Ballard et al. (2010)


SOURCE: Ballard, K. J., Robin, D. A., McCabe, P., & McDonald, J. (2010). A treatment for dysprosody in childhood apraxia of speech. Journal of Speech, Language, and Hearing Research, 53, 1227-1245.




DATE:  11.05.11                               ASSIGNED OVERALL GRADE:  B


Take Away:  Limited  support for improving expressive stress  in 7-10 year-olds diagnosed with childhood apraxia of speech.


1.  What was the focus of the research: Clinical Research


2.  What type of evidence was identified?


2a.  What  type of single subject design was used?  Single Subject Experimental Design:  Multiple Baseline

2b.  What was the level of support associated with the type of evidence?  A-


3.  Was phase of treatment concealed?                                 


3a.  from participants? No                            

3b.  from clinicians? No                                               

3c.  from data analyzers? No                       



4. Were the participants adequately described? Yes


4a.      How many participants were involved in the study? 3


4b.  The following characteristics/variables described:


•           age:  7-10 years

•           gender:  2M, 1F

•           expressive language:  WNL

•           receptive language:  WNL

•           educational level of parents:  Implied—Mom was a teacher


4c.  Were the communication problems adequately described? Yes; Childhood apraxia of speech (CAS)


5.  Was membership in treatment maintained throughout the study?  Yes


5a.  If there was more than one participant, did at least 80% of the participants remain in the study? Yes

5b.  Was any data removed from the study? No


6.  Did the design include appropriate controls? Yes

6a.  Were baseline data collected on all behaviors? Yes

6b.  Did probes include untrained data? Yes

6c.  Did probes include trained data? Yes

6d.  Was the data collection continuous? Yes

6e.  Were different treatment counterbalanced or randomized? Does not apply


7.  Were the outcomes measure appropriate and meaningful? Yes


7a.  List the outcome of interest (dependent variables):

1.  Acoustic: syllable vowel duration, peak vowel intensity, peak fo. These were not analyzed separately. They were used to derive other measures.

#2-4:  Pairwise variability index; measured asymmetry between a 2 syllables


– PVI = WS

2. PVIdur

3. PVIint

4. PVTfo

5. total duration of first 2 syllables (fluency/speech rate)

6.   percentage of  responses judged correct during therapy session

6a = stress  correct

6b = stress and speech rate correct

7.  rating on 5 point scale of 20 CVCVCV strings and 10 real words from  final baseline and first posttest

7b.  Are the outcome measures subjective? 6 & 7 are clearly subjective.  There is a subjective component to 1- (duration measure) , 2 and 5—  Because investigators segmented the samples/

7c.  Are the outcome measure objective? Yes 1 (intensity and fo measures), 3, and 4

7d.  Are the outcome measures reliable? Provided for duration measures:  Interobserver  r = .98; Intraobserver r = .99

8.  Did the target behaviors improve when it was treated? For the most part, yes.

9.  Overall quality of improvement  (M1 = Male patient #1; M2 = Male patient #2; F1 = Female patient )

Strong#6b M1; #2—PVI dur all Ps; #3 PVI int–All Ps; #4 PVIfo–M1; #5 –treated strings–All P; Untreated string–M1, M2;

Moderate; 6b F1;

Ineffective: 6b M2; #5 Untreated strings F1

9a.  Was baseline low and stable?

Yes: M1- #4 (somewhat); F1- #1, M2- #1, 2, 3

No: M1- #2, 3; F1 #2, 3


9b.  What was the percentage of nonoverlapping data (PND)? Not provided

9c.  Does visual inspection of data suggest that the treatment was effective? Yes


10.  What was the magnitude of the treatment effect? effect size (d)


#5 –for all significant comparisons listed d  >.82 (strong effect): .01

treated (all), less complex (M1, M2), more  complex (M1)

#7  = for all significant comparisons listed d >1.5 (large treatment effect) with exception of treated SW for M2 d= .60 (moderate treatment effect); p ≤ .01

treated SW (M1, M2); treated WS (F1, M2)


11.  Was information about treatment fidelity adequate?  Yes  (authors labeled it Interrater Agreement on independent variable.  See p. 1234.  This is a holistic measure that included response accuracy.)

Interobserver:  M1 =85.1%; M2 =87%; F1 = 85%

Intraobserver:  M1 = 96.2%; M2 = 95.6%; F1 = 98%

NOTE:  Also provided generalization and retention data.





PURPOSE:  To produce strong-weak and weak-strong 3 syllable nonword targets with rapid and fluent syllable/segment transitions

ASPECTS OF PROSODY:  Emphasis (lexical stress) [marked by manipulation of duration, loudness, pitch], fluency (rate), loudness



•           acoustic measures:  syllable/word duration; peak vocal intensity; peak fundamental frequency; pairwise variability index (PVI)—normalized measure of durationintensity, fo, Positive PVI  =  strong/weak stress pattern, Negative PVI  = weak/strong stress pattern)total duration of 1st 2 syllables


•           perceptual  measures:  target for judgments of accurate production, stress (syllable duration), rate, pitch variation, intensity variation, speech naturalness, perceptual rating on 5 point scale  (1=  clear fluent SW stress pattern; 5 = clear fluent WS stress pattern)


POPULATION:  childhood apraxia of speech; school-age; 7-10 years

DOSAGE:  60 minutes , 4 times a week for 3 weeks

GOAL ATTACK:  horizontal



  treatment targets were nonsense words (non-referent strings) with  differing levels of complexity:

1.  high = 4 syllables with 3 different consonants and vowels

2.  mid =  3  syllables with 3 different consonants and vowels

3.  low =  3 syllables with 3 different vowels

•  strings were created from only sets of 3 different plosive sounds and 3

different long  vowels

•  resulted in 36 strings that could be produced with either of the stress patterns (WS or SW) for a total of 72 strings

•  randomly selected 25 WS strings and 25 SW strings for treatment

•  other 25 strings were  untreated strings used in  baseline and  probes

•  strings were presented orthographically with stressed syllable highlighted.

•  a feedback sheet was used during pre-practice and was visible during practice

1.  Emphasis = target different durations for unstressed versus stressed syllables

2.  Fluency – use habitual rate with no pauses between syllables

3.  Loudness – change in intensity level of syllables is not exaggerated




Baseline Phase

•  P read aloud 50 word selected randomly in blocks form the list.

•  No feedback was provided

•  Targets were 10 treated strings, 10 untreated same complexity strings, 10 untreated less complexity strings, 10 untreated more complexity strings

Experimental Probes

•  administered every 4th session during treatment and 4 weeks post treatment

Treatment Phase

§ Stimuli


•  Multisyllable nonword targets with

1.  strong/weak or weak/strong stress patterns

2.  randomized ordering of syllables

•  C provided  feedback regarding accuracy provided with a 3-5 second wait time.

•  Target was  providing feedback 50% of time. Range 0-100%

•  Procedures were based on following principles:

1.  high intensity practice

2.  targeting multiple skills

3.  ordering of stimuli is random

4.  focusing oncomplex skills

5.  each session consisting of a prepractice and a practice section



•  C randomly selects 1 set of 10 nonwords from treatment stimuli.

•  P reads aloud target word with a carrier phrases (e.g., He bought a ____; Can you find my   _____).

•  P produces 5 consecutive carrier phrases + target nonword referent with an accurate stress pattern

•  At first, C provides model plus detailed feedback by describes the characteristics of a correct response.  For example, C might focus on one or more of the following as feedback:

1.  relative duration  of syllable1 to syllable2 for SW or WS strings

2.  production of targets using  habitual rate (i.e., no slowing of speech)

3.  elimination of pauses between syllables

4.  avoidance of loudness Increase

5.  avoidance of exaggerated pitch variation

6.  use of rhythmic tapping  and visual aids to facilitate production

•  C also provided a feedback sheet during prepractice.

1.  Emphasis = target duration of different syllables; longer = S; shorter = W

2.  Fluency =  habitual rate with no pauses between syllables

3.  Loudness = intensity level

•  C provided the following when eliciting target sentence:

1.  Cards with the target sentence or word written.

2.  Stressed syllable is highlighted  in bold.


•  Occurs after each prepractice.

•  As needed, C provides 3-5 minute breaks ( for playing a board game).

•  C targets 100-120 practice sentences per session containing (10-12 sets of 10 randomly ordered sentences.)

•  C tells P that feedback will only be concerned with accuracy of the stress pattern; feedback will no longer provided descriptions.  Example:

–  Used   “Good”, “Nice”, or “Not good” feedback for emphasis, fluency, or loudness.

•  C encourages P to self monitor (i.e., listen and evaluate self)

•  C target for feedback: overall target rate is 50%  (beginning = 100% feedback; end (may be last set of the day)  = 10% feedback)

•  C modifies the target behavior for accuracy response as C progresses.  Example:

1.         Stressed syllable is 50% longer than unstressed

2.         Appropriate fluency or loudness, depending on the child’s ability

•  C can place feedback sheet (described in prepractice) on the table with definitions.



Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

%d bloggers like this: