Grossman et al. (2010)

December 15, 2016

 

ANALYSIS

Comparison Research  

NOTES:

  • A summary of an intervention is not included in the review because this in not an investigation about intervention. 

KEY: 

ASD = autism spectrum disorders

eta =   partial eta squared

HFA = high-functioning autism

MLU = mean length of utterance

NA = Not Applicable

P = participant or patient

pmh = Patricia Hargrove, blog developer

SLP = speech-language pathologist

TD = typically developing

  

SOURCE: Grossman, R. B., Bemis, R. H., Skwerer, D. P., & Tager-Flusberg, H. (2010.) Lexical and affective prosody in children with high-functioning autism. Journal of Speech, Language, and Hearing Research, 53, 778- 793.

 

REVIEWER(S): pmh

 

DATE: December 7, 2016

 

ASSIGNED GRADE FOR OVERALL QUALITY:  Not graded, this is an investigation of the nature of prosody within a clinical population; it is not an intervention study.

 

TAKE AWAY: This investigation of lexical and affective prosody in children and adolescents with high functioning autism (HFA) comprises 3 experiments: perception of affective prosody, perception of lexical prosody, and production of lexical prosody. The results indicate that children and adolescents with HFA are similar to TD peers with respect to their perception of affective prosody and lexical stress as well in their ability to mark correctly the stressed syllable in compound words (e.g., greenhouse) and noun phrases (green house.) However, the Ps with HFA differed from their TD peers with respect to how they marked the stressed syllable. Specially, the productions of Ps with HFA were of longer duration which was perceived be slower, more labored, containing longer pauses than TD peers.

 

 

  1. What type of evidence was identified?

                                                                                                           

  • What was the type of design? Comparison; Prospective, Nonrandomized Group Design with Controls

 

  • What was the focus of the research? Clinically Related

           

  • What was the level of support associated with the type of evidence? B+

 

                                                                                                           

  1. Group membership determination:

                                                                                                           

  • If there were groups, were participants randomly assigned to groups? Yes
  • If there were groups and Ps were not randomly assigned to groups, were members of groups carefully matched? Yes

                                                                    

 

  1. Were experimental conditions concealed?

                                                                                                           

  • from participants? No

                                                                    

  • from administrators of experimental conditions? No

                                                                    

  • from analyzers/judges? Unclear

                                                                    

 

  1. Were the groups adequately described? Yes

 

–    How many participants (Ps) were involved in the study?

 

NOTE: The same Ps were involved in all 3 experiments

  • total # of Ps: 31
  • # of groups: 2
  • List names of groups and the number of Ps in each group:

– High-Functioning Autism (HFA) = 16 (except Experiment #3, N = 11 for pitch and intensity analyses

– Typically Developing (TD) = 15 (except Experiment #3, N = 9 for pitch and intensity analyses

 

  • Did all groups maintain membership throughout the investigation? Yes, but data from some Ps were discarded from the pitch and intensity analysis for Experiment #3 due to excessive static on the recordings. This resulted in the following Ns for the 2 groups”

– TD = 9

– HFA = 11

                                                                                

–   CONTROLLED CHARACTERISTICS

                                                                                                                       

  • cognitive skills:

     – HFA = within normal range

     – TD = within normal range

 

  • receptive language:  (performance on the Peabody Picture Vocabulary Test)

     – HFA = within normal range

     – TD = within normal range

 

  • presence of identified genetic disorders: excluded from experiments

 

  • diagnosis of HFA =

     – met criteria in the Diagnostic and Statistical Manual of Mental Disorders; Autism Diagnostic Interview—Revised; Autism Diagnostic Observation Schedule

     – diagnosis noted that the P was in classified as “full range” autism

     – verbal IQ and receive vocabulary were within 2 standard deviations (SD) of the mean but the following challenges still resulted in diagnosis of ASD:

  • social skills
  • communication skills
  • pragmatic skills

     – excluded the following diagnoses:

  • Asperger syndrome
  • Autism, Pervasive Developmental Disorder, not otherwise specified

 

–   DESCRIBED CHARACTERISTICS (* = no significant differences between TD and HFA groups)

 

  • age*:

     HFA = mean – 12;4, range 7;6-17

     TD = mean 12;7, range 7;-18

 

  • gender: Not reported
  • cognitive skills:

HFA =

  • Full range IQ = 106.7 ( mean); 87-123 (range)
  • Verbal IQ* = 101.2 (mean); 83-127 (range)
  • Nonverbal IQ* = 109.6 (mean); 94-127 (range)

     TD =

  • Full range IQ = 108.9 (mean); 87-123 (range)
  • Verbal IQ* = 108.1 (mean); 81-127 (range)
  • Nonverbal IQ* = 106.7 (mean); range 85-116 (range)

 

  • receptive language*: (performance on the Peabody Picture Vocabulary Test)

     HFA = 107.0 (mean); 79-138 (range)

     TD =  111.3 (mean); 79-139 (range)

  • Hearing: Not reported

 

– Were the groups similar? Yes

                                                         

– Were the communication problems adequately described? Unclear.

 

  • disorder type: Ps in the HFA group were described as having social, pragmatic, and communication problems but they were not described. In addition, Ps with HFA were described as “did not at this time exhibit measureable language impairment” (p. 783’)

 

 

NOTE: This study investigated 3 experiments that used the same Ps described in item #4. Items #5 through #13 are presented separately for each of the experiments.

 

EXPERIMENT 1:

Investigation of the perception of affective prosody

 

 

  1. What were the different conditions for this research?

                                                                                                             

  • Subject (Classification) Groups? Yes

HFA

     – TD          

 

  • Experimental Conditions? Yes

 

CONTENT OF STIMULI

– Unfiltered Condition: 6 Sentences produced for each of 3 forms of affective prosody (happy, sad, neutral.) In each case, the content of the sentence and the prosody of the sentence matched the targeted content. For example, happy affective prosody was paired with the following sentence: “When Mike pets the puppy, it’s wagging its tail” (p.792)

– Filtered Condition: The 18 unfiltered sentences were passed through a low-pass filter deleting speech frequencies that resulted in the removal of semantic content but the retention of prosody.

 

EMOTION REPRESENTED ON STIMULI: (p. 781)

– Sad = lower pitch, slower rate, low tone ending

Happy = higher pitch, faster rate, complex tone ending

Neutral = midrange pitch, accent on the main topic, less complex final tone

 

  • Criterion/Descriptive Conditions? No

 

 

  1. Were the groups controlled acceptably? Yes

 

 

  1. Were dependent measures appropriate and meaningful? Yes

 

– OUTCOMES:

 

  • OUTCOME #1: Percentage of correct identifications of depicted affects in unfiltered and filtered sentence stimuli.

 

–   The outcome was subjective.

 

The outcome measures was not objective.

 

 

  1. Were reliability measures provided?

                                                                                                            

  • Interobserver for analyzers? No

 

  • Intraobserver for analyzers? No

 

  • Treatment or test administration fidelity for investigators?

 

 

  1. Description of design:
  • Following a training session that involved only unfiltered sentences, Ps listened to sentences and classified the emotions/affect as happy, sad, or neutral.
  • The validity of prerecorded sentence stimuli wasa ascertained by the independent rating of affect by 10 listeners who were not Ps in the investigation.
  • There were 18 sentences each for the unfiltered and filtered sentences. There were 3 versions of each of the sentence types which difference with respect to the order of sentence. The order of sentences was pseudorandomized and counterbalanced.
  • The verbal stimuli are included in an appendix.
  • Sentences were presented on acomputer via speakers. The filtered sentences were administered first; then the unfiltered sentences were administered.
  • Ps signified their choices by selecting the appropriated button.

 

 

  1. What were the results of the statistical (inferential) testing?

 

– Results:

 

  • OUTCOME #1: Percentage of correct identifications of depicted affects in unfiltered and filtered sentence stimuli.
  • performance above chance for
  • filtered task: both TD and HFA
  • unfiltered task: both TD and HFA
  • performance was better on unfiltered than filtered for both TD and HFA
  • performance on neutral sentences was significantly poorer for neutral sentences for both TD and HFA
  • there was not an overall significant difference between TD and HFA groups

 

– The statistical tests used to determine significance included t-test; ANOVA: repeated measure 2 (Group—TD, HFA) x 2 (Task—filtered, unfiltered) x 3 (Emotion—happy, sad, neutral); repeated measure 2 (Group) x 3 (Emotion) for each task

 

– Were effect sizes provided? Yes

 

  • OUTCOME #1: Percentage of correct identifications of depicted affects in unfiltered and filtered sentence stimuli.
  • performance was better on unfiltered than filtered for both TD and HFA;
  • HFA – Cohen’s d = 1.04 (large effect)
  • TD – Cohen’s d = 0.76 (moderate effect)

 

– Were confidence interval (CI) provided? No

 

 

  1. Summary of correlational results: Not Applicable (NA)

 

 

  1. Summary of descriptive results: Qualitative research NA

 

 

  1. Brief summary of clinically relevant results:

 

  • TD and HFA Ps exhibited similar skills in interpreting affective prosody.
  • Both groups had significantly more difficulty on the filtered task than the unfiltered task. However, they still were able to label emotions above chance in the filtered condition. Therefore, both groups make use of semantic information to interpret affect (in addition to prosody.)
  • Both groups found the neutral affect sentences to be more difficult to interpret.

 

 

ASSIGNED GRADE FOR QUALITY OF EXTERNAL EVIDENCE: B

 

==========================================================

 

EXPERIMENT 2

Investigate the perception of lexical stress for differentiating noun phrases from compound nouns

 

  1. What were the different conditions for this research?

                                                                                                             

  • Subject (Classification) Groups? Yes

– HFA

     – TD

                                                               

  • Experimental Conditions? Yes

     – Noun Phrases (e.g. green house, wet suit, p. 792)

     – Compound Words (e.g., greenhouse, wetsuit, )

     – Foils (e.g., tree house, headphones, p. 792)

 

  • Criterion/Descriptive Conditions? No

 

 

  1. Were the groups controlled acceptably? Yes

 

 

  1. Were dependent measures appropriate and meaningful? Yes

 

–OUTCOMES:

  • OUTCOME #1: Percentage accuracy of responses to the experimental task

 

– The outcome measure was subjective.

 

 The outcome measure was not objective.

 

 

  1. Were reliability measures provided?

                                                                                                            

  • Interobserver for analyzers? No

 

  • Intraobserver for analyzers? No _

 

  • Treatment or test administration fidelity for investigators? No

 

  • Validity of the stimuli was supported by

AUDITORY STIMULI:

  • The investigators analyzed the experimental auditory stimuli using Praat to insure that the first syllable was stressed in the compound word (e.g., HOTdog) and the second syllable was stressed in the noun phrase (e.g., hot DOG.)
  • The acoustic criteria the investigators used to identify stressing were

         ¶ longer duration

         ¶ higher pitch

         ¶ greater intensity

  • For the foil items (i.e., stimuli produced with equal stress), acoustic analysis of the 3 above features did not reveal differences between the first and second syllable.
  • Ten judges listened to the auditory stimuli to determine which words could be properly identified as noun phrases or compound words.

 

VISUAL STIMULI:

  • The investigators insured that judges could properly link the auditory stimuli with pictures of the targeted words.

 

 

  1. Description of design:
  • The experimental stimuli included auditory and visual stimuli.

– Every experimental word (compound word or noun phrase) was paired with a picture presenting the target word and the opposite stress pattern. For example, for the target “wetsuit,” the investigator presented a picture of a “wet suit” and a picture of a suit that was wet.

–  Eleven foil words (i.e., compound words without corresponding noun phrases) were also presented auditorily and visually throughout the testing. For example, ne picture presented the target word (e.g., rainbow) and one picture presented the picture of a hair bow.

– The verbal stimuli are included in an appendix.

 

  • Prior to the experimental task, investigators administered a brief training task insuring that the Ps were able to press a button signifying their picture choice. Investigators provided corrective feedback to the Ps.

 

  • The Ps listened to prerecorded auditory stimuli of the compound words or noun phrases in isolation (i.e., not in a contextualizing sentence.)

 

 

  1. What were the results of the statistical (inferential) testing?

 

– RESULTS:

 

  • OUTCOME #1: Percentage accuracy of responses to the experimental task
  • All Ps from both groups (TD, HFA) performed at the 80% accuracy level on foil stimuli.
  • All Ps from both groups (TD, HFA) performed at the 80% accuracy level for compound word and noun phrase stimuli.
  • Both groups were significantly more accurate on compound word/first syllable stress targets than noun phrase/second syllable stress target.
  • There was no significant differenceS between TD and HFA groups.

 

– The statistical tests used to determine significance included t-test; ANOVA: 2 (Groups: TD, HFA) x 2 (Stress: compound word/first syllable targets, noun phrase/ second syllable targets)

 

–   Were effect sizes provided? Yes

  • OUTCOME #1: Percentage accuracy of responses to the experimental task
  •  Both groups (TD, HFA) were significantly more accurate on compound word/first syllable stress targets than noun phrase/second syllable stress target.

∞ TD Cohen’s d = 0.80 (moderate effect)

∞ HFA Cohen’s d = 1.2 (large effect)

 

 

  1. Summary of correlational results: NA

 

 

  1. Summary of descriptive results: Qualitative research NA

 

 

  1. Brief summary of clinically relevant results:

 

  • HFA and TD groups performed similarly on the task requiring Ps to demonstrate comprehension of compound words (e.g., Yellowstone) compared to noun phrases (yellow stone.) Thus, Ps with HFA and their TD peers are equally competent on this task.

 

 

ASSIGNED GRADE FOR QUALITY OF EXTERNAL EVIDENCE:   B

 

==========================================================

 

EXPERIMENT 3

Investigate the production of lexical prosody

 

 

  1. What were the different conditions for this research?

                                                                                                             

  • Subject (Classification) Groups? Yes

HFA

     – TD

                                                               

  • Experimental Conditions? Yes

 

    – Lexical type (compound words/first syllable stress; noun phrases/second syllable stress)

 

  • Criterion/Descriptive Conditions? Yes

 

 

  1. Were the groups controlled acceptably? Yes

 

 

  1. Were dependent measures appropriate and meaningful? Yes

                                                                                                             

–  Outcome measures:

 

  • OUTCOME #1: Mean pitch of productions

 

  • OUTCOME #2: Mean intensity of productions

 

  • OUTCOME #3: Mean whole word duration

 

– The Outcome measures are not subjective.

 

All of the outcome measures are objective.

 

 

  1. Were reliability measures provided?

                                                                                                            

– Interobserver for analyzers? Yes

 

  • OUTCOME #3: Mean whole word duration—The investigators calculated the mean difference in overall duration length for the original analyzer and a second judge using 15% of the samples. The difference was 4 ms which met the investigators 5 ms or less criterion.

 

– Intraobserver for analyzers? No

 

– Treatment or test administration fidelity for investigators? No

 

 

  1. Description of design:

 

  • Experiment 3 was administered after a break following the administration of Experiments 1 and 2.
  • The investigators used the same picture stimuli as Experiment #2 with Experiment #2 always preceding Experiment #3.
  • In this investigation, there were 27 targets: 11 compound word/first syllable stress, 11 noun phrases/second syllable stress, 5 foils.
  • Scripts of the verbal stimuli are included in an appendix.
  • The investigators taught the procedures of Experiment 3 during a brief training session insuring that P clearly understood the procedures.
  • Investigators read a brief narrative designed to elicit the target word while a picture of that word was displayed.
  • Discarded data due to static on the recordings from the 2 groups resulted in reduced Ns for the intensity and pitch analyses to

– TD = 9

– HFA = 11

 

 

  1. What were the results of the statistical (inferential) testing?

 

– RESULTS

 

  • OUTCOME #1: Mean pitch of productions

– No significant difference between groups (TD, HFA) on the production of

  • compound word/first syllable stressed targets
  • noun phrase/second syllable stressed targets

 

  • OUTCOME #2: Mean intensity of productions

– No significant difference between groups (TD, HFA) on the production of

  • compound word/first syllable stressed targets
  • noun phrase/second syllable stressed targets

 

  • OUTCOME #3: Mean whole word duration

     Compound words/first syllable stressed words were significantly shorter than noun phases/second syllable stressed word

– HFA group produced significantly longer utterances than the TD group

 

– What were the statistical tests used to determine significance? t-test; ANOVA: one way ANOVA; 2 (Group: HFA, TD) x 2 (Stress: compound word/first syllable stressed, noun phrase/second syllable stressed)

  • Other: (List the test and the associated outcomes)

 

– Were effect sizes provided? Yes.

 

  • OUTCOME #3: Mean whole word duration

     Compound words/first syllable stressed words were significantly shorter than noun phases/second syllable stressed word– eta = 0.53 (large effect)

  • for HFA the effect size was — Cohen’s d = 0.90 (large effect)
  • for TD the effect size was—Cohen’s d = 0.80 (large effect)

 

– HFA group produced significantly longer utterances than the TD group eta = 0.16 (large effect)

  • for compound words/first syllable stressed targets the effect size was –Cohen’s d = 0.82 (large effect)
  • for noun phrases/second syllable stressed targets the effect size was –Cohen’s d = 0.75 (moderate effect)

 

– Were confidence interval (CI) provided? No

 

 

  1. Summary of correlational results: NA

 

 

  1. Summary of descriptive results: Qualitative research NA

 

 

  1. Brief summary of clinically relevant results:

 

  • Ps with HFA were able to differentiate compound words/first syllable stress (e.g., highlight) from noun phrases/second syllable stress (e.g., high light) using duration.

 

  • However, their productions of both types of target words were significantly longer than their TD peers.

 

  • The investigators described their perceptions of Ps with HFA longer productions as being labored and slow with exaggerated pauses between syllables. The TD Ps productions, on the other hand, were perceived to be shorter, not labored, with smoother transitions between syllables.

 

 

ASSIGNED GRADE FOR QUALITY OF EXTERNAL EVIDENCE: __B-___

 


Kargas et al. (2016)

August 9, 2016

ANALYSIS

Comparison Research

NOTES: The focus of the investigation is not on intervention. Accordingly, no summary of intervention is included in the review.

KEY:

 ADOS = Autism Diagnostic Observation Scale

ASD = autism spectrum disorders

eta =   partial eta squared

f = female

m = male

MLU = mean length of utterance

NA = Not Applicable

P = participant or patient

pmh = Patricia Hargrove, blog developer

SLP = speech-language pathologist

TD = typically developing

 

 

SOURCE: Kargas, N., López, B., Morris, P., & Reddy, V. (2016). Relations among detection of syllable stress, speech abnormalities, and communicative ability in adults with autism spectrum disorders. Journal of Speech, Language, and Hearing Research, 59, 206-215.

 

REVIEWER(S): pmh

 

DATE: August 4, 2016

 

ASSIGNED GRADE FOR OVERALL QUALITY: No grade assigned. This investigation is concerned with aspects of prosody in a clinical population. Nevertheless, it has application to the clinical practice.

 

TAKE AWAY: The investigators compared adults diagnosed with autism spectrum disorders to typical peers on a task tapping the perception of syllable stress (primary speech perception). The results revealed that adults with ASD have significantly lower scores on the Syllable Stress Perception Task compared to TD peers, although perception varied markedly within the ASD group. Also, within the ASD group, poor stress perception was associated with increased speech production (i.e., stress, intonation, rate) problems but not communication/language problems.

 

 

  1. What type of evidence was identified?
  • What was the type of design? Comparison Research — Prospective, Nonrandomized Group Design

 

  • What was the focus of the research? Clinically Related

                                                                                                           

  • What was the level of support associated with the type of evidence? Level = B

 

                                                                                                           

  1. Group membership determination:

                                                                                                           

  • If there were groups, were participants randomly assigned to groups? No
  • If there were groups and Ps were not randomly assigned to groups, were members of groups carefully matched? Yes

                                                                    

 

  1. Were experimental conditions concealed?

                                                                                                           

  • from participants? No

                                                                    

  • from administrators of experimental conditions? No

                                                                    

  • from analyzers/judges? Yes, the Ps responses were collected on the computer.

 

 

  1. Were the groups adequately described? Variable

 

– How many participants were involved in the study?

 

  • total # of Ps:  42
  • # of groups: 2
  • List names of groups and the number of Ps in each group:

     – Autism Spectrum Disorders (ASD) = 21

– Typically Developing (TD) = 21

  • Did all groups maintain membership throughout the investigation? Yes, there was only one session

                                                                                

– CONTROLLED CHARACTERISTICS                                                  

  • age: adults
  • gender: 18m; 3f in each group
  • diagnosis of Ps with ASD: Asperger’s syndrome using Autism Diagnostic Observation Scale (ADOS)
  • psychiatric or developmental diagnosis of the TD group: Self report of no diagnosis of either of these problems.
  • Hearing: within normal limits based on audiometric testing

 

== DESCRIBED CHARACTERISTICS

  • age:

     – ASD = mean 30.3 years

     – TD = mean 29.5 years

  • gender:
  • cognitive skills:

     – ASD

  • mean Full range IQ = 109.5
  • mean Verbal IQ = 109.8
  • mean Performance IQ 107.2

     – TD

  • mean Full range IQ = 115.9
  • mean Verbal IQ = 113.9
  • mean Performance IQ = 114.2

 

– Were the groups similar? Yes

 

– Were the communication problems adequately described? No

 

 

  1. What were the different conditions for this research?

                                                                                                             

— Subject (Classification) Groups? Yes

  • Diagnosis (ASD, TD)

                                                               

— Experimental Conditions? Yes

  • performance of the Syllable Stress Perception Task (described below in item #9)

 

— Criterion/Descriptive Conditions? Yes

  • performance on Language and Communication Measure of the ADOS

 

 

  1. Were the groups controlled acceptably? Yes

 

 

  1. Was the dependent measure appropriate and meaningful? Yes

 

  • OUTCOME #1: The accuracy of responses on the Syllable Stress Perception Task

 

  • The dependent measure was NOT subjective.

 

  • The dependent measure WAS objective? Ps entered their responses into the computer

 

 

  1. Were reliability measures provided?

                                                                                                            

– Interobserver for analyzers? No _

 

Intraobserver for analyzers? No

 

– Treatment or test administration fidelity for investigators? No

 

 

  1. Description of design:

 

  • Testing for each Ps was accomplished during 3 hour individual sessions.
  • Prior to the Experimental Phase of the testing, an IQ test and the ADOS were administered.
  • The Experimental Phase included:

– tests were administered in a quiet room.

– practice with the Experimental Task.

– the administration of the Experimental Task — Syllable Stress Perception Task in which the P pressed a button via the computer indicating whether the words were same or different

  • The Syllable Stress Perception Task involved

– Ps individually listening to word pairs in which the same 4 syllable word was presented 2 times in one of the 4 following sets:

  • first syllable stressed for both versions (same)
  • first syllable stressed for 1 version, second syllable stressed for the other version (different)
  • second syllable stressed for both versions (same)
  • second syllable stressed for 1 version, first syllable stressed for the other version (different)

 

  • Statistical analyses controlled for overall IQ, Verbal IQ, and Performance IQ.

 

 

  1. What were the results of the statistical (inferential) testing?

 

  • OUTCOME #1: The accuracy of responses on the Syllable Stress Perception Task

– Ps with ASD performed significantly more poorly on the Syllable Stress Perception Task than the TD Ps. The response was stable irrespective of whether the stress was on the first or second syllable.

– Response bias for providing “Same” or “Different” response was similar for the Ps with ASD and the TD group.

– The investigators explored whether there were previously unrecognized subgroups within the 2 major groups (ASD, TD.) They determined that the ASD group had significantly more poor performers (2 standard deviations below the group mean) than the TD group (33% vs 10.5%)

 

  • What statistical tests were used to determine significance? t-test, ANOVA, Chi square

 

  • Were effect sizes provided? No, for the most part.

 

  • Were confidence interval (CI) provided? No

 

 

  1. Summary of correlational results:

 

  • OUTCOME #1: The accuracy of responses on the Syllable Stress Perception Task

– Correlation between accuracy and communication quality:

  • Significant correlation between syllable stress perception accuracy and Speech Abnormalities score on the ADOS (i.e., atypical production of rate, stress, intonation; higher scores signify more problems) — r = – 0.75
  • The correlation between syllable stress perception accuracy and language communication score on the ADOS was not significant

 

 

  1. Summary of descriptive results: Qualitative research NA

 

 

  1. Brief summary of clinically relevant results:
  • Adults with ASD have significantly lower scores on the Syllable Stress Perception Task compared to TD peers, although perception varied markedly within the ASD group.
  • Within the ASD group, poor stress perception was associated with and increased speech production (i.e., stress, intonation, rate) problems.

 

 

 

ASSIGNED GRADE FOR QUALITY OF EXTERNAL EVIDENCE: B