Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

An investigation into the phonological awareness skills in children with Autism Spectrum Disorder (ASD) McGee, Cheryl Lynn 2006

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

Item Metadata

Download

Media
831-ubc_2006-0573.pdf [ 2.85MB ]
Metadata
JSON: 831-1.0092723.json
JSON-LD: 831-1.0092723-ld.json
RDF/XML (Pretty): 831-1.0092723-rdf.xml
RDF/JSON: 831-1.0092723-rdf.json
Turtle: 831-1.0092723-turtle.txt
N-Triples: 831-1.0092723-rdf-ntriples.txt
Original Record: 831-1.0092723-source.json
Full Text
831-1.0092723-fulltext.txt
Citation
831-1.0092723.ris

Full Text

A N INVESTIGATION INTO T H E P H O N O L O G I C A L A W A R E N E S S SKILLS IN C H I L D R E N WITH A U T I S M S P E C T R U M DISORDER (ASD)  by CHERYL LYNN MCGEE B . S c , The University o f Alberta, 2001  A THESIS S U B M I T T E D IN P A R T I A L F U L F I L L M E N T OF THE REQUIREMENTS F O R THE D E G R E E OF  M A S T E R OF SCIENCE  in  T H E F A C U L T Y OF G R A D U A T E STUDIES  (Audiology and Speech Sciences)  T H E U N I V E R S I T Y OF BRITISH C O L U M B I A  May 2006  © Cheryl L y n n M c G e e , 2006  Abstract Phonological awareness is the reflection on or manipulation o f speech sounds within words. Many studies have shown phonological awareness. (PA) to be highly correlated with literacy skills, with children with language impairments often being delayed in acquisition o f both. However, no research has been conducted to date describing the nature o f P A skills for children with Autism Spectrum Disorder ( A S D ) . Thus, the present study was designed to evaluate the phonological awareness skills o f 11 children with A S D , ranging in age from 5 to 9 years. Five phonological awareness tasks targeting syllable, rime, and phoneme levels were administered, along with a pre-literacy (letter-sound identification) task. Four children scored above chance on all P A tasks, three were successful on two tasks, and four children did not complete any of the tasks. Four children achieved some success in subsequent teaching trials. Correlational analyses revealed significant relationships between performance on P A measures and language comprehension scores, percent consonant match (i.e., phonological production), and pre-literacy skills. According to a regression analysis, language comprehension scores and phonological production contributed most to the variance in performance on P A tasks. Children with A S D are thus similar to children with language impairment o f other origins with respect to the interaction o f language, phonology, and phonological awareness skills. Implications for future research and clinical intervention are discussed.  iii  Table of Contents Abstract  ii  Table of Contents  iii  List of Tables  vi  Acknowledgements  vii  Introduction  1  Autism Spectrum Disorder  1  Language skills in children with A S D  1  Phonological skills in children with A S D  3  Phonological Awareness  4  Phonological awareness and phonological skills  4  Phonological awareness and language ability  5  Phonological awareness and links to literacy  6  Phonological Awareness in Children with A S D  6  Research Question  7 '.  Method  8  Participants  8  General Procedures  9  Phonological awareness tasks  10  Rhyme matching I (rime level)  11  Word segmentation (syllable level)  11  Rhyme identification II (rime level)  12  Word-initial sound identification (phoneme level)  12  Word-final sound matching (phoneme level).  :  13  iv  Sound-Letter Correspondence: A Preliteracy Measure  13  Language Measures  14  Phonology  14  Language comprehension  14  Language production  1  Data Analysis  14 15  i  •  Phonology  15  Language comprehension  16  Language production  16  Phonological awareness  17  Results  17 Phonology  17  Language Comprehension  18  Language Production  19  Phonological Awareness Tasks  19  Individual phonological awareness tasks  20  Secondary training trials  21  Letter-Sound Identification: Pre-Literacy Measure  22  Statistical Analyses  22  Correlations between total P A scores and language measures  22  Correlations between individual P A tasks and language measures  23  Correlations among various language measures  24  Regression Analysis Discussion  25 26  V  Phonological Awareness Tasks and Other Language Performance  27  Individual phonological awareness tasks and overall findings  28  Phonological production and total phonological awareness scores  28  Language comprehension and total phonological awareness scores  29  Language production and total phonological awareness scores  30  Regression analysis variables  32  Secondary Training Trials.  32  Literacy  33  Pre-literacy  33  Literacy questionnaire  :  34  Interactions Between Language Measures  34  Explaining the Variability in Phonological Awareness Performance  35  Individual variation  35  Severity o f autism  36  Home language Literacy development Limitations Clinical Implications and Future Research.  :  36 37 :...37 38  References  55  Appendix A : Test Trials and Secondary Training Trials Stimuli Lists  64  Appendix B : Parent Literacy Questionnaire  68  Appendix C : U B C Research Ethics Board Certificate o f Approval  69  vi  List o f Tables Table 1  Participant demographics  40  Table 2  Procedure: Test types and order.  42  Table 3  Phonological production measures across children  43  Table 4  Language comprehension and production measures across participants  44  Table 5  Phonological awareness results across children and P A tasks  46  Table 6  Correlations between total scores on P A tasks and external variables and language measures  48  Table 7  Pre-literacy and language measures as a function o f P A group  50  Table 8  Correlations between rhyme matching I task and language measures  52  Table 9  Correlations among language measures  53  Table 10 Regression analysis for variables predicting performance on phonological awareness tasks  54  Acknowledgements Thank you to Dr. Barbara Bernhardt for agreeing to "start early" with this project. Her enthusiasm with the chosen topic, guidance, patience, support, knowledge, and quick answers to many questions and emails have been greatly appreciated throughout this arduous process. Thank you to Dr. Stefka H . Marinova-Todd and Dr. Pat Mirenda for agreeing to be members o f the thesis committee, specifically for Stefka's support and suggestions with the pilot project and statistical expertise with the data analysis, and for Pat's suggestion o f including the secondary training trials "to make the project more interesting." Thank you to Dr. Judith Johnston for her advice with language sample analysis, S A L T , and suggestions for the final manuscript. Thank you to Julie Scott for lending the P I P A on many occasions. Thank you to Sabrina Driediger for graciously loaning the transportation necessary to collect the data, the most important part o f the project. Thank you to Geeta Modha for her expertise in S A L T transcription and for agreeing to do the reliability testing. Thank you to fellow thesis writers, Erin Albright, Andrea Noland, and Erin Hall, who understand the trials and tribulations of conducting a thesis; their support on the day o f the thesis defence was deeply appreciated. Thank you to all staff, professors, and the class of 2006 in the School o f Audiology and Speech Sciences - these past three years have been well worth it. Thank you to all o f my friends and family all over Canada who have showed endless support, love, and encouragement. Thank you to the Social Sciences and Humanities Research Council of Canada ( S S H R C ) and Human Early Learning Partnership ( H E L P ) for funding that made this project possible. A n d finally, a huge thank you to the families and children who participated in this study, without which, there would be no investigation into the phonological awareness skills in children with Autism Spectrum Disorder. The families and children were truly an inspiration and I genuinely thank them for taking the time to contribute to research and for allowing me to share in their lives, even i f it was for a short period o f time.  1  Introduction Although some studies have measured the remediation o f phonological awareness ( P A ) skills in children diagnosed with Autism Spectrum Disorder ( A S D ) (Heimann, Nelson, Tjus, & Gillberg, 1995; Tjus, Heimann, & Nelson, 1998), no studies have focused specifically on the nature of P A ability in this clinical population. The objective o f the current study was to characterize the phonological awareness skills o f a small group o f children with A S D in order to begin to address this gap in the literature. Autism Spectrum Disorder Autism is a neurodevelopmental disorder, characterized by qualitative impairments in social interaction and communication, and presence o f restricted, repetitive, or stereotyped patterns o f behaviour, interests, and activities (American Psychiatric Association, 2000). Because autism is behaviorally defined and ranges in severity and variability, the term "Autism Spectrum Disorder" or A S D is commonly used. Language skills in children with A S D . Even though a deficit in communication is one o f the main diagnostic features of A S D , there is a wide range o f performance within this domain. M u c h o f the research has focused on the social communication deficits in children with A S D . For example, individuals with autism may exhibit deficits in eye gaze and joint attention (for a review, see Bruinsma, Koegel, & Koegel, 2004), topic initiation (Tager-Flusberg, 1981a), topic maintenance (Hinerman & Channell, 1986), adjusting language based on context and listener, needs (Baltaxe, 1977), and turn taking within a conversation (Loveland, Landry, Hughes, H a l l , & M c E v o y , 1988). Other difficulties in social discourse are exemplified through echolalia, perseveration, pronoun reversal, or off-topic speech (Rapin & Dunn, 2003).  2  Morphosyntactic deficits are variable across children with A S D . In an early study, Bartak, Rutter, and C o x (1977) found that children with autism were comparable to children with developmental language delay on measures o f mean length o f utterance ( M L U ) and grammatical complexity. In Kjelgaard and Tager-Flusberg's (2001) study, it was found that some children with Autism Spectrum Disorder (ages 4 to 14 years) had essentially normal vocabulary and language skills, while others performed significantly below age expectations on a comprehensive battery o f standardized tests. Considerable heterogeneity in syntactic development in children with A S D has also been noted in other studies (e.g., Roberts, Rice, & Tager-Flusberg, 2000). Language comprehension can be particularly affected in children with A S D . Bartak, Rutter, and C o x (1977) found language comprehension in children with autism to be more severely compromised than that of children with specific impairments in language comprehension. Similarly, in Tager-Flusberg's (1981b) study, children with autism had more severe sentence comprehension deficits than a matched control group o f typically developing children. In a later large-scale study, A l l e n and Rapin (1992) found that none o f their 229 subjects with A S D had normal comprehension skills, while 35% o f a group o f 262 children with developmental language disorders had comprehension skills within normal limits. Charman, Drew, Baird, and Baird (2003) found that 16 o f 18 preschool children with A S D fell below the basal level (i.e., 12-month equivalent) on the comprehension scale o f the Reynell Developmental Language Scales (Reynell, 1985). It has been suggested that language comprehension may appear more limited in children with A S D than would be expected based on their vocabulary production (Kjelgaard & Tager-Flusberg, 2001; Lord & Paul, 1997). However, impairments in language comprehension may be secondary to interacting social and/or cognitive factors also  3 manifested in Autism Spectrum Disorder (Lord, 1989; Tager-Flusberg, 1981b), or to lack o f engagement in the testing situation. Phonological skills in children with A S D . Findings have differed regarding phonological development o f children with A S D . Some studies have suggested that phonological development remains relatively intact. For example, from a review of the literature, Tager-Flusberg (1981a) established that children with autism did not differ from children matched for cognitive delay in terms o f phonology, prosody, or syntax. In Kjelgaard and Tager-Flusberg (2001), scores on the Goldman-Fristoe Test o f Articulation ( G F T A , Goldman & Fristoe, 1986) were in the normal range for 72 children grouped into three categories based on language skills: impaired (n = 50), borderline (n = 10), or normal (n = 12). However, the lowest-functioning group did show significantly lower scores than the other two groups. Other research reveals that some children with A S D do have phonological impairments. In a sample o f 229 children with A S D and 262 children with developmental language disorders, A l l e n and Rapin (1992) found evidence for mixed phonology/syntax impairments in 63% o f children with A S D and 50% o f the children with developmental language disorders. Wolk and Edwards (1993) noted unusual sound substitutions (e.g., extensive glottal replacement and segment coalescence) in one eight year-old boy with autism. Gibbon, M c C a n n , Peppe, O'Hare, and Rutherford (2004) found that 6 out o f 30 children with high-functioning autism (aged 6 to 13 years) had articulation disorders that ranged from mild to severe, measured as standard scores less than 85 on the G F T A - 2 (Goldman & Fristoe, 2000). Children in their study also produced more "atypical" substitutions than children in the normal articulation group. Information was not provided regarding the exact nature o f these substitutions, except for "phonetic realizations not seen in normal development after 2;6 years o f age" (p. 2).  /  4  Although some research has investigated the phonological development o f children with A S D , as attested by the above studies, no research to date has examined the nature o f phonological awareness skills in this population. Phonological Awareness Phonological awareness ( P A ) is the reflection on or manipulation o f speech sounds and other phonological units (e.g., words, syllables, and sub-syllabic units such as onset and rime). Phonological awareness tasks require the child to adopt a metalinguistic perspective, that is, to treat language as an object o f conscious reflection (Tunmer, 1991). Tests o f phonological awareness vary in how much implicit and explicit knowledge they require (Cataldo & Ellis, 1988). A variety o f tasks have been designed to assess phonological awareness, ranging from syllable, onset-rime, and phoneme judgments (e.g., Treiman & Zukowski, 1991) to rime and onset matching (e.g., B i r d , Bishop, & Freeman, 1995) to syllable and phoneme segmentation (e.g., Carts, Fey, Zhang, & Tomblin, 2001). Verbal responses are often required, although some tasks allow pointing or other nonverbal responses. A large amount o f research has been conducted into the development o f phonological awareness skills in typically developing children. Preschool, kindergarten, and Grade 1 children typically show the ability to segment speech at the word or syllable level earlier than at the phoneme level (e.g., Liberman, Shankweiler, Fischer, & Carter, 1974). Treiman and Zukowski (1991) revealed a developmental progression from syllable awareness, to onset-rime awareness, to awareness of single phonemes among children in preschool, kindergarten, and Grade 1, respectively. Phonological awareness and phonological skills. P A skills have been found in some studies to correlate significantly with phonological production. In B i r d et al.'.s (1995) study, 31  5  children with phonological impairment scored lower on rime and onset matching, onset segmentation, and reading than a control group matched for age and nonverbal skills. Using similar tasks, Rvachew, Ohberg, Grawburg, and Heyding (2003) found that 13 children with moderate to severe delays in phonological production, but average language comprehension scores, scored lower on P A tasks than 13 typically developing children matched for chronological age (mean age o f 4;7). There is not a one-to-one correspondence between phonological production and awareness skills, however. Some children in a study by Major and Bernhardt (1998) had severe phonological production delays but age-appropriate P A skills, whereas others had relatively mild phonological production impairments and minimal P A skills. Phonological awareness and language ability. Language skills may also be correlated with phonological awareness skills. In a study by K a m h i , Lee, and Nelson (1985), children with delays o f one year in language production and comprehension did not perform as well as typically developing children matched for mental and language age on sentence and word division P A tasks, showing particular difficulty dividing words into phonemes. Similarly, Warrick and Rubin (1992) found that 13 children with language delay, as measured on standardized tests o f comprehension and production, performed significantly below 15 typically developing children on various tasks o f phonological awareness (e.g., rhyming, phoneme segmentation, and initial phoneme isolation). In summary, those children with phonological production impairments (e.g., Bird et al., 1995; Webster & Plante, 1992; Webster, Plante, & Couvillion, 1997) and language impairments (e.g., Catts, 1993; Warrick & Rubin, 1992) appear to be at increased risk for delays in acquisition of phonological awareness.  6  P A and links to literacy. M a n y studies have shown phonological awareness skills to be highly correlated with literacy development (Catts, 1993; Larrivee & Catts, 1999; Liberman & Liberman, 1990; Muter & Snowling, 1998). From the Report o f the National Reading Panel (2000), "phonological awareness and letter knowledge are the two best school-entry predictors o f how well children w i l l learn to read during the first two years o f instruction" (p. 1). Catts, Fey, Zhang, and Tomblin (1999) found that Grade 2 children with limited reading skills were four to five times more likely to have had delays in phonological awareness in kindergarten than gradematched children with good reading skills. These authors also found that the group with limited reading skills in Grade 2 had lower scores on vocabulary, grammar, and narrative tasks in kindergarten, suggesting that early language factors, in addition to P A , may predict later reading scores (e.g., O'Connor & Jenkins, 1999). Phonological Awareness in Children with A S D Because many children with speech and language impairments have delays in P A development, children with A S D and communication deficits might also be expected to show delays in P A development. To date, however, no research has explored the nature o f phonological awareness skills in children with A S D . A few studies have investigated P A as an outcomes measure after specific treatment strategies, suggesting that some children with A S D do show deficits in phonological awareness. In Heimann et al.'s (1995) study, 11 children with autism, 9 children with cognitive impairment, and 10 typically developing children showed an increase in vocabulary skills and word reading after participating in an interactive computer program aimed at teaching basic reading and writing vocabulary. Phonological awareness scores also improved, as measured by a Swedish standardized test that assesses phoneme segmentation, synthesis, and deletion (Torneus, Taube, & Lundberg, 1984). In a related study, Tjus et al. (1998)  7  reported that 13 children with autism (aged from 4 to 11 years) showed gains i n word and sentence reading and phonological awareness skills (e.g., phoneme synthesis and segmentation) after participating in a multimedia computer intervention. They also noted that children with the highest language comprehension levels showed the most improvement in phonological awareness during training and follow-up periods. Research Question The studies o f Heimann et al. (1995) and Tjus et al. (1998) suggest that some children with Autism Spectrum Disorder can benefit from intervention for phonological awareness skills. However, no studies to date have characterized otherwise the phonological awareness development in children with autism, or the potential interaction o f development in phonological awareness with other language development variables. The current study therefore set out to document phonological awareness skills in a group o f verbal children with A S D with a range o f language abilities, investigating potential relationships between P A skills, general language performance, phonological production and pre-literacy skills. It was hypothesized that children with A S D and communication deficits would have difficulty with phonological awareness tasks because communication impairments have been linked to delays to phonological awareness development. Based on the literature, it was further hypothesized that age, phonological production abilities, and language comprehension would be significantly correlated with performance on the P A tasks, with language ability perhaps contributing the greatest variance (Catts et al., 1993; Warrick & Rubin, 1992).  8  Method Participants Eleven children with A S D participated in the study. Selection criteria were (1) use o f at least single-word utterances, (2) English as the dominant language, (3) no physical impairments, and (4) normal hearing. Children were between the ages o f 5;0 and 9;0 years ( M = 6.81, S D = 1.43) and included four girls and seven boys. Level o f mother's education ranged from 12 to 18 years o f formal schooling ( M = 16.18, S D = 2.60). A l l children had diagnoses on the Autism Spectrum, as indicated by pediatrician and/or psychologist reports obtained v i a parent permission (see Table 1). Diagnoses were made based on the Diagnostic and Statistical Manual of Mental Disorders ( D S M - I V , American Psychiatric Association, 2000), the Childhood Autism Rating Scale ( C A R S , Schopler, Reichler, & Renner, 1988), the Autism Diagnostic Interview-Revised ( A D I - R , Rutter, L e Couteur, & Lord, 2003), or a combination o f these measures. Some parents did not provide clinical reports and in these cases, parental reports of severity were utilized.  Insert Table 1 About Here  For five children, English was the only language in the home. In addition to English, three children spoke Cantonese; two spoke Mandarin; and one spoke Spanish. O f these six children, five were considered to be bilingual, based on parental report. A l l children were receiving therapy from either a school-based or private S L P for speech, language, and social skill intervention throughout the data collection, although phonological awareness was not a focus o f intervention with any o f the children. Seven children were enrolled in an Early Intensive Behavioral Intervention (EIBI) program with varying hours o f intervention per week, and  9 individualized program goals. Parents reported that their children had normal hearing as assessed through formal audiological testing. (See Table 1 for details.) General Procedures Data collection occurred in three, one-hour sessions with each child over a series o f three or four weeks. Parent(s) and/or the child's behavioural interventionist were also present during testing sessions, which aided the experimenter in learning the individualized needs, behavioural characteristics, and personalities o f each o f the children. A visual schedule constructed from Boardmaker® picture icons (Mayer-Johnson Co., 1994) was used during testing sessions to show the children the order o f tasks and play breaks. Children were encouraged to assist the experimenter in organizing the visual schedule at the beginning o f each session. A number of strategies were employed to maintain attention and motivation during the testing. Using the PowerPoint® animation feature, cartoon characters (shareware) were randomly displayed between test trials. Verbal reinforcement and encouragement were also given. For three children, edible reinforcers were utilized in order to maintain focus and compliance, because these children presented with attention and behavioural issues during testing. A t the end o f each session, children were given small toys for participating in the study (e.g., stickers, books, puzzles, bubbles, or coloring books).  Insert Table 2 About Here  A l l sessions were videotaped with a Canon Digital Video Camcorder ZR100 for later transcription and analysis. Phonological samples were also audiotaped with a Sony Recording M i n i - D i s c Walkman M Z - R 9 0 0 with an external microphone.  10  Phonological awareness tasks. Similar to studies by Puolakanaho, Poikkeus, Ahonen, Tolvanen and Lyytinen (2003), Rvachew, Nowak, and Cloutier (2004), and Sutherland and G i l l o n (2005), P A tasks were presented by computer. Stimuli were prepared using MayerJohnson Boardmaker® pictures (Version 1.6 for Windows, 1994) and Microsoft PowerPoint® (2000 Edition), and displayed on a Toshiba Portable P C (Version 1.30, 2003). Table 2 displays the specific tasks (explained further below) administered within each testing session. For all phonological awareness tasks, two training trials with corrective feedback were administered, followed by 10 test trials. If the child had three consecutive incorrect responses during a test trial, a second, 10-item training trial was subsequently administered. The objective of the second training trial was to teach the child the P A task, beginning with a high level o f cueing support, and gradually fading the cues. (See Appendix A for specific levels o f cueing and stimuli.) Cues consisted of verbal descriptions and visual supports in the form o f highlighted text (size and colour) demonstrating the phonological awareness concept (e.g., p i g and d i g for the Rhyme Matching I task) and B o a r d m a k e r ® picture icons (Mayer-Johnson Co., 1994) depicting a happy face with two ears for "sounds the same" and a red X for "doesn't sound the same." If the child had five consecutive incorrect answers on one of these training trials, the trial was abandoned. If the child was responding correctly on a secondary training trial (i.e., 5 or more consecutive right responses), the original test trial was re-administered, to determine whether the child could generalize knowledge from the more extensive training trial to the test trial. The types of phonological awareness tasks designed and adapted for the current study have been used extensively in previous research with children with phonological production impairments (e.g., Bird et al., 1995; Rvachew et al., 2003), language impairments (e.g., Warrick & Rubin, 1992), and dyslexia (e.g., Swan & Goswami, 1997). The order and difficulty of the P A  11  tasks followed the developmental sequence in phonological awareness observed by Treiman and Zukowski (1991), i.e., syllable to onset-rime to phoneme awareness. Because nonverbal responses can enhance task completion in children who may not be familiar with test-like situations, who are shy or reluctant to speak (Gillon, 2004), or who have speech production difficulties (e.g., Rvachew et al., 2003), all o f the phonological awareness tasks required only nonverbal responses from the children, such as pointing on the computer screen or selecting with the mouse. Prior to data collection, the P A tasks were piloted with a typically developing child, aged 5;7, who performed at ceiling levels on all tasks. A s a result o f this pilot session, changes were made to the layout and design of the picture stimuli on the computer screen and instructions were simplified. A description o f the individual P A tasks follows (see Appendix A for stimuli). Rhyme matching I (rime level). The rhyme matching task was adapted from Vandervelden and Siegel's (1999) Rhyme Judgment production task. The child was asked to select, from two pictures arranged side by side, the one that rhymed with a target picture situated above. For example, the child was asked, "Here are some pictures: ring, sun. Which one sounds the same/rhymes with king?" The children responded by either pointing with a finger or clicking the correct picture on the computer screen with the mouse. Wording and length o f the instructions varied according to the individual needs o f the children. For instance, some children did not understand "rhyme," but understood "sounds the same" or "sounds different." This task was chosen as the first task because rhyme matching has been found to be one o f the earliest P A skills to develop (Treiman & Zukowski, 1991). Word segmentation (syllable level). A word segmentation task was chosen in accordance with research suggesting that this type o f P A task is predictive o f reading achievement  12  (Torgesen, Wagner, & Rashotte, 1994). Adapted from Catts et al.'s (2001) deletion production task, the word segmentation task required the child to delete a syllable from a compound word and point to the remaining word. For example, the child was shown a picture o f a compound word at the top o f the screen, plus the two pictures comprising the compound word and a semantically related distracter. Instructions were: "This is a cupcake. Here we have: cup, bowl, cake. If we take cup away from cupcake, which one is left behind/left over?" Again, terminology changed depending on the child. A semantic distracter (in the above example, bowl) was included to ensure that the child understood the task. Rhyme identification II (rime level). The rhyme identification task was modified from the Rhyme Awareness subtest of the Pre-Reading Inventory o f Phonological Awareness (PIP A , Dodd, Crosbie, Mcintosh, Teitzel, & Ozanne, 2003), which is normed for children aged 5;0 to 6; 11. The same words and procedures were used as described in the test manual, but stimuli were displayed on the computer screen. The child was asked to select which o f four pictures did not rhyme with the other three. Instructions were: "Here are some pictures: clock, rock, sock, cat. Show me one that does not sound the same/does not rhyme/sounds different." This task was considered to be more demanding than the previous tasks because o f the resources required: the child had to remember four words, and then select the word that was different. Word-initial sound identification (phoneme level). The word-initial sound identification task was modified from the Alliteration Awareness subtest of the P I P A (Dodd et al., 2003) by portraying stimuli on the computer screen. The child was shown four pictures and asked to select the one that did not begin with the same word-initial sound as the other three pictures. For example, "Here are some pictures: fish, feet, fork, boat. Show me the one that does not start with the same sound at the beginning/starts with a different sound at the beginning." Because this task  13 and the Rhyme Identification I task are quite similar, they were administered in different testing sessions. Word-final sound matching (phoneme level). The word-final sound matching task was modified from the Sound Matching-Last Sound subtest o f the Comprehensive Test o f Phonological Processing ( C T O P P , Wagner, Torgeson, & Rashotte, 1999), a subtest normed for children aged 5;0 to 7;5. Using most o f the same stimuli but with computer presentation and adapted instructions, the child was asked to select from three pictures, the one that ended with the same sound as a target word. For example, "Here is a dog. Dog ends in the Igl sound. Here we have: doll, pot, pig. Show me the one that ends in the /g/ sound like dog.'" This task was considered the most difficult phonological awareness task, because judgments about word-final phonemes have been found to be more difficult than judgments about word-initial phonemes (Treiman & Zukowski, 1991). Sound-Letter Correspondence: A Preliteracy Measure A sound-letter identification task was administered, adapted from the Letter-Sound . Knowledge subtest o f the P I P A (Dodd et al., 2003). Children were shown an array o f six lowercase letters on the computer screen and were asked to identify the letter that corresponded to a particular sound. For example, the children were told, "Here are some letters. Show me the letter that makes the /m/ sound." Given the vast amount o f literature on the relationship between P A and pre-literacy skills (e.g., Catts et al., 1999; Muter & Snowling, 1998), this task was intended as a preliminary measure o f the children's literacy level, to determine i f there were any relationships between performance on the previous five phonological awareness tasks and preliteracy skills. Parents also completed a literacy questionnaire consisting o f 12 questions  14  concerning the child's level and participation in reading and phonological awareness activities (Appendix B ) . Language Measures Phonology. The Phonemic Profile o f the Computerized Articulation and Phonology Evaluation System ( C A P E S , Masterson & Bernhardt, 2001) was administered. For this task, children were presented with a series o f 46 coloured pictures on a computer screen, which include a wide range, o f consonants in word-initial, medial, and final positions. Children were encouraged to say the words spontaneously; however, i f a child did not respond, the experimenter provided a model for imitation. A s noted above, this test was both videotaped and audiotaped to ensure adequate sound quality for phonetic transcription. Language comprehension. The Sentence Structure subtest from the Clinical Evaluation o f Language Fundamentals® - Preschool ( C E L F - P , W i i g , Secord, & Semel, 1992) or the Clinical Evaluation o f Language Fundamentals® 4 Edition ( C E L F 4, Semel, W i i g , & Secord, 2003) was th  administered, depending on the child's age. The preschool version is suitable for children aged 3;0 to 6;11, and the C E L F 4 for individuals aged 6;0 to 21 ;11. The Sentence Structure subtest evaluates children's ability to interpret spoken sentences of increasing length and complexity and select the picture that illustrates the meaning o f the sentence. In accordance with the special testing considerations listed in the test manuals, two of the children (Child #7 and C h i l d #8) received the preschool edition o f the C E L F because they appeared to be functioning at a 3- to 6year old level in language development, although they were not in the chronological age range for the CELF-Preschool. Language production. For the first session, the first author visited the child's home with the purpose of familiarizing the. child with an unknown adult. Activities were chosen based on  15  the child's interests, pre-determined upon asking the parent(s). Because all sessions were videotaped, this play session contributed to some o f the spontaneous language sample. The collection o f the language sample continued in the second and third sessions, utilizing the experimenter's activities (e.g., Magna Doodle®, Perfection®, Play-Doh Fun Factory®, bubbles, an interactive dinosaur magnet book, and various other books). The goal was to obtain 100 spontaneous utterances per child across the three testing sessions. A l l language samples were later transcribed and entered into the Systematic Analysis of Language Transcripts (SALT, M i l l e r , 2003) software program. A l l but Child #7 (who produced only 48 utterances) produced at least 100 utterances. Data Analysis Phonology. With C A P E S (Masterson & Bernhardt, 2001), a phonological match/mismatch comparison to the adult targets (i.e., a relational analysis) was conducted to obtain the following measures: percentage o f matching consonants ( P C M ) , percentage o f matching vowels ( P V M ) , percentage o f word shape match ( W S M ) in terms o f C V sequences, and percentage o f stress pattern match (SPM). The author/experimenter transcribed the C A P E S words for each child using the International Phonetic Alphabet (IPA) and narrow transcription after undergoing training by her research supervisor. Interrater reliability was 91.0% for consonants, 87.0% for vowels, and 60.0% for diacritics. The low match in terms o f diacritics was due to abnormal voice quality and/or abnormal prosody o f some o f the children (e.g., devoicing, glottal fry, high and low tones, elongated vowels and consonants). (See Table 3 for a qualitative description o f individual voice characteristics.)  16  Language comprehension. R a w scores from the Sentence Structure subtest o f the C E L F Preschool ( W i i g et al., 1992) and the C E L F 4 (Semel et a l , 2003) were initially transformed into standard scores and percentiles. However, because most o f the children's standard scores were in a very low range and below the 5 percentile, raw scores were used for the statistical analyses. th  Language production. For the language production data, 10% o f the language samples were randomly selected for transcription by a second observer unfamiliar with the objectives o f the research project. Morpheme by morpheme reliability was 85.2% (the number o f agreements divided by the number o f agreements plus disagreements). Although a variety of measures can be derived from the Systematic Analysis o f Language Transcripts ( S A L T , Miller, 2003), the following were used for the purposes o f this study: mean length o f utterance ( M L U ) , M L U 2 (explained below), number o f different words ( N D W ) , and number of total words ( N T W ) . M L U was chosen because it is a common measure o f morphosyntactic development and has been used extensively in research with children with language impairments (e.g., Hewitt, Hammer, Yont, & Tomblin, 2005), and children with autism (Condouris, Meyer, & Tager-Flusberg, 2003; Tager-Flusberg et al., 1990). A n alternative M L U measure, M L U 2, was also calculated as suggested by Johnston (2001), whereby answers to questions and yes/no responses are eliminated from the language sample. N D W and N T W were used because it has been suggested that N D W serves as a strong measure o f semantic diversity and N T W provides an index o f general vocabulary productivity (Miller, Freiberg, Rolland, & Reeves, 1992). N D W has become the preferred measure in child language studies (e.g., Goffman & Leonard, 2000) because it has been found to be a more sensitive estimate o f children's lexical diversity than the Type-Token Ratio (TTR) (Watkins, K e l l y , Harbers, & Hollis, 1995). N D W requires samples o f equivalent length across subjects; thus, calculations were made based on the  17 first 100 utterances for all participants except Child #7 (who was not included in this analysis, due to lower output). Phonological awareness. Both individual task scores and total scores across all P A tasks were used to evaluate the children's P A skills. A number o f correlational analyses were done, using a non-parametric Spearman's correlation coefficient (rho), due to the small sample size and the inability to assume a normal distribution with the children with A S D in this study. A l l statistical analyses were computed using SPSS for Windows 95, Version 10.0. Correlations were performed between the various language measures, between total phonological awareness scores and various language measures and between individual P A tasks and language measures. Results Phonology Although the children were encouraged to produce the Phonemic Profile words spontaneously, percentage o f the samples imitated ranged from 6.5% to 63.0% ( M = 30.0%, S D = 16.6%). (See Table 3 for individual results). The mean percentage of consonant match ( P C M ) was 68.8% (SD = 19.5%), mean percentage o f vowel match ( P V M ) was 71.5% (SD = 17.9%), mean percentage word shape match ( W S M ) was 73.9% (SD =.26.7%), and mean percentage stress pattern match ( S P M ) was 87.0% (SD = 9.3%). Table 3 displays individual data regarding consonant, vowel, word shape, and stress pattern matches.  Insert Table 3 About Here  Due to the unforeseen difficulty in achieving reliability for diacritics, an additional analysis was conducted whereby diacritics were omitted and only matches for consonants and vowels were analyzed. For example in the above analysis, i f a child produced the word pig as  18  [ p i g j , the devoiced / g / would be considered a mismatch. However, in the second analysis, the h  [g] would be considered a match for Igl because the child did produce the correct phoneme (i.e., broad transcription). With diacritics omitted, P C M ranged from 21.9% to 93.3%, with the mean increasing to 76.2% (SD = 21.3%). M e a n values in P V M increased to 83.6% (SD =  9.6%),  with  a range from 58.8% to 91.2% accuracy. Even with diacritics omitted from this analysis, there were still low values in percentage o f matching vowels. This was attributed to the children with abnormal prosody and the five children who spoke Mandarin and Cantonese, which may have affected vowel quality. Severity o f phonological impairment was estimated by comparing the children's Percent Consonant Match ( P C M ) on the C A P E S single word naming task to Austin and Shriberg's (1997) lifespan reference data for Percent Consonants Correct in connected speech samples. Three children (#5, #7, #11) had relatively severe phonological impairments (48.6%, 21.0%, 70.5%, respectively); three children (#6, #8, #9) had moderate impairments (60.0%, 79.1%, 73.3%, respectively); three children (#1, #2, #10) had mild impairments (76.2%, 74.2%, 90.5%, respectively); and Child #3 had normal phonological production ( P C M o f 82.9%). (See Table 3). Language Comprehension Mean raw scores for the eight children that received the Sentence Structure subtest from the CELF-Preschool (Wiig, et al., 1992) were 13.6 (SD = 5.7). Scores ranged from 4 to 22, out o f a possible 22. For the three children given the C E L F 4 (Semel et al., 2003), mean raw scores were calculated at 15.3 (SD = 4.0), and ranged from 11 to 19 out o f a possible 26 correct. Based on the percentile range of performance scale, eight children had scores below the 5  th  percentile.  Child #3 performed in the 'superior' range (between the 9 5 and 9 9 percentile), C h i l d #1 th  th  19 scored in the 'average' range (between the 2 5 and 74 percentile), and C h i l d #8 scored in the th  th  'borderline average range' (between the 6 and 15 percentile). (See Table 4). th  th  Insert Table 4 About Here  Language Production For all but C h i l d #7 (who produced 48 utterances), the average number o f utterances for 10 children was 104.2 (SD = 2.70, range of 101 to 109). (See Table 4). For all participants, the average Mean Length o f Utterance ( M L U ) in morphemes was 2.80 (SD = 1.03), with a range o f 1.31 to 4.66. The alternative measure, M L U 2, was also calculated, in which answers to questions and yes/no responses were omitted (Johnston, 2001). The mean for M L U 2 values was slightly higher (M= 3.05, S D = 1.11, range o f 1.37 to 5.06). The small difference between the two M L U measures suggests that the low M L U values were not a consequence o f the experimenter's interactional style or question asking. Excluding Child #7, who produced 48 utterances, Number o f Different Words ( N D W ) in the first 100 utterances ranged from 57 to 128 ( M = 100.9, S D = 23.8). For all participants, mean Number o f Total Words ( N T W ) in 100 utterances was 253.6 (SD = 81.5). N T W values ranged from 121 to 411, revealing considerable between-subject variability. (See Table 4). Phonological Awareness Tasks At-chance performance on the phonological awareness tasks is a score o f 4 or less (p > .05). Four children (# 5, #6, #7, #11) scored below chance levels on one or two o f the P A tasks. Children #1 and #8 scored above chance levels on two of the tasks (Rhyme Matching I and Word Segmentation), Child # 4 scored above chance on three o f the tasks (Rhyme Matching I, Word Segmentation, and Rhyme Identification II), and four children (#2, #3, #9, #10) scored above  20  chance on all five tasks. For the remainder o f the results and discussion, the four children who had difficulty with the P A tasks w i l l be referred to as the " L o w P A " group, the three children who were successful with some o f the tasks w i l l be referred to as the "Middle P A " group, and the four children who were able to do all five P A tasks w i l l be referred to as the " H i g h P A " group. Table 5 shows results from the phonological awareness tasks.  Insert Table 5 About Here  Individual phonological awareness tasks. Rhyme Matching I was the only task administered to all participants. R a w scores on the Rhyme Matching I task were highly correlated with all language measures. (See Table 8). Seven children attempted all five phonological awareness tasks. For Child #6 and #7, testing was discontinued after administering the first P A task (Rhyme Matching I), because it was believed the remaining P A tasks would be too difficult, given the children's lower language comprehension levels. Attention and behavioral issues also interfered with testing for these two children. For C h i l d #5 and C h i l d #11, testing was discontinued after attempting the Rhyme Matching I and Word Segmentation tasks. Based on the children's responses during testing, it appeared that these four children did not understand what was meant by "rhyme," even when the instructions were simplified (e.g., "sounds the same") and visual cues were added with the second set o f training trials. A s mentioned, children in the Middle P A group either completed two (Rhyme Matching I, Word Segmentation) or three (Rhyme Matching I, Word Segmentation, and Rhyme Identification II) o f the P A tasks. See Table 5 for individual P A task results.  21  Secondary training trials. If the child obtained three incorrect responses in a row during the test trials, the test trials were discontinued and a secondary training trial (maximum 10 tokens) was presented. Secondary training trials were administered to the children in the L o w P A and Middle P A groups only, because the four children in the H i g h P A group did not require these trials. Three o f the children in the L o w P A group scored below chance on the training trials. Child #11 scored 7/10 and 8/10 on the secondary training trials o f the Rhyme Matching I and Word Segmentation tasks, respectively. However, when the test trials were re-administered, C h i l d # 11 scored below chance on both tasks, revealing that generalization had not occurred. The three children in the Middle P A group were able to obtain near-perfect scores on the secondary training trials, but only one showed any generalization to the test trials. C h i l d #8 scored 10/10 on Rhyme Identification II, 8/10 on Word-Initial Sound Identification, and 9/10 on Word-Final Sound Matching secondary training trials, but scored below chance levels during readministration of the test trials. Child #4 scored 8/10 on the Word-Initial Sound Identification secondary training trials, but scored below chance on the re-administration o f the test trials. C h i l d #1 scored 10/10 on the secondary training trials o f the Word-Initial Sound Identification task, and then scored 8/10 when the test trials were re-administered, revealing some generalization of this particular P A skill. However, with the Rhyme Identification II task, C h i l d #1 scored 10/10 on the secondary training trials, but below chance on re-administration o f the test trials. Observations during the secondary training trials revealed that the introduction o f the picture icons helped the children in the Middle P A group learn the teaching stimuli. Some children even used some o f their own strategies. For example, C h i l d #1 used his own fingers and made them in the shape o f an " X " when two words did not rhyme, and pointed to his own ears  22  when words rhymed or had the same initial sound. It is important to note that C h i l d #1 and Child #4 used the picture icons more than the text during the secondary training trials, because they were still developing literacy skills, as indicated by parental report during testing. Letter-Sound Identification: Pre-Literacy Measure A l l children were able to complete the Letter-Sound Identification task. Out o f a possible 10 correct responses, mean scores were calculated at 7.91 (SD = 2.39), and ranged from 3 to 10. Table 5 displays the individual results for this pre-literacy task. From the literacy questionnaires that were returned by the parents (for all but C h i l d #2), parents reported that their child enjoyed reading or looking at books. Parents also commented on their child's current reading level. For children in the L o w P A group, all parents reported a oneyear difference in grade and reading level, except for one Grade 3 student reported to be two years behind in reading. For the Middle P A group, both parents o f kindergarten children reported their children on par with grade level expectations, while one parent reported her Grade 2 child to be one year behind. There was variability among the children in the H i g h P A group. One parent reported a kindergarten child to be two years ahead in reading ability (this appeared to be confirmed according to experimenter observation). One Grade 3 child was reported to be on par with reading expectations, and another Grade 3 student to be one year behind (this was also the participant reported not to enjoy reading activities).  Statistical Analyses Correlations between total P A scores and language measures. Table 6 displays all correlations for total P A scores versus external variables (e.g., age, mother's level o f education, home language), and language measures; only significant results w i l l be discussed in this section.  23  Insert Table 6 About Here  There was a high correlation between performance on phonological awareness tasks and the phonological measure, Percent Consonant Match: rho = 0.83, p = 0.001. Word Shape Match ( W S M ) was moderately correlated with performance on P A tasks: rho = 0.68, p = 0.021. Performance on phonological awareness and language comprehension raw scores were also highly correlated: rho = 0.7\,p = 0.014. For production, M L U 2 in morphemes and P A were moderately correlated, but approached significance: rho = 0.57,p = 0.067. Performance on phonological awareness tasks and the Letter-Sound Identification task revealed a high correlation: rho = 0.84,/? = 0.001. There was also a high correlation between reported reading level from the parent literacy questionnaire and performance on P A tasks: rho = 0.76, p = 0.01 (n = 10 because data was missing for one child). To portray the individual variability across participants, one measure o f phonological production ( P C M ) , one measure o f language production ( M L U 2), and language comprehension raw scores are displayed in Table 7 as a function o f how the child performed on all five phonological awareness tasks (total P A scores). R a w scores on the Letter-Sound Identification task are also included.  Insert Table 7 About Here  Correlations between individual P A tasks and language measures. Individual correlational analyses were conducted between each o f the phonological awareness tasks and all language measures. However, only significant correlations or moderate correlations that approached significance w i l l be discussed in this section. Only three' o f the five P A tasks  24  revealed significant correlations: Rime Matching I, Word Segmentation, and Word-Final Sound Matching. For the Rime Matching I task, there were moderate to high correlations among all language measures (e.g., phonology, comprehension, and production) and scores on the LetterSound Identification task (see Table 8 for specific results).  Insert Table 8 About Here  For the Word Segmentation task, only two significant correlations arose. Scores on Word Segmentation and language comprehension raw scores were highly correlated: rho = 0.77, p = 0.016. Word Segmentation was also highly correlated with scores on the Letter-Sound Identification Task: rho = 0.72, p = 0.030. Although approaching significance, performance on the Word Segmentation task and Percentage Consonant Match ( P C M ) were moderately correlated: rho = 0.66, p = 0.055. Note that n = 9 for this analysis, because C h i l d #6 and C h i l d #7 were not given the Word Segmentation task. For the Word-Final Sound Matching P A task, age was highly correlated with performance on this task, but not significant: rho = 0.75, p = 0.084. Performance on Word-Final Sound Matching and the Letter-Sound Identification was also highly correlated: rho = 0.89, p = 0.017. For this analysis, n = 6 because the Word-Final Sound Matching task was not administered to Children #4, #5, #6, #7, #11. Correlations among various language measures. A n analysis was also conducted among the different language measures to determine any significant relationships (see Table 9). There was a moderate, but insignificant relationship between language comprehension raw scores and M L U 2: rho = 0.57, p = 0.067. For language production, there were very high, significant correlations between M L U 2 and Number o f Difference Words ( N D W ) : rho = 0.981, p = 0.001  25  0 = 10 because C h i l d #7 did not have 100 utterances); and M L U 2 and N T W : rho = 0.982,/? = 0.001 (n= 11). There was also a moderate, significant correlation between Percentage Consonants Match ( P C M ) and other scores: (a) language comprehension raw scores rho = 0.64, p = 0.034, (b) Mean Length o f Utterance ( M L U ) : rho = 0.69, p = 0.019; (c) M L U 2: rho = 0.71, p = 0.015; (d) Number o f Total Words ( N T W ) : rho = 0.71, p = 0.015; and (e) P C M and LetterSound Identification performance: rho = 0.72, p = 0.012.  Insert Table 9 About Here  Regression Analysis Even though it is a parametric method, a regression analysis was conducted to evaluate the potential relationship o f the significantly correlated variables with phonological awareness. It is recommended that the number o f participants be 5 or 10 times the number o f variables entered into the regression analysis (Norman & Streiner, 1998). Therefore, using the minimum recommendation o f five participants, two variables was the maximum that could be entered. The variables entered into the analysis were percentage consonant match ( P C M ) and raw scores on the language comprehension subtest. In M o d e l 1, language comprehension contributed significantly to the model (F [1,9] = 8.31,/? = 0.018, explaining 48% o f the variance (r = 0.48). 2  However, when language comprehension and Percentage Consonant Match were entered into the model, both variables explained approximately 52% o f the variance in P A performance (r = 2  0.524), essentially reaching significance (F [2, 8] = 4.40, p = 0.051). M L U 2 was also entered into the regression for descriptive purposes, even though this did exceed the recommended number o f variables (Norman & Streiner, 1998). M L U 2 did not contribute any change in variance when it was added to the model. Table 10 displays the regression analysis results.  26  Insert Table 10 About Here  Discussion This study endeavored to describe the phonological awareness ( P A ) skills o f 11 children diagnosed with Autism Spectrum Disorder ( A S D ) . Five tasks targeting different levels o f linguistic knowledge were designed to measure phonological awareness: one task at the syllable level (Word Segmentation), two tasks at the rime level (Rhyme Matching I and Rhyme Identification II), one task at the initial phoneme level (Word-Initial Sound Identification), and one task at the final phoneme level (Word-Final Sound Matching). In contrast to the original hypothesis, not all children had difficulty with the P A tasks. Four children scored above chance performance on all five P A tasks (i.e., the H i g h P A group), three children were successful with two or three tasks (i.e., the Middle P A group), and four children had difficulty with even the simplest P A task (i.e., the L o w P A group). For the Middle P A group, tasks that children had some success with were rhyme matching and word segmentation. For C h i l d #1 (5.0 years) and C h i l d #4 (5.8 years), this is consistent with the literature on the development o f phonological awareness. Typically developing preschool and Kindergarten children develop knowledge o f the syllable level and the onset-rime level, respectively, but usually do not develop phoneme awareness until Grade 1 (Treiman & Zukowski, 1991). For the children in the Middle P A group, the introduction of secondary training trials also had some positive effects, especially for C h i l d #1 who showed some generalization with the Word-Initial Sound Identification task. (See further discussion below).  27  Chronological age was not related to total P A performance (unlike what has been found in other studies), or any on the individual phonological awareness tasks, except the Word-Final Sound Matching task. This particular finding supports the developmental progression o f phonological awareness skills as proposed by, e.g., Treiman & Zukowski (1991) for this one task. In general, older children found this task less difficult. It was not too surprising that age did not correlate with performance on phonological awareness tasks for the children in this study. Because no research has been conducted on the P A skills in this population, the P A tasks were designed following previous research with typically developing children (e.g., Treiman & Zukowski, 1991), children with phonological production impairments (e.g., B i r d et al., 1995), and children with language impairments (e.g., Warrick & Rubin, 1992). From the results o f this study, other factors such as language ability seem to affect performance on P A , more so than age. Because all children were diagnosed with Autism Spectrum Disorder, variability and severity in autistic characteristics may have been a more prominent factor than age, and thereby have affected phonological awareness. (See discussion regarding severity o f autism below.) From the three language measures entered into a regression analysis (percentage consonants match for phonology, language comprehension raw scores, and M L U 2 for language production), language comprehension and phonological production contributed most to the variance in P A performance. Overall, results from this preliminary investigation reveal that some children with A S D have the ability to perform metalinguistic tasks requiring conscious manipulation of phonological information. Phonological Awareness Tasks and Other Language Performance Results from this study reveal heterogeneity in phonological production, language comprehension, and language production across the 11 participants. (See also Table 7). This  28  supports previous findings indicating a large amount o f variability i n the language skills of children with A S D (e.g., Kjelgaard & Tager-Flusberg, 2001). In spite o f the variability, there were certain consistent patterns in the data, which are discussed below. Individual phonological awareness tasks and overall findings. Results from the correlational analyses reveal that raw scores on the Rhyme Matching I task were highly correlated with all language measures. (See Table 8). Rhyme Matching I was designed as the easiest P A task, given the reduced demands on language comprehension and memory this task entailed. This was the only task that could be administered to all participants. Findings for other tasks could be an artifact o f performing statistical analyses with small sample sizes. However, scores on the Word Segmentation task and the Letter-Sound Identification task were highly correlated. This finding suggests that word segmentation may be related to sound-letter correspondence teaching. The remainder o f the discussion describes performance on various language and speech tasks and the total score for P A tasks. Phonological production and total phonological awareness scores. Based on Austin and Shriberg's (1997) lifespan data, children in this study ranged from having severe to mild phonological impairment, with only one child (#4) having normal phonological production for his chronological age. These deficits in phonological production are contrary to the literature suggesting that children with A S D have spared articulation skills (e.g., Tager-Flusberg, 1981a) or a delay in phonological development (e.g., Bartolucci & Pierce, 1977). Nine o f the 11 children presented with atypical voice and/or prosodic characteristics, which made phonetic transcription difficult. However, these findings support previous research suggesting that suprasegmental characteristics of speech, such as lexical and syntactic stress, pitch, and intonation may be affected in individuals with A S D (Lord & Paul, 1997; Tager-Flusberg, 2001).  29  There was a significant, high correlation between performance on phonological awareness tasks and percentage consonants match ( P C M ) , thereby confirming one o f the initial hypotheses that children with more severe phonological production impairments may have more difficulty with the P A tasks than children with milder or no phonological impairment. This finding is similar to other studies in the literature in which children with phonological production impairment often show reduced skill on phonological awareness tasks (Bird et al., 1995;^ Rvachew et al., 2003; Webster & Plante, 1992). Language comprehension and total phonological awareness scores. From the percentile scores o f the Sentence Structure subtest, children's scores ranged from the superior range to a very low level in language comprehension. Due to problems with converting raw scores into meaningful standardized scores, descriptive labels (e.g., severe, moderate, high) were assigned based on raw scores (see Table 7). Regardless, both percentiles and raw scores reveal that children in this study had considerately low comprehension skills compared to typically developing children on whom the CELF-Preschool ( W i i g et al., 1992) and the C E L F 4 (Semel et a l , 2003) were standardized. This finding validates previous studies suggesting some children with A S D have difficulties with language comprehension (e.g., Bartak, Rutter, & C o x , 1977; Tager-Flusberg, 1981b). The language comprehension scores were surprisingly lower than expected and possibly not representative of the children's actual understanding o f language. For example, C h i l d #1 refused to do any test items that contained "boys" or "girls" or the pronouns, "he" or "she," which affected the total raw score. It should also be noted that raw scores from the C E L F Preschool and the C E L F 4 were combined together in the analyses. It is possible that the Sentence Structure subtest on the preschool version is simpler to understand that that o f the  30  C E L F 4. However, after examining the raw scores and taking the experimenter's impression o f the children into account, the difference in scores for the three children given the C E L F 4 was considered to be minor. Limitations o f using standardized tests to assess language skills in children with A S D have been noted (e.g., Tager-Flusberg, 2000). Because only one subtest was used to assess language comprehension in this preliminary study, more comprehensive formal and informal testing should be conducted in future studies to obtain a more accurate portrayal o f language comprehension in children with A S D . There was a high, significant relationship between performance on phonological tasks and raw scores from the language comprehension subtest. Based on Catts (1993) and Warrick and Rubin (1992), language comprehension was a possible strong predictor o f phonological awareness performance. Thus, this study adds to that body o f research, but with a small group of children with A S D now included. The language comprehension findings also confirm one o f the initial research hypotheses that children with impairments in language comprehension tended to have more difficulty with the phonological awareness tasks. Findings from this study are also related to Tjus et al.'s (1998) study that found that children with A S D and higher language comprehension levels showed improvement in phonological awareness after a language-based computer intervention. T w o studies now reveal a possible link between P A skills and language comprehension in children with A S D . Language production and total phonological awareness scores. From the language sample analysis, there was a wide range in mean length o f utterance ( M L U ) , number o f different words ( N D W ) , and number o f total words ( N T W ) across the participants. However, all measures were considerably lower compared to what has been suggested in the literature for typically developing children. For instance, a typically developing 5 year-old should have an M L U o f  31  approximately 5.5 (Brown, 1973; Johnston & A m m o n , 1985). For a typically developing 5 yearold producing 100 utterances, N D W should range from 156-206 and N T W should range from 439-602 (Leadholm & M i l l e r , 1992). From examining Table 4, none o f the children met criterion for even the lowest "typical" values for N D W or N T W . However, the language production results were not necessarily unexpected for children with A S D . It was expected that children in this study would possibly have lower levels o f language output because one o f the defining characteristics o f autism is the delay or absence o f spoken language (American Psychiatric Association, 2000). From the correlational analyses, the alternative measure o f Mean Length o f Utterance ( M L U 2, Johnston, 2001) and performance on phonological awareness tasks were moderately, but not significantly related, revealing a moderate trend that some children with lower morphosyntactic skills had more difficulty with the P A tasks. There was no significant correlation between total phonological awareness scores and Number o f Different Words ( N D W ) . This reveals that semantic diversity was not related to performance on phonological awareness tasks for the 11 children in this study. There was a moderate, but insignificant relationship between Number o f Total Words ( N T W ) and phonological awareness scores, suggesting that some children with higher vocabulary production measures had less difficulty with the P A tasks. This finding validates previous research suggesting that as children's vocabularies increase, so does their ability to perform phoneme awareness tasks (Metsala, 1999). However, vocabulary was informally assessed only through language sample analysis in this study. Future research should include standardized measures o f vocabulary comprehension and production to confirm this finding.  32  Regression analysis variables The regression analysis showed that phonological production, as measured by percentage consonants match, and language comprehension raw scores were the two strongest predictors o f P A performance. This finding supports the results from the correlation analyses; percentage consonants match and language comprehension were two o f the language measures highly correlated with P A . This finding also confirms the original hypothesis in that language ability, especially language comprehension and phonological production, contributed the most to performance on P A . However, caution must be warranted when interpreting regression models based on relatively small sample sizes (Norman & Streiner, 1998). This preliminary investigation suggests a trend in that language comprehension and phonological production may be strong predictors of P A in children with A S D . Larger scale studies should be conducted to confirm this finding. Secondary Training Trials One child in the L o w P A group and all children in the Middle P A group used the visual supports during the secondary training trials. For example, with the Word-Initial Sound Identification task, Child #1 initially scored 0/10 on the test trials, achieved a score o f 10/10 with the secondary training trials, and then scored 8/10 when the test trials were re-administered, revealing that some generalization o f this particular phonological awareness skill had occurred. However, when the Rhyme Identification II task was given in the next testing session, C h i l d #1 scored 10/10 on the secondary training trials, but below chance levels when the test trials were re-administered. This shows the variability within one child's performance, across testing days, and between different P A tasks. Interesting results from the secondary training portion o f this study, however was that the children in the Middle P A group used the picture icons, and to a  33  lesser extent, the highlighted text. This became apparent as support was gradually faded out; one child asked, "Where are the ears?" There is a growing body o f literature describing the benefits of using visual strategies to teach children with A S D (e.g., Dettmer, Simpson, Myles, & Ganz, 2000; Tiscott & Evans, 2003). Even though the secondary training trials were administered to all children in the L o w P A group, the additional visual supports may have been detrimental to learning. One parent commented that she thought her child was confused because there were too many distracters on the computer screen. This may be important information to note for children with developing P A skills and lower language comprehension skills. A limitation to the secondary training trials may have been in the initial design. Support may not have been faded as quickly as possible; the least amount o f support (i.e., without the picture icons) only involved Trial 8 and 9. (See Appendix A . ) In hindsight, the picture icons could have been removed after the first five trials to determine i f the children could achieve success on Trials 6 to 10 without the use o f visual supports. Such an alteration in cue reduction could be attempted in future studies. Alternatively, a longer set o f secondary trials might be included. Literacy Pre-literacy. A l l participants, except for one child in the L o w P A group, scored above chance levels on the Letter-Sound Identification task, revealing that most children had adequate sound-letter correspondence, despite having some deficits in phonological awareness. There was also a significant relationship between performance on phonological awareness tasks and lettersound recognition skills, providing additional evidence for the link between phonological awareness and performance on pre-literacy measures, such as sound-symbol correspondence  34  (e.g., Catts, 1993; Catts et a l , 1999; Muter & Snowling, 1998). Informal observation during data collection revealed that six letters on the computer screen might have been too many for some participants. Scores for this task may have been higher if.there had been fewer letters from which to choose. Literacy questionnaire. There was a significant relationship between reading level as reported by parents from the literacy questionnaire and performance on phonological awareness tasks. This study could potentially add to the literature linking phonological awareness to literacy skills (e.g., Catts et a l , 1999; Larrivee & Catts, 1999), but for a group o f children with autism. However, caution must be warranted regarding the anecdotal nature o f this literacy measure. Parents may not have known their child's reading level and could have rendered a guess when completing the literacy questionnaire. Also, results should be interpreted with caution because literacy expectations increase as children enter higher grade levels; it is difficult to compare reading expectations in kindergarten with literacy skills in Grade 3. To alleviate this, future studies investigating literacy in children with A S D should include standardized measures o f reading ability. Interactions Between Language Measures There was a moderate, but insignificant, correlation between language comprehension and morphosyntactic production (i.e., M L U 2) for the 11 children in this study. This supports some previous findings proposing that comprehension skills can be similar to production skills in children with A S D (e.g., Jarrold, Boucher, & Russell, 1997). However, most literature has suggested that language comprehension is more impaired than production (Bartak et al., 1975; Lord & Paul, 1997). Given the small sample size in this study and the wide range in variability, the moderate relationship between language comprehension and production is merely a trend and  35  not representative o f all the children in the group. For example, C h i l d #2 had a "moderate" impairment in language comprehension, but a relatively high M L U 2 (e.g., 4.17), compared to the other children. (See Table 7 for more individualized results). Explaining the Variability in Phonological Awareness Performance Individual variation. Even though results from the correlational analyses suggest potential relationships between phonology and language comprehension and P A skills, there is a considerable amount o f variability in the language ability and phonological awareness skills across the 11 participants (See Table 7). For instance, given the trends from the correlations, one might expect children in the L o w P A group to have " l o w " skills in phonological production, language comprehension, and language production, and so on for the children in the M i d and H i g h P A groups. This is depicted in Table 7 with left arrows (e.g., <-), however this was only a general trend for some o f the children. The asterisks in Table 7 portray certain "anomalies" in the data. For example, given that C h i l d #8 is 7;5, with "moderate" phonological impairment and "high" language comprehension, one might expect to see this child in the H i g h P A group. However, this child was successful with only the first two P A tasks; therefore factors other than age, phonology, and language comprehension might better explain performance on P A tasks, such as characteristics and severity of Autism Spectrum Disorder. With respect to C h i l d #9 with "moderate" phonological impairment and "high" language comprehension skills, one might expect to see P A performance at the intermediate level. However, C h i l d #9 spoke Cantonese and had a lateral lisp characterized by ungrooved sibilants (e.g., [t OU?.] for toes and [haup] for house), which contributed to the h  "moderate" impairment in phonological production. Therefore, phonetic details, not a phonological characteristic, lead to the assignment of "moderate" impairment. C h i l d #2 and  36  Child #10, both in the H i g h P A group, showed anxiety with some o f the testing. A s previously mentioned, C h i l d #2 refused to answer any questions pertaining to boys, girls, or their respective pronouns (e.g., he, she) with the CELF-Preschool, and became visibly upset (e.g., covering his ears and verbally protesting) during administration o f this test. C h i l d #10 also showed anxiety in that he would not respond unless absolutely sure o f the correct answer when given the C E L F 4. Exclamations o f "I don't know" were common throughout language and phonological awareness testing with this child. For both C h i l d #2 and #10, anxiety factors lead to lower raw scores and the subsequent assignment o f "moderate" language comprehension skills. In summary, there did not seem to be a one-to-one relationship between P A skills, phonology, and language comprehension for most o f the participants. Some o f the "anomalies" in the data can be explained by idiosyncrasies in behavior, revealing the individualistic nature o f the children in this study. Severity o f autism. Because there was missing data concerning the level and severity o f autism, and because a variety o f autism diagnosis scales lead to inconsistent results, severity o f autism could not be entered as a variable into the analysis. Based on the diagnoses that were obtained, parental reports, and experimenter observations during testing, severity o f autism did seem to be a contributing factor in how children performed on the phonological awareness and language testing. However, this cannot be confirmed without a standardized assessment. Future research should include a formal measure o f autism, for example, the Childhood Autism Rating Scale (Schopler et al., 1988), to validate these observations. Home language. Because the children's home language revealed little or no correlations with performance on phonological awareness tasks or any o f the language measures, it can be assumed that the presence of a second language did not affect the variability observed in  37  phonological awareness tasks. Again, however, a more formal assessment should be conducted to determine this. T w o o f the children in the L o w P A group had a different home language, as did two children in the H i g h P A group, either o f which could have influenced their ability to do phonological awareness tasks. Literacy development. Children in this study may have been at different points in the developmental phonological awareness sequence (Treiman & Zukowski, 1991), and therefore at different stages in literacy development. The reciprocal nature o f literacy and phonological awareness has been noted in the literature, in that phonemic awareness may develop as a result o f learning to read. It has also been suggested that the ability to identify rime and onset is usually present before children learn to read (Bird et al., 1995). This may or may not have been the case for children in this study. Children in the L o w P A group, who were one to two years behind in reading level in school, performed poorly on all phonological awareness tasks. For children in the Middle P A group, two o f the three children were on par with reading level, but all were successful with at least two o f the P A tasks. Children in the H i g h P A group varied, in being either above, at the same level, or below reading expectations, but able to complete all five P A tasks. Caution must be used when interpreting the children's literacy skills from parental report. However, these findings contribute to the variability in phonological awareness and literacy skills observed for this group o f children with autism. Limitations Because o f the small sample size, this study reveals trends in the relationships between language, phonology, and phonological awareness skills in children with autism. Generalization  38  to the general population o f children with A S D is limited. Future studies should include more participants to confirm the findings o f this preliminary study. To limit the amount o f testing carried out with the participants, neither a test o f nonverbal intelligence nor a formal assessment measuring severity o f autism were administered. Also,' only one pre-literacy measure was administered, because exploration o f literacy skills was not the main objective of the study. Future research should consider such factors. . Due to the investigative and descriptive nature o f this study, a control group o f typically developing children was not included. The phonological awareness tasks were piloted with a typically developing 5-year old, who performed at above chance on all tasks. From the literature on phonological awareness and typically developing children (e.g., Liberman et al.,. 1974; Treiman & Zukowski), it could be assumed that the same results may have been obtained with other typically developing children, with some variability. However, future research should include a control group to validate this assumption. Clinical Implications and Future Research This study has possible implications for children with A S D with phonological and language comprehension deficits. The four children in the L o w P A group scored at or below the 1 percentile on the language comprehension measure and had moderate to severe phonological st  impairment. Based on the premise that children with phonological production difficulties (Rvachew et al., 2003) and language comprehension deficits (Warrick & Rubin, 1992) may be at risk for problems with phonological awareness and later literacy skills, the children in the L o w P A group may also be at risk for future literacy and academic problems. Concern for these children increases once they enter higher grades in school when language demands become increasingly more complex. Their low scores on language comprehension suggest that  39  intervention focusing on more general language development remains important for those children. Given the results from the children in the Middle P A group, this study has positive implications for children with autism who may be at the stage o f developing phonological awareness skills and who have adequate language comprehension skills i n order to understand more abstract concepts such as "rhyme." When provided with additional cues, such as picture icons and written words, these children were able to learn the teaching stimuli and one child was able to generalize initial-phoneme identification. Given the results from the secondary training trials and possible emerging P A skills, these children may be prime candidates for phonological awareness intervention; however, more research is needed in this area. Ultimately, this is a preliminary study that has started to bridge the gap in the literature concerning phonological awareness skills in children with autism, showing that these children may be similar to children with language impairment o f other origins with respect to the interaction o f language and phonological awareness skills. This study also shows that some children with A S D are capable o f performing metalinguistic tasks and may be capable o f being taught certain phonological awareness concepts, provided with appropriate teaching stimuli.  Table 1 Participant demographics  Child  Age  #1  5.0  K  M  #2  5.2  K  #3  5.6  #4  Grade  Gender  Diagnosis  Diagnostic Measure  Home Language  Bilingual  PDD-NOS  ADI-R CARS  Spanish  Yes  Bachelor degree  Yes  M  H F Autism  n/a  English  No  H i g h school diploma  Yes  K  M  H F Autism  DSM-IV, CARS  English & Japanese  No  Bachelor degree  Yes  5.8  K  M  Moderate Autism  n/a  Cantonese  Yes  College diploma  Yes  #5  5.9  K  F  ASD  DSM-IV  English  No  Bachelor degree  Yes  #6  6.7  1  M  Autistic disorder D S M - I V  English  No  Bachelor degree  Yes  #7  7.2  2  F  Severe Autism  n/a  Cantonese  Yes  College diploma  #8  7.5  2  F  Moderate Autistic range  DSM-IV, CARS  English  No  Bachelor degree  No  #9  8.4  3  M  H F Autism  DSM-IV  Cantonese  Yes  H i g h school diploma  No  1  Mother's education  EIBI prograr  •  No  Table 1 (Continued) #10  8.6  3  M  H F Autism  n/a  Mandarin  Yes  Two college diplomas  No  #11  9.0  3  F  Mod-Severe Autism  n/a  Mandarin  No  Bachelor degree  Yes  Note. ' P D D - N O S : Pervasive Developmental Disorder - Not Otherwise Specified; A S D : A u t i s m Spectrum Disorder; H F : H i g h Functioning. 2  D S M - I V : Diagnostic and Statistic Manual o f the American Psychiatric Association, 4  th  Edition (American Psychiatric Association,  2000); C A R S : Childhood Autism Rating Scale (Schopler, Reichler, & Renner, 1988); A D I - R : Autism Diagnostic Interview-Revised (Rutter, L e Couteur, & Lord, 2003); n/a: refers to autism severity description obtained via parental report, not a standardized measure. E I B I : Early Intensive Behavioural Intervention program.  Table 2 Procedure: Test types and order  Session 1  Session 2  1. Language sample 2. Language comprehension measure 3. Rhyme Matching I P A task  Session 3  1. Language sample 1  2. C A P E S phonology measure  1. Language sample 2. Rhyme Identification II P A task  2  3. Word Segmentation P A task  Note.' Sentence Structure subtest from the CELF-Preschool 1  3. Word-Final Sound Matching P A task  3  4. Word-Initial Sound Identification P A task  3  3  4. Letter-Sound Identification task  ( W i i g et al., 1992) or the CELF 4 (Semel et al., 2003).  2  Computerized Articulation and Phonology Evaluation System (CAPES, Masterson & Bernhardt, 2001). Order was randomized across children for these particular P A tasks before each testing session.  3  Table 3 Phonological production measures across children  Child  % Sample Imitated  PCM  PVM  WSM  SPM  Voice/Speech Features  #1  19.6  76.2  89.7  89.1  93.5  #2  30.4  74.3  82.4  87.0  97.8  #3  6.5  82.9  75.0  86.9  82.6  #4  21.7  81.0  83.8  89.1  87.0  Breathy; pharyngealization  #5  41.3  48.6  35.3.  43.5  76.1  Glottal fry  #6  41.3  60.0  57.4  60.9  91.3  Glottal fry; fast rate  #7  63.0  21.0  44.1  8.7  69.6  Breathy; low volume  #8  6.5  79.1  75.0  93.5  89.1  Abnormal prosody  30.4  73.3  85.3  84.8  93.5  Cantonese accent; lateral lisp  #10  28.3  90.5  76.5  95.7  .78.3  #11  41.3  70.5  82.4  73.9  97.8  #9  .  Abnormal prosody; breathy  Mandarin accent; nasality Breathy; l o w volume  Note. P C M : percentage consonant match; P V M : percentage vowels match; W S M : percentage word shape match; S P M : percentage stress pattern match.  Table 4 Language comprehension and production measures across participants  Language Comprehension Child  R a w Score  Percentile  1  Language Production # Utterances  MLU  MLU 2  NDW  NTW  4.66  5.06  128  411  #1  18  25  104  #2  14  2  105  3.84  4.17  128  315  #3  22  91  105  3.54  3.77  121  289  #4  11  1  109  3.55  3.60  109  302  #5  10  103  1.34  1.37  71  121  #6  12  1  102  2.37  2.46  57  178  #7  4  1  48  1.31  1.44  n/a  50  #8  18  104  2.87  3.25  106  261  #9  19  101  2.31  2.55  83  203  #10  16  0.1  108  2.66  3.10  103  243  #11  11  0.1  101  2.31  2.82"  103  213  .  .  4^  Table 4 (Continued) Note. M L U : M e a n Length of Utterance; M L U 2: alternative Mean Length o f Utterance measure (Johnston, 2001); N D W : Number o f Different Words in the first 100 utterances; N T W : Number o f Total Words in the first 100 utterances. 'Participants #1 to #8 received the CELF-Preschool  ( W i i g et al., 1992); participants #9 to #11 received the CELF 4 (Semel et al.,  2003). C h i l d # 7 and #8 were administered the CELF-Preschool, for this test.  even though their chronological ages were not within the age range  Table 5 Phonological awareness results across children and P A tasks  Rhyme Matching I  Word Segmentation  L o w P A Group #5  0  0  n/a  #6  0  n/a  #7  0  #11  Initial Sound Identification  Rhyme Identification II  Final Sound Matching  Total P A  Letter-Sound Identification  n/a  n/a  0  8  n/a  n/a  n/a  0  6  n/a  n/a  n/a  n/a  0  3  0  0  n/a  n/a  n/a  0  7  M i d P A Group # 1  10  7  0  0  0.  17  5  #4  10  8  0  7  n/a  25  10  #8  9  9  0  0  0  18  9  H i g h P A Group #2 '  10  9  10  10  7  46  9  #3  10  10  10  10  8  48  10  #9  10  9  8  10  9  46  10  #10  10  9  10  9  10  48  10  1  Table 5 (Continued) Mean  6.27  6.78  5.43  6.57  5.67  22.5  7.91  SD  4.98  3.93  5.13  4.61  4.50  21.8  2.39  11  9  7  7  6  11  11  n  Note. V a : N o t applicable because Child #4 chose not to do this P A task, not because the task was considered too difficult.  Table 6 Correlations between total scores on P A tasks and external variables and language measures  Variables  Spearman rho coefficient  /7-value  External variables A g e and total P A  -0.12  0.73  Mother's education level and total P A  -0.23  0.49  Home language and total P A  -0.02  0.97  Phonology measures 0.83  0.001**  0.39  0.23  0.68  0.021*  0.04  0.92  Language comprehension measure R a w scores and total P A  0.71  0.014*  Language production measures M L U and total P A  0.51  0.11  M L U 2 and total P A  0.57  0.07  Percent consonant match ( P C M ) and total P A Percent vowel match ( P V M ) and total P A W o r d shape match ( W S M ) and total P A Stress pattern match ( S P M ) and total P A  Table 6 (Continued) N D W and total P A (n= 10)  0.44  0.21  N T W and total P A  0.54  0.09  0.84  0.001*  0.76  0.01*  Literacy measures Letter-sound identification scores and total P A Reported reading level (parent questionnaire) Note. *p < 0.05 two-tailed; **p < 0.01 two-tailed.  Table 7 Pre-literacy and language measures as a function o f P A group  Age  Phonology: PCM  Severity  1  Comprehension: Raw Scores  L o w P A Group #5  5.9  48.6  #6  6.7  60.0  Moderate  #7  7.2  21.0  Severe  4/22  #11  9.0  70.5  Severe  M i d P A Group # 1  5.0  76.2  #4  5.8  #8  3  Production: ' MLU2  Letter-Sound  Total P A  10/22  Moderate  1.37  8  0  12/22  Moderate  2.46  6  0  Severe  1.44  3  0 ^  11/26  Moderate  2.82  7  0  Mild  18/22  High  5.06  5  81.0  Normal  11/22  Moderate  3.60  10  25  7.5  79.1  Moderate*  18/22  High  3.25  9  18  H i g h P A Group #2  5.2  74.3  Mild  14/22  Moderate*  4.17  9  46  #3  5.6  82.9  Normal  22/22  High  3.77  10  48 <-  8.4  73.3  Moderate*  19/26  High  2.55  10  46  8.6  90.5  16/26  Moderate*  3.10  10  48  #9 #10  •  Severe  Severity  Mild  17 <r  Table 7 (Continued) Note. P C M : Percentage Consonants Match; M L U 2: alternative Mean Length o f Utterance measure (Johnston, 2001); Letter-Sound: raw scores on Letter-Sound Identification task; Total P A : sum o f raw scores across all five phonological awareness tasks; Left arrow (<-): depicts children with " l o w " P A skills has also having " l o w " ability in phonological production and language comprehension, children with "middle" P A skills as having "average" ability, and children with "high" P A skills as having "high" ability in phonology and language comprehension; Asterisk (*): depicts certain "anomalies" in the data. 'Severity o f phonological impairment based on Austin and Shriberg's (1997) reference data. R a w scores out o f a possible 22 correct for the CELF-Preschool  ( W i i g et al., 1992); out o f a possible 26 correct for the CELF 4  (Semel et al., 2003). 3  Severity o f language comprehension impairment determined by distribution o f percentages from the raw scores (i.e., " H i g h " = above  73%; "Moderate" = 40-65%; "Severe" = below 30%). Note that due to problems with standardization, these descriptive labels were arbitrarily assigned based on the raw scores.  Table 8 Correlations between rhyme matching I task and language measures  R a w scores on Rhyme Matching I versus:  Spearman rho coefficient  /?-value  Percent consonant match ( P C M )  0.79  0.004**  Percent vowel match ( P V M )  0.70  0.017*  Word shape match ( W S M )  0.72  0.013*  Language comprehension raw scores  0.68  0.022*  MLU  0.73  0.011*  MLU 2  0.76  0.007**  N D W (n=10)  0.67  0.035*  NTW  0.76  0.007**  Letter-sound identification task  0.67  0.024*  Note. *p < 0.05 two-tailed; **p < 0.01 two-tailed. A l l participants were administered the Rhyme Matching I task.  Table 9 Correlations among language measures  Variables  Spearman rho coefficient  /j-value  Language comprehension versus: MLU 2  0.57  0.067  PCM  0.64  0.034*  Percentage consonants match ( P C M ) versus: MLU  0.69  0.019*  MLU 2  0.71  0.015*  NTW  0.71  0.015*  Letter-sound identification task  0.72  0.012*  N D W ( n = 10)  0.981  0.001**  NTW  0.982  0.001**  M L U 2 versus:  Note. *p < 0.05 two-tailed; **p < 0.01 two-tailed.  Table 10 Regression analysis for variables predicting performance on phonological awareness tasks  Model  1  Variables Entered  1  1) Lang. comp. raw scores  r  2  Change in r  2  F-value  Significance  0.48  n/a  8.31  0.018*  2  1) Lang. comp. raw scores 2) P C M  0.52  0.04  4.40  0.051  3  1) Lang comp. raw scores 2) P C M 3) M L U 2  0.52  0.00  2.57  0.137  Note. *Significance =p < 0.05. ' P C M : percentage consonant match; M L U 2: alternative M e a n Length o f Utterance measure (Johnston, 2001).  55  References A l l e n , D . A . and Rapin, I. (1992). Autistic children are also dysphasic. In H . Naruse and E . Ornitz (eds.) Neurobiology  of infantile autism (pp. 73-80). Amsterdam: Excerpta Medica.  American Psychiatric Association. (2000). Diagnostic and statistical manual of mental disorders, DSM-IV(4  ed.). Washington, D C : Author.  th  Austin, D . and Shriberg, L . D . (1997). Lifespan reference data for ten measures of articulation competence using the speech disorders classification system (SDCS). Phonology Project, Waisman Center on Mental Retardation and Human Development. University o f Wisconsin-Madison. Baltaxe, C . A . M . (1977). Pragmatic deficits in the language o f autistic adolescents. Journal of Pediatric Psychology, 2, 176-180. Bartak, L . , Rutter, M . , and C o x , A . (1977). A comparative study o f infantile autism and specific developmental receptive language disorders: III. Discriminant function analysis. Journal of Autism and Childhood Schizophrenia,  7, 383-396.  Bartolucci, G . & Pierce, S. (1977). A preliminary comparison o f phonological development in autistic, normal, and mentally retarded subjects. The British Journal of Disorders of Communication,  12, 137-147.  Bird, J., Bishop, D . V . M . , and Freeman, N . H . (1995). Phonological awareness and literacy development in children with expressive phonological impairments. Journal of Speech and Hearing Research, 38, 446-462. Brown, R. (1973). A first language: The early stages. Cambridge, M A : Harvard University Press.  56  Bruinsma, Y . , Koegel, R . L . , and Koegel, L . K . (2004). Joint attention and children with autism: A review o f the literature. Mental Retardation and Developmental  Disabilities  Research  Reviews, 10, 169-175. Cataldo, S. and Ellis, N . (1988). Interactions in the development o f spelling, reading and phonological skills. Journal of Research in Reading, 11, 86-109. Catts, H . W . (1993). The relationship between speech-language impairments and reading disabilities. Journal of Speech and Hearing Research, 36, 948-958. Catts, H . W . , Fey, M . E . , Zhang, X . , and Tomblin, J . B . (1999). Language basis o f reading and reading disabilities: Evidence from a longitudinal investigation. Scientific Studies of Reading, 5,331-361. Catts, H . W . , Fey, M . E . , Zhang, X . , and Tomblin, J . B . (2001). Estimating the risk o f future reading difficulties in kindergarten children: A research-based model and its clinical implementation. Language, Speech, and Hearing Services in Schools, 32, 38-50. Charman, T., Drew. A . , Baird, C , and Baird, G . (2003). Measuring early language development in pre-school children with autism spectrum disorder using the MacArthur Communicative Development Inventory (Infant Form). Journal of Child Language, 30, 213-236. Condouris, K . , Meyer, E . , and Tager-Flusberg, H . (2003). The relationship between standardized measures o f language and measures o f spontaneous speech in children with autism. American Journal of Speech-Language  Pathology, 12, 349-358.  Dettmer, S., Simpson, R . L . , Myles, B . , and Ganz, J. (2000). The use o f visual supports to facilitate transitions o f children with autism. Focus on Autism and Other Disabilities,  15, 163-176.  Developmental  57  Dodd, B . , Crosbie, S., Mcintosh, B . , Teitzel., T., and Ozanne, A . (2003). inventory of phonological  Pre-reading  awareness. San Antonio, T X : The Psychological Corporation.  Gibbon, F . E . , M c C a n n , J., Peppe, S., O'Hare, A . , and Rutherford, M . (2004). Articulation disorders in children with high functioning autism. Congress Proceedings Association  of Logopedics  and Phoniatrics.  of The  International  Brisbane, Australia.  G i l l o n , G . (2004). Phonological awareness assessment. In G . G i l l o n (ed.) Phonological awareness: From research to practice (pp. 109-132). N e w York, N Y : Guilford Publications. Goffman, L . and Leonard, J. (2000). Growth o f language skills in preschool children with specific language impairment: Implications for assessment and intervention. American  Journal  of Speech-Language Pathology, 9, 151-161. Goldman, R . and Fristoe, M . (1986). Goldman-Fristoe  Test of  Articulation-Revised.  Circle Pines, M N : American Guidance Service. Goldman, R . and Fristoe, M . (2000). Goldman-Fristoe  2 Test of Articulation.  Circle  Pines, M N : American Guidance Service. Heimann, M . , Nelson, K . E . , Tjus, T., and Gillberg, C. (1995). Increasing reading and communication skills in children with autism through an interactive multimedia computer program. Journal of Autism and Developmental Disorders, 25, 459-480. Hewitt, L . E . , Hammer, C.S., Yont, K . M . , and Tomblin, J.B. (2005). Language sampling for kindergarten children with and without S L I : M e a n length o f utterance, I P S Y N , and N D W . Journal of Communication  Disorders, 38, 197-213.  Hinerman, P. and Channell, R. (1986). Conversational abilities o f autistic individuals. Australian Journal of Human Communication  Disorders,  14, 103-108.  58  Jarrold, C , Boucher, J., and Russell, J. (1997). Language profiles in children with autism: Theoretical and methodological implications. Autism, 1, 57-76. Johnston, J. (2001). A n alternate M L U calculation: Magnitude and variability o f effects. Journal ofSpeech, Language, and Hearing Research, 44, 156-164. Johnston, J. and A m m o n , M . (1985). The Chart: Order of acquisition for selected linguistic forms. Unpublished manuscript. K a m h i , A . G . , Lee, R . F . , and Nelson, L . K . (1985). Word, syllable and sound awareness in language disordered children. Journal of Speech and Hearing Disorders, 50, 207-212. Kjelgaard, M . M . and Tager-Flusberg, H . (2001). A n investigation o f language impairment in autism: Implications for genetic subgroups. Language and Cognitive  Processes,  16, 287-308. Larrivee, L . and Catts, H . (1999). Early reading achievement in children with expressive phonological disorders. American Journal of Speech-Language  Pathology, 8, 118-128.  Leadholm, B . and M i l l e r , J. (1992). Language sample analysis: The Wisconsin guide. Madison, W I : Wisconsin Department o f Public Instruction. Liberman, I. and Liberman, A . (1990). Whole language vs. code emphasis: Underlying assumptions and their implications for reading instruction. Annals of Dyslexia, 40, 51-76. Liberman, I.Y., Shankweiler, D . , Fischer, F . W . , and Carter, B . (1974). Explicit syllable and phoneme segmentation in the young child. Journal of Experimental Child Psychology, 201-212. Lord, C . (1989). Language comprehension in autism. Nimhans Journal, 7, 169-172.  18,  59  Lord, C . and Paul, R. (1997). Language and communication in autism. In D . J . Cohen and F.R. Volkmar (eds.) Handbook of autism and pervasive developmental disorders (pp. 195-225). N e w York, N Y : W i l e y . . Loveland, K . A . , Landry, S.H., Hughes, S.O., H a l l , S.K., and M c E v o y , R . E . (1988). Speech acts and the pragmatic deficits o f autism. Journal of Speech and Hearing Research,  31,  593-604. . Major, E . M . and Bernhardt, B . H . (1998). Metaphonological skills o f children with phonological disorders before and after phonological and metaphonological intervention. International Journal of Language & Communication  Disorders, 33, 413-444.  Masterson, J. and Bernhardt, B . (2001). Computerized Articulation  and  Phonology  Evaluation System (CAPES). Austin, T X : Pro-Ed. Mayer-Johnson C o . (1994). Boardmaker®  for Windows. (Version 1.6). Solano Beach,  C A : Author. Metsala, J.L. (1999). Young children's phonological awareness and nonword repetition as a function o f vocabulary development. Journal of Educational Psychology, 91, 3-19. Miller, J.F. (2003). Systematic Analysis of Language Transcripts (SALT). W i n d o w s ® Student V 8 . Madison, W I : Language Analysis Laboratory, University o f Wisconsin. M i l l e r , J.F., Freiberg, C , Rolland, M . , and Reeves, M . (1992). Implementing computerized language sample analysis in the public school. Topics in Language Disorders,  12,  69-82. Muter, V . and Snowling, M . (1998). Concurrent and longitudinal predictors o f reading: The role o f metalinguistic and short-term memory skills. Reading Research Quarterly, 33, 320337.  60  National Institute o f C h i l d Health and Human Development. (2000). Report of the National Reading Panel. Teaching children to read: An evidence-based assessment of the scientific research literature on reading and its implications for reading instruction. Retrieved March 15, 2006 from http://www.nichd.nih.gov/publications/nrp/smallbook.htm. Norman, G . R . and Streiner, D . L . (1998). Biostatistics:  The bare essentials. Hamilton:  B . C . Decker Inc. O'Connor, R . E . and Jenkins, J.R. (1999). Prediction o f reading disabilities in kindergarten and first grade. Scientific Studies of Reading, 3, 159-197. Puolakanaho, A . , Poikkeus, A . , Ahonen, T., Tolvanen, A . , and Lyytinen, H . (2003). Assessment of three-and-a-half-year-old children's emerging phonological awareness in a computer animation context. Journal of Learning Disabilities, 36, 416-423. Rapin, I. and Dunn, M . (2003). Update on the language disorders o f individuals on the autistic spectrum. Brain & Development, 25, 166-172. Reynell, J . K . (1985). Reynell Developmental  Language Scales, 2  nd  edition. Windsor:  N F E R Nelson. Roberts, J.A., Rice, M . L . , and Tager-Flusberg, H . (2004). Tense marking in children with autism. Applied Psycholinguistics,  25, 429-448.  Rutter, M . , Le Couteur, A . , and Lord, C . (2003). Autism Diagnostic  Interview-Revised  ADI-R. Los Angeles, C A : Western Psychological Services. Rvachew, S., Nowak, M . , & Cloutier, G . (2004). Effect o f phonemic perception training on the speech production and phonological awareness skills o f children with expressive phonological delay. American Journal of Speech-Language  Pathology, 13, 250-263.  61  Rvachew, S., Ohberg A . , Grawburg., and Heyding, J. (2003). Phonological awareness and phonemic perception in 4-year-old children with delayed expressive phonology skills. American Journal of Speech-Language  Pathology, 12, 463-471.  Schopler, E . , Reichler, R.J., and Renner, B . R . (1988). The Childhood Autism Rating Scale (CARS). Los Angeles, C A : Western Psychological Services. Semel, E . , W i i g , E . , and Secord, W . (2003). Clinical Evaluation of Language Fundamentals®  - Fourth Edition. San Antonio, T X : The Psychological Corporation, Ptarcourt  Brace & C o . Sutherland, D . and Gillon, G . T . (2005). Assessment o f phonological representations in children with speech impairment. Language, Speech, and Hearing Services in Schools, 36, 294307. Swan, D . and Goswami, U . (1997). Phonological awareness deficits in developmental dyslexia and the phonological representations hypothesis. Journal of Experimental Psychology,  Child  66, 18-41.  Tager-Flusberg, H. (1981a). O n the nature o f linguistic functioning in early infantile v  autism. Journal of Autism and Developmental  Disorders, 11, 45-56.  Tager-Flusberg, H . (1981b). Sentence comprehension in autistic children. Applied Psycholinguistics,  2, 5-24.  Tager-Flusberg, H . (2000). The challenge o f studying language development in children with autism. In L . M e n n and N . Bernstein Ratner (eds.) Methods for studying production  (pp. 313-332). Mahwah, N J : Lawrence Erlbaum Associates Inc.  language  62  Tager-Flusberg, H . (2001). Understanding the language and communicative impairments in autism. In L . M . Glidden (ed.), International review of research in mental retardation:  Autism  (pp. 185-205). San Diego, C A : Academic Press. Tager-Flusberg, H . , Calkins, S., N o l i n , T., Baumberger, T., Anderson, M . , and Chadwick-Dias, A . (1990). A longitudinal study o f language acquisition in autistic and D o w n syndrome children. Journal of Autism and Developmental Disorders, 20, 1-21. Tiscott, C . and Evans, R. (2003). Visual teaching strategies for children with autism. Early Childhood Development, 173, 425-433. Tjus, T., Heimann, M . , and Nelson, K . E . (1998). Gains in literacy through the use o f a specially developed multimedia computer strategy: Positive findings from 13 children with autism. Autism, 2, 139-156. Torgesen, J.K., Wagner, R . K . , and Rashotte, C A . (1994). Longitudinal studies o f phonological processing and reading. Journal of Learning Disabilities, 27, 276-286. Torneus, M . , Taube, K . , and Lundberg, I. (1984). Fonologisk medvetenhet: Handledning for kartlaggning och utveckling (in Swedish). Stockholm: Psykologiforlaget. Treiman, R. and Zukowski, A . (1991). Levels o f phonological awareness. In S. Brady and D . Shankweiler (eds.), Phonological processes in literacy (pp. 67-84). Hillsdale, N J : Erlbaum. • Tunmer, W . E . (1991). Phonological awarenessand literacy acquisition. In L . Rieben and C A . Perfetti (eds.). Learning to read: Basic research and its implications (pp. 105-119). Hillsdale, N J : Erlbaum. Vandervelden, M . and Siegel, L . (1999). Phonological processing and literacy in A A C users and students with motor speech impairments. Augmentative Communication,  15, 191-211.  and Alternative  63  Wagner, R., Torgeson, J., and Rashotte, C . (1999). Comprehensive  Test of  Phonological  Processing. Austin, T X : Pro-Ed. Warrick, N . and Rubin, H . (1992). Phonological awareness: Normally developing and language delayed children. Journal of Speech-Language  Pathology and Audiology, 16, 11-20.  Watkins, R., K e l l y , D . , Harbers, H . , and Hollis, W . (1995). Measuring children's' lexical diversity: Differentiating typical and impaired language learners. Journal of Speech and Hearing Research, 38, 476-489. Webster, P.E. and Plante, A . S . (1992). Effects o f phonological impairment on word, syllable, and phoneme segmentation and reading. Language, Speech, and Hearing Services in Schools, 23, 176-182. Webster, P.E., Plante, A . S . , and Couvillion, L . M . (1997). Phonological impairment and prereading: Update on a longitudinal study. Journal of Learning Disabilities, 30, 365-375. W i i g , E . H . , Secord, W . , and Semel, E . (1992). Clinical Evaluation of Language Fundamentals®  - Preschool. San Antonio, T X : The Psychological Corporation, Harcourt Brace  & Co. Wolk, L . and Edwards, M . L . (1993). The emerging phonological system o f an autistic child. Journal of Communication  Disorders, 26, 161-177.  64  Appendix A Test Trials and Secondary Training Trials Stimuli Lists 1. Rhyme Judgment I Task Test Trials  Secondary Training Trials  1. snake: corn, cake  Most Support: Boardmaker picture icons & color-coded text  2. clock: cat, rock  1. dig: pig, stove  3. car: star, fan  2. mat: bear, rat  4. bat: hose, hat  3. tail: shirt, pail  5. king: ring, sun  4. sun:fun, shoe  6. dog: log, feet  5. lick: light, stick  7. rose: cup, nose  Intermediate Support: Picture icons; text without color-coding  8. mouse: house, chair  6. coat: orange, goat  9. box: fox, bed  7. rug: stop, mug  10. van: pear, man  8. bag: flag, glove Least Support: No picture icons, text without color-coding 9. cook: book, brush 10. sand: hand, bird  2. Word Segmentation Task Test Trials  Secondary Training Trials  1. hotdog: mouse, hot, dog  Most Support  2. blackboard: board, black, school  1. pancake: pan, cake, bowl  3. cowboy: boy, cow, sheep  2. ice cream: cookie, cream, ice  4. toothbrush: tooth, brush, comb  3. playground: ground, rope, play  65  5. haircut: comb, cut, hair  4. sunset: set, sun, clock  6. doughnut: muffin, nut, dough  5. beanbag: bag, carrot, bean  7. duck pond: duck, pond, fish  Intermediate Support  8. junk food: food, fork, junk  6. headlight: head, tire, light  9. baseball: run, ball, base  7. bedroom: room, chair, bed  10. baby: crib, bay, bee  8. horseshoe: horse, cow, shoe Least Support 9. popcorn: pop, corn, salt 10. firetruck: truck, fire, rain  3. Word-Initial Sound Identification Task Test Trials  Secondary Training Trials  1. Idl: duck, door, dog, cake  Most Support  2. /t/: table, balloon, turtle, toothbrush  1. /b/: book, bird, ball, leaf  3. Ibl: baby, button, hammer, balloon  2. /h/: head, house, bird, hat  4. HI: fish, feet, fork, boat  3. Iml: map, mask, pot, moon  5. /k/: cage, cup, sun, cow  4. Ifl: fox, fall, cat, fish  6. /s/: hat, seat, sock, sun  5. Ill: man, light, leg, leaf  7. /p/: pig, bike, pin, pea  Intermediate Support  8. /k/: feet, key, kite, corn  6. lyl: juice, coat, jump, jam  9. /k/: kangaroo, telephone, computer, camera  7. Igl: goat, boot, game, gum  10. /b/: bone, bed, cup, bike  8. Ill: lemon, leopard, laundry, popcorn Least Support 9. Ik/: table, castle, cupcake, canoe  66  10./p/: pencil, picture, towel, penguin 4. Pvhyme Identification II Task Test Trials  Secondary Training Trials  1. corn, snake, cake, rake  Most Support  2. clock, rock, sock, cat  1. pink, sink, wink, pot  3. car, jar, fan, star  2. pie, tie, toe, lie  4. hose, cat, hat, bat  3. van, dog, man, can  5. swing, ring, king, sun  4. goat, coat, fish, boat  6. feet, frog, log, dog  5. shoe, tail, pail, mail  7. rose, nose, clock, hose  Intermediate Support  8.  bed, shoe, head, bread  6. hand, ship, lip, zip  9.  key, bow, pea, tree  7. bug, mug, rug, house  10. bear, chair, hair, kite  8. pan, ape, tape, grape Least Support 9. rat, fat, gum, bat 10. tree, bee, see, pail  5. Word-Final Sound Matching Task Test Trials  Secondary Training Trials  1. IM bowl: doll, hat, cup  Most Support  2. /p/ cap: car, lip, fan  1. /ml gum: goat, bear, mom  3. /g/ dog: pear, pot, pjg_  2. Igl leg: leaf, rug, pen  4. Ill cat: sit, comb, back  3. Ik/ book: cat, rain, duck  5. Ibl bib: rat, rain, tub  4. Idl bed: dad, mouse, sun  6. Inl pin: pot, bone, crib  5. /!/ seal: pan, mail, nose  7. / d / l i d : bed, light, pig  Intermediate Support  8. /k/ sack: bat, kick, soap  6. /g/ bag: peach, bug, chair  9. /s/ face: cake, fish, mice  7. /p/ cup: lip, star, bee  10./f/leaf: line, knife, teeth  8. /JVfish: bear, wash, pear Least Support 9. /n/ van: pen, coat, head 10. Ixl star: bed, door, rug  Letter-Sound Identification Task Test Trials 1.  b:/b/  2. 1:/1/ 3. a:/ae/ 4.  sh:/J7  5.  z:/z/  6. i : / l /  7. th:/0/ 8.  o:/a/  9. d:/dV 10. k : / k / Note. Underlined words refer to correct responses.  68  Appendix B Parent Literacy Questionnaire  1. Does your child like to read or look at books? 2. Does your child like to read with adults or siblings, or does he/she prefer to read alone?  3. Has your child been exposed to phonological awareness tasks (e.g., rhyming, beginning sounds o f words, final sounds o f words, tapping out syllables o f words, etc.) before this research project? 4. Have you worked on phonological awareness tasks with your child? 5. Have others, such as behavioral interventionists or a Speech-Language Pathologist, worked on phonological awareness with your child?  .__  6. About how often are phonological awareness tasks worked on? 7. Can your child identify letters and numbers (e.g., "Where's the letter B")? 8. Does your child sound out words when he/she reads (i.e., does your child decode the sounds o f the words)?  '  9. What method o f teaching was used when your child was learning to read, either at school or at home? (Circle all that apply).  .  1. Phonics (i.e., teaching individual letter sounds) 2. Whole word (e.g., teaching the entire word and/or core vocabulary) 10. Does your child understand the majority o f what he/she reads?  _  11. What grade is your child in?. 12. What grade level is your child currently reading at?  ,  

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
http://iiif.library.ubc.ca/presentation/dsp.831.1-0092723/manifest

Comment

Related Items