studies: (1) infant BC thresholds were reported by Small and Stapells (2006) for infants aged 0–8 months; (2) infant AC thresholds were published by Van Maanen & Stapells (2009) for infants aged 0.5–66.2 months; (3) adult AC thresholds (converted from dB SPL to dB HL) are from Herdman & Stapells (2001) and; (4) adult BC thresholds from Ishida et al. (2011) study B.  22  Vander Werff et al. (2009) assessed both AC and BC tone-evoked ABR at 500 and 2000 Hz in infants and adults; they found the familiar infant threshold pattern, but adults showed no threshold differences between frequencies or sound conduction mode. In the absence of a conductive component, normal hearing infants presented with almost a 15 dB ABG at 500 Hz, while adults had a nominal average ABG of 2.5 dB at the same frequency. In infants with true conductive hearing losses, the average ABG was even greater than that of the normal hearing infants. Similar studies, assessing both AC and BC ASSRs in the same infant, have not yet been conducted for the ASSR. The apparent ABG in normal hearing infants was termed the ―maturational‖ ABG by Hulecki and Small (2011). Hulecki and Small (2011) found that behaviourally infant BC thresholds show the same pattern evident in BC ASSR— thresholds are better in the low frequencies compared to the high frequencies. This is opposite to the trend for behavioural AC MRLs observed by Parry et al. (2003) — thresholds were worse in the low frequencies compared to the high frequencies. Given the differences between AC and BC VRA MRLs, Hulecki and Small (2011) concluded that the maturational ABG at low frequencies continues to be present behaviourally until at least 30 months of age. The authors emphasize that maturational influences confound infant ABGs, such that they are not necessarily indicative of conductive hearing impairment. These infant-adult differences in BC hearing sensitivity must be accounted for when interpreting infant audiological assessments. Clinically, the estimation of a conductive component, or lack of a conductive component, is important to direct future medical management and if needed, to provide an accurate prescription for amplification.  1.5 Accounting for Infant-Adult Differences There are a number of sensory and non-sensory immaturities that might account for the differences between infant and adult auditory thresholds measured (Nozza, 1995; Parry et al., 2003; Werner et al., 1993). These factors also help explain the threshold differences between behavioural and physiological methods. Sensory factors reflect the development of the auditory pathway up to and including the brainstem and consist of anatomical changes, such as skull growth and middle-ear changes, that affect sound transmission to the cochlea (Nousak & Stapells, 1992; Werner et al., 1993). Non-sensory factors include any other possible influences, such as test or procedural variables and more central processes (e.g., attention and motivation) (Nozza, 1995; Parry et al., 2003). This section will primarily address sensory factors as they relate to the development of the outer/middle ear, cochlea,  23  brainstem and skull. It is important to keep in mind that each sensory factor influences air- and boneconduction thresholds to a different degree.  1.5.1 Outer/Middle Ear Saunders, Doan and Cohen (1993) concluded that changes in the middle ear with development affect auditory sensitivity across species, including humans. Both the external auditory canal and middle ear undergo significant anatomical changes in the first few years of life and continue to mature into adulthood (Keefe, Bulen, Arehart, & Burns, 1993). Thus, as part of the auditory pathway, these maturing elements of the outer and middle ear in infants and children have the potential to alter the acoustic properties and efficiency of sound transmission to the cochlea (Saunders, Kaltenback, & Relkin, 1983). Infant-adult differences observed with measures of tympanometry (e.g., Alaerts, Luts, & Wouters, 2007; Holte, Margolis, & Cavanaugh, 1991; Paradise, Smith, & Bluestone, 1976) and wideband reflectance (e.g., Keefe et al., 1993; Keefe, Bulen, Campbell, & Burns, 1994; Keefe & Levi, 1996) are attributed to the growth in size of the outer and middle ear as well as the compliance of structural components. Developmental changes to the outer ear occur until approximately seven years of age and include thickening of the ear canal walls and increased length and diameter of the ear canal (Northern & Downs, 1991; Saunders et al., 1983). The tympanic membrane undergoes maturational changes until approximately two years of age, becoming thinner and due to changes in the temporal bone, changing orientation (Anson & Donaldson, 1981; Eby & Nadol, 1986; Saunders et al., 1983). Also attributable to temporal bone growth, is the enlargement of the middle-ear cavity, which is 1.5 times smaller in an infant less than one year of age than it is in an adult (Ikui, Sando, Haginomori, & Sudo, 2000; Saunders et al., 1983). In a study of the development of the temporal bone, Eby and Nadol (1986) found that the middle ear primarily grows in length post-natally. This change in the middle-ear cavity volume also corresponds to increased compliance of the tympanic membrane (Saunders et al., 1983). With development, there are also ossicular changes relating to joints, orientation and composition (Anson & Donaldson, 1981; Saunders et al., 1983). Developmental tissue called mesenchyme may also be present in the middle ear cavity of infants and normally disappears by around the end of the 1st year of life (Takahara, Sando, Hashida, & Shibahara, 1986). These changes decrease the mass, and thus the resistance, of the middle ear system (Saunders et al., 1983). Keefe et al. (1993) assessed the efficiency of the energy flow through the outer and middle ear in groups of infants from 1 to 24 months of age as well  24  as adults. The authors found that due to structural differences within the middle ear, compared to adults, infants have lower middle-ear compliance and higher middle-ear resistance, which decreases the efficiency of low-frequency sound transmission to the cochlea. As the infant middle ear is massdominated, it also exhibits a lower resonant frequency than an adult middle ear (Holte et al., 1991). This is in contrast to the ear canal which exhibits a higher resonant frequency in infants than adults (Kruger & Ruben, 1987; Kruger, 1987). These resonant properties may partially account for the frequencydependent differences observed between the auditory thresholds of infants and adults, as some frequencies will be more effectively transmitted to the cochlea than others. Due to ear-canal size (Feigin, Kopun, Stelmachowicz, & Gorga, 1989; Jirsa & Norris, 1978) and resonant properties, for frequencies greater than approximately 2000 Hz (Keefe et al., 1993), an identical auditory stimulus will reach a much greater level at the tympanic membrane of an infant relative to that of an adult (Kruger & Ruben, 1987). Differences between the sound pressure at the tympanic membrane and the calibrated stimulus level complicate the interpretation of threshold measurements. By measuring sound pressure at the tympanic membrane as peak-equivalent SPL (dB peSPL) the contributions of the external ear can be removed from threshold estimation. However, even when the stimuli are calibrated with this measurement, infants still show auditory immaturity physiologically, especially at higher frequencies (Rance & Tomlin, 2006; Sininger et al., 1997). Despite clear maturational changes that occur in the outer and middle ear, there is doubt that these changes can fully account for the variance between infant and adult auditory thresholds (Rance & Tomlin, 2006; Sininger et al., 1997). Additionally, although maturation of the outer and middle ear feasibly contribute to infant-adult threshold differences to air-conducted stimuli, the outer ear itself, or osseotympanic pathway, is not a significant pathway in infant BC hearing (Small et al., 2007; Small & Hu, 2011) and thus cannot fully explain infant-adult differences to bone-conducted stimuli (Rance & Tomlin, 2006; Sininger et al., 1997; Stenfelt & Goode, 2005).  1.5.2 Cochlea Much of the research on cochlear maturation indicates that it is largely mature by full-term birth, both structurally and functionally (Abdala & Sininger, 1996; Abdala, Sininger, Ekelid, & Zeng, 1996; Abdala, 2000; Bredberg, 1968; Lavigne-Rebillard & Pujol, 1987; Lavigne-Rebillard & Pujol, 1988;  25  Lavigne-Rebillard & Bagger-Sjöbäck, 1992; Pujol, 1985; Pujol, Lavigne-Rebillard, & Uziel, 1991). Such early development of auditory function is also evidenced by the presence of auditory evoked potentials as early as the 25th fetal week (Starr et al., 1977). However, recent research on the differences between infant and adult DPOAEs suggests that while much of the difference is a consequence of the conductive auditory system, slight immaturities in the newborn cochlea may account for a portion of the difference, but have yet to be fully described (e.g., Abdala & Keefe, 2006; Abdala & Dhar, 2010; Dhar & Abdala, 2007). Despite the fact the cochlea is mostly mature, as it relates to auditory threshold, the auditory system central to the cochlea is still developing and is a more probable source of the higher thresholds seen in infants relative to adults.  1.5.3 Brainstem The neural pathways within the auditory brainstem continue to develop post-natally, as evidenced by research involving physiological assessment. Werner, Folsom, and Mancl (1994) compared ABR latencies for wave V and identified significant correlation with behavioural thresholds for 3-month-old infants, such that longer wave V latencies corresponded to high behavioural thresholds. This relationship was not found for 6-month-old infants or adults. These results support that brainstem immaturity is a common contributing factor to both physiological and behavioural testing methods in young infants, but perhaps not in older ones. The authors conclude that behaviourally, the combination of non-sensory factors, as well as sensory factors that are peripheral and central to the brainstem must explain infant-adult threshold differences. Further research involving the ABR helps define the maturational changes occurring to the neurons within the brainstem. For example, ABR latencies may be affected by myelination, which begins with the cochlear nerve at approximately 27 weeks of gestation and continues until at least one year of age (Moore, Perazzo, & Braun, 1995; Moore & Linthicum, 2001). Additionally, in modelling the structures and neural pathways associated with the waves of ABR, researchers Ponton, Moore and Eggermont (1996) suggested that into the second year of life, the synapses of the cochlear nuclei also mature and affect ABR latencies. Increased myelination and development of neural synapses could both feasibly increase the efficiency with which a sound can travel through the auditory pathway. Because the detection of physiological responses is dependent on the synchronous firing of neurons in response to a stimulus, some studies also suggest elevated thresholds may be due to immature neural synchrony (Eggermont &  26  Salamy, 1988; Gorga, Kaminski, Beauchaine, Jesteadt, & Neely, 1989; Rance et al., 1995; Sininger et al., 1997).  1.5.4 Skull The infant skull is composed of a number of smaller unfused bones that are held together by fibrous sutures, which close by approximately one year of age (Anson & Donaldson, 1981). As the infant develops, the skull grows in size and the junctions between multiple bones become more rigid when the connecting tissue is replaced by bone (Pierce & Mainen, 1977). The temporal bone itself continues to grow into adolescence (Eby & Nadol, 1986). The transmission of bone-conducted stimuli is affected by these maturing skull characteristics. For example, Stuart, Yang and Stenstrom (1990) posited that bone-conducted stimuli are more efficient on the infant skull because the soft sutures connecting the cranial bones attenuate the energy transfer between the bones. This lack of energy dispersion to other cranial bones may result in a bone-conducted stimuli being more intense when coupled to an infant skull relative to an adult skull. The authors also suggest that the relative intensity of a bone-conducted stimulus may also be greater given the fact that the infant temporal bone is no more than half the size it is in an adult. However, these properties of the infant skull suggest that stimuli across the frequency range should be more intense on the infant skull than an adult skull. The skull does not seem to account for the frequency-dependent threshold differences seen in infants, where low frequencies tend to elicit lower thresholds than higher frequencies do.  1.5.5 Non-Sensory Factors Non-sensory factors, also called nuisance variables, apply to both behavioural and physiological testing and can relate to the variability caused by internal noise and equipment and procedural variables (Werner et al., 1993), which are discussed in a later physiological section. Due to the nature of behavioural testing, non-sensory effects on threshold estimation are especially important to consider (Nozza, 1995). Behavioural responses in children may be influenced by limitations on attention span, interest in the task, and behavioural state. However, the degree to which non-sensory and sensory factors each contribute to infant-adult differences is not well understood. Nozza (1995) and Nozza and Henson  27  (1999) compared behavioural thresholds of infants, with an age range of approximately 7–11 months, and adults. They determined that infant-adult differences at 500, 1000, and 2000 Hz are primarily attributable to sensory factors. Non-sensory factors were found to only explain approximately 4 dB of variance, independent of frequency and the amount of infant-adult difference, which ranged from 5–14 dB across frequency. In contrast, Werner et al. (1993) found young infants and adults had similar ABR thresholds to 1000-, 4000-, and 8000-Hz tonal stimuli, despite elevated behavioural results in infants compared to adults. This finding suggests that sensory factors play only a minor role in infant-adult differences, and thus non-sensory factors must be a major contributor to behavioural response disparities. Even though the degree of their influence remains inconclusive, it is clear that both nonsensory and sensory factors influence the behavioural responses of infants, as neither alone can fully account for the variance found between infants and adults. These factors are also relevant in considering threshold differences across testing methods.  1.6 Comparing Threshold Estimates across Test Methods The value in objective physiological tests in infants is their ability to accurately predict behavioural hearing thresholds in infants who are too young or can otherwise not be tested behaviourally. Defining the hearing status of these infants is critical in meeting the objectives of newborn hearing screening programs. In defining hearing status, it is important to recognize that behavioural measures of threshold are estimated from and not directly equivalent to physiological thresholds. Picton, Dimitrijevic, Perez-Abalo and Van Roon (2005) recognize that differences can stem from the following: (1) methodological variations involving the stimulus type and frequency, as well as the duration of the recording; and (2) subject factors such as age and degree of hearing loss. Physiological thresholds are also inherently different than perceptual thresholds and are often elevated in comparison by 10 dB or more (Picton et al., 2005; Stapells, 2010a). This may be due to the involvement of different neuronal pathways and/or the physiological measures‘ inadequacies in detecting nearthreshold responses (Picton et al., 2003). Identifying consistent differences between physiological and behavioural thresholds allows one to estimate the behavioural hearing level (eHL) either by the use of regression equations (e.g., Rance et al., 2005) or the subtraction of correction factors (e.g., BCEHP, 2008; OIHP, 2008). Both of these methods for reaching eHL are established by research comparing the behavioural and physiological thresholds for large groups of individuals with normal hearing and  28  different types of hearing loss. Current estimated eHL correction factors are summarized in Table 1.1 below, although more conservative (i.e., larger) values may be used for ABR (Stapells, 2010a; Hatton et al., submitted). The correction factors for the ASSR are preliminary and are thus provided as a range of values from least to most conservative. For example, if the AC ASSR threshold at 500 Hz is 45 dB nHL, the eHL would be ASSR threshold (45 dB nHL) minus the correction value (e.g. less conservative value of 10 dB), which is 35 dB eHL. Table 1.1 Threshold Correction Factors (in dB) for Infant Tone-evoked ABR and ASSR. Adapted from Stapells, 2010a.  ABR ASSR  500 Hz AC 10 10-20  BC na na  1000 Hz AC 5 10-15  BC na na  2000 Hz AC 0 10-15  BC 0* na  4000 Hz AC -5 5-15  BC na na  Note: ‗na‘ indicates that a correction factor is not available *This correction factor was posited by Hatton et al. (submitted).  The high correlation between behavioural and ABR measures is well established and is evidenced by its current use as a gold-standard hearing assessment technique. Many studies in infants and young children with normal hearing and hearing loss have found high correlations (r>0.86) between AC ABR and VRA (e.g., Lee et al., 2007; Stapells, 2000b; Vander Werff et al., 2009; Rodrigues & Lewis, 2010). Typically, AC ABR thresholds are within 10 dB of AC VRA thresholds (Stapells, 2010a). Stueve and O‘Rourke (2003) studied young children and infants with hearing loss and also found strong positive correlations between physiological (both the tone-evoked ABR and the ASSR) and behavioural (VRA) AC thresholds. With research involving ASSRs in infants, correlations can be made with the tone-evoked ABR or VRA (or other behavioural measures), as they are both gold standards for hearing evaluation in infants. There are high correlations between frequency-specific infant behavioural thresholds and ASSR thresholds to single-stimuli (Rance et al., 1995; Rance & Rickards, 2002; Rance et al., 2005; Stueve & O'Rourke, 2003) and multiple-stimuli (Chou et al., 2012; Han et al., 2006; Luts et al., 2006; Rodrigues & Lewis, 2010). However, it is advantageous in research to compare the ASSR with the tone-evoked ABR instead, as there is the potential for both tests to be conducted in the same day, which controls for the variables of age and hearing status (Van Maanen & Stapells, 2010). Rance et al. (2006) investigated physiological thresholds to 500 and 4000 Hz stimuli in very young infants and found that, although thresholds of the two methods were highly correlated, the ASSR 29  thresholds (calibrated in dB HL) were higher than the ABR thresholds (calibrated in dB nHL). This difference between the two methods is also captured in their recommended normal maximum levels, for which ASSRs are expected to be up to 10-15 dB higher than the tone-evoked ABR (BCEHP, 2008; Stapells, 2010a; Van Maanen & Stapells, 2009). Both Rodrigues and Lewis (2010) and Van Maanen and Stapells (2010) found that thresholds estimated using mASSRs and tone-evoked ABR were highly correlated (r= 0.76 to 0.97) in infants with SNHL. Compared to the tone-evoked ABR, there are many fewer studies on the ASSR, especially involving infants with hearing loss. Until further data are obtained to inform estimates of behavioural hearing level, the ASSR can only be used as a screening tool or supplement to the tone-evoked ABR (Stapells et al., 2005). Criteria for normal BC levels for the tone-evoked ABR at 500 and 2000 Hz were posited by Stapells and Ruben (1989). Hatton et al. (submitted) confirmed that these normal levels identified both normal and elevated BC thresholds in children with conductive and sensorineural hearing loss with 97 and 98% accuracy at 500 and 2000 Hz, respectively. For BC ASSR normal levels have been posited for 500, 1000, 2000 and 4000 Hz (Small & Stapells, 2008a). As is the case with AC, thresholds tend to be higher for the ASSR than the tone-evoked ABR. Because there are few BC studies in general, there is limited understanding of the relationships between the tone-evoked ABR, the ASSR, and VRA to boneconducted stimuli. This is an area requiring future research, especially in children with hearing loss. Because of this limitation, there are currently no eHL correction values for BC ASSRs.  1.7 The Auditory Steady-State Response In 1981, research by Galambos, Makeig and Talmachoff demonstrated that the 40-Hz auditory evoked potential was a promising technique to obtain auditory thresholds in adults. The evoked potentials they recorded are now more commonly known as the ASSR. The ASSR can be evoked by any periodic stimulus, such as repeating transient stimuli or continuous amplitude- and/or frequencymodulated tonal stimuli. Recording an ASSR involves presenting the stimuli at such a rate that the response to each individual stimulus overlaps with the response to the previous stimulus, evoking a sustained or steady-state response from the brain (reviewed in Picton et al., 2003). This sustained response has frequency components that remain constant over time in regards to amplitude and phase, and is thus most appropriately analyzed in the frequency-domain (Regan, 1989 p. 35). The response is assessed at the carrier frequency‘s stimulus rate.  30  1.7.1 Rate/EEG The frequency at which a stimulus‘ spectral energy is centred is called the carrier frequency. The rate, also called the modulation frequency indicates how often (i.e., how many times per second) that stimulus repeats or varies in its amplitude and/or frequency. As noted by Picton et al. (2003), the rate affects the response in predictable ways. It is a very important variable in the detection of ASSRs as it affects the signal-to-noise ratio: both the amplitude of the response as well as the background (EEG) noise. With increasing rate, the amplitude of a response generally decreases. Although in some regions, specifically around 40 Hz (Galambos, Makeig, & Talmachoff, 1981; Stapells, Linden, Suffield, Hamel, & Picton, 1984) and 80 (Aoyagi et al., 1993) or 90 Hz (Cohen, Rickards, & Clark, 1991), the amplitude is actually enhanced; however, the higher rate still elicits smaller responses relative to 40 Hz. A caveat of this is that the enhancement varies with age (see Factors of Age and Sleep section). This enhancement increases the detectability of a response; therefore, using rates in these regions is common. ASSRs are often named from the rate used to evoke them, such as the 40-Hz response, which suggests rates from 30-50 Hz, and the 80-Hz response, which corresponds to the range of 70-110 Hz (Stapells et al., 2005). Another important effect of rate is that as it increases, the EEG noise decreases such that at a 40Hz modulation frequency the EEG noise will be greater than it is at an 80-Hz modulation frequency (Cohen et al., 1991). As mentioned previously, the EEG represents the electrical activity of the brain (Picton et al., 2003). However, it also reflects other electrical noise extrinsic and intrinsic to the subject (Cone & Dimitrijevic, 2009). Two of the largest contributors to EEG noise are electromagnetic interference and biological noise (e.g., the subject‘s alpha activity, muscle movement and vocalizations) (Regan, 1989). Although the 80-Hz response is smaller in amplitude than responses to lower modulation rates, the lower EEG noise at this frequency allows its smaller responses to remain detectable (Picton et al., 2003).  1.7.2 ASSR Generators It is important to know what the source of a response is to fully understand the reasons and implications behind its presence or absence (Herdman et al., 2002a). Studies in both humans and animals have together identified multiple generators for the ASSR (Herdman et al., 2002a; Johnson, Weinberg, Ribary, Cheyne, & Ancill, 1988; Kuwada et al., 2002). Subcortical and cortical areas are  31  involved to different extents, depending on the stimulus rate (Herdman et al., 2002a; Kuwada et al., 2002). The brainstem is active across a large range of stimulus rates and is the dominant generator for the 80-Hz response (Herdman et al., 2002a; John & Picton, 2000b; Kuwada et al., 2002). Further research to isolate the specific generators within the brainstem suggests the inferior colliculus and cochlear nucleus contribute to the 40-Hz and 80-Hz responses respectively (Herdman et al., 2002a; Kuwada et al., 2002). The cortex is thought to be the main source of slower stimulus rate responses, specifically the 40-Hz ASSR. Although it is possible that other generators are involved to a smaller extent, including the brainstem (Herdman et al., 2002a; Kuwada et al., 2002) and possibly the thalamus (Johnson et al., 1988).  1.7.3 Factors of Age and Sleep In infants, the response amplitude is enhanced at stimulus rates around 80 Hz, but in contrast to adults, it is not enhanced at 40 Hz; thus the 40 Hz response cannot be recorded reliably in infants (Levi, Folsom, & Dobie, 1993; Levi, Folsom, & Dobie, 1995; Maurizi et al., 1990; Stapells, Galambos, Costello, & Makeig, 1988; Suzuki & Kobayashi, 1984). Picton and colleagues (2003) review two reasons for why detection is difficult using a 40-Hz stimulus rate. Firstly, it is believed that the main generator of the 40-Hz response, the cortex, is immature and unable to maintain a continual response to steady-state stimuli at such a fast rate. Thus for infants, enhancement around a 40-Hz stimulus rate is not physiologically possible. Secondly, it is much more realistic to obtain recordings in an infant who is asleep, yet sleep diminishes the amplitude of the 40-Hz response at all ages, which makes response detection even more difficult. The 80-Hz response, however, can be reliably recorded at all ages; although responses in young infants generally have lower amplitudes, but similar latencies compared to adults (Lins et al., 1996; Pethe, Muhler, Siewert, & von Specht, 2004). Additionally, the 80-Hz response is not significantly affected by sleep compared to the lower stimulus rates (e.g., 40 Hz) (Cohen et al., 1991) and has been used successfully in many previous studies to assess thresholds in normal-hearing infants (e.g., Cone-Wesson et al., 2002a; Cone-Wesson, Dowell, Tomlin, Rance & Ming, 2002b; John et al., 2004; Lins et al., 1996; Luts et al., 2006; Rance et al., 2005; Rance & Tomlin, 2006; Ribeiro et al.,  32  2010; Savio et al., 2001; Small et al., 2007; Small & Stapells, 2006, 2008a, 2008b; Van Maanen & Stapells, 2009). To obtain reliable infant ASSRs in the present study, stimulus rates ranged from 78 to 101 Hz across carrier frequencies.  1.7.4 Stimuli In general, the carrier frequency specifies the region(s) of the basilar membrane that will be activated by the stimulus (Lins & Picton, 1995). The frequency specificity of a stimulus varies depending on the type of stimulus used and there is a trade-off between the frequency specificity of a stimulus and the amplitude of the response that it elicits. For example, a click stimulus contains energy across a wider frequency range than does a brief tone and thus it activates more of the basilar membrane. Greater stimulation of the basilar membrane results in larger response amplitudes. Although larger responses enable faster detection, some frequency specificity is sacrificed (Picton et al., 2003). For threshold estimation purposes, it is preferred to have a more frequency-specific stimulus to obtain responses that better reflect auditory sensitivity for specific frequencies, which provides greater accuracy in fitting amplification (JCIH, 2007). Some stimuli that have been used to evoke ASSRs include clicks, brief tones, sinusoidally amplitude-modulated (AM) tones and noise, exponentially amplitude-modulated (AM2) tones, frequency-modulated (FM) tones, and mixed-modulated (MM) tones, which combine amplitude- and frequency-modulation. ASSR studies commonly use modulated tones of some sort because of their desirable frequency specificity (Herdman, Picton, & Stapells, 2002b; Picton et al., 2003). AM tones are frequency-specific with their acoustic spectra showing energy at only three frequencies: the carrier frequency and two sidebands, which deviate from the carrier frequency by plus and minus the modulation frequency (i.e., stimulus rate) (John, Dimitrijevic, van Roon, & Picton, 2001a; Picton et al., 2003). There is evidence that MM and AM2 stimuli evoke larger responses than AM stimuli in both adults and infants (Cohen et al., 1991; John, Dimitrijevic, & Picton, 2002a; John et al., 2004; Wood, 2009). It has been suggested that combining MM and AM2 stimuli could provide the greatest increase in amplitude (John et al., 2001a; John et al., 2002a; John et al., 2004). D‘Haenens et al. (2007) tested this theory in adults and found that response amplitudes were significantly larger for MM/AM2 stimuli than MM stimuli alone. However, since the authors did not compare these amplitudes with responses elicited  33  by AM2 alone, the study does not reveal whether combining MM and AM2 provides further amplitude enhancements than that provided by AM2. A MM stimulus is thought to combine the separate responses of its AM and FM components, thus eliciting a larger response than either an AM or FM stimulus could evoke on its own, although only when their responses overlap (John et al., 2001a). This means that to obtain the maximal amplitude response, the phases of the stimuli must be adjusted so that the amplitude and frequency maxima occur simultaneously (Cohen et al., 1991). John et al. (2002a) identify this as a weakness of MM stimuli compared to AM2 stimuli, as it makes setting them up more difficult and requires further research regarding the optimal relative phases across age groups. On the other hand, an AM tone that has been modified to follow an exponential envelope (e.g., AM2) decreases the duration of the tone, which results in a decreased rise time (i.e., increases the rise slope) (John et al., 2002a). When a neuron is stimulated by a steady-state stimulus, it is most sensitive to the rise time of the stimulus; a steep rise slope will result in improved synchrony and thus evoke larger amplitude responses than an AM stimulus (John et al., 2002a, 2004). This is thought to be the main mechanism responsible for the amplitude enhancement (John et al., 2002a). Response amplitudes are only optimal for exponential stimuli that are AM2, as it has been shown that larger exponents result in much poorer frequency specificity with only slightly enhanced amplitude (John et al., 2002a). Although a faster rise time does decrease frequency specificity relative to an AM stimulus (Picton et al., 2003), AM2 stimuli still have better frequency specificity than MM stimuli, as shown in Figure 1.2. As investigated by Mo and Stapells (2008), brief-tone stimuli can also be shortened to enhance response amplitudes; however, the authors found that for mASSR amplitudes significantly increase only for very brief tones at 2000 Hz and not at 500 Hz. They concluded that due to the reduced frequency specificity and resulting interactions, brief-tone stimuli may not be optimal for threshold estimation using the mASSR. Frequency specificity and response amplitude also have implications for determining the relative efficiency of presenting a single versus a multiple stimulus when eliciting the ASSR.  34  Figure 1.2 Acoustic spectra of AM/FM and AM2 stimuli for the carrier frequencies of 500, 1000, 2000, and 4000 Hz. Y-axis ticks represent 20 dB intervals (Adapted with permission from Wood, 2010).  1.7.4.1  Single vs. Multiple Stimuli When an ASSR is recorded to a single stimulus at one stimulus rate, the response is also  measured at that stimulus rate (or modulation frequency) in the frequency domain. This type of response detection makes it possible to present multiple stimuli at once, where each carrier frequency is modulated at a different rate and the corresponding response is also measured at that rate (i.e., mASSR; Picton, Skinner, Champagne, Kellett, & Maiste, 1987; Regan & Regan, 1988). This feature provides ASSR the potential to obtain information about an individual‘s auditory sensitivity faster than a method that assesses only one frequency at a time, such as the ABR (Herdman & Stapells, 2001; John, Lins, Boucher, & Picton, 1998; Lins & Picton, 1995). However, mASSR may or may not improve efficiency depending on the degree of interactions between the individual responses. An interaction exists when the amplitude of a response is smaller when there are multiple-simultaneous stimuli compared to a single stimulus due to processes of the auditory pathway (Picton et al., 2003). For the use of simultaneous multiple stimuli to be advantageous compared to multiple single stimulus presentations, it is important that reductions of the response amplitudes are less than 1/√N, where N is the number of stimuli  35  presented simultaneously (John et al., 1998). This is because the time advantage provided by testing with simultaneous multiple stimuli is reduced as the response amplitudes decrease and more time is needed for the EEG noise to be reduced by averaging enough to detect the response (John et al., 1998). John et al. (1998) found that the degree of interaction for the mASSR using lower modulation frequencies was much greater than for the 80-Hz mASSR. The focus of this study is the 80-Hz mASSR, for which it is known that the degree of stimulus interaction changes with stimulus level and type (Herdman & Stapells, 2001; John et al., 1998; Lins & Picton, 1995; Wood, 2009). In adults, ASSRs to multiple simultaneous AM stimuli presented at or above 73 dB SPL show significant decreases in response amplitudes at some carrier frequencies, partially due to masking (John et al., 1998; Picton, van Roon, & John, 2009). No significant amplitude reductions occur with multiple AM or AM2 stimuli presented at or below approximately 60 dB SPL (Herdman & Stapells, 2001; John et al., 1998; Lins & Picton, 1995; Picton, van Roon, & John, 2009; Wood, 2009). In contrast, for AM/FM stimuli, Wood (2009) found ASSR amplitudes were reduced for the multiple versus single stimulus conditions at both 60 and 80 dB HL; despite this finding, the multiple stimuli condition remained at least as efficient as the single stimulus condition. The findings reported for the lower intensities suggest that at near-threshold levels the use of multiple stimuli is at least equal to or more efficient than single stimuli. Some adult studies have also revealed that mASSR efficiency is decreased with certain hearing loss configurations (Herdman & Stapells, 2003; John, Purcell, Dimitrijevic, & Picton, 2002b). Hatton and Stapells (2011) compared single versus multiple simultaneous AM stimuli in normalhearing infants (6-38 weeks of age) and found that, unlike adults, infants display significant amplitude reductions at a suprathreshold level (60 dB SPL) as the number of stimuli increases. However, the study suggests that the mASSR is still more efficient compared to the single-stimulus ASSR. Additionally, there were no significant amplitude differences in the ASSRs between monotic multiple stimuli and dichotic multiple stimuli. Given the amplitude similarities, the dichotic-multiple condition, which allows assessment of both ears at once, was found to be more efficient than the monotic-multiple condition. A second experiment in the study also assessed normal-hearing infants and compared the 500-Hz threshold and amplitude, as well as the EEG noise in the single-stimulus condition and the dichotic-multiple condition. Mean thresholds were greater than 50 dB SPL and there were no significant differences in thresholds or amplitudes found between the stimulus conditions. However, the EEG noise was greater in  36  the dichotic multiple condition. This finding suggests that multiple stimuli may also be more efficient than a single stimulus at threshold levels, at least at 500 Hz. The authors caution that there is still little known about mASSR interactions at threshold levels in infants with either normal hearing or hearing loss. Efficiency also varies with type of stimuli. Wood (2009) found that with AM and AM2 stimuli, the mASSR was more efficient than the single-stimulus ASSR. Stimuli that were less frequencyspecific, such as MM stimuli, had greater interactions and were less efficient compared to more frequency-specific stimuli. Additionally for mASSRs, AM2 stimuli elicited significantly larger response amplitudes than either AM or MM stimuli at 60 dB HL. Given the discussion of stimuli type and multiple versus single stimulus ASSR, the current study used AM2 stimuli to elicit the mASSR for optimal efficiency. Despite the theoretical advantages of AM2 stimuli, there are only a few studies for adults (D‘haenens et al., 2007; John et al., 2001a; John et al., 2002a, 2003; Wood, 2009) and only one study for infants (John et al., 2004) to support its use for mASSR. No studies have assessed its use in subjects with hearing loss. However, for their research using the Intelligent Hearing Systems SmartEP-ASSR, Van Maanen and Stapells (2009, 2010) created cosine3-windowed sinusoidal stimuli, which closely resemble continuous stimuli and are similar in frequency specificity to AM2 stimuli. Their research investigated AC mASSR in infants and children with normal-hearing and hearing loss. The present study will provide additional AM2 threshold data for both infants and adults with normal hearing. 1.7.4.2  Stimuli and ASSR Systems Clinically, the recommended use of different types of stimuli can pose problems. Firstly, the  equipment available must be capable of testing with that stimulus. There are a number of systems available for recording multiple ASSRs (for comments on equipment see Stapells, 2010b); however, these systems differ in flexibility and may or may not be able to accommodate a research-recommended stimulus or a desired protocol (for other concerns regarding ASSR systems see Stapells et al., 2005). The Bio-logic Systems MASTER system has a number of stimulus options, including AM, AM/FM, AM2 for both bone- and air-conduction. The Intelligent Hearing Systems (IHS) SmartEP system also can assess bone- and air-conduction and although it uses brief tones by default, the system is flexible and the  37  stimuli can be modified (Intelligent Hearing Systems, 2011). The Interacoustic ASSR system uses only AC narrowband chirps (Interacoustics A/S, n.d.). GN Otometrics offers AM and AM/FM tone-burst stimuli through air- or bone-conduction, but uses a relatively new detection algorithm that has not been thoroughly studied (GN Otometrics, 2009; Stapells, 2010b). Secondly, there must be sufficient evidence to support the use of a certain stimulus when testing the population of interest. Interpretation of research findings is not straightforward because of the added variability introduced by the specifications of the ASSR equipment used to conduct the research. As a result, comparing results across studies is an ongoing challenge. It follows that research findings must be interpreted cautiously, as they may or may not be generalizable across ASSR systems.  1.7.5 Reducing EEG Noise The detection of a response depends on a favourable signal-to-noise ratio (SNR), which may be improved by an increase in the amplitude of the response and/or a reduction in noise. The amplitude of a response can be enhanced by certain modulation rates, types of stimuli and greater amplitude modulation depths, but also by increasing the stimulus intensity (Galambos et al., 1981; Stapells et al., 1984) and using the optimal electrode montage (see Electrode Montage section). Regan (1989) points out that, ―It is noise (defined as undesired signal) and not the maximum available amplification that determines how small a signal can be detected‖ (p. 30). There are three well-known techniques used to reduce noise and thereby enhance the SNR: averaging, weighted-averaging, and artifact refection. An ASSR recording is made up of a number of shorter recordings, sweeps, which are the targets for averaging; sweeps can be further segmented into ―epochs‖ (John & Picton, 2000a). Averaging relies on two assumptions: that the signal is time-locked to the stimulus and the noise is essentially random throughout a recording (Glaser & Ruchkin, 1976; Picton et al., 2003). Given these assumptions, averaging over time will reduce the amplitude of the noise and have little effect on the amplitude of the signal—thus longer recording times provide smaller amplitude noise and greater SNRs. If the noise levels vary between sweeps, averaging on its own becomes less effective. Weightedaveraging (Dobie & Wilson, 1994; Lütkenhöner, Hoke, & Pantev, 1985) and artifact rejection can both improve the SNR compared to normal averaging alone (John et al., 2001a). Weighted-averaging is a technique in which each epoch affects the overall average to a different extent based on its variance, such that a noisier epoch would have less of an influence compared to a less noisy epoch (Hoke, Ross,  38  Wickesberg, & Lütkenhöner, 1984; John, Dimitrijevic, & Picton, 2001b; Lütkenhöner et al., 1985). This technique is advantageous compared to normal averaging only if there is variance in the noise between epochs, otherwise the averages will be the same (Glaser & Ruchkin, 1976; Picton et al., 2003). Artifact rejection is used in conjunction with averaging and occurs before the averaging process. It involves removing epochs of a recording that show higher than expected noise levels (Picton, Linden, Hamel, & Jagdish, 1983). This is often based on a preset maximum amplitude value for the noise level.  1.7.6 Detecting a Response One previously mentioned advantage of the ASSR is that its measurement is more objective than ABR. This is because the repetitive stimuli used for recording ASSRs allow it to be measured in the frequency domain, thus eliminating the need to subjectively select peaks and valleys of a waveform in the time domain, such as is the case with ABR measurement. ASSR is measured by transforming EEG recordings (time domain) into the frequency domain, most commonly by using a Fast Fourier Transform (FFT; Cooley & Tukey, 1965). The resulting information from the FFT is used to detect the response through objective statistical measures of frequency amplitude (strength of the response) and/or phase (timing of the response). The present study uses the two-channel version of the Multiple Auditory Steady State Response system (MASTER), which was developed and described by John and Picton (2000a). The system is capable of generating single or multiple simultaneous steady-state auditory stimuli and recording the electrophysiological response to these stimuli. It accumulates averaged EEG activity (both the response and background noise) and statistically determines whether or not a response is present. The following describes how the MASTER system detects a response. 1.7.6.1  Amplitude One of the outputs of a FFT is an estimate of the energy present over a range of frequencies in  the overall recording of the response—viewed as a frequency spectrum (Picton et al., 2003). Information on the response to each carrier frequency is measured at the modulation frequency (John & Picton, 2000a). The amplitude of the response at the modulation frequency is compared to the background noise present at other adjacent frequencies by calculating an F-statistic (Picton et al., 2003). If the amplitude is significantly greater (usually p<0.05) than that of the noise, the response is considered present.  39  1.7.6.2  Phase Another way to present the output of a FFT is with a polar plot, which displays information on  the amplitude and onset phase of the response, as well as the EEG noise level (Picton et al., 2003).These are similar components to those that are measured for wave V of the ABR. Measures of latency (in milliseconds) and phase (in degrees) both reflect neural transmission time of the response for ABR and ASSR, respectively. However, because ABRs and ASSRs are assessed in different domains, these measures cannot be interpreted in the same way; latency is a linear measure, but phase is a circular measure (as reviewed in John and Picton, 2000b). To be more meaningful, phase measurements can be converted into the time domain, but as reviewed in John and Picton (2000b) and Picton et al. (2003), this conversion is not without problems and in certain cases requires additional modifications. Although phase statistics can be used to detect responses, the MASTER system uses amplitude alone2 and thus phase will be addressed only briefly here. For example, one common phase statistic is phase coherence, which varies in value from 0 (random) to 1 (very consistent) (Jerger, Chmiel, Frost, & Coker, 1986; Stapells, Makeig, & Galambos, 1987). When a response is present, the phase at the modulation frequency is consistent between measurements and said to be ―phase-locked‖.  1.7.7 Stopping Criteria Stopping criteria answer the question of when to stop recording, either because a response is clearly present or absent. These criteria influence threshold estimation and thus are important to consider in order to obtain the most accurate results within a reasonable amount of time. Criteria used and recommended vary depending on the study and researchers, but often include minimum/maximum requirements related to noise level, number of sweeps and recording time, and consecutive sweeps with significance (reviewed in D‘haenens et al., 2010; Luts, Van Dun, Alaerts, & Wouters, 2008). Although noise will lessen with repeated averaging and make any present response more clear, it takes time. Recording time cannot be unlimited and it is necessary to establish criteria for when it is reasonable to stop recording.  2  It is important to recognize that although the MASTER system does not use phase to statistically detect significant responses, the ASSR is averaged in the time-domain and thus response amplitudes still rely on the response being phaselocked (i.e., occurring at a predictable latency relative to the stimulus).  40  The EEG noise-level criterion of studies is generally considered acceptable if it is well below (at least <11nV) the amplitudes expected for near-thresholds responses, and thus an examiner can be reasonably certain that the response is absent (Hatton & Stapells, 2011; Small & Stapells, 2006, 2008a; Van Maanen & Stapells, 2009). For the present study the same EEG noise-level criterion (<11 nV) was used. A criterion of recording a minimum number of sweeps is useful to reduce detection errors that may result from the ASSR amplitude variability present early in the recording (D‘haenens et al., 2010). A minimum number of consecutive sweeps confirms the response and thus also reduces the error rate (Luts et al., 2008). D‘haenens and colleagues (2010) assessed error rates for adult mASSRs and recommended a minimum of eight consecutive significant sweeps to reduce errors in detection (i.e., to provide error rates of <5% for 500, 1000 and 4000 Hz and 10% for 2000 Hz), especially when recording times are variable. In infants especially, it is recognized that very low error rates (<5%) may not be as efficient because of time constraints. Luts et al. (2008) and D‘haenen‘s et al. (2010) found that the error rate was reduced by requiring more consecutive sweeps for variable recording times. For example, at 4000 Hz the error rate was 36, 16, 11, 6 and 4% for 0, 2, 4, 6, and 8 consecutive sweeps, respectively (D‘haenen‘s et al., 2010). Luts et al. (2008) suggest that when using a variable recording time a minimum of two consecutive significant sweeps should be used to avoid false positive identification of a response. Small and Stapells (2006, 2008a) and Hatton & Stapells (2011) used this criterion of two minimum consecutive sweeps to stop recording when a response was detected. The present study used 3 or 4 sweeps as a compromise between amount of error and testing duration. More stringent criteria can provide improved threshold estimates; however they also require longer recording times (John et al., 2002b; John et al., 2004; Luts & Wouters, 2004; Picton et al., 1983; Picton et al., 2003; Van Maanen & Stapells, 2009). Further research is needed in infants regarding the most efficient combination of stopping criteria.  1.7.8 Electrode Montage Response amplitudes to auditory stimuli are maximized when the surface electrodes are positioned appropriately in relation to the generator of the response (reviewed in Cone & Dimitrijevic, 2009). This occurs when the non-inverting and inverting electrodes are aligned with the direction of the generator‘s dipoles. For example, the 80-Hz ASSR is largely generated by the brainstem, which has  41  vertically oriented dipoles. Therefore, for an optimal response, a common electrode montage uses a noninverting electrode (such as Cz or Fpz) paired with inverting electrodes at the mastoid(s), inion, or nape of the neck. Van der Reijden, Mens and Snik (2004) investigated the effect of various electrode montages on the SNR of the 80-Hz response in adults and showed that the inverting electrodes placed at the mastoids (two-channel) or the inion provided the largest SNR when paired with Cz. In contrast, their follow-up study on infants (<6 months of age) found that the best SNR was obtained by recording with electrodes using Cz and the mastoid ipsilateral to the test ear, with no benefit from a Cz and inion placement (van der Reijden et al., 2005). Following these recommendations as well as currently published clinical protocols, the current study used four electrodes with similar placements; the noninverting electrode was placed midline on the forehead, just below the hairline; the inverting electrodes were placed behind the ears, each on the lower portion of a mastoid; and the ground electrode was placed on the low forehead (BCEHP, 2008; OIHP, 2008). 1.7.8.1  Two-Channel Recordings Studies that have investigated both air- and bone-conduction ASSR (Small & Stapells, 2008b;  van der Reijden et al., 2005; Van Maanen & Stapells, 2009) and ABR (Edwards et al., 1985; Foxe & Stapells, 1993; Stapells & Ruben, 1989) have described asymmetries in two-channel recordings, which capture the EEG activity ipsilateral and contralateral to the stimulus ear (Small & Stapells, 2008b). As discussed earlier, the ASSR studies reveal that when the ipsilateral EEG channel shows a present response, the response in the contralateral EEG channel is either present, but smaller and with a longer phase delay, or absent. As the response in the contralateral EEG channel is variable, it is not recommended for assessing thresholds. However, these asymmetries have the potential to be clinically useful during BC testing to determine which cochlea is responding to a stimulus without the need for masking (Small & Stapells, 2008b). In this study, two-channels were recorded, however, only the ipsilateral channel was analyzed. Currently, only three studies have investigated two-channel ASSRs and only one of those studies has included BC ASSRs (Small & Stapells, 2008b; van der Reijden et al., 2005; Van Maanen & Stapells, 2009). Because asymmetries between the two-channels are helpful in recording BC ABRs, further research is necessary to determine if the same is true for BC ASSRs.  42  1.8 Visual Reinforcement Audiometry An infant‘s behavioural response to sound (i.e., eye widening, startle, localization) changes over time and affects an examiner‘s ability to reliably estimate auditory thresholds using behavioural test methods. Behavioural observation audiometry (BOA) is used clinically to provide information on the auditory function of children who are developmentally too young to participate in other behavioural techniques (Northern & Downs, 1991). This method involves the examiner observing behavioural responses that are time-locked to the presentation of an auditory stimulus. Unfortunately, it is not a very reliable indicator of hearing sensitivity as it only elicits suprathreshold responses (Northern & Downs, 1991), has high variability (Thompson & Weber, 1974) and rapid habituation (Widen, 1990). For these reasons, BOA is generally supplemented or replaced by physiological measures until the child is able to provide more definitive behavioural results that indicate whether the child‘s hearing is or is not within the normal range of 0 to 20 or 25 dB HL (depending on the protocol). When the child is a suitable age, it is more appropriate to include conditioned behavioural testing methods in the audiological assessment. Behavioural conditioning in infants was first applied to audiology by Suzuki and Ogiba (1961) and Liden and Kankkunen (1969).The work of these authors was further developed by several publications from the University of Washington (Moore, Thompson, & Thompson, 1975; Moore et al., 1977; Thompson & Folsom, 1984; Thompson & Wilson, 1984; Thompson & Folsom, 1985), which lead to a clinically-accepted procedure that is now commonly known as visual reinforcement audiometry (VRA). It can be used to obtain reliable responses to lowintensity stimuli (20 dB HL or less) in most infants of 5 to 6 months up to approximately 24 months developmental age (Gravel & Traquina, 1992; Moore et al., 1977; Northern & Downs, 1991; Widen, 1990; Widen et al., 2000). Typically VRA is more difficult in older infants and toddlers; although they learn the task quickly they respond less, habituate faster, and are more apt to reject assessment with earphones (Gravel & Traquina, 1992; Northern & Downs, 1991; Primus & Thompson, 1985; Thompson, Thompson, & Vethivelu, 1989; Widen et al., 2000). VRA is a testing method that relies on operant conditioning of an infant‘s natural tendency to turn his/her head towards a sound source. During VRA, an infant learns to associate the presentation of an auditory stimulus with the illumination of an attractive visual reinforcement, such as an illuminated toy or video clip. In response, the child will turn his/her head anytime the sound is heard so that he/she  43  may view the rewarding reinforcer. Once the infant is conditioned, he/she will continue to respond to the stimulus if it is heard; the tester can then adjust the stimulus presentation level and estimate the infant‘s behavioural thresholds (Moore & Thompson, 1976, as cited in Widen, 1990). Although reinforcement following the response delays the rapid habituation that occurs in BOA (Moore et al., 1977), even in VRA head-turn responses are not unlimited. There are many age-dependent factors that influence the reliability and effectiveness of VRA, including behaviour, habituation and interest in the task, the type of reinforcement and the type of stimulus.  1.8.1 Behaviour Although typical adults can follow instructions and maintain attention throughout an auditory assessment, infants can be challenging to assess because of their immature behaviour. Infants vary in their cooperativeness, attention span and activity state. This variability in behaviour can substantially influence the time and effort taken to obtain an adequate amount of reliable information (Parry et al., 2003). For example, when beginning VRA, some infants are intolerant of earphone or bone-oscillator placement and so can only be assessed in the soundfield, which lacks ear specificity and thus limits the conclusions that can be drawn from the results. As testing progresses, the infant‘s attention will wane. Because VRA is a flexible procedure that can accommodate infant behaviour, ideally, the examiner can re-engage the infant; this is discussed further in the following section on habituation. Given the attention requirements of VRA, it follows that testing would be most effective when the child is alert and calm. However, an infant‘s state may change quickly and testing is not always possible if the child becomes too active or fussy; this may lead to incomplete results and necessitate a return visit. Additionally, these results may vary in reliability if the child‘s activity state fluctuates during testing. Considering all of these behavioural uncertainties, the priority of clinical VRA is typically to confirm that hearing sensitivity is ―within normal limits‖ using MRLs, as opposed to searching for thresholds. This means that clinically, and in research, responses to VRA may not always be assessed below a certain ―normal‖ level. A MRL is a more appropriate term if the protocol used does not involve a bracketing procedure that verifies that there is no behavioural response below the level considered to be the MRL. However, even if a search procedure is used, the term MRL is still applicable as it also reflects the fact that infant behavioural responses are undergoing maturation and may improve with age (Matkin, 1977). Essentially, the evaluation of infant auditory sensitivity is limited by developmental factors and as such, MRLs may be equivalent to or elevated compared to true auditory thresholds. For simplicity, the term 44  threshold will be used to describe responses for both behavioural and physiological measures used in this study.  1.8.2 Habituation After repeated stimulus presentations, an infant will eventually habituate and cease to respond; this habituation may stem from a lack of interest in either the visual reinforcer or auditory stimulus (Culpepper & Thompson, 1994). Habituation occurs more rapidly in toddlers (21-26 months) than in younger infants (under 13 months) (Gravel & Traquina, 1992; Primus & Thompson, 1985; Thompson et al., 1989; Werker, Polka, & Pegg, 1997). For example, Primus and Thompson (1985) found that on average 1-yr-olds provided 31.3 responses (i.e., stimulus-locked head-turns) before habituation, while 2yr-olds only provided 20.0 responses. However, there are strategies that can delay habituation. Depending on the situation, the examiner may choose to switch transducers, stimuli or reinforcers or even take short breaks to arouse the infant, extending the testing session and increasing the number of responses. Culpepper and Thompson (1994) found that, in 2-year-olds, presenting the reinforcer for a shorter duration can help engage the child longer. For younger infants (11-13 months) specifically, a small break provided during the testing session can renew the infant‘s interest when the task is resumed (Thompson, Thompson, & McCall, 1992; Widen et al., 2000). There is also evidence that introducing a novel reinforcer to the task may help maintain the infant‘s responses (Diefendorf & Gravel, 1996; Gravel, 1989; Karzon & Banerjee, 2010; Primus & Thompson, 1985; Werker et al., 1997). The impact of reinforcer novelty can be maximized by having multiple reinforcers available and alternating their presentation (Diefendorf & Gravel, 1996). It has also been suggested that any type of variety is also effective, whether it involves changing the type of stimuli, test frequency, or test ear (Primus & Thompson, 1985; Primus, 1987; Widen et al., 2000; Widen & O'Grady, 2002).  1.8.3 Reinforcement Type As the reinforcement drives the infant‘s head-turn, it is important to consider how different types of reinforcers affect an infant‘s responsiveness to the task. The reinforcement options for VRA often include an animated mechanical toy and/or flashing lights or brief video clips. Moore, Thompson and Thompson (1975) and Moore, Wilson and Thompson (1977) concluded that the number of head-turns  45  was greater with increased complexity of the visual reinforcement; for example, flashing lights resulted in fewer head-turns than flashing lights accompanied by an animated toy. There is conflicting research on the effectiveness of an animated mechanical toy versus video reinforcement. Schmida, Peterson and Tharpe (2003) reported that on average, children 19-24 months of age responded with a greater number of head-turns to a video reinforcement than to the animated mechanical toy. In contrast, Karzon and Banerjee (2010) determined that in similarly aged infants (1624 months), the animated toy elicited significantly more head-turns compared to the video reinforcement. A critical difference between these two studies is that Karzon and Banerjee (2010) had at least three animated toys (with a maximum of six) available for reinforcement, but Schmida et al. (2003) used only one. It follows that employing multiple animated toys likely increased the novelty of the mechanical reinforcers, eliciting more head-turns than could a single mechanical toy. Lowery, von Hapsburg, Plyler, and Johnstone (2009) investigated comparatively younger infants (7-16 months) and found no significant differences in the quality and number of responses between the two types of reinforcers; although they also only used a single mechanical toy. All of the discussed studies noted that some infants were ―high-responders‖, suggesting that individual preferences also plays a role in the effectiveness of a reinforcer (Karzon & Banerjee, 2010). Lowery et al. (2009) and Karzon and Banerjee (2010) both propose that clinically it may be more useful to recognize that both the animated toy and video are rewarding to infants and toddlers and may be most efficient when used in combination.  1.8.4 Stimulus Type The stimuli used for behavioural assessment varies and depending on the goal of the assessment may include any of the following: speech, music, sound effects, broadband or narrowband noise, pure tones, and/or frequency-modulated tones (warble-tones). The stimuli used for assessment are of interest for their potential to affect the behavioural response. For BOA in particular, both frequency and bandwidth affect the infant‘s interest and responsiveness to the stimulus. A number or studies provide evidence that broadband or more complex stimuli (e.g., speech or noise) are more effective than more frequency-specific stimuli (e.g., pure-tones, warble-tones and narrowband noise) in evoking a behavioural response in newborns (Eisenberg, 1965; Ling, Ling, & Doehring, 1970) as well as young infants (3-11 months) (Hoversten & Moncur, 1969; Mendel, 1968). Additionally, the probability of obtaining a response is greater with the use of a low- to mid-frequency broadband stimulus compared to  46  a stimulus that is narrowband or high-frequency broadband (Moore et al., 1975; Thompson & Folsom, 1985). In contrast to BOA, for a conditioned response, the stimulus has no significant influence, as it is just a means to signal the actual event of interest—the reinforcer (Primus & Thompson, 1985; Thompson & Folsom, 1985; Thompson & Thompson, 1972). The findings of both Thompson and Thompson (1972) and Thompson and Folsom (1985) support that stimulus type affects the number of responses for BOA, but does not significantly influence conditioned responses. However, the stimulus preferences infants exhibit for unconditioned responses are still applicable to the initial conditioning phase of VRA, where the infant is trained in the task. Thompson and Folsom (1985) and Shaw and Nikolopoulos (2004) advise examiners not to begin VRA with a warbled tone, but rather use a stimulus with a broader bandwidth (i.e., narrowband noise) because it is more likely to generate an unconditioned initial head-turn. Once the infant demonstrates conditioned head-turns it is prudent to continue the assessment with more frequency-specific (i.e., warbled-tone or narrowband noise) stimuli to obtain threshold estimates. The frequency-specific stimuli used for infants (warbled tone or narrowband noise) differ from that of adults (pure tones). In the sound field, standing wave interactions occur with pure tones and introduce unwanted variability of the stimulus (Dillon & Walker, 1982); the use of warbled tones (with limited frequency deviation) and/or narrowband noise (NBN) is more suitable to the sound field as they minimize these interactions (Morgan, Dirks, & Bower, 1979). Because VRA may be conducted in the sound field, warbled tones and NBN are the preferred stimuli for frequency-specific VRA both clinically and in research. In normal-hearing infants, there is no significant difference between the MRLs obtained with warbled tones and NBN (McDermott & Hodgson, 1982). However, warbled tones are the slightly favoured stimuli as NBN is less frequency-specific and may underestimate the degree of certain types of hearing losses (Walker, Dillon, & Byme, 1984).  1.9 Rationale for Thesis In summary, this study aims to assess BC hearing in a group of normal-hearing infants with both ASSR and VRA techniques, as no previous studies assess the same child with both measures. Secondarily, AC thresholds for both measures will be collected to allow further estimation of the maturational air-bone gap within the same infant. The participants will be 6–20 months of age, which is a reasonable age range for predicting AC and BC thresholds using either testing method. A group of  47  normal-hearing adults will also be assessed both physiologically and behaviourally, and will function as a control group for making age-related comparisons. Infants exhibit immaturities that influence auditory threshold estimation. For accurate classification of type and magnitude of hearing loss it is critical to understand BC threshold maturation. By comparing BC thresholds of normal-hearing infants across different testing techniques and in relation to AC thresholds, we can increase our understanding of what constitutes a normal response. It follows that this study will help better define the relationship between physiological and behavioural BC thresholds in infants, as the tests will be conducted in the same child within a near equivalent maturational time period. With this information, the diagnosis of hearing loss may improve in accuracy and lead to more informed treatment. These data will contribute to the initial step towards establishing an ―eHL‖ for air- and bone-conducted stimuli and will provide a more accurate measure of the maturational gap by comparing BC and AC sensitivity in the same subject.  48  CHAPTER 2: Comparisons of Auditory Steady-State Response and Behavioural Air- and Bone-Conduction Thresholds in Infants and Adults with Normal Hearing  49  2.1 Introduction The first goal of early hearing programs is to identify permanent childhood hearing impairment, whether it is sensorineural, conductive or mixed in nature (e.g., BCEHP, 2008; OIHP, 2008). To distinguish between these three types of hearing loss, it is essential to assess BC hearing sensitivity when AC thresholds are elevated. To accurately interpret BC thresholds in an infant, one must acknowledge that maturation influences a child‘s responses and thus what is considered a ―normal‖ auditory response differs from an adult and varies with age and assessment method. For example, there is evidence that there are frequency-dependent differences in the BC thresholds of infants, such that the thresholds at lower frequencies are significantly better than the thresholds at higher frequencies (Small & Stapells, 2006, 2008a; Foxe & Stapells, 1993; Nousak & Stapells, 1992; Stapells & Ruben, 1989; Vander Werff et al., 2009; Hulecki & Small, 2011). Additionally, as children mature, the technique that provides the most reliable information about their auditory sensitivity changes. Although behavioural estimates of hearing thresholds are ideal, they cannot be obtained reliably in certain populations. Visual reinforcement audiometry (VRA) is a gold-standard behavioural technique that can often be reliably obtained in infants aged approximately six months to 2.5 years. In cases where VRA is not feasible, physiological estimates of hearing sensitivity are substituted. The current physiological gold-standard measure is the brief-tone auditory brainstem response (ABR) (JCIH, 2007). The auditory steady-state response (ASSR) is another physiological method that may be more efficient than ABR and employs more objective statistical response detection measures (Picton et al, 2003; Stapells et al, 2005; Cone and Dimitrijevic, 2009). For these reasons, the ASSR is currently being researched for its potential to supplement, or even replace, the tone-evoked ABR. Although physiological thresholds are elevated in comparison to behavioural thresholds, they can be used to estimate behavioural thresholds. Applying a correction factor to the physiological estimation results in an estimated behavioural hearing level, eHL, which is the closest approximation of the infant‘s hearing thresholds (Picton et al., 2003; Stapells, 2010a). These eHLs are needed to prescribe amplification and compare thresholds over time. Establishing correction factors requires defining the relationship between the actual thresholds, as confirmed by a gold-standard measure (either VRA or tone-evoked ABR) and those predicted by the ASSR in both normal hearing infants and infants with  50  hearing loss (Stapells, 2010a). Data are scarce for BC ASSR, as there is currently no study that directly assesses the association between BC ASSR and either tone-evoked ABR or VRA, using bone-conducted stimuli, in infants with normal or impaired hearing, and as such there are no eHL correction factors available for BC ASSR. Comparing BC ASSR and VRA thresholds in normal infants will demonstrate their relationship and contribute to the initial step in establishing eHL correction values for BC ASSR. Additionally, the relationship between BC and AC ASSR is also important in meeting early hearing program objectives. In adults, a gap between the AC and BC thresholds is indicative of a conductive component and is commonly called an air-bone gap (ABG) (Lierle & Reger, 1946; Carhart, 1950). Elevated BC thresholds and the absence of an ABG indicates a sensorineural hearing loss (Lierle & Reger, 1946; Carhart, 1950). The relationship between AC and BC thresholds is thus an important component of an audiological assessment. However, studies involving the assessment of BC thresholds, both physiologically (ASSR: Small & Stapells, 2006, 2008a; ABR: Foxe & Stapells, 1993; Nousak & Stapells, 1992; Stapells & Ruben, 1989; Vander Werff et al., 2009) and behaviourally (Hulecki & Small, 2011), provide evidence that this relationship differs in infants. Specifically, infants exhibit an ABG in dB HL that results from their different sensitivity to bone-conducted stimuli and is not a result of conductive pathology. It follows that to more accurately determine the existence of an air-bone gap, we must define the normal relationship found between AC and BC thresholds in infants for both behavioural and physiological threshold estimation methods. Without accounting for maturation, conductive hearing loss may be over-diagnosed, leading to unnecessary use of healthcare resources. The present study assesses AC and BC sensitivity for both VRA and ASSR testing methods in normal hearing infants, as well as AC and BC ASSR sensitivity and pure-tone audiometry in adults. While, individually, assessment of BC ASSR and VRA thresholds in normal hearing infants will provide normative data for a narrow age range for both techniques, the greater value in this study is comparing results between different sound conduction pathways and techniques. These comparisons will provide a broader understanding of what constitutes a normal response. The study has the following two objectives related to infants: (1) to compare BC responses between VRA and ASSR to further define their relationship and; (2) to compare BC and AC sensitivity in both methods to better quantify the maturational ABG. Adult subjects will serve as a control group for making maturational comparisons.  51  2.2 Methods and Materials 2.2.1 Participants There were 33 infants who participated in this study, of which 9 infants were excluded because they did not complete any of the tests and 1 infant was excluded due to test results of poor reliability in the only test completed. Partial or complete results for VRA and/or ASSR were obtained for 23 normalhearing infants (mean age: 10.5 months; range: 6.5–19 months; 12 female). Of the 21 infants who provided reliable BC VRA thresholds, most gave thresholds for at least three frequencies. Only 14 of these 21 participants also provided thresholds for AC, all of whom provided at least two AC thresholds. Although most of these infants required only one visit to complete VRA, four infants did return for a second VRA visit. Table 2.1 summarizes the count and percentage of infants who gave each number of AC and BC thresholds. From these infants, 102 thresholds were obtained: 60% were by BC and 40% were by AC. Mean thresholds and standard deviations for each frequency are provided in Table 2.4. Table 2.1 Count and Percentage of infants who provided air- or bone-conduction behavioural thresholds for a given number of frequencies.  Conduction Mode Bone Air  Number of Behavioural Thresholds Obtained 1 2 3 4 2 (9.5%) 1 (4.8%) 15 (71.4%) 3 (14.3%) 0 (0%) 5 (35.7%) 5 (35.7%) 4 (28.6%)  The mean number of head turns was 27.9 (standard deviation=9.6), with a range of 15–45. The mean number of control trials was 13.4. The false-response rate of each infant was calculated by dividing the number of false responses by the number of control trials. The mean reliability of the infant responses was 88.1% and ranged from 78.3 to 100%, which was considered to be good overall. Infants were included in the study if they were considered to be at low risk for hearing loss. Most parents/guardians of the infant subjects indicated their child‘s hearing had been screened at birth with passing results3. To verify, hearing was screened again bilaterally with transient evoked otoacoustic 3  For one infant, the parent reported a failed infant hearing screening. However, results from follow-up testing indicated normal hearing. The parent has had no further hearing-related concerns with her infant.  52  emissions (TEOAE) using a Madsen AccuScreen and Bio-logic pediatric foam tips. The TEOAE stimulus level ranged from 70–84 dB SPL as the equipment self-calibrated for the infant‘s ear canal size. Of the 23 participants, 19 infants passed TEOAEs bilaterally and two infants passed unilaterally (i.e., only one ear was tested due to excessive movement4). If a child did not pass or could not complete the TEOAEs, tympanometry was attempted to assess and document middle-ear integrity. Tympanometry was only needed for three infants and was carried out using a GSI Tympstar (Version 2) run in the screening mode using a 226-Hz probe tone. A tympanogram was considered abnormal (i.e., suggestive of middle-ear pathology) if it appeared flat and lacked a point of maximum admittance and ear canal volume was within normal limits. Two of the infants had normal tympanograms bilaterally, but one of the infants presented with a flat tympanogram unilaterally5. Other inclusion criteria included parental report on the absence of neurological problems and developmental concerns and no history of middleear pathology. A total of 12 normal-hearing adults participated in the study (mean age: 29.9 years; range: 17–50 years; 10 female) also participated in the study. Adults were considered to have normal hearing if, by standard technique, their pure-tone bone- and air-conduction thresholds were better than 25 dB HL for 500, 1000, 2000 and 4000 Hz, with no significant (>10 dB HL) air-bone gaps. However, distortionproduct otoacoustic emissions (DPOAEs), with an f2 frequency range from 2–5 kHz, were substituted for three adults due to time constraints; these adults all passed bilaterally at the four frequencies assessed and were considered to be at low risk for hearing loss. Seven adult participants completed both ASSR and behavioural testing, four completed only ASSR testing and one was tested only behaviourally.  2.2.2 Stimuli 2.2.2.1  Bone Conduction (BC) The ASSR stimuli were bone-conducted tones with the carrier-frequencies 500, 1000, 2000 and  4000 Hz generated by the two-channel Multi-MASTER Research System. The acoustic spectra of the AM2 stimuli used are shown in Figure 2.1. They were exponential envelope modulated (AM2) at modulation frequencies of 78, 85, 93, and 101 Hz for 500, 1000, 2000 and 4000 Hz, respectively, and 4  For both infants the ear that passed TEOAEs was used for testing This particular infant passed TEOAEs at the first visit, but due to parental report of a recent cold was re-screened at the second visit using tympanometry. Select data from this infant were still included in the study. 5  53  were presented simultaneously. The bone-conducted stimuli used to estimate behavioural thresholds were frequency-modulated (warble) tones at 500, 1000, 2000 and 4000 Hz, generated by a GSI-61 audiometer. Frequency-modulation was applied to +5% of the centre frequency at a rate of 5 Hz (Grason-Stadler, n.d.). Each stimulus was presented individually for 1-2 seconds, with at least five seconds between stimulus presentations.  Figure 2.1 Acoustic spectra of air-conducted AM2 stimuli for the carrier frequencies of 500, 1000, 2000 and 4000 Hz. Y-axis ticks represent 10 dB intervals.  All bone-conducted stimuli were presented to the subject using a B-71 bone oscillator. The bone oscillator was positioned on the temporal bone slightly posterior to the upper part of the pinna (Small and Stapells, 2008b). The transducer was coupled to the head using approximately 4.0-4.5 Newtons (N) of force. Depending on the test method and the infant‘s preference the bone-oscillator was held in place by either a steel or elastic headband, or by hand. For VRA, the steel headband was the first choice for testing, with the bone-oscillator attached to one end of the headband and a small foam sponge placed at the other end to provide a more comfortable experience for the child. If the steel headband was not tolerated, the elastic headband was used instead (N=9). This involved placing the elastic headband  54  around the infant‘s head, securing it in place with Velcro ™ and positioning the bone-oscillator appropriately. The hand-held method was only used for a portion of the behavioural testing if the infant would not tolerate the headband choices (N=2)6, as it is difficult to maintain constant force and positioning on an alert and active child. For ASSR testing, the hand-held technique was always used as it was easier to place on a child in various sleeping positions and the least disruptive to initiate during sleep; Small et al. (2007) found that the hand-held coupling technique, as performed by a trained assistant, provides BC ASSR thresholds equivalent to those obtained using an elastic headband. For adults, the bone oscillator was coupled with an elastic headband for ASSR testing and with a metal headband for screening with pure-tone audiometry. 2.2.2.2  Air Conduction (AC) The AC stimuli used for ASSR and behavioural testing were the same as those described above  for BC testing. The AC stimuli were presented via an ER-3A insert earphone on the same side used to estimate BC thresholds.  2.2.3 Calibration All bone-conducted stimuli were calibrated in Reference Equivalent Threshold Force Levels (RETFL) in dB re: 1 µN corresponding to 0 dB HL for the mastoid (ANSI S3.6-1996) using the Larson and Davis system 824 sound level meter and B&K Mastoid 4930 artificial mastoid. The oscillator was coupled to the artificial mastoid with 5.5 N of force. The air-conducted stimuli were calibrated in dB HL (ANSI, 1996) using a Larson Davis system 824 sound level meter and G.R.A.S. RA0113 2-cc coupler with 1-inch microphone. The four sinusoidal air- and bone-conducted tones were calibrated separately in dB HL, then combined.  2.2.4 Recording The two-channel Multi-MASTER Research system (John & Picton, 2000) was used to both generate the stimuli and record the electrophysiologic responses. Four electrodes were placed on the subject to record the ASSRs: a non-inverting electrode, a ground electrode, and two inverting electrodes. 6  One infant required the bone oscillator to be handheld for only one threshold. The second infant required the handheld method for three thresholds. Both infants had thresholds comparable to the other participants.  55  The non-inverting electrode was centred on the high forehead, with the ground electrode placed directly below it on the low forehead. The inverting electrodes were positioned behind the ears, each on the lower portion of a mastoid. The skin was prepared for electrode placement with a mild abrasive to obtain electrode impedances under 5 kOhms and inter-electrode impedances of less than 2 kOhms at 30 Hz (using a GRASS Model F-EZM5 Electrode Impedance Meter). The EEG was filtered by a 30-Hz high-pass filter (-12 dB/octave slope) and a 300-Hz low pass filter (-24 dB/octave slope), and amplified 100, 000 times [10,000X in GRASS LP511 AC Amplifier; 10X in National Instruments DAQmx (USB-6259)]. The EEG was further filtered using a 400-Hz lowpass anti-aliasing filter via the Stanford Research Systems Model SR650 (115 dB/ octave slope). The EEG was then processed using a 1250-Hz analog-to-digital conversion rate (Small & Stapells, 2004). Each EEG recording sweep was made up of 16 epochs of 1024 data points (0.819 seconds per epoch) and lasted a total of 13.107 seconds. The artifact rejection was set to eliminate epochs of electrophysiologic activity that exceeded 80 µV in amplitude to reduce contributions to the EEG from muscle artifact. ASSRs were averaged in the time domain and then analyzed online in the frequency domain using a Fast Fourier Transform (FFT). Weighted averaging was used over a frequency range of 70–110 Hz (John, Dimitrijevic & Picton, 2001b). The FFT resolution, over a range of 0 to 625 Hz, was 0.076 Hz. Amplitudes were measured baseline-to-peak and expressed in nano Volts (nV). An F-ratio was calculated by Multi-MASTER to estimate the probability that the amplitude of the ASSR at the modulation frequency for each carrier frequency was significantly different from the average amplitude of the background noise in adjacent frequencies within 60 bins of the modulation frequency (―noise‖) (John & Picton, 2000a). ASSRs were recorded for a minimum of 10 consecutive sweeps. After these initial 10 sweeps, recording was continued either until a response was detected (p< .05) or until the EEG noise criterion was reached (circle radius of < 20 nV) and no response detected (p >0.05). A response was considered to be present if the F-ratio, compared to the critical values for F (2, 240), was significant at a level of p< 0.05 for at least three consecutive sweeps. A response was also considered to be absent when response amplitude was < 10 nV and the p-value >0.30. The mean EEG noise of absent responses for both air-  56  and bone-conducted stimuli were calculated using the EEG noise at one intensity level below the accepted threshold.  2.2.5 Procedure The study involved two sessions per child, with one session allocated for each procedure. The study involved only one testing session for adults. This section describes the procedure overall and is followed by detailed descriptions of the two test procedures. Each adult participant and the parent/legal guardian of infant participants signed a consent form before initiation of testing and was paid an honorarium at the end of each session. For adults, hearing screening procedures (either pure-tone audiometry or  DPOAEs) and recording of the ASSR data were conducted during the same visit. For infants, the first session was scheduled for recording of the ASSR data; however, if the infant did not sleep, VRA was conducted in the first session and ASSR testing was completed in the second session. The session lengths were approximately two hours for ASSR and one hour for VRA testing. Most often, less than two weeks elapsed between the visits (mean days between visits: 11.2 days; range: 0–48 days)7. If threshold data were obtained for either testing technique during the first session, all participants were required to pass a transient evoked otoacoustic emissions (TEOAE) screening test bilaterally. If TEOAEs were not obtainable, tympanometry was substituted. All testing was completed in double-walled sound-attenuating booths, either the Pediatric Audiology Laboratory (ASSR) or the Pediatric Teaching Booth (VRA) at the University of British Columbia‘s School of Audiology and Speech Sciences. Average ambient noise levels for were 0, 2, 4, and 6 dB SPL in the booth used for VRA and 1, 3, 6, and 7 dB SPL in the booth used for ASSR, for oneoctave bands centred at the test frequencies of 500, 1000, 2000, and 4000 Hz, respectively. Thresholds to air- and bone-conducted stimuli for the group of normal-hearing infants were obtained for the ASSR and VRA testing methods. The lowest level tested was -10 dB HL. A group of normal-hearing adults were also assessed using air- and bone-conduction ASSR and pure-tone audiometry. Each test maximally assessed four bone- and air-conducted stimuli at 500, 1000, 2000, and 4000 Hz in one ear of the subject. BC thresholds were considered a priority and always completed prior to attempting to obtain AC thresholds for both ASSR and VRA (i.e., selection of AC versus BC 7  One infant returned for the second session 3 months after the first session and was excluded from these calculations.  57  condition was not randomized). Due to the infant‘s state and attrition, not all conditions were obtained for each subject. 2.2.5.1  VRA Two testers were required to be involved in the VRA test: the examiner, who presented the  stimuli and visual reinforcement while recording the child's responses, and the assistant, who was in the room with the child and acted as a distracter. The examiner conducted the test from an adjacent booth and observed the infant through a shared window between the booths. Each alert infant was seated in his/her parent‘s lap, facing forward in the centre of the testing booth. This arrangement allowed for the four visual reinforcers to be positioned approximately 90° to both the left and right of the infant, two per side. Both animated toys, as well as video reinforcement, were available to be used on either side of the infant. This arrangement has been used in other studies on VRA and allows for greater flexibility in testing as there is an option to switch sides or type of reinforcer (Hulecki & Small, 2011; Schmida et al., 2003; Widen et al., 2000). All infants were reinforced with both types of reinforcer, though some children responded preferentially to one or the other. Each parent was instructed on the procedure and the importance of not distracting the child with noises or cueing the child with their own movement. As an extra precaution, the parent was asked to wear Peltor Optime 101 ear muffs to provide enough attenuation to make the bone-conducted stimuli inaudible to the parent. An assistant remained in the booth with the parent and child during testing, sitting at eye-level at a 45° angle (on the opposite side of the test ear) in front of the child. Both the assistant and the tester had an unobstructed view of the infant‘s reactions to the stimuli, but only the examiner was responsible for judging responses. The assistant was responsible for bringing the infant‘s attention back to between stimulus presentations using small toys or books. The assistant moderated his/her distraction techniques to be only interesting enough to distract the infant from the reinforcer, but not enough to distract the infant from the task. This allowed for easier observation of the head-turn response. The examiner communicated with the assistant through a frequency-modulated (FM) communication system, which informed the assistant of stimulus presentation and conditioning status and aided in the appropriate timing for distraction. The assistant also occasionally repositioned the bone oscillator if it shifted or came off during testing.  58  The protocol used to obtain the infant thresholds was similar to Hulecki and Small‘s (2011) adaptation of the protocol presented by Widen et al. (2003) and incorporated probe, conditioning and test trials. The protocol was used as an efficient means for obtaining reliable thresholds in one visit, while also providing enough structure to maintain testing consistency and the flexibility needed when testing infants. The examiner began the VRA testing session by presenting a warble tone through the boneoscillator8 at 30 dB HL to probe for a response. This probe trial was done prior to either the conditioning or test trials. If a head turn did not result from the probe tone, the stimulus level was increased to 50 dB HL and the examiner again presented a probe tone. If the infant responded with a head turn within four seconds of the stimulus presentation, he/she was rewarded with the activation of a reinforcer on the same side he/she turned towards. After two correct head turns to the probe tone the examiner began the test protocol in search for the threshold at each test frequency. If there was no head turn the examiner proceeded to the conditioning phase. Conditioning Phase The conditioning phase typically involved pairing the bone-conducted stimulus with the presentation of the reinforcer. The assistant pointed out the toy/video to the infant to condition him/her to turn when the stimulus was heard. This conditioning involved two trials and was followed by another probe trial, where the reinforcement was activated only after a head turn was observed in response to the probe stimulus. When an infant did not condition well initially or ceased to respond at any point in testing, the examiner adapted the protocol by changing the stimulus level, type of stimulus and/or mode of stimulus presentation to regain the infant‘s attention (Hulecki & Small, 2011; Parry et al, 2003; Widen et al., 2000). Additional conditioning trials were sometimes necessary when changing the test frequency or type of stimulus and conducted in the same manner described previously. Threshold search continued after two correct head turn responses to two consecutive probe trials at the same level.  8  In some cases, the probe trials were also conducted with insert earphones and occasionally in the soundfield, depending on the information still needed from the infant and his/her tolerance of the other transducers.  59  Test Phase The testing protocol began with a 2 kHz bone-conducted warble-tone stimulus of 1-2 seconds in duration. The threshold search used the bracketing method with 10 dB step-size (Widen et al., 2000). This was obtained by following a 20 dB down and 10 dB up threshold search. False responses were minimized by varying the inter-stimulus interval, with a minimum of five seconds between stimulus presentations. Three-second long control trials were randomly recorded during these inter-stimulus intervals. The lowest level at which 2 responses out of 3 presentations were obtained was considered the threshold, which is the same criterion used by many previous studies (Hulecki & Small, 2011; Karzon & Banerjee, 2010; Parry et al, 2003; Widen et al., 2000). The same procedure was applied to the three other test frequencies. The subsequent test frequency was presented at the level of the previous response or at a minimum of 10 dB HL. This was done to reduce the number of supra-threshold head-turns, thus obtaining responses closer to threshold sooner and increasing testing efficiency. Testing priority was given to 2000 and 500 Hz, followed by 4000 Hz and lastly 1000 Hz, following the British Columbia Early Hearing Program protocol (BCEHP, 2008). This testing order and/or stimuli for some subjects was altered to avoid habituation of the subject. To regain the infant‘s attention, the test protocol varied by order of test frequency, type of stimulus (warbled tone, narrowband noise and speech) and level or stimulus. Extra conditioning trials were also sometimes necessary after a stimulus type change or frequency change and for infants who had habituated to the stimuli. As suggested by Thompson, Thompson and McCall (1992), small breaks of 5-10 minutes were also sometimes used to refresh the infant‘s interest in the task and thus obtain more head-turns once testing was resumed. Testing continued until reliable thresholds were obtained for all four frequencies for both air- and bone-conduction or until testing was no longer possible. The latter was the most common and due to poor conditioning, habituation, restlessness and crying, and/or repeated objection to wearing either the insert earphone or bone oscillator. Following Widen et al. (2000), if the infant falsely responded in fewer than 30% of the control trials, the test was considered reliable9. The examiner also noted reliability subjectively by observing the 9  Only one infant was excluded based on poor response reliability.  60  infant‘s response to probe trials and consistency of responses around threshold (Widen et al., 2003). Three infants did not provide any VRA data: two infants could not be conditioned to the task and one did not return for the second session. However, the ASSR data for all three of these infants were still included in the study. The sample size for each frequency and conduction mode is summarized in Table 2.2, below. Thresholds were most often not obtained at all frequencies for both air- and boneconduction, rather most infants provided just a sub-set of thresholds before testing was no longer possible. Table 2.2 Infants sample size of VRA thresholds obtained by frequency and mode.  Mode  500 Hz  1000 Hz  2000 Hz  4000 Hz  Bone conduction  19  4  20  18  Air conduction  12  4  13  10  2.2.5.2  ASSR ASSR testing was conducted in a double-walled sound-attenuated booth. Infants were required to  sleep for the duration of testing, while adults were instructed to relax or sleep. An assistant was always present in the booth with the infant, and the parent was present at their own discretion. The assistant held the bone oscillator during BC testing. For AC testing, the assistant observed the infant throughout testing to ensure proper earphone placement. Adults did not require an assistant for testing. Responses were obtained for multiple-simultaneous stimuli at 500, 1000, 2000 and 4000 Hz, with an initial randomized intensity level of 10, 20, or 30 dB HL. Threshold was determined using a bracketing technique with 10 dB steps10. ASSR threshold was the lowest intensity at which a response was present, with responses absent 10 dB HL below that intensity level. For some subjects, not all responses at levels above ―threshold‖ reached significance. If only one level above the threshold was absent, it was considered a false negative and the lowest level at which the response was present was accepted as the threshold. However, if more than one level above ―threshold‖ was absent, the lowest level of response was considered a false positive and the next level where a response was present was deemed the threshold. If response presence/absence was questionable that level was repeated to acquire a lower noise level and/or greater number of sweeps, time permitting. For two infants, a threshold at one 10  The intensity at which at least 90% of the participants had responses is considered to be the ―normal level‖ for both AC and BC AM2 stimuli.  61  frequency was not obtained due to poor reliability or no response, these thresholds were not included in further analysis. The number of infants who completed AC and BC ASSR testing for each frequency are shown in Table 2.3. Table 2.3 Infants sample size of ASSR thresholds obtained by frequency and mode.  Mode  500 Hz  1000 Hz  2000 Hz  4000 Hz  Bone conduction  19  20  20  20  Air conduction  14  13  13  14  2.2.6 Data and Statistical Analyses The 90th percentile was calculated by expressing the number of responses present at each intensity as a cumulative percent for infant and adult BC and AC ASSRs. Thresholds were averaged across frequency and between testing method, stimulus presentation mode and age. Individual infant threshold data were also presented for BC ASSRs and VRA. Infant difference scores (ASSR-VRA thresholds) were calculated for air- and bone-conducted stimuli. Differences scores (AC-BC thresholds) were averaged for infant ASSRs and VRA and adult ASSRs. The difference between AC and BC thresholds were also expressed as the percentage of occurrence of ABGs ranging from 5-30 dB. The ASSR amplitudes and phase delays were also averaged and compared across frequency and age for 10 and 20 dB HL for bone-conducted stimuli and for 20 and 30 dB HL for air-conducted stimuli. Statistical analyses were made using repeated-measures and mixed-model analysis of variances (ANOVAs). Using one-way ANOVAs, infant ASSR thresholds were compared across frequency for both BC and AC separately. The same analysis was carried out for AC VRA. A two-way ANOVA compared infant BC thresholds between the ASSR and VRA testing methods. Using a three-way analysis of variance (ANOVA) infant and adult ASSR thresholds were compared for both AC and BC across frequency. One-way ANOVAs compared infant AC thresholds between the ASSR and VRA testing methods at individual frequencies. A one-way ANOVA was used to compare the ABG across frequencies for VRA. One-way ANOVAs were also used to compare the infant ABG found for ASSR and VRA, as well as the ABG for ASSR between infants and adults at individual frequencies.  62  Comparisons between infant and adult ASSR amplitudes were made using a three-way ANOVA for both AC (20 and 30 dB HL) and BC (10 and 20 dB HL). The phase values provided by MASTER were adjusted by adding 90º to provide the onset phase (John & Picton, 2000b). Conversion to phase delay involved subtracting the onset phase value from 360º. Phase-delay values that differed ≥180º from similar measures were ―unwrapped‖ by adding 360º to their value (John & Picton, 2000b). Mean phase delays were calculated using only significant phase values. Missing data were dealt with by using case-wise deletion. Thus, individual frequencies were analyzed separately in cases where it would optimize the number of subjects included in the analysis. To compensate for the multiple separate one-way ANOVAs the significance criterion was corrected using the Bonferroni adjustment technique. For analyses that were conducted on individual frequencies the criterion was p<0.0125. For analyses conducted only across frequency, the criterion was corrected to p<0.025. The criterion for statistical significance was p<0.05 for multi-factorial ANOVAs. GreenhouseGeisser epsilon adjustments for repeated measures were made where appropriate. Newman-Keuls posthoc comparisons were carried out for significant main effects and interactions with a criterion of p<0.05. Results were only analysed if at least five subjects contributed to a mean.  2.3 Results 2.3.1 The 90th Percentile Figure 2.2 shows the cumulative percent of present AC and BC ASSR responses across frequency for both infants and adults. The intensity at which at least 90% of the participants had responses is considered to be the ―normal level‖ for both AC and BC AM2 stimuli. For infants, the normal levels were lower for BC compared to AC at 500 and 1000 Hz, but were similar at 2000 and 4000 Hz. For adults the normal levels for BC were better than AC at 500 Hz but similar at 1000, 2000 and 4000 Hz. The results showed that at least 90% of infants had thresholds at or better than 21, 14, 30 and 25 dB HL for BC ASSR at 500, 1000, 2000 and 4000 Hz, respectively. Similarly, 90% of infants had thresholds at or better than 32, 29, 27 and 28 dB HL for AC ASSR at the same frequencies, respectively. For the adults, 90% had thresholds at or better than 24, 28, 16 and 16 dB HL for BC ASSR and 35, 26, 18 and 18 dB HL for AC ASSR at 500, 1000, 2000 and 4000 Hz, respectively. VRA  63  thresholds were not included as their range across individuals for both air- and bone-conducted stimuli was much smaller, with the majority of responses present at 10 dB HL.  Figure 2.2 The cumulative percent of responses present for infant and adult BC and AC ASSRs across intensity at each frequency. The dotted lines represent the intensities at which 90% of the subjects for each condition have present responses.  2.3.2 ASSR Thresholds The mean and standard deviations of both air- and bone-conduction thresholds for behavioural and physiological methods are shown in Table 2.4. The findings for infants and adults for each testing method and for air- and bone-conducted stimuli are described below in separate sections.  64  2.3.2.1  Infants  2.3.2.1.1  Infant BC ASSR  Infant BC ASSR thresholds were lowest (i.e., best) for 1000 Hz, followed by 500 and 4000 Hz, with the highest thresholds occurring at 2000 Hz. Mean ASSR thresholds at 500, 1000, and 4000 Hz were 9.6, 11.5, and 6 dB better, respectively, than the mean threshold at 2000 Hz (Table 2.2). Results of a one-way ANOVA comparing mean BC ASSR thresholds across carrier frequency indicated a significant frequency effect [F (3, 54)= 7.803, p=0.002]. Neuman-Keuls post-hoc comparisons revealed that ASSR thresholds at 2000 Hz were significantly worse compared to ASSR thresholds at 500 Hz (p=0.003), 1000 Hz (p<0.001) and 4000 Hz (p=0.016). There were no significant differences for 500 vs. 4000 Hz (p=0.346) or 500 vs. 1000 Hz (p=0.230). The difference in ASSR thresholds between 1000 and 4000 Hz approached significance (p=0.086).  65  Table 2.4 Bone- and Air-conduction mean thresholds (in dB HL) and standard deviations by age, method and frequency.  Age  Method  Infants  ASSR  Mode Bone  Air  VRA  Bone  Air  Adults  ASSR  Bone  Air  Audiometry1 Bone  Air  Variable Mean SD N Mean SD N Mean SD N Mean SD N Mean SD N Mean SD N Mean SD N Mean SD N  500 11.1 11.0 19 27.9 8.0 14 -0.5 9.1 19 5.8 6.7 12 20.0 8.9 11 28.2 9.8 11 1.3 6.4 8 2.5 6.6 8  Frequency (Hz) 1000 2000 9.0 20.5 9.1 10.5 20 20 22.3 20.0 8.3 7.8 13 13 2.5 2.5 12.6 8.5 4 20 5.0 5.7 5.8 6.5 4 13 15.5 12.7 12.1 7.9 11 11 22.7 18.2 4.7 6 11 11 1.3 2.5 4.4 4.6 8 8 4.4 3.1 5.6 4.6 8 8  4000 14.5 11.4 20 22.3 9.0 14 5.6 7.0 18 2.7 9.0 10 8.5 10.4 11 18.0 8.9 11 0 6.0 8 1.3 3.5 8  1  The warbled-tone audiometry results of 8 adults are included in this table. They were not analysed, but instead provide a behavioural comparison, using the same equipment and stimuli, for the infant VRA results.  2.3.2.1.2  Infant AC ASSR  AC ASSR thresholds were highest for 500 Hz. The thresholds were very similar for 1000, 2000 and 4000 Hz (Table 2.4). Results of a one-way ANOVA comparing mean AC ASSR thresholds across frequencies revealed a significant frequency effect [F(3,33)=3.997, p=0.016], with a criterion of p<0.025. Post-hoc testing showed that the thresholds at 500 Hz were significantly higher than the thresholds at 1000 Hz (p=0.017) and 2000 Hz (p=0.017) and approached significance at 4000 Hz  66  (p=0.032). There were no significant differences in pair-wise comparisons between the 1000, 2000 and 4000 Hz (p=0.617 to 0.901). 2.3.2.2  Adults  2.3.2.2.1  Adult BC ASSR  For adult BC ASSRs, mean thresholds decreased with increasing carrier frequency. The threeway ANOVA comparing age, stimulus presentation mode and frequency for ASSR identified significant interactions [F(3,60)=3.902, p=0.013)]. Post-hoc comparisons revealed that the thresholds at 500 Hz were significantly worse than at 4000 Hz (p=0.002) and that the higher threshold at 1000 vs. 4000 Hz approached significance (p=0.056). No significant differences were found when comparing 500 Hz vs. 1000 Hz (p=0.547), 500 vs. 2000 Hz (p=0.117), 1000 vs. 2000 Hz (p=0.316) or 2000 vs. 4000 Hz (p=0.260). 2.3.2.2.2  Adult AC ASSR  For adult AC ASSRs, mean thresholds were worse for 500 Hz compared to 2000 and 4000 Hz. Thresholds at 1000 Hz were midway between the lowest and higher frequency. Post-hoc comparisons for the significant age x stimulus presentation mode x frequency interaction showed that the 500-Hz mean adult threshold was significantly worse than the adult thresholds at both 2000 Hz (p=0.013) and 4000 Hz (p=0.013); thresholds at 1000 Hz were not significantly different from 500 Hz (p=0.116), 2000 Hz (p=0.628) or 4000 Hz (p=0.653). 2.3.2.3  Comparing Adult and Infant ASSRs The overall threshold pattern for bone-conducted stimuli between infant and adults (see Table  2.4) suggests that infant ASSR thresholds were better in the low frequencies, and worse in the high frequencies compared to adults, whereas AC thresholds for infants and adults appear to be very similar. Despite this trend in the mean values, results of a three-way ANOVA comparing BC and AC ASSR thresholds at 500, 1000, 2000 and 4000 Hz between infants and adults indicated no significant main effect of age [F(1,20)= 0.0831, p=0.776]. There were significant main effects of both stimulus presentation mode [F(1,20)= 25.511, p<0.001] and frequency [F(3,60)= 4.235, p=0.009]. Significant interactions were also present between stimulus presentation mode x age [F(1,20)=0.081, p=0.779], frequency x age [F(3,60)=5.676, p=0.002), stimulus presentation mode x frequency [F(3,60)=5.734,  67  p=0.002], as well as stimulus presentation mode x frequency x age [F(3,60)=3.902, p=0.013)]. Post-hoc testing revealed that the significant three-way interaction was mainly due to frequency-dependent differences between AC and BC thresholds within each age group, as there were no significant differences between infant and adult thresholds at any carrier frequency for either stimulus presentation mode. The p-values for infant-adult threshold comparisons at the same frequency (e.g., adult vs. infant 500-Hz threshold) ranged from 0.282 to 1.000 for air- and bone-conducted stimuli.  2.3.3 VRA Thresholds Mean behavioural BC thresholds for infants were frequency dependent (500 < 2000 < 4000 Hz). However, the mean threshold at 500 Hz was only 6 dB better than that at 4000 Hz. A two-way ANOVA comparing BC thresholds and testing method revealed a significant effect of frequency [F(2,26)= 3.725, p=0.038]. Post-hoc analyses indicated that the difference between BC VRA thresholds at 500 and 4000 Hz was significant (p=0.009). The difference between thresholds at 500 vs. 2000 Hz (p=0.089) and 2000 vs. 4000 Hz (p=0.153) approached significance. The mean behavioural AC thresholds for infants were similar at 500 and 2000 Hz and slightly better at 4000 Hz. A separate one-way ANOVA comparing AC mean thresholds at 500, 2000 and 4000 Hz indicated no significant effect of frequency [F(2,16)= 1.600, p=0.233].  2.3.4 The Relationship between ASSR and VRA Thresholds The effects of test method are consistent across the infant data. As shown in Table 2.5, infant ASSR thresholds are 7–22 dB higher (i.e., poorer) than VRA thresholds at 500, 2000 and 4000 Hz. Figure 2.3 compares ASSR and VRA BC thresholds in each subject indicating the range of difference scores across individuals. The difference between the ASSR and VRA threshold of an individual varied from -10 to 50 dB.  68  Table 2.5 Descriptive statistics of difference scores (in dB) between individual infant ASSR and VRA thresholds.  Difference Score in dB (ASSR threshold-VRA threshold)  Conduction Variable Mode Bone Mean SD Median N Air Mean SD Median N  500 11.9 16.0 10 16 22.0 6.3 20 10  Frequency (Hz) 2000 4000 16.5 7.0 14.1 13.5 10 13 17 15 14.5 19.1 11.3 16.2 20 28 11 9  Figure 2.3 BC ASSR and BC VRA threshold comparisons in infants at 500, 2000 and 4000 Hz.  The differences between the VRA and ASSR thresholds were significant between the stimulus presentation modes, as well as across frequency. A two-way ANOVA comparing BC ASSR and VRA thresholds found significant main effects of testing method [F(1,13)= 13.844, p=0.003] and frequency  69  [F(2,26)= 4.499, p=0.021], as well as a significant interaction between method x frequency [F(2,26)= 3.725, p=0.038]. Post-hoc comparisons indicated that VRA provides significantly better thresholds than ASSR at 500 (p<0.001), 2000 (p<0.001), and 4000 Hz (p=0.031). As discussed previously, there are also frequency-dependent threshold differences between the testing methods that also account for the results of the ANOVA. Likewise, using one-way ANOVAs for individual frequencies, VRA also provided better AC thresholds than ASSR at 500 (p<0.001), 2000 (p=0.002), and 4000 Hz (p=0.008).  2.3.5 The Air-Bone Gap (ABG) 2.3.5.1  AC vs. BC As discussed earlier, a three-way ANOVA comparing BC and AC ASSR thresholds of infants  and adults across frequencies revealed a significant interaction between stimulus presentation mode, frequency and age [F (3,60)=3.902, p=0.013]. Post-hoc comparisons revealed a number of sources for this interaction. Infant AC thresholds at 500 and 1000 Hz were significantly higher than the BC thresholds at the same frequencies. There were no significant differences between infant AC and BC thresholds at 2000 (p=0.907) and 4000 Hz (p=0.372). For adults at 500 Hz, AC thresholds were close to being significantly higher than BC thresholds (p=0.052). The adult thresholds at 4000 Hz were also significantly different between stimulus presentation modes (p=0.010). There were no significant differences between adult AC and BC thresholds at 1000 (p=0.199) and 2000 Hz (p=0.268). For VRA, AC thresholds also tended to be slightly elevated compared to BC thresholds at 500 and 2000 Hz, although the 4000-Hz threshold was actually worse by BC than AC. Given these differences seen between AC and BC thresholds between methods and age groups, further analyses investigated the ABG. 2.3.5.2  The Presence of ABGs Figure 2.4 summarizes the mean ABGs for infant ASSR and VRA thresholds, as well as adult  ASSR thresholds. Only for infant ASSR thresholds at 500 and 1000 Hz, does the mean ABG exceed 10 dB-- the commonly accepted criterion for an ABG of clinical significance. However, a number of individuals across conditions had ABGs greater than 10 dB, as indicated by the large standard deviations. The percentages of participants, who presented with an ABG of any size, as well as the median ABGs across frequency, are shown in Table 2.6.  70  *  Figure 2.4 Comparisons of the mean air-bone gap (AC minus BC thresholds; ±1 SD) at each frequency for adult ASSR and infant ASSR and VRA thresholds. Note: A ―*‖ indicates that N<5 for that condition and thus the data were not included in the graph. Table 2.6 Air-Bone Gap (ABG; ≥ 5 dB) for infants and adults by method: reported as the percentage of participants who presented with an ABG of a given size (percentages shown are not cumulative). Median ABGs are shown for each condition.  Frequency (Hz) 500  1000  2000  4000  Age  Method/ABG  10  20 or 30  10  20 or 30  10  >10 5 -10 15- 25  Infant  ASSR  31%  62%  33%  50%  33%  0%  Median ABG  20 dB  VRA  60%  Median ABG  10 dB  ASSR  45%  Median ABG  10 dB  Adult  15 dB 10%  --  0 dB --  -18%  36% 10 dB  36%  5 dB 0%  0 dB 18%  55% 10 dB  23% 46%  8%  8%  0 dB 9%  18%  55%  15 dB  Note: ‗--‗ indicates that N<5 for the condition and thus data were not included in the graph.  71  The mean ABGs for ASSR thresholds were compared at each individual frequency to maximize the number of infant subjects included in the analysis. At 500 Hz, the mean ABG was nearly significantly larger for infants compared to adults (p=0.038) at a criterion of p<0.012511. There were no significant differences between infant and adult ABGs at the other frequencies (p= 0.115 to 0.650). The adult behavioural data were not analyzed or compared to the other data as the mean ABGs were so small, ranging from 1–3 dB across frequency. Most subjects had ABGs of 0 or 5 dB, a few had 5 dB ABGs (BC>AC), and only one subject presented with an ABG of 10 dB at one frequency. Another one-way ANOVA, comparing mean infant VRA ABGs across 500, 2000 and 4000 Hz revealed a significant main effect of frequency. The mean ABG at 500 Hz was significantly larger than the mean ABG at 4000 Hz (p=0.009) but not significantly different compared to 2000 Hz (p=0.403). The difference in the ABG between 2000 and 4000 Hz approached significance (p=0.020). The infant ABG can also be compared between testing methods, although the number of subjects changed with frequency making it necessary to analyze frequencies individually. A one-way ANOVA at 500 Hz showed that the mean ASSR ABG was nearly significantly higher than the ABG for VRA (p=0.087). There were no significant differences in ABGs found between methods for either 2000 (p=0.619) or 4000 Hz (p=0.106).  11  The Fisher exact test was also conducted for 500 Hz and revealed a significantly greater number of occurrences of ABGs in infants relative to adults (p=0.036). The outcomes for 1000–4000 Hz were similar to the ANOVA results.  72  2.3.6 ASSR Amplitudes  Figure 2.5 Mean ASSR amplitudes (±1 SD) at each carrier frequency for infants and adults.  Figure 2.5 displays the mean BC and AC ASSR amplitudes for infants and adults at all frequencies. It is clear that the mean amplitudes generally increase with increasing stimulus intensity. There were no significant differences in BC or AC ASSR mean amplitudes between infants and adults at any of the carrier frequencies; however, there were a few notable trends. Results of a three-way ANOVA comparing mean BC ASSR amplitudes of infants and adults for each carrier frequency at 10 and 20 dB HL revealed significantly larger amplitudes at 20 dB HL compared to 10 dB HL [F(1,29)=44.604, p<0.001], a significant main effect of frequency [F(3,87)=5.812, p=0.001] and a significant interaction of frequency x age [F(3,87)=15.517, p=0.026]. Post-hoc testing indicated that the significant frequency effect, as well as the interaction between frequency and age, was due to the tendency for infant amplitudes to be greater for the low frequencies compared to the high frequencies. For infants, significantly greater amplitudes were found at 500  73  compared to 2000 Hz at 10 dB HL, and compared to both 2000 and 4000 Hz at 20 dB HL. Infant mean amplitudes at 1000 Hz were also greater than the mean amplitudes for 2000 Hz at 10 and 20 dB HL and for 4000 Hz at 20 dB HL. Results of a three-way ANOVA comparing mean AC ASSR amplitudes of infants and adults for each carrier frequency at 20 and 30 dB HL revealed significantly larger amplitudes at 30 dB HL compared to 20 dB HL [F(1,23)=15.517, p<0.001], a significant main effect of frequency [F(3,69)=7.052, p<0.001] and no significant interactions. Post-hoc comparisons indicated that the significant frequency effect was due to the amplitudes for 2000 Hz being significantly higher compared to the other frequencies (p=0.001 to 0.030). A two-way ANOVA was also performed to compare infant response amplitudes to bone- and airconducted stimuli at 20 dB HL. This analysis showed a significant main effect of frequency [F(3,36)=5.364, p=0.004] and stimulus presentation mode approached significance [F(1,12)=4.733, p=0.050]. There was a significant interaction between frequency x conduction mode [F(3,36)=3.963, p=0.015]. Post-hoc testing found that the mean response amplitude at 500 Hz was significantly larger for BC than for AC (p=0.002). The same trend was present at 1000 Hz, but it did not reach significance (p=0.052). There were no significant differences between BC and AC mean amplitudes at 2000 (p=0.972) or 4000 Hz (p=0.168).  2.3.7 Phase Figure 2.6 shows the mean phase delays (±1 SD) for infants and adults with a few notable trends. Previous literature suggests that phase delay becomes shorter with increasing frequency and intensity. In the current data, there was a tendency for phase delay to decrease with increasing intensity. The phase delays at 1000 Hz also appear to be longer than the delays at either 2000 or 4000 Hz, with little difference between 2000 and 4000 Hz. This trend does not appear to carryover to 500 Hz. The difference between 1000 Hz and the higher frequencies however, may be confounded by the circularity of phase measurements and corrections involving unwrapping. Phase delays were unwrapped if they differed by more than 180 degrees from similar measures (John & Picton, 2000b). Similar measures were identified as significant phase delays at other frequencies for the individual or of neighbouring subjects. Unwrapping was most often necessary for the 1000-Hz phase delays. For example at 1000 Hz,  74  13 out of 69 and 24 out of 45 significant phase delays were adjusted for infants and adults, respectively. No further phase delay analyses were conducted.  Figure 2.6 Mean ASSR phase delays (±1 SD) at each carrier frequency for infants and adults.  75  CHAPTER 3: Discussion and Conclusion  76  3.1 Discussion 3.1.1 ASSR 3.1.1.1  Infant ASSR Thresholds This is the first study to estimate BC ASSR thresholds to AM2 stimuli. The mean BC thresholds  of the current study were 9 and 11 dB HL at 500 and 1000 Hz and 21 and 15 dB HL at 2000 and 4000 Hz. The 2000-Hz mean threshold was significantly worse than the thresholds at the other frequencies. Similar to previous studies, as shown in Table 3.1, frequency-dependent differences in BC sensitivity were noted. For example, the present study‘s mean BC ASSR thresholds fall within the range of means for 1000–4000 Hz reported by Small and Stapells (2008a) for a younger (0–11 months) and older group (12–24 months) of infants (the age range of the present study overlaps both of these age groups). Mean BC ASSR thresholds in the present study are slightly better than the values reported by Small and Stapells (2008a) at 500 Hz, which may be a consequence of differing sample size, stimuli and EEG noise. The findings of the current study replicate the frequency-dependent pattern that is now well established in the literature for both ABR and ASSR BC thresholds—the 500 and 1000 Hz thresholds are typically better than the 2000-Hz threshold (Foxe & Stapells, 1993; Nousak & Stapells, 1992; Small & Stapells, 2006, 2008a; Stapells & Ruben, 1989; Vander Werff et al., 2009).  77  Table 3.1 Comparisons of mean ASSR thresholds (dB HL) for both infants and adults across frequency from the present and previously published studies.  BC Study  Stimulus  Small & Stapells, 2008a  MM  Age 0–11 months  Frequency (Hz) 500 1000 2000 14 5 26  4000 14  12–24 months  22  13  26  13  Present  AM2  6.5–18 months  11  9  21  15  Ishida et al., 2011 study A  MM  Adult  22  14  22  28  MM  Adult  22  16  16  11  Small & Stapells, 2005  MM  Adult  22  26  18  18  Small & Stapells, 2008a  MM  Adult  31  24  20  16  Present  AM2  Adult  20  16  13  9  Lins et al., 1996*  AM  1–10 months  34  22  17  20  Swanepoel & Steyn, 2005  MM  3–8 weeks  37  34  34  30  Rance et al., 2005  MM  0.5–3 months  32  33  24  28  Luts et al., 2006*  MM  < 3 months  37  35  29  32  Rance & Tomlin, 2006  MM  6 weeks  40  --  --  32  Van Maanen & Stapells, 2009  cos3 sinusoids°  6–66 months  41  37  31  22  Ribeiro et al., 2010*  MM  term: 1–8 days  39  28  24  27  Alaerts et al., 2010*  MM  < 3 months  40  38  30  33  Hatton & Stapells, 2011*§  AM  1–9.5 months  48  --  --  --  Present  AM2  6.5–18 months  28  22  20  22  D’haenens et al., 2008†  AM2/FM  Adult  23  17  13  17  Alaerts et al. 2010  MM  Adult  29  21  18  23  Present  AM2  Adult  28  23  18  18  study B  AC Study  °Van Maanen & Stapells (2009) used cosine3-windowed sinusoids, which are nearly equivalent to AM2 stimuli. The younger infants they assessed were found to have similar thresholds, but these results were not compared to the present study. * Threshold data were converted from dB SPL to dB HL using standard ANSI-1996 adjustment values. Lins et al. (1996) thresholds converted using RETSPLs for TDH-39 earphones (ANSI, 1996). § Threshold estimate based on average of stimulus conditions (dichotic multiple and monotic single) †As D‘Haenens et al. (2008) reported two sets of mean thresholds (test and re-test values), the present study‘s means were compared to the average of these values.  78  Although this is the first study to estimate AC ASSR thresholds to AM2 stimuli using the MASTER system, two other studies by Van Maanen and Stapells (2009, 2010) have assessed infant thresholds using comparable stimuli with the Intelligent Hearing Systems SmartEP-ASSR. The mean AC ASSR thresholds reported in the current study were 28 dB HL at 500 Hz and between 20 and 22 dB HL for 1000-4000 Hz (Table 3.1). The mean threshold at 500 Hz was significantly poorer compared to the higher frequencies. The results of this study are similar to the thresholds reported by Lins et al. (1996), but they used different stimuli and earphones (TDH-39). The results are also comparable to the thresholds reported by Ribeiro et al. (2010) and Rance et al. (2005) for some frequencies. However, Van Maanen and Stapells (2009) may be the most appropriate comparison for the AC ASSR results of this study, as the authors used similar stimuli and procedures and conducted mASSR in infants of a similar age range (i.e., the older group consisted of infants who were > 6 months and was used for comparison). Compared to Van Maanen and Stapells (2009), the mean thresholds of the current study are at least 11 dB lower (i.e., better) at all frequencies except at 4000 Hz, for which the studies found equivalent mean thresholds (Table 3.1). Given the size of the differences in thresholds (at 500–2000 Hz) between Van Maanen and Stapells (2009) and the present study, they are likely the result of a combination of differences between the studies. For example, the sample size is 50% larger in Van Maanen and Stapells (2009), which could account for a portion of the difference. Another source for variation may be the stimuli, which are similar between the studies but not identical and as such may have influenced response detection in a frequency-dependent manner. The present study also has better thresholds compared to many of the other comparison studies across frequency (Swanepoel & Steyn, 2005; Luts et al., 2006; Rance & Tomlin, 2006; Alaerts et al., 2010; Hatton & Stapells, 2011). Overall, these interstudy differences may be attributable to age range, stimuli, number of stimuli (i.e., four for monotic versus eight for dichotic presentation), EEG noise (also see Infant ASSR Amplitudes), recording system (e.g., detection algorithms) and procedures (e.g. calibration and stopping criteria). It is also possible, that the infant participants of the present study had atypically better auditory sensitivity compared to the participants of the other studies. Both the present study and Lins et al., (1996) found only the thresholds at 500 Hz to be significantly higher compared to the other frequencies. This pattern of results is in agreement with some ASSR studies that have reported no significant difference between thresholds at 1000, 2000 and 4000 Hz (Luts et al., 2006; Rance et al., 2005), but differs somewhat from Van Maanen and Stapells (2009), who found that mean thresholds improved significantly with increasing frequency. Despite some 79  variability in thresholds at 1000–4000 Hz, the literature supports that the 500-Hz threshold is consistently elevated compared to the other frequencies for a variety stimulus types (Lins et al., 1996; Luts et al., 2006; Ribeiro et al., 2010; Swanepoel & Steyn, 2005; Van Maanen & Stapells, 2009; Alaerts et al., 2010; Hatton & Stapells, 2011; Alaerts et al., 2010). However, it should be noted that the 500-Hz thresholds for both the present study, Lins et al. (1996) and Rance et al. (2005) are more than 10 dB better than the highest thresholds estimated by these other studies. This may be accounted for by higher EEG noise at 500 Hz for these other studies. Additionally, detection of the threshold-level response using MM stimuli (used by the majority of studies shown in Table 3.1) may be negatively impacted by MM stimulus factors, such as increased interactions and phase settings, at 500 Hz. Brennan, Brooke and Stevens (2012) investigated MM phase settings in neonates and found that they significantly affect ASSR response amplitudes and that optimal settings may differ by frequency. Overall, the source of the 500-Hz threshold elevation in infants is thought to be related to poorer neural synchronization in the lower-frequencies (Rance et al., 1995) and/or structural limitations of the outer and middle-ear, which decrease the efficiency of lower-frequency sound transmission to the cochlea (Keefe et al., 1993). 3.1.1.2  Infant ASSR Amplitudes Frequency-dependent differences are also seen for BC ASSR amplitudes. There was a tendency  for infant mean response amplitudes (at 10 and 20 dB HL) to be larger for the low frequencies, which ranged from 12–18 nV, compared to the higher frequencies, which ranged from 5–10 nV. ASSR amplitudes were larger at 500 and 1000 Hz compared to 2000 Hz at 10 dB HL, and 4000 Hz at 20 dB HL. Mean amplitudes were also larger at 500 compared to 2000 Hz at 20 dB HL. For both the present study and Small and Stapells (2008a), who reported mean BC amplitudes for younger and older infants, amplitudes tended to be larger for the lower frequencies compared to the higher frequencies. Averaging the response amplitudes of the younger and older infant groups, Small and Stapells (2008a) found that amplitudes were approximately 20, 30 and 40 nV for 10, 20 and 30 dB HL respectively at 500 Hz. The mean amplitudes were similar for 4000 Hz, nearly 10 nV larger for 1000 Hz, and 10 nV smaller at 2000 Hz at each intensity. For the present study, the mean amplitudes at and below an intensity level of 30 dB HL were all less than 20 nV. These amplitude differences between the studies are interesting because AM2 stimuli are expected to evoke slightly larger or at least similar amplitude responses to MM stimuli (John et al., 2004; John et al., 2001a; John et al., 2002a). Despite these smaller BC response amplitudes, the mean thresholds of the current study are similar or better than  80  the thresholds reported by Small and Stapells (2008a). Lower overall EEG noise levels may have enabled better identification of near-threshold responses in the higher frequencies for the present study. Even though the studies used the same no-response criterion for noise (<11nV), the mean EEG noise levels of the present study were less than 5 nV across frequency. The frequency-dependent trends noted for the mean BC amplitudes were also evident in comparisons between BC and AC amplitudes of the present study. The 500-Hz mean amplitude was significantly larger for BC compared to AC. Also, the mean amplitude was nearly significantly larger for BC compared to AC at 1000 Hz. For AC ASSR amplitudes, there was a tendency for infant mean response amplitudes (at 20 and 30 dB HL) to be larger at 2000 Hz, which ranged from 11–16 nV, compared to the other frequencies, which ranged from 5–11 nV. Van Maanen and Stapells (2009) elicited responses with mean amplitudes of approximately 15 nV–20 nV, across frequency for stimuli presented at 30 and 40 dB HL, respectively. The present study found similar mean AC amplitudes at the same stimulus levels. However, even with similar response amplitudes, the present study found better AC thresholds. Van Maanen and Stapells (2009) used an equivalent no-response criterion for noise; however, the actual mean EEG noise levels across frequency for the present study ranged from 4–6 nV at the intensity level below threshold. These EEG noise levels may be lower than those found by Van Maanen and Stapells (2009) which could have allowed, in part, for the detection of smaller amplitude responses and better mean AC thresholds. However, this may not have been the case as the older subjects tested in Van Maanen and Stapells (2009) were sedated and were unlikely to be noisier than the participants of the present study. Also, as found by John and colleagues (2004), the standard deviation from the mean amplitudes decreased with increasing carrier frequency for infants for both air- and bone-conducted stimuli. However, it should be noted that there was large inter-subject variability in amplitudes for both infants and adults, as has been reported elsewhere in the literature (e.g., Picton et al., 2005; D‘haenens et al., 2008), and as such the mean may over-estimate the central tendency of the amplitude data. 3.1.1.3  Adult ASSR Thresholds The adult BC ASSR mean thresholds of this study decreased with increasing frequency and  ranged from 20 (at 500 Hz) to 9 (at 4000 Hz) dB HL. The threshold at 500 Hz was significantly higher  81  compared to 4000 Hz which is consistent with the thresholds reported by Ishida et al. (2011). The mean thresholds of the present study differ from Ishida et al. (2011) study B (N=10) by only 1–3 dB at each frequency. Although study A of Ishida et al. (2011) reported thresholds for a greater number of adults (N=16), the higher frequency thresholds seem elevated in comparison to their study A, Small and Stapells (2005, 2008a) and the present study and thus were not used for comparison. The thresholds of the present study are 2–10 dB and 7–11 dB better than thresholds reported by Small and Stapells (2005) and Small and Stapells (2008a), respectively (Table 3.1). As discussed previously in regards to infant BC and AC thresholds, the source of these inter-study differences is difficult to parse out, but may reflect differences in stimuli and residual noise levels. The variability of the mean thresholds seems similar across studies as well, as the standard deviations overlap substantially. Similar to the BC findings, the adult AC ASSR mean thresholds of this study decreased with increasing frequency, with a range of 28 (at 500 Hz) to 18 (at 4000 Hz) dB HL. The threshold at 500 Hz was significantly poorer compared to the higher frequency thresholds. Comparable thresholds are found in the literature. For example, Alaerts, Luts, Van Dun, Desloovere, and Wouters (2010) reported adult AC ASSR thresholds that ranged from 18–29 dB HL (when converted from dB SPL using standard adjustment values; Table 3.1). Using more comparable stimuli to the present study, D‘Haenens et al. (2008) found thresholds that ranged from 13– 23 dB HL12, which differ from the present study‘s thresholds by approximately 5, 6, 5, and 1 dB for 500, 1000, 2000 and 4000 Hz, respectively (Table 3.1). The standard deviations were also comparable. The present study, and the previously discussed AC ASSR studies, found that the 500-Hz thresholds tend to be elevated compared to thresholds for higher frequencies. There is also a tendency for the 500-Hz BC ASSR threshold to be elevated compared to the higher frequencies, as seen in the current study and Ishida et al. (2011). There are three reasons that have been proposed to account for elevated physiological thresholds at 500 Hz in adults (John et al., 2004; D‘haenens et al., 2007). Firstly, neural synchrony is thought to be less efficient at lower frequencies and thus may reduce the size of a response at 500 Hz (Rance et al., 1995). Secondly, ambient noise may mask the response at 500 Hz and make detection more difficult (Lins et al., 1996). Lastly, the 500-Hz response may be more optimally recorded with different stimulus parameters than the other frequencies (Herdman & Stapells, 2003).  12  This range is based on the calculated average of the Test and Retest ASSR thresholds reported by D‘haenens et al. (2008)  82  3.1.1.4  Adult ASSR Amplitudes For the present study, adult BC ASSR amplitudes were similar across frequency and ranged from  8–18 nV. These results are consistent with Small and Stapells (2008a), who reported adult amplitudes ranging from approximately 15–30 nV at comparable intensities. For adult AC ASSR, amplitudes ranged from approximately 7–19 nV and were significantly larger at 2000 Hz compared to the other frequencies. This finding is similar to that reported by D‘haenens et al. (2008), who found adult AC ASSR amplitudes that ranged from 11–30 nV at comparable stimulus levels. However, they also noted significantly lower amplitudes at 4000 Hz when compared across a larger intensity range. The explanation for these frequency-dependent differences in amplitudes between studies may simply be related to the large variability seen with amplitudes in general.  3.1.2 Behavioural 3.1.2.1  Infant VRA The mean BC VRA thresholds of the current study ranged from approximately -1 to 6 dB HL.  The mean BC thresholds at 500 Hz were significantly lower (i.e., better) than thresholds at 4000 Hz. As shown in Table 3.2, this relationship between the 500- and 4000-Hz mean thresholds was also found by Hulecki and Small (2011). However, in contrast to the present study, Hulecki and Small (2011) also found that thresholds for 500 and 1000 Hz were better compared to 2000 and 4000 Hz. Another difference between the studies is that the mean thresholds across frequency are between approximately 7 and 12 dB better for the present study compared to the younger infant (7-15 months) mean thresholds reported by Hulecki and Small (2011).  83  Table 3.2 Comparisons of mean behavioural thresholds (dB HL) for both infants and adults across frequency from the present and previously published studies.  BC Behavioural Study  7-15 months  Frequency (Hz) 500 1000 2000 9.0 9.8 14.3  4000 14.0  18-30 months  10.4  8.6  13.0  16.5  Present  6.5-18 months  -0.5  2.5  2.5  5.6  Haughton & Pardoe, 1981  Adult  -3  0  -2  -2  Present  Adult  1.3  1.3  2.5  0.0  Parry et al., 2003  8–12.5 months  16.4  13.3  7.1  6.4  Nozza & Henson, 1999  6.7–9.4 months  17.2  --  6.8  --  Present  6.5–18 months  5.8  5.0  5.7  2.7  (pure tone)  Adult  3.4  2.1  0.4  2.2  (warbled tone)  Adult  4.1  1.9  -0.5  0.8  Franklin et al., 2011*  Adult  7.5  4.0  4.5  3.0  Franklin et al., 2009  Adult  13.8  9.8  10.2  8.2  Present  Adult  2.5  4.4  3.1  1.3  Hulecki & Small, 2011  Age  AC Behavioural Study  Dockum & Robinson, 1975  Notes: ―--― indicates that no data are available for this condition. Data at 1000 Hz were not compared to other studies as fewer than five participants completed this condition. * Values for Franklin et al. (2011) were estimated from a figure.  The present study found mean AC behavioural thresholds that ranged from 3–6 dB HL across frequency, with no apparent frequency-dependent trends. The mean AC thresholds of the present study are fairly similar to those found by Parry et al., (2003) for both 2000 and 4000 Hz and Nozza and Henson (1999) at 2000 Hz (Table 3.2) but differed significantly at 500 Hz. The 500-Hz thresholds are better for the present study by approximately 10 and 11 dB compared to Parry et al., (2003) and Nozza and Henson (1999), respectively, resulting in little difference in threshold across frequency. The findings of the current study also appear to agree with Rance et al. (2005), who investigated the relationship between single-stimuli ASSR and VRA AC thresholds for both normal-hearing infants (N=285) and infants with sensorineural hearing loss or auditory neuropathy (N=290). The VRA  84  thresholds of Rance et al. (2005) were not specifically reported, but can be calculated from other reported measurements. For normal-hearing infants, Rance et al. (2005) showed that the mean difference scores (ASSR minus VRA thresholds) were nearly equivalent to their reported ASSR thresholds. By subtracting the difference scores from the mean ASSR values, VRA thresholds were found to be approximately 1.9, 1.3, 1.9, and 0.6 dB HL at 500, 1000, 2000 and 4000 Hz, respectively. Despite this consistency with the present study, Rance and colleagues (2005) did find a larger range of AC thresholds. It is reasonable to suspect that these between-study differences in mean behavioural thresholds may be due to a number of possible differences between the studies, related to (1) sample characteristics (2) procedural differences, and (3) stimuli (including type, duration, and calibration). The ages of the current study‘s participants are comparable to both of the studies discussed earlier. Parry et al. (2003) assessed infants of a more restricted age range, between 8–12.5 months, Hulecki and Small (2011) tested infants who were 7–30 months of age; importantly, they did not find any significant threshold differences in younger infants (7–15 months) and older infants (18–30 months). However, the sample sizes for both of the other studies are each nearly twice the size of the corresponding sample size in the present study. This suggests that the other studies may more accurately represent the large response variability of the normal-hearing infant population. It is conceivable that for the present study, the smaller group of infants may have had better auditory thresholds, which was also noted as a potential factor for the physiological findings of this study. Additionally, the infants of the present study may also have had greater task responsiveness compared to the greater population. Similar to the present study, all three of the previously discussed infant behavioural studies used either insert earphones or a B-71 bone oscillator and employed a threshold search that was not restricted by a lower search boundary, such as 10 or 20 dB HL. The procedure for the present study was the same as that used by Hulecki and Small (2011) -- they both employed manual VRA and used a common clinical procedure adapted from Widen (2000). Nozza and Henson (1999) used an automated auditory test system (IVRA system by Intelligent Hearing Systems) and Parry et al. (2003) used manual VRA with two experienced pediatric audiologists, but the findings between the studies were comparable. The step-size was 10 dB for the current study and Hulecki and Small (2011), but was 5 dB for Parry et al. (2003) and Nozza and Henson (1999). Each study also recorded response reliability and required more than one response at each threshold level for confirmation. Bias of the tester is always a concern for  85  VRA, but the strict protocol, recording of control trials, and evaluation of infant response reliability are designed to reduce the possible subject influence on audiological assessment (Gravel, 2002). The procedure used by the present study and Hulecki and Small (2011) allowed for use of warbled-tone as well as narrow-band noise to better maintain the infant‘s interest, while Parry et al., (2003) used exclusively warbled-tone stimuli. These stimulus variations are of no consequence to thresholds, as research shows that the stimulus type has no impact on an already conditioned response (Primus & Thompson, 1985; Thompson & Folsom, 1985; Thompson & Thompson, 1972). All three studies used stimuli of durations ranging from 1-4 seconds; however, it is unlikely that this variability in stimulus duration would affect thresholds. Stimulus calibration may be one source of inter-study threshold variability. The ANSI S3.6-2004 tolerances for AC and BC measurements are ±3 dB, therefore stimulus calibration might differ by as much as 6 dB. The range of thresholds found in this study should be considered in conjunction with other studies to provide an improved understanding of the potential range of responses that can be acquired from normal-hearing infants of this age group. The VRA thresholds in the study ranged from -10 to 20 dB HL with mean thresholds ranging from -1 to 6 dB HL for both air- and bone-conduction. These findings overlap with the ranges of responses reported by previous studies. Parry et al. (2003) identified the range of thresholds as -5 to 25 dB HL for AC and Hulecki and Small (2011) found thresholds ranging from -10 to 30 dB HL for BC. In Gravel (1989), the range of normal hearing for AC, using TDH-50 earphones, for 6 to 12-month old infants was 5–30 dB HL. The most likely explanation for the differences between mean thresholds between this study and other studies appears to be related to better auditory sensitivity and/or task responsiveness of the participants. This conclusion is supported by the typical adult findings described in the following section, which rule out any significant calibration issues. 3.1.2.2  Adult Behavioural For the current study, the mean adult behavioural thresholds to air- and bone-conducted warbled-  tone stimuli ranged from 0–4 dB HL, which are within the normal range expected for normal-hearing young adults. Despite the fact that there are no published studies on behavioural BC thresholds to warbled-tone stimuli in adults, there are studies using pure-tone stimuli, which provide a reasonable  86  comparison13. Consistent with the current study, previous studies have found that mean BC thresholds hover around 0 dB HL for mastoid placement (for review of studies see Haughton and Pardoe, 1981). For example, Haughton and Pardoe (1981) reported mean BC thresholds of -3.0, -0.5, -2.0 and -2.5 dB HL at 500, 1000, 2000 and 4000 Hz, respectively14. This range of thresholds is also comparable to previously reported AC findings for both warbled-tone and pure-tone stimuli. For the frequencies of interest (500 to 4000 Hz), Dockum and Robinson (1975) found mean AC warble-tone thresholds of adult subjects (N=198) that ranged from -1 to 4 dB HL and pure-tone thresholds ranging from 0–3 dB HL. Likewise, Franklin et al. (2011) reported AC thresholds to pure-tone stimuli for normal hearing young adults (N=49) from 250-8000 Hz with means of approximately 4–7 dB HL at the frequencies of interest. Franklin et al., (2009) found somewhat higher mean warbled-tone thresholds ranging from 8–14 dB HL, for smaller group of 25 adult participants.  3.1.3 Comparing Infants to Adults In the present study, there was a tendency for infant and adult mean thresholds to differ for BC ASSRs, but not for AC ASSRs or for behavioural AC or BC testing. Infant mean threshold values to bone-conducted stimuli were 7–9 dB better for 500 and 1000 Hz and 6–8 dB worse for 2000 and 4000 Hz compared to adults, as shown in Table 3.3. Compared to adults, infants tended to have better thresholds in the lower frequencies and poorer thresholds at 2000 Hz, which is consistent with the findings of Small and Stapells (2006, 2008a). However, these infant-adult differences did not reach significance in the present study. In contrast, Small and Stapells (2008a) found that significant infantadult threshold differences existed in the low frequencies, but not in the high frequencies.  13  Studies that have assessed AC thresholds to warbled-tone stimuli in normal-hearing adults suggest there is only a small and clinically insignificant difference in mean thresholds acquired using pure-tone versus warbled-tone stimuli (Franklin et al., 2011; Franklin et al., 2009; Dockum & Robinson, 1975). 14 Converted from dB re:1 μN to dB HL using values from ANSI (1996)  87  Table 3.3 Within study comparisons of infant-adult differences in ASSR thresholds (dB; infant minus adult) from the present and previously published studies.  BC ASSR Study  Frequency (Hz) 500  1000  2000  4000  Small & Stapells, 2008a*  -13  -15  6  2.5  Small & Stapells, 2006†  -6  -24  8  4  Present  -8.9  -6.5  7.8  6  Lins et al., 1996  6  0  3  -2  Present  -0.3  -0.4  1.8  4.3  AC ASSR Study  *The averages of the two infant age groups from Small & Stapells (2008a) were calculated for each frequency for comparisons to the adult data. †Small & Stapells (2006) compared infant data to previously collected adult data originally published in Small & Stapells (2005).  In contrast to the BC results, the AC results for infants and adults were very similar, differing by only 0–4 dB across frequency. Lins et al. (1996) also compared the AC thresholds (in dB SPL) of infants (N=21; age range: 1–10 months) and adults and found them to be similar across frequency, except at 500 Hz, where infant thresholds were significantly higher15. In contrast, findings of some studies propose that adult thresholds are at least 10 dB better than the thresholds of infants (Van Maanen & Stapells, 2009; Rance & Rickards, 2002; Luts et al., 2006; Alaerts et al., 2010). Luts et al. (2006) reported adult data that are very similar to the present study across frequencies; however, their infant data are elevated by approximately 8–13 dB compared to the present study. One potential reason for this is that their infant participants were very young (<3 months of age) and possibly more affected by developmental factors. Developmental changes in threshold can also be found in other studies involving newborns less than six weeks of age (Rance & Tomlin, 2006; Rickards et al., 1994; Savio et al., 2001; John et al., 2004). This conclusion is challenged by Van Maanen and Stapells (2009) who did not find large differences between their younger (≤ 6 months) and older (>6 months) infant subjects or a significant correlation between age and threshold. Qian et al. (2010) reached similar conclusions in their longitudinal study comparing AC threshold changes in ABR and ASSR (both calibrated in dB nHL). They found that there were no significant threshold changes during the period from 0–6 months of age.  15  The study assessed two groups of infants, one of which had considerably higher ambient noise during testing (the Havana group). Thus only the Ottawa infant and adult data were used for comparisons to the current study.  88  It may be that changes during this time, such as sound pressure at the eardrum and transmission through the middle-ear system, interact with each other in such a way that there is no overall difference in thresholds (in dB HL or dB nHL) between infants and adults (Sininger & Hyde, 2009). The behavioural thresholds for infants and adults were similar in the present study, a finding that differs from the literature which suggests that behavioural thresholds of infants are elevated by 10–15 dB compared to the thresholds of adults (e.g., McDermott & Hodgson, 1982; Berg & Smith, 1983; Olsho et al., 1988; Parry et al., 2003). This suggests that the expected non-sensory influences of VRA may have been minimized for this group of infants, which allowed for a more effective test and lower threshold estimates or that these particular infant participants had better than average hearing sensitivity.  3.1.4 The Relationship between ASSR and VRA Although subject factors (e.g., age) and methodological factors (e.g., step-size), were controlled in this study, the direct comparisons between ASSR and VRA are challenging because of their inherent differences in stimuli, response detection, and even the neuronal pathways involved (Picton et al., 2003). It is however still valuable to explore their relationship to gain further insight as to what the results of the ASSR might predict about the responses to VRA. Difference scores between ASSR and VRA thresholds are used to establish correction factors that serve this purpose. Although it is necessary to examine the relationship in infants with hearing loss to actually posit correction factors, determining the difference scores for normal-hearing infants is the first step in defining the relationship between these testing methods. The present study found infant difference scores (ASSR minus VRA thresholds) that ranged from 7–16 dB for bone-conducted stimuli for individual infants. Previously, Hulecki and Small (2011) hypothesized infant difference scores by comparing BC VRA thresholds and BC ASSR thresholds (Small & Stapells, 2006, 2008a) in different groups of infants and found comparable difference scores of 5–13 dB across frequency. For the present study, using air-conducted stimuli, ASSR-VRA difference scores ranged from 15–22 dB. These ranges are consistent with previous findings on the relationship between ASSR and behavioural audiometry in normal-hearing infants. Additionally, they overlap with the meta-analyses on adult AC ASSR by Tlumak, Rubinstein and Durrant (2007), who reported mean AC threshold differences (ASSR minus behavioural thresholds) ranging from 11-17 dB across frequency. Typically, AC ABR thresholds are within 5–10 dB of VRA thresholds (Stapells, 2010a).  89  Compared to AC ABR thresholds, AC ASSR thresholds can be elevated by 9–17 dB in normal-hearing infants (Van Maanen & Stapells, 2010). This leads to the conclusion that AC ASSR thresholds are conceivably up to 14–27 dB higher than VRA thresholds; this range of difference scores overlaps considerably with the present study. Rance et al. (2005) also found comparable AC ASSR-VRA difference scores, ranging from 22–31 dB. Also of interest is that Chou et al. (2011) compared AC ASSR thresholds for normal-hearing infants under 13 months of age to conditioned play audiometry thresholds at 23–48 months of age and reported similar mean differences (19–28 dB for 500-4000 Hz). Many studies have examined the relationship between ASSR and behavioural thresholds in normalhearing and hearing-impaired infants (e.g., Cone-Wesson et al., 2002b; Alaerts et al., 2010; Rance & Rickards, 2002; Steuve & O‘Rourke, 2003; Luts et al., 2006), but did not present the findings for the normal-hearing infants separately from the infants who had hearing loss, making comparisons with the current study difficult. The relationship between testing methods may differ depending on hearing status such that normal-hearing individuals may have larger difference scores than individuals who have a hearing loss (e.g., Rance et al., 1995, Dimitrijevic et al., 2002; Tlumak et al., 2007; Van Maanen & Stapells, 2010; Chou et al., 2011). Difference scores reported for infants with a wide range of hearing thresholds tend to be lower than the differences scores of the present study; for example, they were found to range from 6–9 dB and 12–17 dB by Luts et al. (2006) and Alaerts et al. (2010), respectively. This is because response detection at threshold in normal-hearing individuals is more likely to be obscured by low-level EEG and ambient noise compared to an individual with a hearing loss, who requires higher intensity levels to elicit threshold responses. Additionally, in individuals with cochlear hearing loss, physiological recruitment increases the amplitude of near-threshold responses, making them easier to detect (Picton et al., 2005). For these reasons, correction factors cannot be determined solely from ASSR-VRA difference scores in normal-hearing infants. For the present study, the relationship between ASSR and VRA thresholds was not constant. For example, an individual with a relatively higher ASSR threshold did not necessarily present with a similarly elevated VRA threshold. Figure 2.2 shows that this relationship between BC ASSR and VRA thresholds for individual subjects was highly variable across frequency. For example, at 500 Hz, 16 infants had the following ASSR-minus-VRA difference scores (number of infants shown in brackets): 10(3), 0(3), 10(4), 20(3), 30 (1) and 40(2) dB. This suggests that for the present study, ASSR thresholds in normal-hearing infants were often not predictive of their behavioural thresholds and the group means may not accurately reflect individual variability. This finding is of interest because ASSR-minus-VRA  90  difference scores for BC have not been assessed in previous studies; however, it is recognized that difference scores are most valuable when they represent a wide range of hearing sensitivities and inform the estimation of behavioural thresholds from ASSR thresholds in individuals with hearing loss.  3.1.5 The Air-bone Gap This is the first study to directly assess the differences between AC and BC thresholds (i.e., the air-bone gap) in the same infants by both behavioural and physiological measures. The present study found a mean ABG of 17 and 14 dB at 500 Hz and 1000 Hz, respectively, compared to -1 and 7 dB at 2000 and 4000 Hz16, respectively, for infant ASSR thresholds. In comparison, adult ASSR ABGs ranged from 5–8 dB across frequency. Although, the infant-adult differences in ASSR ABGs did not reach statistical significance, it is arguable that a 14–17 dB infant ABG has practical significance. The ABGs for infant behavioural thresholds were smaller compared to ASSR thresholds and there were no significant differences between infants and adults. The current study suggests that a maturational ABG can also be found at 500 and 1000 Hz for ASSR in older infants aged 6.5 to 18 months. Given that the clinical standard for a significant ABG is a difference between AC and BC thresholds of greater than 10 dB, the percentage of participants in this study who presented with an ABG of greater than 10 dB is notable (Table 2.6). Of the 13 infants who provided both AC and BC ASSR thresholds at 500 and 1000 Hz, 50–62% of them presented with a 20 or 30 dB ABG and 31–33% of them presented with a 10 dB ABG. This relationship between infant AC and BC ASSR in the low frequencies has confirmed what has previously been inferred by comparing frequency-dependent threshold differences in BC and AC ASSRs obtained in different studies (e.g., Small and Stapells, 2006, 2008; Van Maanen & Stapells, 2009; Lins et al., 1996). These ASSR findings are also consistent with ABR research conducted by Vander Werff et al. (2009) who used tone-evoked ABRs to estimate thresholds at 500 and 2000 Hz in infants under three months of age. They found that in the absence of a conductive component, normal-hearing infants presented with almost a 15 dB ABG 16  The 4000-Hz ASSR ABG that appears for both infants and adults might be partially explained by acoustic radiation of the bone-conducted signal into the unoccluded ear canal, such that the sound is also heard by AC. Lightfoot (1979) found that the acoustic output of a B-71 bone vibrator for a 0 dB HL audiometer dial setting results in a measurable 6 dB HL acoustic output at the concha in adults. Hearing a bone-conducted signal by AC makes the BC threshold appear better than it actually is and thus creates a false ABG equivalent to the size of the air-conducted portion of the stimulus. However, variability in the ASSR ABGs of individuals suggests that acoustic radiation either varies across individuals or that there are other sources for this apparent ABG. This ABG could also be a result of individual variability around the average dB HL values. Additionally, it is not clear why neither the infants nor adults in the current study present with a similar gap behaviourally.  91  at 500 Hz, whereas adults had only a small mean ABG (2.5 dB) at the same frequency. They also did not find an ABG at 2000 Hz for either infants or adults. For the present study, the magnitude of the ASSR ABG gap was not reproduced in the infant behavioural findings, for which 60% of the infants had an ABG of 10 dB and only one of the 10 infants had an ABG larger than 10 dB at 500 Hz. In contrast to infant findings, only 18% of adults had ABGs larger than 10 dB for ASSR and no adult had an ABG of larger than 10 dB for behavioural testing. However, the behavioural results suggest that infants actually may exhibit a maturational ABG that is smaller than 10 dB. The infant behavioural ABG at 500 Hz (8 dB) was significantly larger than at 4000 Hz (-2 dB), but not significantly different from 2000 Hz (3 dB). The ABGs at 2000 and 4000 Hz were not significantly different. It is notable that despite statistically similar mean ABGs across frequency, at 500 Hz, the ABG tended to be larger compared to the higher frequencies which is consistent with the prediction made by Hulecki and Small (2011), that infants exhibit a behavioural ABG of approximately 6–8 dB at 500 Hz. They also suggested that there is the potential for an ABG as large as 17 dB, but as mentioned the ABGs did not exceed 10 dB for the majority of infants in the present study. As Hulecki and Small (2011) proposed, the maturational ABG may be reduced behaviourally because threshold differences between frequencies are smaller for behavioural versus physiological testing methods. The frequency-dependent thresholds reported for VRA to bone-conducted (Hulecki & Small, 2011) and airconducted (Nozza and Henson,1999; Parry et al., 2003) stimuli not only suggest an ABG at the lower frequencies, but also imply the existence of a reverse ABG (BC>AC) at the higher frequencies. The findings of the present study provide limited statistical support for the existence of a low-frequency maturational VRA ABG and did not show a reverse ABG in the higher frequencies. It is difficult to parse out why the ASSR assessment demonstrated a larger ABG compared to VRA because, aside from methodological differences between the testing methods, there are inherent differences. For example, as noted by Picton et al. (2003, 2005), the neuronal activity present at perceptual threshold may not be detectable by physiological measures which could partially or fully explain why physiological thresholds are elevated compared to behavioural thresholds. Additionally, VRA is susceptible to the influence of non-sensory factors, such as behaviour and motivation, which combined with a 10-dB step size and a small sample size, may minimize the frequency-dependent differences in AC and BC sensitivity. This could account for the maturational ABG being smaller behaviourally than it is for  92  ASSR. Thus, the ABG disparity between ASSR and VRA most likely stems from the limitations of VRA. As addressed by Hulecki and Small (2011), the presence of a maturational ABG has clinical implications. Firstly, it is possible that normal-hearing infants may be misdiagnosed as having a conductive hearing loss, given that the size of the ABG commonly seen in normal-hearing infants exceeds the current criterion for a significant ABG (>10 dB). However, this seems improbable as for both behavioural and physiological testing BC is only conducted in the case of elevated AC thresholds (e.g. BCEHP, 2008; OIHP, 2008). Secondly, there is the risk that infants who have conductive hearing loss and use amplification may be over-amplified if the maturational component of their loss is not considered. In such cases, it may be prudent to adjust the prescription amplification to more conservatively estimate the conductive component, by excluding the maturational contribution at 500 Hz. For the ASSR, the determination of an ABG that reflects a conductive loss may be more accurate if the criterion is adjusted to reflect the fact that many normal-hearing infants have an ABG larger than 10 dB at 500 and 1000 Hz. This is an important detail to consider if the ASSR is used clinically to quantify ABGs in the future. However, currently, early hearing programs in Canada test BC when AC thresholds are elevated to determine whether or not there is sensorineural involvement (OIHP, 2008; BCEHP, 2008). As addressed by Stapells (2010a), the goals of the programs are not to initially quantify a conductive component, but to direct management. Stapells (2010a) acknowledges that the precise quantification of the ABG is restricted in multiple ways for clinical ABR. These restrictions include (1) the limited data on the relationship between AC and BC thresholds and effective masking levels for isolating the cochleae, (2) the limited range of output from the bone oscillator, and (3) the limited precision offered by using a 10 dB step-size for threshold search. All of these reasons also apply to ASSR. Although the findings of this study have improved our understanding of the relationship between AC and BC thresholds using two different assessment methods, other limitations, combined with clinical priorities and time constraints, may restrict clinical feasibility of estimating the precise physiological ABG in infants with hearing loss. If quantification of the ABG becomes a goal of ASSR assessments, future research on the ABG would be necessary in infants with hearing loss, especially conductive loss, to better inform estimates of the true ABG.  93  3.1.6 Infant-Adult Differences Although the present study did not find a significant difference between the ASSR and behavioural thresholds of infants and adults, there are patterns in the data that suggest infants and adults differ in some respects. It is a common expectation that the developmental changes that occur between infancy and adulthood influence tests of auditory sensitivity. These influences can be separated into two categories: sensory factors, which refer to anatomical and physiological features of the auditory system that have the potential to effect the transmission of sound to the cochlea, and; non-sensory factors, which relate to testing variables and central processes, such as behaviour and attention. As there is more research on the sensory-related influences of physiological testing compared to behavioural testing methods, this section will focus on physiological testing. It should be kept in mind that these sensory factors are likely also applicable to behavioural testing to some extent, but are more difficult to substantiate. 3.1.6.1  Sensory Factors One of the clear trends found in this study was for infant BC thresholds. Bone-conducted stimuli  are thought to reach the cochlea by numerous pathways, each of which is involved to a different extent (for review see Stenfelt and Goode, 2005). In infants, less is understood about the dominant mechanisms responsible for BC hearing. One theory is that bone-conducted stimuli are more intense when coupled to an infant skull compared to an adult skull (Stuart, Yang, Stenstrom, 1990). In adults, sound energy is dissipated across the cranial bones. However, for the infant skull, energy transfer between the cranial bones is thought to be attenuated by the presence of soft sutures. This leads to a more intense stimulus because the sound energy is concentrated on the temporal bone. Foxe and Stapells (1993) estimated that bone-conducted stimuli were more intense by 9-17 dB for 500 Hz and 12.8 dB for 2000 Hz and hypothesized this was due to the smaller size of an infant temporal bone relative to an adult. However, these skull differences suggest that all frequencies are more intense for an infant than an adult, but may be more so in the low frequencies. It is feasible that skull characteristics at any point along the BC pathways may alter the stimulus in a frequency-dependent way and account for the threshold patterns seen in infant BC thresholds.  94  Development of the outer and middle ear is also thought to contribute to infant-adult threshold differences, although it is not known whether or not their development substantially affects infant BC hearing. In research on the occlusion effect in infants, it has been suggested that the outer ear has only a very minor role in BC hearing, in contrast to adult findings (Small et al., 2007; Small & Hu, 2011). By definition, AC hearing includes transmission through the outer and middle ear are thus is likely more affected by their maturational changes. Differing resonant properties of the middle ear suggest that the middle ears of infants and adults may transmit some frequencies more effectively than others, but this is a less pronounced maturational difference in infants older than 6 months of age (Holte et al., 1991; Keefe et al., 1993). The infant ear canal is also smaller (Feigin et al, 1989; Jirsa & Norris, 1978) and has unique resonant properties of its own (Keefe et al., 1993) that make an auditory stimulus more intense at the tympanic membrane of an infant relative to that of an adult (Kruger & Ruben, 1987). However, even when the effects of the outer ear canal are removed by calibrating the stimulus at the tympanic membrane, infants still show auditory immaturity, primarily in the higher frequencies (≥ 4000 Hz) (Rance & Tomlin, 2006; Sininger et al., 1997). Although middle-ear immaturity attenuates the higher frequencies in very young infants (Keefe et al., 1993), it is not considered to be of sufficient influence to fully explain the high-frequency immaturity. The cochlea is also largely mature by full-term birth, both structurally and functionally and is thus unlikely to be the source of infant-adult threshold differences (Abdala & Sininger, 1996; Abdala et al., 1996; Abdala, 2000; Bredberg, 1968; Lavigne-Rebillard & Pujol, 1987; Lavigne-Rebillard & Pujol, 1988; Lavigne-Rebillard & Bagger-Sjöbäck, 1992; Pujol, 1985; Pujol et al., 1991). Instead, Rance and Tomlin (2006) and Sininger et al., (1997) argue that neural immaturity, particularly in the higher frequencies, is the source of these infant-adult threshold differences. Developmental elements of the neural system include increased myelination and synaptic efficiency (Moore et al., 1995; Moore & Lithicum, 2001; Ponton et al., 1996). Changes in these elements may improve the auditory system‘s efficiency. Additionally, this inefficiency may also extend to neural synchronization. As both the toneevoked ABR and the ASSR depend on neural synchronization to detect a response, some studies suggest that its immaturity could elevate physiological thresholds (Eggermont & Salamy, 1988; Gorga et al., 1989; Rance et al., 1995; Sininger et al., 1997). Rance and Tomlin (2006) and Sininger et al. (1997) concluded that neural factors contribute to infant-adult differences to air-conducted stimuli by 3–28 dB at 500 Hz and by 24–38 dB at 4000 Hz.  95  Because AC and BC have the same neural pathways, the same neural factors must also contribute to BC hearing. Small and Stapells (2008a) explain that, compared to adults, bone-conducted stimuli presented to infants must be increased in intensity across the frequency range, but even more so in the lower frequencies to account for the frequency-dependent findings of infant BC hearing. The equivalent thresholds found in this study for infant-adult AC ASSRs do not appear to reflect the relationship between age and thresholds that is typically reported for studies calibrated at the tympanic membrane (dB pe SPL). As discussed in the review of the literature, infant-adult differences may be obscured when using adult-referenced calibration and thus not accounting for infant-adult differences in sound pressure in the ear canal. Sininger et al. (1997) investigated ABR thresholds in infants and found that when thresholds account for sound pressure in the ear canal, the infant-adult relationship substantially changes at 4000 and 8000 Hz. Higher frequency stimuli are higher in SPL in the ear of an infant compared to an adult. However, both Sininger et al. (1997) and Rance and Tomlin (2006) suggest that infant-adult differences above 500 Hz remain even when sound pressure in the ear canal is accounted for; they suggest that this is due to neural immaturity. There is evidence that the 4000-Hz threshold for physiological measures matures later than lower frequency thresholds (e.g., Marcoux et al., 2011, Sininger et al., 1997; Klein, 1984). Despite this, the present study did not find an infant-adult difference at 4000-Hz, which may be attributable to variability or the age of the subjects. The present study was also conducted with older infants, who may have had greater neural maturity, resulting in similar infant-adult thresholds. 3.1.6.2  Non-sensory Factors Behavioural testing methods are more affected by non-sensory factors than physiological  methods, and even more so for infants. However, Nozza (1995) and Nozza and Henson (1999) found frequency-dependent differences between infant MRLs and adult thresholds and concluded that the majority of the difference in the low frequencies was actually accounted for by sensory processing. This conclusion is consistent with the findings of Parry et al. (2003). But this theory does not explain the results of the current study, which found only one frequency-dependent difference in thresholds and no differences between infant and adult thresholds. In this case, it appears that either sensory factors tended to contribute less to infant-adult differences than found in previous studies and/or that the infant participants of the present study had better than average hearing sensitivity.  96  An ongoing challenge in the audiological assessment of the pediatric population is that often the information required must be obtained within a limited, and often unknown, time frame. With a change of state, a child may quickly change from testable to untestable, whether he/she is required to sleep or attend to a task. As a result, test protocols are prioritized for efficiency. As such, a 10 dB step-size for VRA is commonly recommended for clinical use (e.g., Diefendorf & Gravel, 1996; Widen et al., 2000; BCEHP, 2008; OIHP, 2008) and is considered to provide a reasonably accurate threshold estimation for this population. This step size is also reasonable for clinical physiological assessments (BCEHP, 2008; OIHP, 2008). However, a 10 dB step-size limits the accuracy of threshold measurements and may not reveal more subtle threshold differences. For this reason, the findings of this study are limited by the 10dB step size used for both testing methods. The standard deviations of the means for the current study were often around 10 dB. It follows that differences between means that are smaller than 10 dB may not be quite large enough to become statistically significant. However, these differences may be real effects with potential clinical implications.  3.2 Conclusion The present study compared the thresholds of normal-hearing infants and adults to AM2 stimuli using different sound conduction modes (AC and BC) and testing methods (ASSR and behavioural). The adult group functioned as a control group for making age-related comparisons. Comparing BC thresholds across different testing techniques, and in relation to AC thresholds, has the potential to increase our understanding of what constitutes a normal response for an infant. In addition to providing normal response levels, the study achieved the following two objectives: (1) to compare BC and AC sensitivity for both methods to better quantify the maturational ABG and; (2) to compare BC responses between VRA and ASSR testing methods to further define their relationship. These comparisons have not been explored in previous studies. Mean thresholds to AM2 stimuli were found for normal-hearing infants and adults. Similar to previous ASSR studies (Small & Stapells, 2008a, 2006), frequency-dependent differences in BC sensitivity were found—the 500- and 1000-Hz thresholds were better than the 2000-Hz threshold. For AC ASSR, only the thresholds at 500 Hz were higher than the other frequencies, as was also found by Lins et al., (1996). In contrast, Van Maanen and Stapells (2009) found significant differences at all frequencies, with mean thresholds that improved with increasing frequency. For the present study, there  97  was a tendency for infant and adult mean ASSR thresholds to differ for BC, but not for AC. The adult BC thresholds of the current study were comparable to the thresholds reported by Ishida et al. (2011) and Small and Stapells (2005) and were worse in the lower frequencies and better in the higher frequencies compared to infants. Similar to infants, adult AC thresholds were poorer for 500 Hz compared to the higher frequencies, which is a comparable trend to that found by D‘Haenens et al. (2008) and Alaerts et al. (2010). For the lower frequencies, infants tended to have better BC ASSR thresholds and larger response amplitudes compared to adults. AC ASSR thresholds and response amplitudes were similar between infants and adults across frequency. Mean behavioural thresholds to air- and bone-conducted stimuli ranged from -1 to 6 dB for infants and adults, with little difference across frequency. For infants, the BC and AC thresholds of the current study are better at most frequencies compared to previous VRA studies (Hulecki & Small, 2011; Parry et al., 2003; Nozza and Henson, 1999). The objective to compare AC and BC thresholds for both testing methods in the same study was motivated by the frequency-dependent threshold differences reported in previous studies which suggested the existence of a maturational ABG for both the ASSR (e.g. Small & Stapells, 2008a, 2006; Van Maanen & Stapells, 2009) and VRA (Hulecki & Small, 2011; Parry et al., 2003; Nozza & Henson, 1999), but only assessed either AC or BC. The present study supports the presence of a clinically significant maturational ABG in the low frequencies for infant ASSRs. This is consistent with Vander Werff et al. (2009) who found a large infant ABG at 500 Hz using tone-evoked ABRs. In the present study, the infant behavioural ABG also tended to be largest at 500 Hz, as was posited by Hulecki and Small (2011), but was too small to be practically significant. As expected, there was no significant ABG found for adults with either testing method. Directly assessing the association between BC ASSRs and BC VRA at a similar maturational time period in individual infants was the other main objective of the present study. The findings indicated that BC ASSR thresholds were, on average, 7–16 dB poorer compared to BC VRA thresholds. The results have confirmed the previously reported 10–15 dB offset between physiological and behavioural estimates and have provided a better understanding of the large variability in the relationship between ASSR and VRA thresholds in normal-hearing infants.  98  The findings of this study improve upon the current knowledge regarding normal levels, the maturational ABG and the BC ASSR-VRA relationship in normal-hearing infants. The clinical implications of these findings should encourage future research in infants with hearing loss.  3.2.1 Clinical Implications The current study is the first to compare behavioural and physiological responses to both air- and bone-conducted stimuli in normal-hearing infants and adults. These threshold data contribute to our understanding of what constitutes a normal response for infants and adults for the procedures used in this study. These results support the use of AM2 stimuli for threshold estimation using mASSR in infants as they provide thresholds that are equivalent to or better than the thresholds reported in the current literature using AM or MM stimuli. Based on the findings of this study, normal levels are posited. It should be noted that these are preliminary normal levels that are based on a small sample size and are much more strict that previously published normal levels for infants, especially at 500 Hz (e.g., BC: Small & Stapells, 2008a; AC: Van Maanen & Stapells, 2009). The normal levels for AC and BC AM2 stimuli, for infants 6.5–18 months of age, at 500, 1000, 2000, and 4000 Hz, respectively are as follows: (i) BC ASSR: 21, 14, 30 and 25 dB HL, and (ii) AC ASSR: 32, 29, 27, and 28 dB HL. For adults, the normal levels are as follows: (i) BC ASSR: 24, 28, 16, and 16 dB HL, and (ii) AC ASSR: 35, 26, 18 and 18 dB HL. The present study likely has more strict normal levels due to the small range of ASSR thresholds for the infants in this sample. It follows that the infant participants of this study may have atypically better auditory sensitivity compared to the larger population. These normal levels for AM2 stimuli should be confirmed in a larger sample size and evaluated for accuracy in infants with confirmed hearing loss. They could then be used in conjunction with these other studies for an initial screening stage of a diagnostic protocol to determine whether an infant‘s AC and/or BC thresholds are within the normal limits or elevated. Normal levels were not posited for infant VRA as the threshold search for the present study was more rigorous than is typically efficient clinically. The search for VRA threshold usually ends when the infant has responded to the stimulus at the standard normal criterion of 20 dB HL; thus normal levels based on a search procedure below 20 dB HL, as in the present study, would be too strict and of no practical application. The VRA results for the present study do suggest that some normal-hearing infants are capable of  99  responding reliably at normal adult threshold levels, and even as low at -10 dB, across frequency for both AC and BC stimuli. This study found that mean BC difference scores for ASSRs and VRA ranged from 7–16 dB with standard deviations of 13–16 dB. This relationship varied greatly across individuals, which suggests that there is a wide range of physiological results that correspond to normal behavioural responses. The findings of the present study also indicate a mean maturational ABG for infant ASSRs of 17 and 14 dB at 500 and 1000 Hz, respectively; this suggests that although an ABG greater than 10 dB is the standard criterion consistent with a conductive loss, it is not an appropriate criterion for ASSRs in infants. The majority of infants (13/16) had ABGs of at least 20 dB and 3/16 had ABGs of 30–40 dB. The results also suggest that behaviourally it is not necessary to account for infant-adult differences in AC and BC thresholds or to account for a maturational gap when using a 10 dB step-size. Given these findings, in infants with normal hearing, clinical consideration of the maturational ABG seems warranted when using the ASSR, but not for VRA.  3.2.2 Future Research This study provides valuable information about the relationships between various threshold measures and testing techniques for normal-hearing infants and functions as preliminary step to further exploration of these topics. Additional research is needed to independently verify the relationships reported in this study for infants with varying types and degrees of hearing loss. The relationship between testing methods needs to be investigated in these infant populations to reach the ultimate goal of establishing eHL correction values for BC ASSR thresholds. Also, the results of this study suggest that the current clinically significant ABG criterion (>10 dB) is not appropriate to use for infant ASSR. However, to posit a more suitable ABG criterion, the ABG must be explored in the population of primary interest—that is, in infants with confirmed conductive and sensorineural hearing loss.  100  References Abdala, C. (2000). Distortion product otoacoustic emission (2f1-2f2) amplitude growth in human adults and neonates. Journal of the Acoustical Society of America, 107, 446-456. Abdala, C., & Dhar, S. (2010). Distortion product otoacoustic emission phase and component analysis in human newborns. Journal of the Acoustical Society of America, 127(1), 316-325. Abdala, C., & Keefe, D. H. (2006). Effects of middle-ear immaturity on distortion product otoacoustic emission suppression tuning in infant ears. Journal of the Acoustical Society of America, 120(6), 3832-3841. Abdala, C., Sininger, Y., Ekelid, M., & Zeng, F. (1996). Distortion product otoacoustic emission suppression tuning curves in human adults and neonates. Hearing Research, 98, 38-53. Abdala, C., & Sininger, Y. S. (1996). The development of cochlear frequency resolution in the human auditory system. Ear & Hearing, 17(5), 374-385. Alaerts, J., Luts, H., & Wouters, J. (2007). Evaluation of middle ear function in young children: Clinical guidelines for the use of 226- and 1,000-hz tympanometry. Otology & Neurotology, 28(6), 727-732. Alaerts, J., Luts, H., Van Dun, B., Desloovere, C., & Wouters, J. (2010). Latencies of Auditory SteadyState Responses Recorded in Early Infancy. Audiology & Neurotology, 15(2), 116-127. Allen, G. W., & Fernandez, C. (1960). The mechanism of bone conduction. Annals of Otology, Rhinology, & Laryngology, 69, 5-25. ANSI. (1996). American National Standard Specifications for Audiometers (ANSI S3.6-1996). New York: ANSI. Anson, B. J., & Donaldson, J. A. (1981). The temporal bone. In B. J. Anson, & J. A. Donaldson (Eds.), Surgical anatomy of the temporal bone (pp. 3-25). Philadelphia: W.B. Saunders Company.  101  Aoyagi, M., Kiren, T., Kim, Y., Suzuki, Y., Fuse, T., & Koike, Y. (1993). Optimal modulation frequency for amplitude-modulation following response in young children during sleep. Hearing Research, 65(1-2), 253-261. Bagatto, M., Moodie, S., Scollie, S., Seewald, R., Moodie. S., Pumford, J., Liu, K.P. (2005). Clinical protocols for hearing instrument fitting in the Desired Sensation Level method. Trends in Amplification, 9 (4), 199-226. Balfour, P. B., Pillion, J. P., & Gaskin, A. E. (1998). Distortion product otoacoustic emission and auditory brain stem response measures of pediatric sensorineural hearing loss with islands of normal sensitivity. Ear & Hearing, 19(6), 463-472. Barany, E. (1938). A contribution to the physiology of bone conduction. Acta Oto-Laryngologica, 26(Suppl. 26), 1-233. Berg, D., & Smith, M. (1983). Behavioral thresholds for tones during infancy. Journal of Experimental Child Psychology, 35, 409-425. Bredberg, G. (1968). Cellular pattern and nerve supply of the human organ of corti. Acta Otolaryngologica, Suppl. 236, 10-135. Brennan, S. K., Brooke, R. E., & Stevens, J. C. (2012). The effect of varying stimulus phase between frequency and amplitude modulation on auditory steady-state responses in neonates. International Journal of Audiology, 51(2), 116-123. British Columbia Early Hearing Program (BCEHP) (2008). Diagnostic audiology protocol. Retrieved October 21, 2011, from http://www.phsa.ca/NR/rdonlyres/06D79FEB-D187-43E9-91E48C09959F38D8/40115/aDAAGProtocols1.pdf. Carhart, R. (1950). Clinical application of bone conduction audiometry. Archives of Otolaryngology— Head & Neck Surgery, 51(6), 798-808. Carhart, R. (1962). Effect of stapes fixation on bone conduction response. In H. Schuknecht (Ed.), Otosclerosis (pp. 175-197). Boston: Little, Brown and Company.  102  Chou, Y. F., Chen, P. R., Yu, S. H., Wen, Y. H., & Wu, H. P. (2012). Using multi-stimulus auditory steady state response to predict hearing thresholds in high-risk infants. European Archives of OtoRhino-Laryngology, 269(1), 73-79. Cohen, L. T., Rickards, F. W., & Clark, G. M. (1991). A comparison of steady-state evoked potentials to modulated tones in awake and sleeping humans. The Journal of the Acoustical Society of America, 90(5), 2467-2479. Cone, B., & Dimitrijevic, A. (2009). The auditory steady-state response. In J. Katz, L. Medwetsky, R. Burkard & L. Hood (Eds.), Handbook of Clinical Audiology (6th ed., pp. 322-350). Philadelphia: Lippincott, Williams & Wilkins. Cone-Wesson, B., Parker, J., Swiderski, N., & Rickards, F. (2002a). The auditory steady-state response: full-term and premature neonates. Journal of the American Academy of Audiology, 13(5), 260-269. Cone-Wesson, B., Dowel, R. C., Tomlin, D., Rance, G., & Ming, W.J. (2002b). The auditory steadystate response: Comparisons with the auditory brainstem response. Journal of the American Academy of Audiology, 13(4), 173-187. Cone-Wesson, B., & Ramirez, G. M. (1997). Hearing sensitivity in newborns estimated from ABRs to bone-conducted sounds. Journal of the American Academy of Audiology, 8(5), 299-307. Cooley, J. W., & Tukey, J. W. (1965). An algorithm for the machine calculation of complex fourier series. Mathematics of Computation, 19(90), 297-301. Cornacchia, L., Martini, A., & Morra, B. (1983). Air and bone conduction brain stem responses in adults and infants. International Journal of Audiology, 22(5), 430-437. Culpepper, B., & Thompson, G. (1994). Effects of reinforcer duration on the response behavior of preterm 2-year-olds in visual reinforcement audiometry. Ear & Hearing, 15(2), 161-167. D‘haenens, W., Dhooge, I., De Vel, E., Maes, L., Bockstael, A., & Vinck, B. M. (2007). Auditory steady-state responses to MM and exponential envelope AM2/FM stimuli in normal-hearing adults. International Journal of Audiology, 46(8), 399-406.  103  D‘haenens, W., Vinck, I., De Vel, E., Maes, L., Bockstael, A., Keppler, H., . . . Dhooge, I. (2008). Auditory steady-state responses in normal hearing adults: a test-retest reliability study. International Journal of Audiology, 47(8), 489-498. D‘haenens, W., Vinck, B. M., Maes, L., Bockstael, A., Keppler, H., Philips, B., . . . Dhooge, I. (2010). Determination and evaluation of clinically efficient stopping criteria for the multiple auditory steadystate response technique. Clinical Neurophysiology, 121(8), 1267-1278. de Jong, M., Perez, R., Adelman, C., Chordekar, S., Rubin, M., Kriksunov, L., & Sohmer, H. (2011). Experimental confirmation that vibrations at soft tissue conduction sites induce hearing by way of a new mode of auditory stimulation [Abstract]. Journal of Basic and Clinical Physiology and Pharmacology, 22(3), 55-58. Dhar, S., & Abdala, C. (2007). A comparative study of distortion-product-otoacoustic-emission fine structure in human newborns and adults with normal hearing. Journal of the Acoustical Society of America, 122(4), 2191-2202. Diefendorf, A. O., & Gravel, J. S. (1996). Behavioral observation and visual reinforcement audiometry. In S. E. Gerber (Ed.), The Handbook of Pediatric Audiology (pp. 55-83). Washington, DC: Gallaudet Press. Dillon, H., & Walker, G. (1982). Comparison of stimuli used in sound field audiometric testing. The Journal of the Acoustical Society of America, 71(1), 161-172. Dimitrijevic, A., John, M. S., Van Roon, P., Purcell, D. W., Adamonis, J., Ostroff, J., . . . Picton, T. W. (2002). Estimating the audiogram using multiple auditory steady-state responses. Journal of the American Academy of Audiology, 13(4), 205-224. Dobie, R. A., & Wilson, M. J. (1994). Objective detection of 40 hz auditory evoked potentials: Phase coherence vs. magnitude-squared coherence. Electroencephalography & Clinical Neurophysiology/Evoked Potentials Section, 92(5), 405-413. Durrant, J. D., & Hyre, R. (1993). Relative effective frequency response of bone versus air conduction stimulation examined via masking. Audiology, 32, 175-184.  104  Eby, T. L., & Nadol, J. B. (1986). Postnatal growth of the human temporal bone. Annals of Otology, Rhinology, & Laryngology, 95, 356-364. Edwards, C. G., Durieux-Smith, A., & Picton, T. W. (1985). Neonatal auditory brainstem responses from ipsilateral and contralateral recording montages. Ear & Hearing, 6(4), 175-178. Eggermont, J. J., & Salamy, A. (1988). Maturational time course for the ABR in preterm and full term infants. Hearing Research, 33(1), 35-47. Eisenberg, R. B. (1965). Auditory behavior in the human neonate. I. methodologic problems and the logical design of research procedures. Journal of Auditory Research, 5, 159-177. Everberg, G. (1968). Congenital absence of the oval window. Acta Oto-Laryngologica, 66(1-6), 320332. Feigin, J. A., Kopun, J. G., Stelmachowicz, P. G., & Gorga, M. P. (1989). Probe-tube microphone measures of ear-canal sound pressure levels in infants and children. Ear & Hearing, 10(4), 254-258. Foxe, J. J., & Stapells, D. R. (1993). Normal infant and adult auditory brainstem responses to boneconducted tones. International Journal of Audiology, 32(2), 95-109. Franklin, C.A., Franklin, T.C., & Franklin, J.L. (2011). Comparing pure-, pulsed, and warbled tone thresholds in adults 3000 and 6000 Hz. American Journal of Audiology, 20, 42-47. Franklin, C., Johnson, K., Smith-Olinde, L., & Nicholson, N. (2009). The relationship of audiometric thresholds elicited with pulsed, warbled and pulsed-warbled tones in adults with normal hearing. Ear & Hearing, 30(4), 485-487. Freeman, S., Sichel, J. Y., & Sohmer, H. (2000). Bone conduction experiments in animals – evidence for a non-osseous mechanism. Hearing Research, 146(1-2), 72-80. Fria, T. J., & Doyle, W. J. (1984). Maturation of the auditory brain stem response (ABR): Additional perspectives. Ear & Hearing, 5(6), 361-365.  105  Galambos, R., Makeig, S., & Talmachoff, P. J. (1981). A 40-hz auditory potential recorded from the human scalp. Proceedings of the National Academy of Sciences, 78(4), 2643-2647. Glaser, E. M., & Ruchkin, D. S. (1976). Principles of neurobiological signal analysis. New York: Academic Press. GN Otometrics (2009). ICS chartr EP 200 international brochure. Retrieved November 30, 2011, from http://www.otometrics.com/SiteCollectionDocuments/Product%20brochures/Chart%20EP%20int%2 0brochure.pdf. Gorga, M. P., Johnson, T. A., Kaminski, J. R., Beauchaine, K. L., Garner, C., A., & Neely, S. T. (2006). Using a combination of click- and tone burst-evoked auditory brain stem response measurements to estimate pure-tone thresholds. Ear & Hearing, 27(1), 60-74. Gorga, M. P., Kaminski, J. R., Beauchaine, K. L., Jesteadt, W., & Neely, S. T. (1989). Auditory brainstem responses from children three months to three years of age. Journal of Speech & Hearing Research, 32, 281-288. Gravel, J. S. (1989). Behavioral assessment of auditory function. Seminars in Hearing, 10, 216-228. Gravel, J. S. (2002). Potential pitfalls in the audiologic assessment of infants and young children. In R. C. Seewald & J. S. Gravel (Eds.), A sound foundation through early amplification 2001. (pp. 63-73). Basel: Phonak AG. Gravel, J. S., & Traquina, D. N. (1992). Experience with the audiologic assessment of infants and toddlers. International Journal of Pediatric Otorhinolaryngology, 23(1), 59-71. Grason-Stadler. (n.d.). GSI 61 clinical audiometer product specifications. Retrieved 12/08, 2011, from http://grason-stadler.us/products/gsi-61. Han, D., Mo, L., Liu, H., Chen, J., & Huang, L. (2006). Threshold estimation in children using auditory steady-state responses to multiple simultaneous stimuli [Abstract]. ORL; Journal for Oto-RhinoLaryngology and its Related Specialties, 68(2), 64-68.  106  Hatton, J., & Stapells, D. R. (2011). The efficiency of the single- versus multiple-stimulus auditory steady state responses in infants. Ear & Hearing, 32(3), 349-357. Haughton, P.M., Pardoe, K. (1981). Normal pure tone thresholds for hearing by bone conduction. British Journal of Audiology, 15, 113-121. Hecox, K., & Galambos, R. (1974). Brain stem auditory evoked responses in human infants and adults. Archives of Otolaryngology--Head & Neck Surgery, 99(1), 30-33. Herdman, A. T., Lins, O., Van Roon, P., Stapells, D. R., Scherg, M., & Picton, T. W. (2002a). Intracerebral sources of human auditory steady-state responses. Brain Topography, 15(2), 69-86. Herdman, A. T., Picton, T. W., & Stapells, D. R. (2002b). Place specificity of multiple auditory steadystate responses. The Journal of the Acoustical Society of America, 112(4), 1569-1582. Herdman, A. T., & Stapells, D. R. (2001). Thresholds determined using the monotic and dichotic multiple auditory steady-state response technique in normal-hearing subjects. Scandinavian Audiology, 30(1), 41-49. Herdman, A. T., & Stapells, D. R. (2003). Auditory steady-state response thresholds of adults with sensorineural hearing impairments. International Journal of Audiology, 42(5), 237-248. Hoke, M., Ross, B., Wickesberg, R., & Lütkenhöner, B. (1984). Weighted averaging — theory and application to electric response audiometry. Electroencephalography & Clinical Neurophysiology, 57(5), 484-489. Holte, L., Margolis, R. H., & Cavanaugh, R. M. (1991). Developmental changes in multifrequency tympanograms. International Journal of Audiology, 30, 1-24. Hood, J. D. (1962). Bone conduction: A review of the present position with especial reference to the contributions of dr. georg von békésy. The Journal of the Acoustical Society of America, 34(9b), 1325-1332.  107  Hoversten, G. H., & Moncur, J. P. (1969). Stimuli and intensity factors in testing infants [Abstract]. Journal of Speech & Hearing Research, 12(4), 687-702. Huizing, E. H. (1960). Bone conduction: The influence of the middle ear. Acta Oto-Laryngologica, Suppl. 155, 1-99. Hulecki, L. R., & Small, S. A. (2011). Behavioral bone-conduction thresholds for infants with normal hearing. Journal of the American Academy of Audiology, 22(2), 81-92. Humes, L. E. (1979). The middle ear inertial component of bone-conduction hearing in man. International Journal of Audiology, 18(1), 24-35. Ikui, A., Sando, I., Haginomori, S., & Sudo, M. (2000). Postnatal development of the tympanic cavity: A computer-aided reconstruction and measurement study. Acta Oto-Laryngologica, 120(3), 375-379. Intelligent Hearing Systems. (2011). SmartEP-ASSR product brochure. Retrieved November 30, 2011, from http://www.ihsys.com/Brochures/BROCHURE_SEP-ASSR.pdf. Interacoustics A/S. (n.d.). Interacoustics ASSR specifications. Retrieved November 30, 2011, from http://www.interacoustics.dk/com_en/pages/product/abr/_index.htm?prodid=60840. Ishida, I. M., Cuthbert, B. P., & Stapells, D. R. (2011). Multiple auditory steady state response thresholds to bone conduction stimuli in adults with normal and elevated thresholds. Ear & Hearing, 32(3), 373-381. Jeng, F. C., Brown, C. J., Johnson, T. A., & Vander Werff, K. R. (2004). Estimating air-bone gaps using auditory steady-state responses. Journal of the American Academy of Audiology, 15(1), 67-78. Jerger, J., Chmiel, R., Frost, J. D., Jr., & Coker, N. (1986). Effect of sleep on the auditory steady state evoked potential. Ear & Hearing, 7(4), 240-245. Jerger, J., and Hayes, D. 1976. The crosscheck principle in pediatric audiometry. Archives of Otolaryngology, 102(10), 614–620.  108  Jiang, Z. D., Zheng, M. S., Sun, D. K., & Liu, X. Y. (1991). Brainstem auditory evoked responses from birth to adulthood: Normative data of latency and interval. Hearing Research, 54(1), 67-74. Jirsa, R. E., & Norris, T. W. (1978). Relationship of acoustic gain to aided threshold improvement in children. Journal of Speech & Hearing Disorders, 43, 348-352. John, M. S., Brown, D. K., Muir, P. J., & Picton, T. W. (2004). Recording auditory steady-state responses in young infants. Ear & Hearing, 25(6), 539-553. John, M. S., Dimitrijevic, A. & Picton, T. W. (2003). Efficient stimuli for evoking auditory steady-state responses. Ear & Hearing, 24, 406-423. John, M. S., Dimitrijevic, A., & Picton, T. W. (2001b). Weighted averaging of steady-state responses. Clinical Neurophysiology, 112(3), 555-562. John, M. S., Dimitrijevic, A., & Picton, T. W. (2002a). Auditory steady-state responses to exponential modulation envelopes. Ear & Hearing, 23(2), 106-117. John, M. S., Dimitrijevic, A., van Roon, P., & Picton, T. W. (2001a). Multiple auditory steady-state responses to AM and FM stimuli. Audiology & Neuro-Otology, 6(1), 12-27. John, M. S., Lins, O. G., Boucher, B. L., & Picton, T. W. (1998). Multiple auditory steady-state responses (MASTER): Stimulus and recording parameters. Audiology : Official Organ of the International Society of Audiology, 37(2), 59-82. John, M. S., & Picton, T. W. (2000a). MASTER: A windows program for recording multiple auditory steady-state responses. Computer Methods and Programs in Biomedicine, 61(2), 125-150. John, M. S., & Picton, T. W. (2000b). Human auditory steady-state responses to amplitude-modulated tones: Phase and latency measurements. Hearing Research, 141(1-2), 57-79. John, M. S., Purcell, D. W., Dimitrijevic, A., & Picton, T. W. (2002b). Advantages and caveats when recording steady-state responses to multiple simultaneous stimuli. Journal of the American Academy of Audiology, 13(5), 246.  109  Johnson, B. W., Weinberg, H., Ribary, U., Cheyne, D. O., & Ancill, R. (1988). Topographic distribution of the 40 hz auditory evoked-related potential in normal and aged subjects. Brain Topography, 1(2), 117-121. Joint Committee on Infant Hearing (JCIH) (2007). Year 2007 position statement: Principles and guidelines for early hearing detection and intervention programs. Pediatrics, 120(4), 898-921. Karzon, R. K., & Banerjee, P. (2010). Animated toys versus video reinforcement in 16-24-month-old children in a clinical setting. American Journal of Audiology, 19(2), 91-99. Keefe, D. H., Bulen, J. C., Arehart, K., & Burns, E. M. (1993). Ear-canal impedance and reflection coefficient in human infants and adults. The Journal of the Acoustical Society of America, 94(5), 2617-2638. Keefe, D. H., Bulen, J. C., Campbell, S. L., & Burns, E. M. (1994). Pressure transfer function and absorption cross section from the diffuse field to the human infant ear canal. The Journal of the Acoustical Society of America, 95(1), 355-371. Keefe, D. H., & Levi, E. (1996). Maturation of the middle and external ears: Acoustic power-based responses and reflectance tympanometry. Ear & Hearing, 17(5), 361-373. Kennedy, C. R., McCann, D. C., Campbell, M. J., Law, C. M., Mullee, M., Petrou, S., . . . Stevenson, J. (2006). Language ability after early detection of permanent childhood hearing impairment. The New England Journal of Medicine, 354(20), 2131-2141. King, E. R. (2007). In search of conductive hearing loss: A comparison of behavioral audiometric testing, ASSR, & SAL technique. (Unpublished Doctoral dissertation). James Madison University, Harrisonburg, VA. Kirikae, I. (1959). An experimental study on the fundamental mechanism of bone conduction. Acta Otolaryngologica, Suppl. 145, 1-111. Kruger, B. (1987). An update on the external ear resonance in infants and young children. Ear & Hearing, 8(6), 333-336.  110  Kruger, B., & Ruben, R. J. (1987). The acoustic properties of the infant ear. A preliminary report [Abstract]. Acta Oto-laryngologica, 103(5-6) 578-585. Krumholz, A., Felix, J. K., Goldstein, P. J., & McKenzie, E. (1985). Maturation of the brain-stem auditory evoked potential in premature infants. Electroencephalography & Clinical Neurophysiology, 62(2), 124. Küçük, B., Abe, K., Ushiki, T., Inuyama, Y., Fukuda, S., & Ishikawa, K. (1991). Microstructures of the bony modiolus in the human cochlea: A scanning electron microscopic study. Journal of Electron Microscopy, 40(3), 193-197. Kuwada, S., Anderson, J. S., Batra, R., Fitzpatrick, D. C., Teissier, N., & D'Angelo, W. R. (2002). Sources of the scalp-recorded amplitude-modulation following response. Journal of the American Academy of Audiology, 13(4), 188-204. Lavigne-Rebillard, M., & Bagger-Sjöbäck, D. (1992). Development of the human stria vascularis. Hearing Research, 64(1), 39-51. Lavigne-Rebillard, M., & Pujol, R. (1988). Hair cell innervation in the fetal human cochlea. Acta Otolaryngologica, 105(5-6), 398-402. Lavigne-Rebillard, M., & Pujol, R. (1987). Surface aspects of the developing human organ of corti. Acta Oto-laryngologica, Suppl. 436, 43-50. Lee, C. Y., Hsieh, T. H., Pan, S. L., & Hsu, C. J. (2007). Thresholds of tone burst auditory brainstem responses for infants and young children with normal hearing in taiwan. Journal of the Formosan Medical Association, 106(10), 847-853. Levi, E. C., Folsom, R. C., & Dobie, R. A. (1993). Amplitude-modulation following response (AMFR): Effects of modulation rate, carrier frequency, age, and state. Hearing Research, 68(1), 42-52. Levi, E. C., Folsom, R. C., & Dobie, R. A. (1995). Coherence analysis of envelope-following responses (EFRs) and frequency-following responses (FFRs) in infants and adults. Hearing Research, 89(1-2), 21-27.  111  Liden, G., & Kankkunen, A. (1969). Visual reinforcement audiometry. Acta Oto-laryngologica, 67(2-6), 281-292. Lierle, D. M., Reger, S. M. (1946). Correlations between bone and air conduction acuity measurements over wide frequency ranges in different types of hearing impairments. The Laryngoscope, 56(5), 187-224. Lightfoot, G. R. (1979). Air-borne radiation from bone conduction transducers. British Journal of Audiology, 13, 53-56. Ling, D., Ling, A. H., & Doehring, D. G. (1970). Stimulus, response, and observer variables in the auditory screening of newborn infants. [Abstract]. Journal of Speech & Hearing Research, 13(1), 918. Lins, O. G., & Picton, T. W. (1995). Auditory steady-state responses to multiple simultaneous stimuli. Electroencephalography & Clinical Neurophysiology, 96(5), 420-432. Lins, O. G., Picton, T. W., Boucher, B. L., Durieux-Smith, A., Champagne, S. C., Moran, L. M., . . . Savio, G. (1996). Frequency-specific audiometry using steady-state responses. Ear & Hearing, 17(2), 81-96. Lowery, K. J., von Hapsburg, D., Plyler, E. L., & Johnstone, P. (2009). A comparison of video versus conventional visual reinforcement in 7- to 16-month-old infants. Journal of Speech, Language & Hearing Research, 52(3), 723-731. Lütkenhöner, B., Hoke, M., & Pantev, C. (1985). Possibilities and limitations of weighted averaging. Biological Cybernetics, 52(6), 409-416. Luts, H., Desloovere, C., & Wouters, J. (2006). Clinical application of dichotic multiple-stimulus auditory steady-state responses in high-risk newborns and young children. Audiology & NeuroOtology, 11(1), 24-37. Luts, H., Van Dun, B., Alaerts, J., & Wouters, J. (2008). The influence of the detection paradigm in recording auditory steady-state responses. Ear & Hearing, 29(4), 638-650.  112  Luts, H., & Wouters, J. (2004). Hearing assessment by recording multiple auditory steady-state responses: The influence of test duration. International Journal of Audiology, 43(8), 471-478. Marcoux, A. M. (2011). Maturation of auditory function related to hearing threshold estimations using the auditory brainstem response during infancy. International Journal of Pediatric Otorhinolaryngology, 75(2), 163-170. Matkin, N. D. (1977). Assessment of hearing sensitivity during the preschool years. In F. H. Bess (Ed.), Childhood deafness: Causation, assessment and management (pp. 127-134). New York: Grune and Stratton. Maurizi, M., Almadori, G., Paludetti, G., Ottaviani, F., Rosignoli, M., & Lucianob, R. (1990). 40-hz steady-state responses in newborns and in children. International Journal of Audiology, 29(6), 322328. McDermott, J. C., & Hodgson, W. R. (1982). Auditory thresholds in children for narrow-band noise and warble tones in sound field. British Journal of Audiology, 16(4), 221-225. Mendel, M. (1968). Infant responses to recorded sounds [Abstract]. Journal of Speech & Hearing Research, 11 811-816. Mochizuki, Y., Go, T., Ohkubo, H., & Motomura, T. (1983). Development of human brainstem auditory evoked potentials and gender differences from infants to young adults. Progress in Neurobiology, 20(3-4), 273-285. Møller, A. R. (2000). Sound-conducting apparatus. In Hearing: Its physiology and pathophysiology (pp. 379-394). San Diego: Academic Press. Moore, J. K., & Linthicum, F. H. J. (2001). Myelination of the human auditory nerve: Different time courses for schwann cell and glial myelin. Annals of Otology, Rhinology, & Laryngology, 110(7 Pt 1), 655-661. Moore, J. K., & Linthicum, F. H. J. (2007). The human auditory system: A timeline of development. International Journal of Audiology, 46(9), 460-478.  113  Moore, J. K., Perazzo, L. M., & Braun, A. (1995). Time course of axonal myelination in the human brainstem auditory pathway. Hearing Research, 87(1-2), 21-31. Moore, J. K., Ponton, C. W., Eggermont, J. J., Wu, B. J. C., & Huang, J. Q. (1996). Perinatal maturation of the auditory brain stem response: Changes in path length and conduction velocity. Ear & Hearing, 17(5), 411-418. Moore, J. M., Thompson, G., & Thompson, M. (1975). Auditory responses of infants as a function of reinforcement conditions. Journal of Speech & Hearing Disorders, 40, 29-34. Moore, J. M., Wilson, W. R., & Thompson, G. (1977). Visual reinforcement of head-turn responses in infants under 12 months of age. Journal of Speech & Hearing Disorders, 42(3), 328-334. Morgan, D. E., Dirks, D. D., & Bower, D. R. (1979). Suggested threshold sound pressure levels for frequency-modulated (warble) tones in the sound field. Journal of Speech & Hearing Disorders, 44, 37-54. Muchnik, C., Neeman, R. K., & Hildesheimer, M. (1995). Auditory brainstem response to boneconducted clicks in adults and infants with normal hearing and conductive hearing loss. Scandinavian Audiology, 24(3), 185-191. Northern, J. L., & Downs, M. P. (1991). Hearing in children (4th ed.). Baltimore: Williams & Wilkins. Nousak, J. M. K., & Stapells, D. R. (1992). Frequency specificity of the auditory brain stem response to bone-conducted tones in infants and adults. Ear & Hearing, 13(2), 87-95. Nozza, R. J. (1995). Estimating the contribution of non-sensory factors to infant-adult differences in behavioral thresholds. Hearing Research, 91(1-2), 72-78. Nozza, R. J., & Henson, A. M. (1999). Unmasked thresholds and minimum masking in infants and adults: Separating non-sensory contributions to infant-adult differences in behavioral thresholds. Ear & Hearing, 20, 483-496.  114  Nozza, R. J., & Wilson, W. R. (1984). Masked and unmasked pure tone threhsolds of infants and adults: Development of auditory frequency selectivity and sensitivity. Journal of Speech & Hearing Research, 27, 613-622. Olsho, L. W., Koch, E. G., & Carter, E. A. (1988). Pure-tone sensitivity of human infants. Journal of the Acoustical Society of America, 84(4), 1316-1324. Ontario Infant Hearing Program (OIHP) (2008). Ontario infant hearing program audiological assessment protocol. Retrieved October 15, 2011, from http://www.mountsinai.on.ca/care/infanthearing-program/documents/IHPAudiologicAssessmentProtocol3.1FinalJan2008.pdf. Paradise, J. L., Smith, C. G., & Bluestone, C. D. (1976). Tympanometric detection of middle ear effusion in infants and young children. Pediatrics, 58(2), 198-210. Parry, G., Hacking, C., Bamford, J., & Day, J. (2003). Minimal response levels for visual reinforcement audiometry in infants. International Journal of Audiology, 42(7), 413-417. Perez-Abalo, M. C., Savio, G., Torres, A., Martin, V., Rodriguez, E., & Galan, L. (2001). Steady state responses to multiple amplitude-modulated tones: An optimized method to test frequency-specific thresholds in hearing-impaired children and normal-hearing subjects. Ear & Hearing, 22(3), 200211. Pethe, J., Muhler, R., Siewert, K., & von Specht, H. (2004). Near-threshold recordings of amplitude modulation following responses (AMFR) in children of different ages. International Journal Audiology, 43(6), 339-345. Picton, T. W. (2011). Human auditory evoked potentials. San Diego: Plural Publishing. Picton, T. W., Dimitrijevic, A., Perez-Abalo, M. C., & Van Roon, P. (2005). Estimating audiometric thresholds using auditory steady-state responses. Journal of the American Academy of Audiology, 16(3), 140-156. Picton, T. W., Durieux-Smith, A., & Moran, L. M. (1994). Recording auditory brainstem responses from infants. International Journal of Pediatric Otorhinolaryngology, 28(2-3), 93-110.  115  Picton, T. W., John, M. S., Dimitrijevic, A., & Purcell, D. (2003). Human auditory steady-state responses: Respuestas auditivas de estado estable en humanos. International Journal of Audiology, 42(4), 177-219. Picton, T. W., Linden, D. R., Hamel, G., & Jagdish, T. M. (1983). Aspects of averaging. Seminars in Hearing, 4(4), 327-340. Picton, T. W., Skinner, C. R., Champagne, S. C., Kellett, A. J., & 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), 165-178. Picton, T. W., Stapells, D. R., & Campbell, K. B. (1981). Auditory evoked potentials from the human cochlea and brainstem. Journal of Otolaryngology, 10(Suppl. 9), 1-41. Picton, T. W., van Roon, P., & John, M. S. (2009). Multiple auditory steady state responses (80-101 hz): Effects of ear, gender, handedness, intensity and modulation rate. Ear & Hearing, 30(1), 100-109. Pierce, R. H., & Mainen, M. W. (1977). The cranium of the newborn infant: An atlas of tomography and anatomical sections. Bethesda, Md.: Dept. of Health, Education, and Welfare, Public Health Service, National Institutes of Health, National Institute of Dental Research, Diagnostic Radiology Dept., Clinical Center. Retrieved January 11, 2012, from http://catalog.hathitrust.org/Record/000750944. Ponton, C. W., Moore, J. K., & Eggermont, J. J. (1996). Auditory brain stem response generation by parallel pathways: Differential maturation of axonal conduction time and synaptic transmission. Ear & Hearing, 17(5), 402-410. Primus, M. A. (1987). Response and reinforcement in operant audiometry. Journal of Speech & Hearing Disorders, 52, 294-299. Primus, M. A. (1988). Infant thresholds with enhanced attention to the signal in visual reinforcement audiometry. Journal of Speech & Hearing Research, 31, 480-484. Primus, M. A., & Thompson, G. (1985). Response strength of young children in operant audiometry. Journal of Speech & Hearing Research, 28, 539-547.  116  Pujol, R. (1985). Morphology, synaptology and electrophysiology of the developing cochlea. Acta OtoLaryngologica, 99, 5-9. Pujol, R., Lavigne-Rebillard, M., & Uziel, A. (1991). Development of the human cochlea. Acta OtoLaryngologica, 111, 7-13. Qian, L., Yi, W., Xingqi, L., Yinsheng, C., Wenying, N., Lili, X., & Yinghui, L. (2010). Development of tone-pip auditory brainstem responses and auditory steady-state responses in infants aged 0-6 months. Acta Oto-Laryngologica, 130(7), 824-830. Rance, G., & Briggs, R. J. (2002). Assessment of hearing in infants with moderate to profound impairment: The melbourne experience with auditory steady-state evoked potential testing [Abstract]. Annals of Otology, Rhinology, & Laryngology Supplement, 189, 22-28. Rance, G., & Rickards, F. W. (2002). Prediction of hearing threshold in infants using auditory steadystate evoked potentials. Journal of the American Academy of Audiology, 13(5), 236-245. Rance, G., Rickards, F. W., Cohen, L., De Vidi, S., & Clark, G. (1995). The automated prediction of hearing thresholds in sleeping subjects using auditory-steady state evoked potentials. Ear & Hearing, 16, 499-507. Rance, G., Roper, R., Symons, L., Moody, L. J., Poulis, C., Dourlay, M., & Kelly, T. (2005). Hearing threshold estimation in infants using auditory steady-state responses. Journal of the American Academy of Audiology, 16(5), 291-300. Rance, G., & Tomlin, D. (2006). Maturation of auditory steady-state responses in normal babies. Ear & Hearing, 27(1), 20-29. Rance, G., Tomlin, D., & Rickards, F. W. (2006). Comparison of auditory steady-state responses and tone-burst auditory brainstem responses in normal babies. Ear & Hearing, 27(6), 751-762. Regan, D. (1989). Human brain electrophysiology: Evoked potentials and evoked magnetic fields in science and medicine. New York: Elsevier.  117  Regan, M. P., & Regan, D. (1988). A frequency domain technique for characterizing nonlinearities in biological systems. Journal of Theoretical Biology, 133(3), 293-317. Ribeiro, F. M., Carvallo, R. M., & Marcoux, A. M. (2010). Auditory steady-state evoked responses for preterm and term neonates. Audiology & Neuro-Otology, 15(2), 97-110. Rodrigues, G. R. I., & Lewis, D. R. (2010). Threshold prediction in children with sensorineural hearing loss using the auditory steady-state responses and tone-evoked auditory brain stem response. International Journal of Pediatric Otorhinolaryngology, 74(5), 540-546. Sabo, D. L., Paradise, J. L., Kurs-Lasky, M., & Smith, C. G. (2003). Hearing levels in infants and young children in relation to testing technique, age group, and the presence or absence of middle-ear effusion. Ear & Hearing, 24(1), 38-47. Salamy, A. (1984). Maturation of the auditory brainstem response from birth through early childhood. Journal of Clinical Neurophysiology, 1(3), 293-329. Salamy, A., Eldredge, L., & Wakeley, A. (1985). Maturation of contralateral brain-stem responses in preterm infants. Electroencephalography & Clinical Neurophysiology, 62(2), 117-123. Salamy, A., & McKean, C. M. (1976). Postnatal development of human brainstem potentials during the first year of life. Electroencephalography & Clinical Neurophysiology, 40(4), 418-426. Salamy, A., McKean, C. M., & Buda, F. B. (1975). Maturational changes in auditory transmission as reflected in human brain stem potentials. Brain Research, 96(2), 361-366. Saunders, J. C., Doan, D. E., & Cohen, Y. E. (1993). The contribution of middle-ear sound conduction to auditory development. Comparative Biochemistry & Physiology Part A: Physiology, 106(1), 7-13. Saunders, J. C., Kaltenback, J. A., & Relkin, E. M. (1983). The structural and functional development of the outer and middle ear. In R. Romand (Ed.), Development of auditory and vestibular systems (pp. 3-25). New York: Academic Press.  118  Savio, G., Cardenas, J., Abalo, M. P., Gonzalez, A., & Valdes, J. (2001). The low and high frequency auditory steady state responses mature at different rates. Audiology & Neuro-Otology, 6(5), 279-287. Schmida, M. J., Peterson, H. J., & Tharpe, A. M. (2003). Visual reinforcement audiometry using digital video disc and conventional reinforcers. American Journal of Audiology, 12(1), 35-40. Shaw, P., & Nikolopoulos, T. (2004). The effect of initial stimulus type for visual reinforcement audiometry. International Journal of Audiology, 43, 193-197. Sininger, Y. S., Abdala, C., & Cone-Wesson, B. (1997). Auditory threshold sensitivity of the human neonate as measured by the auditory brainstem response. Hearing Research, 104, 27-38. Sininger, Y. S., & Hyde, M. L. (2009). Auditory brainstem resposne in audiometric threshold prediction. In J. Katz, L. Medwetsky, R. Burkard & L. Hood (Eds.), Handbook of Clinical Audiology (6th ed., pp. 293-321). Philadelphia: Lippincott, Williams & Wilkins. Small, S. A., Hatton, J. L., & Stapells, D. R. (2007). Effects of bone oscillator coupling method, placement location, and occlusion on bone-conduction auditory steady-state responses in infants. Ear & Hearing, 28(1), 83-98. Small, S. A., & Hu, N. (2011). Maturation of the occlusion effect: A bone conduction auditory steady state response study in infants and adults with normal hearing. Ear & Hearing, 32(6), 708-719. Small, S. A., & Stapells, D. R. (2008a). Maturation of bone conduction multiple auditory steady-state responses. International Journal of Audiology, 47(8), 476-488. Small, S. A., & Stapells, D. R. (2008b). Normal ipsilateral/contralateral asymmetries in infant multiple auditory steady-state responses to air-and bone-conduction stimuli. Ear & Hearing, 29(2), 185-198198. Small, S. A., & Stapells, D. R. (2006). Multiple auditory steady-state response thresholds to boneconduction stimuli in young infants with normal hearing. Ear & Hearing, 27(3), 219-228.  119  Small, S. A., & Stapells, D. R. (2005). Multiple Auditory Steady-State Responses to Bone-Conduction Stimuli in Adults with Normal Hearing. Journal of the American Academy of Audiology, 16(3), 172183. Small, S. A., & Stapells, D. R. (2003). Normal brief-tone bone-conduction behavioral thresholds using the B-71 transducer: three occlusion conditions. Journal of the American Academy of Audiology, 14(10), 172-183. Sohmer, H., & Freeman, S. (2004). Further evidence for a fluid pathway during bone conduction auditory stimulation. Hearing Research, 193(1-2), 105-110. Sohmer, H., Freeman, S., Geal-Dor, M., Adelman, C., & Savion, I. (2000). Bone conduction experiments in humans – a fluid pathway from bone to ear. Hearing Research, 146(1-2), 81-88. Stapells, D. R. (2000a). Frequency-specific evoked potential audiometry in infants. In R.C. Seewald (Ed.), A Sound Foundation through Early Amplification (pp. 13-31). Basel: Phonak AG. Stapells, D. R. (2000b). Threshold estimation by the tone-evoked auditory brainstem response: A literature meta-analysis. Journal of Speech-Language Pathology & Audiology, 24(2), 74-83. Stapells, D. R. (2010a). Frequency-specific threshold assessment in young infants using the transient ABR and the brainstem ASSR. In R. C. Seewald, & A. M. Tharpe (Eds.), Comprehensive Handbook of Pediatric Audiology (pp. 409-448). Plural Publishing. Stapells, D. R. (2010b). Tips for clinicians: Auditory steady state responses. Retrieved November 30, 2011, from http://www.courses.audiospeech.ubc.ca/haplab/ASSR.html. Stapells, D. R., Galambos, R., Costello, J. A., & Makeig, S. (1988). Inconsistency of auditory middle latency and steady-state responses in infants. Electroencephalography & Clinical Neurophysiology, 71(4), 289-295. Stapells, D. R., Gravel, J. S., & Martin, B. A. (1995). Thresholds for auditory brain stem responses to tones in notched noise from infants and young children with normal hearing or sensorineural hearing loss. Ear & Hearing, 16(4), 361-371.  120  Stapells, D. R., Herdman, A., Small, S. A., Dimitrijevic, A., & Hatton, J. (2005). Current status of the auditory steady-state responses for estimating an infant‘s audiogram In R. C. Seewald, & J. Bamford (Eds.), A Sound Foundation through Early Amplification 2004 (pp. 43-59). Basel: Phonak AG. Stapells, D. R., Linden, D., Suffield, J. B., Hamel, G., & Picton, T. W. (1984). Human auditory steady state potentials. Ear & Hearing, 5(2), 105-113. Stapells, D. R., Makeig, S., & Galambos, R. (1987). Auditory steady-state responses: Threshold prediction using phase coherence. Electroencephalography & Clinical Neurophysiology, 67(3), 260270. Stapells, D. R., & Mosseri, M. (1991). Maturation of the contralaterally recorded auditory brain stem response. Ear & Hearing, 12(3), 167-173. Stapells, D. R., & Oates, P. (1997). Estimation of the pure-tone audiogram by the auditory brainstem response: A review. Audiology & Neuro-Otology, 2(5), 257-280. Stapells, D. R., & Ruben, R. J. (1989). Auditory brain stem responses to bone-conducted tones in infants. The Annals of Otology, Rhinology, & Laryngology, 98(12 Pt 1), 941-949. Starr, A., Amlie, R. N., Martin, W. H., & Sanders, S. (1977). Development of auditory function in newborn infants revealed by auditory brainstem potentials. Pediatrics, 60(6), 831. Stenfelt, S., & Goode, R. L. (2005). Bone-conducted sound: Physiological and clinical aspects. Otology & Neurotology, 26(6), 1245-1261. Stenfelt, S., Wild, T., Hato, N., & Goode, R. L. (2003). Factors contributing to bone conduction: The outer ear. The Journal of the Acoustical Society of America, 113(2), 902-913. Stockard, J. J., Stockard, J. E., & Sharbrough, F. W. (1978). Nonpathologic factors influencing brainstem auditory evoked potentials. American Journal of EEG Technology, 18, 1771-209. Stuart, A., Yang, E. Y., & Botea, M. (1996). Neonatal auditory brainstem responses recorded from four electrode montages. Journal of Communication Disorders, 29(2), 125-139.  121  Stuart, A., Yang, E. Y., & Green, W. B. (1994). Neonatal auditory brainstem response thresholds to airand bone-conducted clicks: 0 to 96 hours postpartum [Abstract]. Journal of the American Academy of Audiology, 5(3), 163-172. Stuart, A., Yang, E. Y., & Stenstrom, R. (1990). Effect of temporal area bone vibrator placement on auditory brain stem response in newborn infants. Ear & Hearing, 11(5), 363-369. Studebaker, G. A. (1967). Clinical masking of the non-test ear. Journal of Speech & Hearing Disorders, 32, 360-371. Stueve, M., & O'Rourke, C. (2003). Estimation of hearing loss in children: Comparison of auditory steady-state response, auditory brainstem response, and behavioral test methods. American Journal of Audiology, 12(2), 125-136. Suzuki, T., & Kobayashi, K. (1984). An evaluation of 40-hz event-related potentials in young children. International Journal of Audiology, 23(6), 599-604. Suzuki, T., & Ogiba, Y. (1961). Conditioned orientation reflex audiometry: A new technique for puretone audiometry in young children under 3 years of age. Archives of Otolaryngology--Head & Neck Surgery, 74(2), 192-198. Swanepoel, D. W., Ebrahim, S., Friedland, P., Swanepoel, A., & Pottas, L. (2008). Auditory steady-state responses to bone conduction stimuli in children with hearing loss. International Journal of Pediatric Otorhinolaryngology, 72(12), 1861-1871. Swanepoel, D. W., & Steyn, K. (2005). Short report: Establishing normal hearing for infants with the auditory steady-state response. South African Journal of Communication Disorders, 52, 36-39. Takahara, T., Sando, I., Hashida, Y., & Shibahara, Y. (1986). Mesenchyme remaining in human temporal bones [Abstract]. Otolaryngology-Head & Neck Surgery, 95(3) 349-357. Thompson, G., & Folsom, R. C. (1984). A comparison of two conditioning procedures in the use of visual reinforcment audiometry (VRA). Journal of Speech & Hearing Disorders, 49, 241-245.  122  Thompson, G., & Folsom, R. C. (1985). Reinforced and nonreinforced head-turn responses of infants as a function of stimulus bandwidth. Ear & Hearing, 6, 125-129. Thompson, G., Thompson, M., & McCall, A. (1992). Strategies for increasing response behavior of 1and 2-year-old children during visual reinforcement audiometry (VRA). Ear & Hearing, 13(4), 236240. Thompson, G., & Weber, B. A. (1974). Responses of infants and young children to behavior observation audiometry (BOA). American Speech-Language-Hearing Association, 39(2), 140-147. Thompson, G., & Wilson, W. R. (1984). Clinical application of visual reinforcement audiometry. Seminars in Hearing, 5, 85-99. Thompson, M., & Thompson, G. (1972). Responses of infants and young children as a function of auditory sitmuli and test methods. Journal of Speech & Hearing Research, 15, 699-707. Thompson, M., Thompson, G., & Vethivelu, S. (1989). A comparison of audiometric test methods for 2year-old children. Journal of Speech & Hearing Disorders, 54, 174-179. Tlumak, A. I., Rubinstein, E., & Durrant, J. D. (2007). Meta-analysis of variables that affect accuracy of threshold estimation via measurement of the auditory steady-state response (ASSR). International Journal of Audiology, 46(11), 692-710. Tonndorf, J. (1966). Quantitative evaluation of bone conduction components in cats. Acta Otolaryngologica, Suppl. 213, 1-132. Trehub, S. E., Schneider, B. A., & Endman, M. (1980). Developmental changes in infants' sensitivity to octave-band noises. Journal of Experimental Child Psychology, 29(2), 282-293. van der Reijden, C. S., Mens, L. H., & Snik, A. F. (2004). Signal-to-noise ratios of the auditory steadystate response from fifty-five EEG derivations in adults. Journal of the American Academy of Audiology, 15(10), 692-701.  123  van der Reijden, C. S., Mens, L. H., & Snik, A. F. (2005). EEG derivations providing auditory steadystate responses with high signal-to-noise ratios in infants. Ear & Hearing, 26(3), 299-309. Van Maanen, A., & Stapells, D. R. (2009). Normal multiple auditory steady-state response thresholds to air-conducted stimuli in infants. Journal of the American Academy of Audiology, 20(3), 196-207. Van Maanen, A., & Stapells, D. R. (2010). Multiple-ASSR thresholds in infants and young children with hearing loss. Journal of the American Academy of Audiology, 21(8), 535-545. Vander Werff, K. R., Prieve, B. A., & Georgantas, L. M. (2009). Infant air and bone conduction tone burst auditory brain stem responses for classification of hearing loss and the relationship to behavioral thresholds. Ear & Hearing, 30(3), 350-368. Vento, B. A., & Durrant, J. D. (2009). Assessing bone conduction thresholds in clinical practice. In J. Katz, L. Medwetsky, R. Burkard & L. Hood (Eds.), Handbook of Clinical Audiology (6th ed., pp. 50-63). Philadelphia: Lippincott, Williams & Wilkins. Walker, G., Dillon, H., & Byme, D. (1984). Sound field audiometry: Recommended stimuli and procedures. Ear & Hearing, 5(1), 13-21. Werker, J. F., Polka, L., & Pegg, J. E. (1997). The conditioned head turn procedure as a method for testing infant speech perception. Early Development & Parenting, 6, 171-178. Werner, L. A. (1996). The development of auditory behavior (or what the anatomists and physiologists have to explain). Ear & Hearing, 17(5), 438-446. Werner, L. A., Folsom, R. C., & Mancl, L. R. (1993). The relationship between auditory brainstem response and behavioral thresholds in normal hearing infants and adults. Hearing Research, 68(1), 131-141. Werner, L. A., Folsom, R. C., & Mancl, L. R. (1994). The relationship between auditory brainstem response latencies and behavioral thresholds in normal hearing infants and adults. Hearing Research, 77(1-2), 88-98.  124  Wever, E., & Lawrence, M. (1952). Sound conduction in the cochlea. Transactions of the American Otological Society, 40, 168-179. Widen, J. E. (1990). Behavioural screening of high-risk infants using visual reinforcement audiometry. Seminars in Hearing, 11(4), 342-356. Widen, J. E., Folsom, R. C., ConeWesson, B., Carty, L., Dunnell, J. J., Koebsell, K., . . . Norton, S. J. (2000). Identification of neonatal hearing impairment: Hearing status at 8 to 12 months corrected age using a visual reinforcement audiometry protocol. Ear & Hearing, 21(5), 471-487. Widen, J. E., & O'Grady, G. M. (2002). Using visual reinforcement audiometry in the assessment of hearing in infants. Hearing Journal, 55(11), 28-36. Wood, L. L. (2009). Multiple brainstem auditory steady-state response interactions for different stimuli. (Master of Science - MSc, University of British Columbia). Retrieved from http://hdl.handle.net/2429/15532. Yacullo, W. S. (2009). Clinical masking. In J. Katz, L. Medwetsky, R. Burkard & L. Hood (Eds.), Handbook of Clinical Audiology (6th ed., pp. 80-115). Philadelphia: Lippincott, Williams & Wilkins. Yang, E. Y., Rupert, A. L., & Moushegian, G. (1987). A developmental study of bone conduction auditory brain stem response in infants. Ear & Hearing, 8(4), 244-251. Yoshida, M., & Uemura, T. (1991). Transmission of cerebrospinal fluid pressure changes to the inner ear and its effect on cochlear microphonics. European Archives of Oto-Rhino-Laryngology, 248(3), 139-143. Yoshinaga-Itano, C., Sedey, A. L., Coulter, D. K., & Mehl, A. L. (1998). Language of early- and lateridentified children with hearing loss. Pediatrics, 102(5), 1161-1171.  125  Appendix A: Individual Infant Threshold Data  126  Infant BC and AC ASSR thresholds for the ipsilateral channel Infant # BC AC Age (months) 500 Hz 1000 Hz 2000 Hz 4000 Hz 500 Hz 1000 Hz 19 1 --30 30 33 ----14 2* ------------10 3 0 10 30 33 30 20 9 4 10 10 40 13 20 30 10 5 30 10 30 33 20 20 15 6 20 10 30 13 ----18 7 30 20 20 23 ----16 8 20 0 20 3 30 10 9 9 ------------9 10 10 0 10 13 ----7 11 --------30 30 9 12 10 0 10 3 30 20 8 13 -10 10 20 13 10 10 11.5 14 0 0 10 13 30 --7.5 15 10 -10 10 3 20 20 16 16 ------------10 17 0 10 10 -7 ----9.5 18 0 10 10 3 ----7.5 19 10 10 10 3 ----7 20 20 10 40 13 30 30 6.5 21 20 20 20 23 40 40 7 22 20 20 20 23 40 20 8 23 10 0 30 23 30 20 12 24 0 10 10 13 30 20 10.5 Mean 11.05 9.00 20.50 14.50 27.86 22.31 3.73 SD 11.00 9.12 10.50 11.37 8.02 8.32 23 N 19 20 20 20 14 13 Note: ―---― indicates that data was not obtained for this individual for that method, mode and/ or frequency  2000 Hz  4000 Hz  ----20 --30 ----30 ----20 10 10 20 10 --------30 30 10 20 20 20.00 7.84 13  ----28 28 28 ----28 ----8 8 8 28 18 --------28 38 18 28 18 22.29 9.03 14  ―*‖ indicates that the data was not used due to poor reliability 127  Infant AC and BC VRA thresholds Infant # BC 500 Hz 1000 Hz 2000 Hz  4000 Hz  Bone-oscillator Coupling Method  AC 500 Hz  1000 Hz  1 --------------2* *20 --*10 --metal *30 --3 10 --10 10 elastic/1 handheld 10 10 4 -10 --10 0 elastic 0 0 5 -10 0 0 0 elastic -10 0 6 --------------7 -10 -10 -10 0 metal ----8 -10 --10 10 metal *20 --9 -10 ---10 -10 metal 0 --10 -10 ---10 0 elastic ----11 0 --0 10 elastic 10 --12 0 0 0 0 metal ----13 ------0 metal ----14 10 20 10 --handheld ----15 10 --0 10 metal/elastic 10 --16 10 --10 10 metal ----17 -10 ---10 0 metal 0 --18 10 --10 10 metal ----19 10 --10 10 metal ----20 0 ---10 --metal 10 --21 10 --10 10 metal/elastic 10 --22 ----10 --elastic 10 --23 0 --0 10 elastic 10 10 24 -10 --10 20 metal 10 --Note: ―---― indicates that data were not obtained for this individual for that method, mode and/ or frequency  2000 Hz --*10 10 10 0 ----10 0 --0 --0 --10 --0 ----0 10 10 0 20  4000 Hz ----10 0 -10 ----0 0 --10 --20 --0 ---10 ------10 --0 ---  ―*‖ indicates that the data was not used due to poor reliability  128  Appendix B: Individual Adult Threshold Data  129  Adult AC and BC ASSR thresholds Adult # Age (years) 1 2 3 4 5 6 7 8 9 10 11 Mean SD  46 26 43 20 29 20 26 23 50 17 29 29.91 11.30  BC 500 Hz  1000 Hz  2000 Hz  4000 Hz  30 20 0 30 20 10 20 30 20 20 20 20.00 8.94  10 40 10 30 20 10 0 10 20 20 0 15.45 12.14  10 20 10 20 20 10 0 20 20 10 0 12.72 7.86  13 13 -7 13 23 3 13 23 -7 3 3 8.45 10.36  Adult AC and BC behavioural thresholds Adult # 17 Age (years) BC 500 Hz 1000 Hz 1 28 5 5 2 26 -5 0 3 20 0 5 4 20 -10 -5 5 23 10 5 6 29 5 -5 7 50 0 0 8 17 5 5 26.63 Mean 1.25 1.25 10.34 SD 6.41 4.43 17  AC 500 Hz  1000 Hz  2000 Hz  4000 Hz  20 20 20 30 20 20 30 20 20 20 30 22.73 4.67  20 10 20 10 20 20 20 30 20 20 10 18.18 6.03  38 -2 18 18 18 18 18 18 18 18 18 18.00 8.94  40 20 20 40 30 10 30 30 30 20 40 28.18 9.82  2000 Hz  4000 Hz  AC 500 Hz  -5 0 5 0 10 5 5 0 2.5 4.63  5 0 -5 -10 5 -5 5 5 0 5.98  5 -5 5 -10 10 5 5 5 2.5 6.55  1000 Hz  2000 Hz  4000 Hz  10 0 0 0 15 0 5 5 4.38 5.63  0 0 -5 5 10 5 5 5 3.13 4.58  5 0 0 -5 5 0 0 5 1.25 3.54  It should be noted that the subject numbers do not represent the same individual adults across testing methods.  130  Appendix C: Individual Infant Amplitude Data  131  Infant amplitude (nV) data at 500 Hz for each intensity by sound presentation mode and subject  Subject # 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24  BC -10 ----4 --------5 ----3 22 6 1 --5 9 2 ------13 6  0 ----10 4 7 2 --10 --5 --2 15 10 1 --6 23 4 2 ----4 15  10 2 --17 12 7 4 7 9 --20 --9 31 15 4 --10 43 12 2 4 3 19 16  AC 20 12 --21 11 4 9 10 42 --19 --10 41 20 8 --16 56 10 4 9 5 26 25  30 6 --26 29 13 9 34 41 ------12 24 ------18 ----8 ----36 ---  40 6 ----28 13 ----65 ----------------------10 ---------  -10 -------------------------------------------------  0 --------------9 ----8 6 8 --2 ------------3 -----  10 ----5 5 3 ----8 ----7 0 17 5 1 ------------3 7 3  20 ----4 21 18 ----8 ----8 3 34 6 3 --------4 4 3 2 7  30 ----18 8 5 ----12 ----26 13 19 11 6 --------4 4 2 13 11  40 ----25 * 15 ------------------------------16 4 --16  Note: All amplitude values that were significant are italicized. Amplitude values that were considered to be false-negatives are in bolded font, and values deemed to be false-positives are both bolded and italicized.  132  Infant amplitude (nV) data at 1000 Hz for each intensity by sound presentation mode and subject  Subject # 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24  BC -10 ----3 --------8 ------2 4 1 3 --4 9 2 ------8 5  0 ----2 5 1 2 19 --14 --4 4 5 1 --3 8 2 2 ----14 5  10 1 --7 9 10 9 7 29 --32 --4 22 4 8 --8 26 6 3 4 3 24 17  AC 20 3 --8 13 13 11 15 19 --44 --10 17 7 10 --7 41 9 9 11 7 42 20  30 10 --23 21 8 7 15 27 ------14 7 ------8 ----8 ----77 ---  40 11 ----16 9 ----61 ----------------------11 ---------  -10 -----------------------------------------------  0 --------------6 ----3 1 4 --2 ------------1 -----  10 ----4 6 11 ----17 ----5 2 20 5 2 ------------1 5 6  20 ----8 7 13 ----13 ----4 9 15 2 5 --------4 10 4 17 16  30 ----17 12 8 ----22 ----12 14 6 1 9 --------7 9 3 32 24  40 ----20 * 18 ------------------------------13 12 --33  Note: All amplitude values that were significant are italicized. Amplitude values that were considered to be false-negatives are in bolded font, and values deemed to be false-positives are both bolded and italicized.  133  Infant amplitude (nV) data at 2000 Hz for each intensity by sound presentation mode and subject  Subject # 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24  BC  AC  -10 -----  0 -----  10  20  30  40  5  4  9  10  ---  ---  ---  1  1 4 2 1  4 6 4 3 17 26  11 8 5 8 14 51  -----  -------  ---------  ---  9  3  2 5 1 2 1 6  -------  ---  ---  ---  6  10  18  ---  ---  ---  1 3 2 1  1 4 2 1  4 4 3 2  5 20 3 4  ---  ---  ---  ---  2 8 4  2 3 3 2  6 36  -------  12 8  ----52  -----------------------  -10 -------------------------------------------------  0 ---------------  10 -----  20 -----  30 -----  40 -----  1 14 1  6 7 7  15 13 5  14 * 9  -----  -----  -----  4  6  4  9  -----  -----  -----  -----  3 1 1 2  6 4 4 1 5  27 7 28 6 9  38 11 35 8 13  -------------  -------------  ---------  ---------  -------------------------------  ---  7 7 8 --11 38 --5 5 --2 2 1 6 3 5 --------4 10 11 17 18 --------2 7 1 5 9 12 16 ------8 6 7 5 20 6 18 31 ------1 2 9 13 5 10 14 26 Note: All amplitude values that were significant are italicized. Amplitude values that were considered to be false-negatives are in bolded font, and values deemed to be false-positives are both bolded and italicized.  134  Infant amplitude (nV) data at 4000 Hz for each intensity by sound presentation mode and subject  Subject # 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24  BC  AC  -10 -----  0 -----  10  20  30  40  3  5  8  11  ---  ---  ---  2  1 3 5 2  1 16 2 7 11 25  6 15 41 10 17 38  -----  -------  ---------  ---  4  9  2 11 3 6 3 17  --------  ---  ---  ---  7  20  23  ---  ---  ---  1 4 1 2  3 6 2 2  5 10 3 2  5 11 5 4  ---  ---  ---  ---  5 8 3  2 12 6 2  7 19  -------  8 13  ----31  -----------------------  -10 -------------------------------------------------  0 ---------------  10 -----  20 -----  30 -----  40 -----  2 1 3  2 10 5  5 5 4  5 * 9  -----  -----  -----  3  5  7  14  -----  -----  -----  -----  6 2 5 2  12 2 10 1 1  17 6 16 1 5  21 5 20 4 6  -------------  -------------  ---------  ---------  -------------------------------  ---  4 5 4 --12 6 --5 10 --2 4 6 8 1 2 --------1 6 2 8 26 --------1 3 1 1 3 4 5 ------10 4 1 17 18 5 3 10 ------7 3 14 15 3 5 14 19 Note: All amplitude values that were significant are italicized. Amplitude values that were considered to be false-negatives are in bolded font, and values deemed to be false-positives are both bolded and italicized.  135  Appendix D: Individual Adult Amplitude Data  136  Adult amplitude (nV) data at 500 Hz for each intensity by sound presentation mode and subject  Subject #  BC -10  1 2 3 4 5 6 7 8 9 10 11  ----0 ----7 8 9 4 2 11  0 2 17 5 2 --9 4 3 16 8 4  10  AC 20  1 9 5 7 2 22 9 3 11 8 13  1 21 11 8 7 40 40 4 21 9 34  30 5 39 2 9 18 ----9 11 -----  40 8 114 -------------------  -10 --7 -------------------  0 --5 --4 --14 ---------  10  20  4 8 2 2 2 19 7 12 8 4 4  3 23 5 1 3 6 6 4 11 12 4  30 1 29 5 6 12 12 13 11 16 15 11  40 14 --9 7 ------34 ----35  Note: All amplitude values that were significant are italicized. Amplitude values that were considered to be false-negatives are in bolded font, and values deemed to be false-positives are both bolded and italicized.  Adult amplitude (nV) data at 1000 Hz for each intensity by sound presentation mode and subject  Subject # 1 2 3 4 5 6 7 8 9 10 11  BC -10 ----3  AC  0  10  20  30  40  1 4 2 4  7 18 4 5 4 30 17 13 13 4 31  10 5 10 6 5 18 50 14 27 18 25  7 6 12 11 5  16 29  -----  ---  5 5 4 6 3 8  7 19 3 14 5 22  ----17 25  -----  -------------------  -10 ---  0 ---  10  20  30  40  ---  ---  8  1 7 7 4 3 8  6 29 6 3 5 18 4 10 31 13 3  10 27 2 7 4 14 14 20 24 17 27  20  1  3 2 0 1  6  2  -------------------  ---  --------8  --8 16 5  ----28  ----41  Note: All amplitude values that were significant are italicized. Amplitude values that were considered to be false-negatives are in bolded font, and values deemed to be false-positives are both bolded and italicized.  137  Adult amplitude (nV) data at 2000 Hz for each intensity by sound presentation mode and subject  Subject # 1 2 3 4 5 6 7 8 9 10 11  BC -10 ----2  ----3 6 1 1 2 8  AC  0  10  20  30  40  2 10 1 6  2 15 5 2 4 12 24 5 6 8 37  8 23 6 15 5 19 21 7 21 9 25  18 67 16 7 9  27 30  --7 19 0 7 3 31  ----17 48  -----  -------------------  -10 ---  0 ---  2  3  -------------------  2  --8  --------5  10  20  30  40  4 17 4 3 1 8 2 5 8 6 29  3 32 9 5 4 13 12 1 35 12 27  9 27 11 10 7 21 22 10 35 19 39  15  --11 12  ------26  ----29  Note: All amplitude values that were significant are italicized. Amplitude values that were considered to be false-negatives are in bolded font, and values deemed to be false-positives are both bolded and italicized.  Adult amplitude (nV) data at 4000 Hz for each intensity by sound presentation mode and subject  Subject # 1 2 3 4 5 6 7 8 9 10 11  BC -10 ----7  ----5  --5 15 3 2  AC  0  10  20  30  40  3 11 6 3  5 21 10 9 1 12 5 4 19 13 17  6 26 9 9 5 27 15 12 14 16 29  9 47 13 10 15  17 77  --13 10 4 25 9 18  --39 20 30  -----  -------------------  -10 ---  0 ---  13  13  -------------------  --2  --2  --------1  10  20  30  2 5 1 2 5 1 5 5 5 3 5  1 25 4 5 4 12 8 8 16 18 19  2 27 7 7 5 17 12 10 20 13 15  40 9  --8 9  ------26  ----24  Note: All amplitude values that were significant are italicized. Amplitude values that were considered to be false-negatives are in bolded font, and values deemed to be false-positives are both bolded and italicized.  138  Appendix E: Individual Infant Phase Data  139  Infant phase (degrees) data for 500 and 1000 Hz by sound presentation mode and subject  500 Hz Subject # 10 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24  BC 20  30  196.04 277.98 268.58  ---  ---  306.97 284.11 250.30 255.00 60.65 245.43  268.58 232.08 287.66 260.04 224.92 203.84  ---  ---  284.28 264.86  --313.21 217.36 260.44 338.54  --263.82 236.84 252.65 348.34 23.01 277.92 256.60 309.20  ---  ---  20 -----  AC 30 -----  40 -----  251.28 351.26 314.87 289.32 --209.28 172.04 309.72 295.28 260.84 257.18 242.51 ------256.49 ------226.52 --224.52 44.04 325.59  -------  -------  -------  262.68 222.34 350.17 364.27 205.61 238.50 193.18 189.28 --221.60 15.79 274.82 --296.43 309.89 292.93  ---  1000 Hz Subject #  ---  ---------  ---------  -------------------------  247.27 218.10 --207.33 --239.30 261.88 243.89 16.59 418.19 --317.22 358.76 100.53 400.20 --235.35 135.48 228.99 257.12 --217.01 218.22 283.88 304.85 --277.52 17.91 413.55 343.35  10 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24  97.32  BC 20  338.71 296.43  ---  ---  274.19 323.35 372.70 357.39 93.48 348.28  275.46 269.27 266.57 292.70 398.25 343.29  ---  ---  246.98 229.73  --389.06 469.07 280.55 330.97  --187.22 312.47 433.09 438.47 289.61 325.19 283.02 320.32  30  ---  ---  20 -----  AC 30 -----  238.44 299.12 285.08 229.16 10.81 324.04 210.54 340.08 268.75 ----277.40 ----332.75 270.70 338.37 314.02  -------  -------  -------  294.25 277.98 394.76 316.14 430.16 269.21 434.98 326.33 --244.74 50.34 236.21 --289.21 359.11 320.26  ---  ---  ---------  ---------  195.13 175.19 --270.30 --325.42 389.08 316.42 50.63 364.39 --323.70 145.57 6.57 --295.22 409.14 344.84 256.89 232.08 357.90 305.88 --315.33 383.35 332.01  40 ----264.45 * 207.90  ----------------------------324.84 284.45  --285.25  140  Infant phase (degrees) data for 2000 and 4000 Hz by sound presentation mode and subject  2000 Hz Subject # 10 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24  BC 20  30  349.48 305.59 292.53  ---  ---  ---  251.51 316.31 86.09 311.49 201.60 178.00  4.16 318.43 249.61 266.57 314.19 190.31  235.63 265.26 253.40 212.03 283.02 177.36  ---  ---  -------  318.94 277.52  --312.30 47.93 287.54 428.56  --268.24 210.83 314.24 297.34 276.43 310.69 221.77 309.83  ---  20 -----  AC 30 -----  ---  40 -----  289.49 240.85 227.33 --314.47 284.45 222.69 228.82 229.27  -----  -----  283.99 228.87  -------  -------  262.62 236.26 283.08 270.41 326.33 288.12 340.94 300.04 --262.11 266.06 277.98 --381.29 322.90 287.54  ---  4000 Subject #  ---------  ---------  -------------------------------  231.22 213.69 --207.50 --309.20 326.68 17.62 101.22 280.27 --264.34 263.60 241.82 228.36 --278.78 291.90 259.47 246.06 --326.33 275.17 277.12 253.63 --283.30 282.56 268.69 230.36  10 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24  BC 20  30  321.01 256.55 245.55  ---  ---  ---  323.76 288.40 179.43 265.08 326.39 251.68  113.19 240.96 356.82 231.57 284.16 274.48  223.03 293.68 267.43 215.41 265.66 270.24  ---  ---  -------  232.77 233.29  --278.26 338.65 244.06 372.07  --344.84 205.55 278.66 288.17 246.35 286.46 181.55 233.46  ---  20 -----  AC 30 -----  310.00 254.77 214.84 288.98 235.75 --246.58 228.93 260.44  -----  -----  306.45 259.47  -------  -------  270.76 245.78 328.28 271.90 296.43 262.22 304.73 293.27 --209.91 255.46 263.74 --289.38 322.27 290.41  ---  ---  40 -----  ---------  ---------  -------------------------------  341.57 300.04 --247.38 --277.29 282.79 229.39 44.09 282.79 --342.09 294.99 256.15 248.58 --303.72 286.23 296.20 274.08 --274.25 245.66 316.02 263.14 --217.70 218.10 231.80 253.05  141  Appendix F: Individual Adult Phase Data  142  Adult phase (degrees) data for each frequency by sound presentation mode and subject 500 Hz Subject # 1 2 3 4 5 6 7 8 9 10 11  10  BC 20  57.33 164.82 193.75 74.12 338.65 136.34 266.46 304.22 259.58 125.91 336.70  356.53 189.11 195.01 287.49 321.35 101.51 224.00 288.17 310.23 131.81 245.72  2000 Hz Subject # 1 2 3 4 5 6 7 8 9 10 11  10  BC 20  325.93 250.99 355.04 343.24 251.05 117.43 322.61 245.32 353.95 259.76 290.93  336.82 267.09 17.85 270.30 316.99 244.74 293.33 258.78 279.87 264.45 296.02  30  20  AC 30  233.74 136.05 262.85 226.30 282.16  135.83 245.03 165.91 180.86 275.17 128.09 167.62 136.97 36.25 122.82 230.54  107.23 182.46 186.13 173.18 227.21 457.27 153.42 3.07 18.31 151.24 230.25  ----306.97 265.77  -----  30  20  AC 30  310.41 231.05 336.02 246.86 275.34  328.80 294.25 316.88 262.16 342.66 288.23 326.39 7.83 331.55 257.24 299.92  277.46 233.17 287.49 263.48 265.20 260.33 277.40 307.77 297.80 244.69 272.42  ----255.92 289.44  -----  1000 Hz Subject # 40 261.59  --390.97 289.61  ------339.91  ----224.06  1 2 3 4 5 6 7 8 9 10 11  10  BC 20  365.06 403.87 431.42 80.25 160.18 370.29 402.61 449.47 0.32 192.72 337.39  337.91 86.55 431.88 278.15 356.19 400.26 309.72 428.85 314.07 367.77 360.95  4000 Hz Subject # 40 251.05  --299.58 253.97  ------203.38  ----244.40  1 2 3 4 5 6 7 8 9 10 11  10  BC 20  310.23 232.77 305.25 313.67 235.81 283.48 250.70 266.17 270.87 283.76 255.29  321.29 214.38 312.12 281.76 251.91 231.51 239.53 276.60 248.01 290.18 246.23  30  20  294.133 102.536 360.607 226.639 310.405  362.04 461.39 429.88 331.72 383.81 --414.01 --113.08 364.961 438.13 332.865 370.92 --383.75 --11.89  30 281.59 202.52 262.74 241.08 115.60  20  335.96 259.81 334.07 332.12 322.27 --275.68 --299.46 252.19 311.67 255.97 357.85 --285.02 --235.86  AC 30 352.06 161.67 11.26 216.61 362.78 354.35 153.42 440.82 344.78 335.10 359.68  AC 30 312.35 269.73 317.17 322.04 284.34 276.09 254.49 252.54 295.22 281.41 275.91  40 264.63  --384.84 260.73  ------402.26  ----304.91  40 213.72  --299.18 236.21  ------196.85  ----231.28 143  

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}]}"
                            data-media="{[{embed.selectedMedia}]}"
                            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:
https://iiif.library.ubc.ca/presentation/dsp.24.1-0072812/manifest

Comment

Related Items