AJSLP
Research Article
Feedback Frequency in Treatment for
Childhood Apraxia of Speech
Edwin Maas,a Christine E. Butalla,a,b and Kimberly A. Farinellaa,c
Purpose: To examine the role of feedback frequency in
treatment for childhood apraxia of speech (CAS). Reducing
the frequency of feedback enhances motor learning, and
recently, such feedback frequency reductions have been
recommended for the treatment of CAS. However, no published
studies have explicitly compared different feedback frequencies
in this population.
Method: Using an alternating treatments single-subject
design with multiple baselines across behaviors, retention and
transfer of learning were compared following high-frequency
feedback and low-frequency feedback in 4 children with CAS.
Feedback frequency was manipulated in the context of an
integral stimulation treatment. Changes in perceptual accuracy
were quantified with effect sizes and were compared across
conditions.
L
earning a new motor skill or improving an existing
motor skill typically requires feedback regarding the
accuracy and performance of the movement during
practice. Such feedback enables the learner to determine
which aspects of the movement were successful, which were
not, and how the motor plan can be modified for success on
a future attempt. For children with childhood apraxia of
speech (CAS), learning the movements of speech is thought
to be particularly difficult (e.g., American Speech-Language-
Hearing Association [ASHA], 2007). As such, feedback
may be especially critical for children with CAS, and a close
examination of different aspects of feedback implemented
during treatment may contribute to improving speech motor
learning in this population. The present study was part of
a larger research program that systematically investigated
the effects of various conditions of practice in treatment for
CAS (e.g., Maas & Farinella, 2012). In the current study,
a putational modeling simulations (e.g., Terband & Maassen,
University of Arizona, Tucson
b 2010).
Panacea Therapeutics, Tucson, AZ
c Although there are few published controlled treatment
Northern Arizona University, Flagstaff
Correspondence to Edwin Maas: emaas@arizona.edu
Editor: Carol Scheffner Hammer
Associate Editor: Ken Bleile
Received August 31, 2011
Revision received January 9, 2012
Accepted March 9, 2012
DOI: 10.1044/1058-0360(2012/11-0119)
American Journal of Speech-Language Pathology • Vol. 21 • 239–257 • August 2012 • A American Speech-Language-Hearing Association
Results: Findings were mixed, with 2 children showing an
advantage for low-frequency feedback, 1 child showing a
small advantage for high-frequency feedback, and 1 child
showing no clear improvement in either condition.
Conclusion: These findings suggest that reducing the
frequency of feedback may be beneficial for some children
with CAS, although this may vary with the child’s age or
severity of apraxia. Caution is warranted in extrapolating from
the nonspeech motor learning literature to speech treatment
for CAS. Finally, this study contributes another replication
to the literature on the efficacy of integral stimulation treatment
for children with CAS.
Key Words: childhood apraxia of speech, treatment,
motor learning, feedback, speech disorders
we manipulated the amount of feedback provided during
treatment and examined the effects on retention and transfer
of learning.
Speech Motor Learning in CAS
The current consensus in the literature is that CAS re-
flects an impairment of speech motor programming and
planning, which is evidenced by difficulties moving from
one speech sound to the next, inconsistent errors on vowels
and consonants, and prosodic problems (e.g., ASHA, 2007;
Grigos & Kolenda, 2010; Nijland et al., 2002; Skinder,
Strand, & Mignerey, 1999; Terband, Maassen, van Lieshout,
& Nijland, 2011). In addition, it has been suggested that
children with CAS also have difficulties with speech motor
learning, based on their often noted slow, limited prog-
ress in treatment (e.g., Ballard, Robin, McCabe, & McDonald,
2010; Campbell, 1999; Shriberg, Aram, & Kwiatkowski,
1997; Strand, Stoeckel, & Baas, 2006) and on recent com-
efficacy studies that focus on CAS, the clinical observations
of slow and limited progress in treatment have motivated a
search for ways to optimize speech motor learning in chil-
dren with CAS. The literature on motor skill learning pro-
vides some potentially relevant ideas in this regard, in that
some conditions of practice and feedback have been found
optimal in promoting motor learning (sometimes referred to
as “principles of motor learning”; see Maas et al., 2008, for
239
review). Various authors have advocated the use of these
conditions of practice and feedback in CAS treatment (e.g.,
Hall, 2000; Iuzzini & Forrest, 2010; McCauley & Strand,
1999; Robin, Maas, Sandberg, & Schmidt, 2007), and sev-
eral studies have incorporated these conditions into CAS
treatment (e.g., Ballard et al., 2010; Edeal & Gildersleeve-
Neumann, 2011; Iuzzini & Forrest, 2010; Maas & Farinella,
2012; Strand & Debertine, 2000; Strand, Stoeckel, & Baas,
2006). However, only two studies to date have explicitly
compared the effects of different practice conditions in CAS
treatment to determine whether the same advantages apply
to this population (Edeal & Gildersleeve-Neumann, 2011;
Maas & Farinella, 2012).
Edeal and Gildersleeve-Neumann (2011) examined the
role of practice amount, arguably the most powerful factor in
learning (Schmidt & Lee, 2005), in an integral stimulation
treatment for two children with CAS. Integral stimulation
treatment was originally developed for adults with apraxia of
speech (Rosenbek, Lemme, Ahern, Harris, & Wertz, 1973)
and involves auditory and visual models provided by the
clinician (“listen to me, watch me, and say what I say”) and
principles of motor learning (see Edeal & Gildersleeve-
Neumann, 2011, and Strand & Skinder, 1999, for excellent
reviews of the integral stimulation treatment literature for
CAS). Edeal and Gildersleeve-Neumann defined practice
amount in terms of the number of practice trials and used an
alternating treatments single-subject design in which differ-
ent sound targets were assigned to different practice con-
ditions (100–150 vs. 30–40 trials per session). They observed
improvements in speech sound accuracy for both practice
conditions, but retention and transfer were greater for speech
sounds that were practiced 100–150 times per session than
for speech sounds that were practiced 30–40 times per ses-
sion. These findings are consistent with the motor learning
literature (e.g., Shea, Kohl, & Indermill, 1990; see Maas
et al., 2008, for review).
In an integral stimulation treatment with four children
with CAS, Maas and Farinella (2012) investigated another
promising variable, namely, practice schedule. Practice
schedule refers to the order in which different targets are
practiced, and a random practice schedule has been found to
enhance motor learning compared to a blocked practice
schedule (in which different targets are practiced in separate
blocks). Maas and Farinella compared random and blocked
practice in an alternating treatments design with multiple
baselines across behaviors. Treatment was provided in two
phases, each lasting 4 weeks (È12 hr of treatment per phase)
and separated by a 2-week maintenance phase following
each treatment period (same design as present study). Indi-
vidually tailored treatment targets and probe items were
selected for each child depending on his or her particular
difficulties, and treatment involved the dynamic temporal
and tactile cuing (DTTC) approach (Strand, Stoeckel, &
Baas, 2006). The findings were mixed: Three of the four
children showed a clear response to treatment (in both
conditions) and retention following treatment in terms of
overall utterance accuracy, and two of those demonstrated
a slight net advantage for blocked over random practice,
whereas the third child showed a clear and consistent advan-
tage for random over blocked practice. Transfer to untreated
240 American Journal of Speech-Language Pathology • Vol. 21 • 239–257 • August 2012
items was negligible in all cases. These findings suggest that
although practice factors derived from the motor learning
literature may influence speech motor learning in children
with CAS, the direction of those influences may be different.
In other words, caution is warranted in extrapolating practice
principles from the motor learning literature to CAS treatment.
Thus, although the motor learning literature can provide
guidance for efforts to improve speech motor learning
outcomes in children with CAS, the few direct empirical
investigations of various practice factors in this population
have yielded somewhat mixed results. As such, the con-
tinued direct study of various practice factors in treatment
for CAS is necessary to determine optimal practice con-
ditions for children with CAS. The present study represents
a continuation along these lines by examining the role of
feedback frequency in CAS treatment.
Feedback Frequency Effects in Motor Learning
Models of motor control and learning assume that pro-
duction of a skilled movement involves exploiting knowl-
edge of the relations between various sources of information
about a movement, such as motor commands, sensory con-
sequences of the movement, and movement outcome (e.g.,
Bullock, Grossberg, & Guenther, 1993; Schmidt, 1975;
Willingham, 1998; Wolpert, Ghahramani, & Flanagan,
2001). Feedback is considered critical to allow the learner
to determine these relations and figure out which combina-
tions of motor commands produce the desired outcome in
a given situation. Intrinsic feedback refers to information
about a movement that is inherent in performance of the
motor act. For example, in throwing a dart, the visual system
provides information about where the dart landed, and the
somatosensory system provides information about the grip
force and arm excursion used. In many situations, learners
can use intrinsic feedback to improve their performance, but
there are situations in which intrinsic feedback is not suf-
ficient or optimal. For example, a learner may not have
access to all relevant intrinsic information (e.g., in learning a
novel dance routine, the learner may not have access to, or be
able to focus on, information about arm movements when
focusing on foot movements), or a learner may not have a
frame of reference that is calibrated to some standard of
correctness (e.g., the learner may not know how to interpret
the scores on a dart board). In such situations, externally
provided (augmented) feedback from an instructor may be
necessary to improve the skill (Wulf & Shea, 2004).
Considerable research effort has been devoted to finding
ways to optimize augmented feedback in the learning of
motor skills (see Maas et al., 2008; Wulf & Shea, 2004,
for reviews). One feedback variable known to affect motor
learning is feedback frequency, which refers to how often
learners receive augmented feedback on their motor perfor-
mance during practice. A typical finding is that high-frequency
feedback (e.g., feedback after every practice trial) enhances
practice performance but degrades motor learning, as com-
pared to reduced frequency feedback (e.g., feedback only
after half of all practice trials) (e.g., Bruechert, Lai, & Shea,
2003; Nicholson & Schmidt, 1991; Winstein & Schmidt,
1990).
 Probably the most influential explanation for the learning
benefit of reduced frequency feedback has been the guidance
hypothesis (Salmoni, Schmidt, & Walter, 1984; cf. Wulf &
Shea, 2004, for a more recent review). According to the
guidance hypothesis, learners come to rely too much on the
guiding role of feedback when it is present during practice
(perhaps even integrating feedback with the task; Schmidt
& Bjork, 1992), thereby impeding the development of in-
ternal response evaluation and error correction mechanisms
based on intrinsic feedback (cf. Bruechert et al., 2003).
Reducing the frequency of feedback would encourage the
learner to use intrinsic information to evaluate a response in
no-feedback trials and would allow the learner to calibrate
this internal judgment of correctness to the external standard
provided by the instructor. It is this ability to use intrinsic
feedback to evaluate movement performance that presum-
ably facilitates ongoing learning after practice ends (analo-
gous to the well-known proverb, “give hungry people fish
and they will be fed for a day; teach them how to fish
and they will be fed for a lifetime”). Bruechert et al. (2003)
showed that participants were better at estimating their errors
after practice with 50% feedback than after practice with
100% feedback. An alternative explanation for the learn-
ing benefit of reduced frequency feedback is the stability
hypothesis (e.g., Lai & Shea, 1998; Schmidt & Bjork, 1992),
which states that frequent feedback may result in attempts
at correction of small, relatively insignificant errors, creat-
ing response instability that prevents the formation of
a clear, stable movement pattern. These explanations are
not mutually exclusive, and in fact, the importance of
response stability was already noted by Salmoni et al.
(1984).
Investigation of feedback frequency in speech motor
learning has been limited (Maas et al., 2008). One study
compared 100% feedback and 60% feedback in treatment for
four adults with apraxia of speech (Austermann Hula, Robin,
Maas, Ballard, & Schmidt, 2008). Findings suggested that
reduced frequency feedback produced greater learning
(evident on transfer and /or retention tests) in two of the
four participants; data for the other two participants were
confounded with target-complexity effects and could not be
interpreted relative to feedback frequency. Benefits for re-
ducing feedback frequency were also reported by Steinhauer
and Grayhack (2000) in a task requiring speakers to achieve
a particular nasalance target for sustained vowels. Thus,
initial findings for speech motor learning are promising.
Although reducing the frequency of augmented feedback
appears to enhance motor learning, there are several reasons
to be cautious in extending these findings to treatment for
children with CAS. First, feedback frequency may interact
with task complexity. For example, Wulf, Shea, and Matschiner
(1998; Wulf & Shea, 2002) showed that learning of complex
motor tasks appeared to benefit from more frequent (i.e.
100%) feedback compared to 50% feedback. If speech is
considered a complex motor task, then reducing feedback
frequency may be detrimental to learning. However, initial
findings from speech motor learning (e.g., Austermann
Hula et al., 2008) argue against this idea because reduced
frequency feedback did appear to enhance learning, at least
for adults.
Second, feedback frequency effects may differ for dif-
ferent movement aspects (e.g., Lai & Shea, 1998; Wulf,
Schmidt, & Deubel, 1993). For example, Wulf et al. (1993)
reported that reduced frequency feedback enhanced learning
of underlying movement patterns (i.e., relative timing and
amplitude), whereas high-frequency feedback enhanced
learning of scaling parameters (i.e., absolute timing and
amplitude). Perceptual judgments of speech accuracy do not
separate these aspects (Maas et al., 2008), thus potentially
masking learning effects (although Austermann Hula et al.,
2008, did report feedback frequency effects based on per-
ceptual judgments).
Third, and perhaps most relevant to the present study,
there are indications that feedback frequency has different
effects in children than in adults (e.g., Chiviacowsky, Wulf,
Laroque de Medeiros, Kaefer, & Wally, 2008; Sullivan,
Kantak, & Burtner, 2008). For example, Sullivan et al.
(2008) compared 100% feedback and 62% feedback for an
arm movement task and found that children (average age
10 years) in the 100% feedback group performed better on a
retention task than children in the 62% feedback group.
The Present Study
Feedback on accuracy is a critical component of virtually
all behavioral treatment programs in speech-language pa-
thology, including treatments for CAS. Decisions have to be
made, either explicitly or implicitly, regarding the frequency
of clinician-provided feedback. Currently, there are no
published studies that have systematically studied feedback
frequency effects in children with CAS. Thus, the main
research question of the present study was: Does feedback
frequency affect the retention and transfer of speech motor
learning in children with CAS?
The preceding review of the literature indicates that
reduced frequency feedback appears to enhance learning of
speech and nonspeech motor tasks for adults. However,
recent research suggests that high-frequency feedback may
be better for learning of nonspeech motor tasks in children.
As such, the expectations regarding feedback frequency
effects in treatment for children with CAS are unclear. If
children with CAS pattern with adults (for speech and non-
speech motor learning), then we would expect greater
learning with reduced frequency feedback. If, on the other
hand, children with CAS pattern with children in nonspeech
motor learning, then we would expect greater learning with
high-frequency feedback.
The main purpose of the present study was to test these
opposing predictions by comparing the effects of high- and
low-frequency feedback on speech motor learning in four
children with CAS. We implemented the feedback frequency
manipulation in the context of an integral stimulation
treatment modeled after the DTTC approach (Strand &
Debertine, 2000; Strand, Stoeckel, & Baas, 2006) for two
main reasons. First, DTTC has replicated evidence to support
its efficacy, thus making this approach the most likely to
produce gains (minimizing potential floor effects and max-
imizing potential gain for participants). Second, this ap-
proach explicitly incorporates principles of motor learning,
and therefore allowed us to manipulate one factor (feedback
Maas et al.: Feedback Frequency in Treatment for CAS 241
frequency) while retaining the essential ingredients of the
overall approach as much as possible.
A secondary purpose of the study was to provide an-
other systematic replication of this integral stimulation treat-
ment (cf. Maas & Farinella, 2012). As noted by Edeal and
Gildersleeve-Neumann (2011), integral stimulation is cur-
rently the approach with the most evidence to support its use.
However, there are very few peer-reviewed treatment studies
for CAS overall (ASHA, 2007), and even fewer studies
that were primarily based on integral stimulation (Edeal &
Gildersleeve-Neumann, 2011; Maas & Farinella, 2012;
Strand & Debertine, 2000; Strand, Stoeckel, & Baas, 2006;
see Strand & Skinder, 1999, for review). Thus, additional
replications across participants and research groups are nec-
essary to establish the generalizability of the efficacy of
integral stimulation treatment (cf. ASHA, 2007); the present
study adds to the evidence base in this regard by contributing
another well-controlled single-subject design study involv-
ing four participants with CAS.
Method
Participants
This study included four monolingual English children
with a diagnosis of CAS who were recruited from the
University of Arizona’s Grunewald-Blitz Clinic for Com-
munication Disorders in Children and from local schools.
Diagnosis of CAS was confirmed by a licensed and certified
speech-language pathologist (SLP; third author) with ex-
tensive experience in differential diagnosis of childhood
speech disorders. Diagnosis of CAS was based on the pres-
ence of all three characteristics proposed by the ASHA
technical report (ASHA, 2007); namely, inconsistent errors
on vowels and consonants, difficulty transitioning between
speech sounds, and prosodic errors including incorrect
or equal stress and segmented speech. In the absence of
validated standardized assessment instruments for diagnos-
ing and/or quantifying CAS (McCauley & Strand, 2008), the
current gold standard is clinical expert opinion. For the
present study, CAS features and severity were determined
perceptually by an experienced clinician based on connected
speech samples (e.g., narrative samples) as well as dynamic
assessment procedures in structured, single-word or single-
phrase production tasks, including the Goldman-Fristoe Test
of Articulation 2 (GFTA–2; Goldman & Fristoe, 2000) and
the Dynamic Evaluation of Motor Speech Skill (DEMSS;
Strand, McCauley, & Stoeckel, 2006). The DEMSS requires
the child to directly imitate stimulus items of increasing
complexity, beginning with very simple consonant–vowel
(CV) and CVC words (e.g., up, day). Using a dynamic
scoring procedure (scores depend on level of support needed),
the DEMSS specifically measures vowel accuracy, move-
ment gesture accuracy (moving from one articulatory position
to another), consistency of sound production, and prosody.
Judgments of inconsistent errors were based on between
two and five repetitions of the same utterance and required
at least one production that differed from the others. Tran-
sitioning difficulties were judged based on the presence of
inter- and intrasyllable pauses and /or noted improvement
242 American Journal of Speech-Language Pathology • Vol. 21 • 239–257 • August 2012
in segmental accuracy with reductions in speaking rate. Se-
verity judgments were based on the expert clinician’s
perceptual judgment of speech motor control difficulty and
intelligibility using a 5-point verbal scale (i.e., mild, mild-
moderate, moderate, moderate-severe, severe). In addition,
the percentage of consonants correct (PCC) and percent-
age of vowels correct (PVC; Shriberg, Austin, Lewis,
McSweeny, & Wilson, 1997) based on narrow transcription
of repetition of lists of words and phrases are provided as
a more quantified metric of the severity of each child’s
speech impairment.
Formal testing of receptive language was based on the
Concepts & Following Directions subtest of the Clinical
Evaluation of Language Fundamentals, Fourth Edition
(CELF–4; Semel, Wiig, & Secord, 2003); the Auditory
Comprehension subtest of the Preschool Language Scale,
Fourth Edition (PLS–4; Zimmerman, Steiner, & Pond,
2002); and /or the Peabody Picture Vocabulary Test, Fourth
Edition (PPVT–4; Dunn & Dunn, 2007). Expressive lan-
guage was assessed based on core language subtests
(Word Structure, Recalling Sentences, and Formulated Sen-
tences) of the CELF–4 (Semel et al., 2003) and language
sampling.
Children ranged in age from 5;4 (years;months) to 8;4
at study onset, and all had normal hearing based on a pure-
tone hearing screening or parent report. All children demon-
strated speech sound production difficulties as reflected in
standard scores on the GFTA–2 <85 (<12th percentile; see
Table 1). None of the children had a neurological or medical
diagnosis at the time of the study. As is not uncommon in
this population, the children had varying degrees of CAS
severity and concomitant diagnoses; however, CAS was a
major diagnosis in all cases and formed the focus of treat-
ment. Three of the four children (CAS001, CAS002, and
CAS005) had participated in a previous treatment study
(Maas & Farinella, 2012).1 Each child is described sepa-
rately below; a summary is presented in Table 1. All study
procedures were approved by the University of Arizona’s
Institutional Review Board.
CAS001. CAS001 was a non-Hispanic White girl, age
8;4, with moderate-severe CAS. Her speech exhibited incon-
sistent vowel and consonant substitutions and distortions,
segmentation of sounds and syllables, equal and incorrect
stress in multisyllabic words, and intermittent hypernasality,
all of which were consistent with a diagnosis of CAS. No
formal intelligibility assessment was undertaken, but clin-
ical judgment indicated severely reduced intelligibility.
CAS001 demonstrated persistent use of age-inappropriate
phonological error patterns, including consistent prevocalic
voicing, final consonant deletion, and fronting, as well as
inconsistent deaffrication, velar assimilation (e.g., gog for
dog), and weak syllable deletion. Based on analysis of a
word repetition list, CAS001’s PCC was 53% (particular
problems with fricatives, affricates, and velar sounds), and
1
For all three children, the time between studies was 1 week (5 weeks
between treatment sessions). To enable fair comparisons between condi-
tions, entirely new speech targets, transfer items, and control items were
selected and baselined for each child. The same participant numbers refer to
the same individuals between studies.
TABLE 1. Study participant information.

Age Childhood apraxia

ID Sex (years;months) of speech Dysarthria Language Speech

CAS001 Female 8;4 Moderate-severe None Moderate expressive delay; GFTA–2 SS: <40
low average receptive skills (<1st percentile)
PPVT–4 SS: 97 (42nd percentile) PCC: 53%
CELF–4 Core SS: 76 (5th percentile) PVC: 46%
Concepts & Following Directions = 9/10
Word Structure = 4/10
Recalling Sentences = 4/10
Formulated Sentences =7/10
CAS002 Male 5;4 Severe Very mild unilateral Expressive not testable; receptive GFTA–2 SS: <40
upper motor neuron language within normal limits (<1st percentile)
PPVT–4 SS: 92 (30th percentile) PCC: 42%
PLS–4 SS: 95 (37th percentile) PVC: 11%
CAS005 Female 7;3 Moderate-severe Moderate Severe expressive delay; GFTA–2 SS: 62
spastic-flaccid mild-moderate receptive delay (<1st percentile)
PPVT–4 SS: 79 (8th percentile) PCC: 55%
CELF–4 Core SS: 46 (< 0.1st percentile) PVC: 34%
Concepts & Following Directions = 4/10
Word Structure = 1/10
Recalling Sentences = 1/10
Formulated Sentences = 1/10
CAS012 Male 6;4 Moderate None Severe expressive delay; moderate GFTA–2 SS: 84
to severe receptive delay (12th percentile)
CELF–4 Core SS: 46 (<0.1st percentile) PCC: 59%
Concepts & Following Directions = 4/10 PVC: 62%
Word Structure = 1/10
Recalling Sentences = 1/10
Formulated Sentences = 1/10
Sentence Structure = 3/10

Note. SS = standard score; PCC = percentage of consonants correct; PVC = percentage of vowels correct; PPVT–4 = Peabody Picture
Vocabulary Test, Fourth Edition (Dunn & Dunn, 2007); CELF–4 = Clinical Evaluation of Language Fundamentals, Fourth Edition (Semel, Wiig, &
Secord, 2003); GFTA–2 = Goldman-Fristoe Test of Articulation 2 (Goldman & Fristoe, 2000); PLS–4 = Auditory Comprehension subtest of the
Preschool Language Scale, Fourth Edition (Zimmerman, Steiner, & Pond, 2002).

her PVC was 46% (particular difficulties with diphthongs
and rhotic vowels). A structural–functional oral mechanism
exam (Duffy, 2005) revealed no neuromuscular deficits
suggestive of dysarthria.
CAS001’s expressive language as assessed with the
CELF–4 revealed a moderate delay. She demonstrated
difficulty with correct production of various morphological
markers (e.g., possessive pronouns, regular and irregular past
tense). Use of complex sentence structure was noted on
occasion (e.g., The girl who has a soccer ball was third). Her
receptive language was evaluated using the CELF–4 and the
PPVT–4 and was determined to be in the low average range.
CAS002. CAS002 was a non-Hispanic White boy, age
5;4, with severe CAS and normal receptive language. Ex-
pressive language was not formally tested due to his severely
reduced intelligibility, but informal observation suggested
a severe expressive language delay. His CAS was charac-
terized by inconsistent errors on vowels and consonants,
difficulty achieving initial articulatory configurations, trial
and error groping behaviors, inconsistent productions following
repetition of stimulus items, intermittent hypernasality, and
regular use of a stereotypical nonword (e.g., dunch). His
speech was facilitated by verbal, visual, and tactile cues.
Performance on a word repetition list revealed a PCC of
42% (particular difficulty with fricatives and liquids) and a
PVC of 11% (difficulties with most vowels and diphthongs,
except nonrhotic central vowels and /u/). A mild breathy/
harsh voice quality was noted, and a structural–functional
oral mechanism exam (Duffy, 2005) revealed very mild left
facial asymmetry, suggesting a possible mild unilateral upper
motor neuron dysarthria. Spontaneous utterances and ver-
balizations were characterized by adequate respiratory func-
tion for purposes of speech breathing and by grossly normal
loudness, pitch, and rate.
CAS002’s expressive language could not be formally
assessed due to his markedly severe CAS, but language
sample analysis revealed evidence of expressive language
delay. There was little evidence for correct use of age-
appropriate morphological markers (e.g., absence of third
person singular, copula omission). When asked to describe
and tell a story about a picture book, CAS002 verbalized
with high frequency and clearly was “telling” the story;
however, the majority of his utterances were completely
unintelligible. Occasionally, an intelligible word was noted
(e.g., jumping), indicating that he likely had the concept of
certain grammatical markers (i.e., verb + ing). His pragmatic
language skills and play schemes were age appropriate.
Receptive language as assessed using the PPVT–4 and the
Auditory Comprehension subtest of the PLS–4 was deter-
mined to be in the low average range. Difficulties were noted

Maas et al.: Feedback Frequency in Treatment for CAS 243

with understanding qualitative concepts (i.e., longest, pointed,
thin), quantity concepts (e.g., half, whole, each), passive
voice, rhyming sounds, and adding and subtracting numbers
to five. He did not seem to have difficulty understanding
contextually based questions and comments; however, he
was often unable to respond using intelligible words, often
relying on the stereotypical nonword dunch.
CAS005. CAS005 was a non-Hispanic White girl, age
7;3, with moderate-severe CAS, moderate-severe dysarthria
(mixed spastic–flaccid), a mild-moderate receptive language
delay, and a severe expressive language delay. Diagnosis
of CAS was based primarily on the presence of inconsistent
errors on consonants and vowels (especially diphthongs).
Her PCC was 55% (particular difficulties with fricatives,
affricates, and liquids), and her PVC was 34% (particular
difficulties with diphthongs and rhotic vowels). Other speech
features consistent with a CAS diagnosis were also noted
(including speech sound distortions, sound and syllable
segmentation, intermittent hypernasality, prosodic abnor-
malities, and reduced speech rate), although these features
may also be attributable to her dysarthria. Patterned speech
errors included deaffrication, labialization, stopping, gliding,
and consonant cluster reduction (especially /l/ and /a/
clusters). Deletion of final /z/ and final /v/ was also noted.
CAS005 also exhibited a markedly severe frontal lisp and
consistently distorted /s/ and /z/ due to marked tongue
protrusion on these sounds.
CAS005’s referral stated a history of hypotonia and
developmental delay but no specific medical–neurological
diagnosis. Her voice quality was intermittently hoarse and
breathy, and difficulties with other fine motor tasks (e.g.,
writing) were noted. Spontaneous utterances were charac-
terized by mildly reduced respiratory drive for purposes
of speech production, slow speech rate, hypernasality,
monoloudness, and monopitch. A structural–functional oral
exam (Duffy, 2005) revealed evidence of moderate tongue
weakness in lateral tongue movements and tongue protrusion
and moderately restricted range of motion on tongue pro-
trusion. No tongue fasciculations or atrophy were observed;
however, fasciculations of the chin were noted on the left
side. Movement of the soft palate was symmetrical but re-
duced during production of /a/. Speech alternating motion
rates ( papapa, tatata, kakaka) were slow, and sequencing
difficulties were noted during speech sequential motion
rates ( pataka).
CAS005’s expressive language was evaluated using
the CELF–4 expressive core language subtests. Many age-
appropriate morphological markers were absent (e.g., regular
plural, third person singular) or were used inconsistently
(e.g., irregular plural, auxiliary + ing). CAS005 also con-
sistently used me for I. These difficulties were also noted in
her spontaneous speech.
CAS005’s receptive language was assessed using the
CELF–4 and the PPVT–4. She had difficulty following
directions containing many different linguistic concepts
(e.g., before/after, closest to), and the PPVT–4 revealed a
moderate to severe delay in her receptive vocabulary.
CAS012. CAS012 was a Hispanic left-handed boy, age
6;4, with moderate CAS, a markedly severe expressive
language delay, and a moderate to severe receptive language
244 American Journal of Speech-Language Pathology • Vol. 21 • 239–257 • August 2012
delay. His referral indicated a history of ear infections and
difficulties with sequential tasks (e.g., auditory and working
memory), which may have contributed to the severity of his
speech and language difficulties. He had no medical or
neurological diagnoses at the time of the study, but he did
display some features consistent with attention deficit/
hyperactivity disorder based on parent and teacher rating
scales and educational psychological testing that had been
completed previously.
CAS012’s CAS was characterized by frequent frank
vowel errors, inconsistent errors, prosodic errors (e.g., in-
correct lexical stress), difficulty achieving initial articulatory
configurations, greater difficulty with longer utterances, and
intermittent hypernasality. Of note, his productions im-
proved with slow, simultaneous clinician models. CAS012’s
PCC based on a word/phrase repetition list was 59% (par-
ticular difficulties with fricatives, affricates, and liquids), and
his PVC was 62% (particular difficulties with diphthongs
and rhotic vowels). He demonstrated idiosyncratic errors
(e.g., distortion of initial /w/) and several phonological
patterned errors, including gliding and consonant cluster
reduction. His spontaneous utterances were characterized
by mildly abnormal pitch variability, high pitch, inconsis-
tent hypernasality, and normal loudness and voice quality.
Direct observation of his oral articulators revealed grossly
normal range of motion, strength, and speed of the jaw,
lips, and tongue. Based on a structural–functional oral
exam (Duffy, 2005), there was no evidence of nonverbal
oral apraxia and no evidence of frank weakness, nor was
there evidence of dysarthria. Speech alternating motion
rates ( papapa, tatata, kakaka) were adequate; sequential
motion rates ( pataka) were sequenced and produced
incorrectly.
CAS012’s expressive language was assessed using the
CELF–4, which revealed a markedly severe delay. During
contextual interactions, CAS012 tended to respond using
short, simple, unelaborated utterances (e.g., The potato good;
Me have lots of toys). The only grammatical markers noted
included verb + ing, plurals, articles, and irregular third
person (e.g., does). He consistently substituted me for I and
did not use any complex sentence structure.
CAS012’s receptive language was assessed using the
Concepts & Following Directions subtest and the Sentence
Structure subtest of the CELF–4, which revealed a severe
delay. He had difficulty understanding directions containing
several linguistic concepts (e.g., farthest, before/after, be-
tween, all but one, separated by, neither). In addition, he
struggled with longer, multipart directions, and sometimes
also with understanding contextually based questions and
comments during communicative interactions. Based on
educational psychological testing completed previously,
his auditory processing/memory impairments possibly con-
tributed to the severity of his language and speech deficits,
particularly his receptive language.
Design
We used an alternating treatments design with multiple
baselines across behaviors over two phases (Kearns, 1986).
The independent variable was the amount of feedback, with
two conditions: high-frequency feedback (HFF; feedback on
all trials) and low-frequency feedback (LFF; feedback on
È60% of trials; see Treatment Procedures for more detail).
Both conditions were administered every treatment session,
with the order of conditions counterbalanced across sessions.
This design feature eliminated order/sequence effects
without relying on between-participant counterbalancing.
An initial baseline period with a minimum of three probe
sessions over 2 weeks was implemented before any treat-
ment. To assess retention, a 2-week maintenance period
followed each treatment phase. A final probe was adminis-
tered 1 month after the last treatment session. Thus, the total
study duration from baseline to follow-up was 16 weeks (i.e.,
2 weeks baseline, 4 weeks treatment, 2 weeks maintenance,
4 weeks treatment, 2 weeks maintenance, follow-up after
another 2 weeks).
The main reason for implementing two phases of
treatment was to control for unanticipated differences in
difficulty or learnability between the target sets in Phase 1.
Given that each condition was paired with a different set of
targets, any differential learning might be attributed to the
targets rather than the feedback frequency condition. Im-
plementing a second phase with a new set of targets in each
condition enabled us to provide a within-participant repli-
cation of any effects (i.e., the multiple-baseline-across-
behaviors component of the design). Although the use of
multiple phases necessarily introduces the possibility of
order and sequence effects, we attempted to minimize this
possibility by reversing the pairing between condition and
general target type in Phase 2 (e.g., clusters or adjacent stress
in LFF in Phase 1, then clusters or adjacent stress in HFF in
Phase 2) and demonstrating stable baselines before Phase 2
treatment.
Weekly probes were administered throughout the study to
track retention and transfer of learning. The study included
both a performance-based criterion (80% correct on targets
on two consecutive probes) and a time-based treatment
termination criterion (4 weeks per phase). Because none of
the children reached the performance-based criterion, treat-
ment phases lasted 4 weeks for all children,2 constituting
È10 hr of treatment per phase for each child. There were
three weekly 50-min treatment sessions during treatment
phases for all children except CAS012, whose schedule
involved two weekly 1-hr sessions due to scheduling
constraints.
Treatment Target Selection
Comparing conditions or treatments in a single-subject
design context requires careful selection of different item
sets. Important requirements for treatment target sets include
(a) low baseline performance, to justify treatment and enable
detection of treatment effects; (b) equal difficulty level
before treatment, to enable a fair comparison between
conditions; and (c) independent targets, to minimize the
2 & Farinella, 2012) were also included (but will not be
Note that due to technical difficulties, some probe sessions during treatment
were not recorded for CAS002 and CAS012. Thus, Figures 2 and 4 display
fewer probe points during treatment periods, even though four probes
were administered in both phases as planned.
possibility of transfer between item sets. The first two
requirements were met by selecting target sets for each
child that were equally difficult as determined based on a
pretreatment word repetition task. This word repetition task
was similar to the baseline and weekly probes but included
a wider range of words and phrases expected to reveal
difficulties, as well as items expected to be within a given
child’s capacity. As an additional control for possible con-
founds due to the difficulty or learnability of targets, the
design included a multiple baseline replication in a second
treatment phase in which target-condition pairings were
reversed where possible.
The third requirement was addressed primarily on theo-
retical grounds. In particular, we chose different target sets
involving fundamentally different speech planning opera-
tions (e.g., sequencing consonants vs. sequencing syllables;
cf. Bohland & Guenther, 2006, for empirical support for
this distinction) or different relative timing (stress) patterns.
The motor learning literature suggests that different relative
timing patterns are governed by different underlying motor
programs (e.g., Schmidt, 1975; Shea & Wulf, 2005) and
are thus considered independent. Empirical support for the
notion of independence comes from motor learning studies
in which limited or no transfer was observed between move-
ments with different relative timing patterns (e.g., Sakai,
Kitaguchi, & Hikosaka, 2003; Wilde & Shea, 2006). Given
that stress patterns are defined in part by their relative timing
structure, we conjectured that different stress patterns
would be governed by different underlying motor programs
(Maas et al., 2008; cf. also Skinder, Connaghan, Strand, &
Betz, 2000) and therefore would not show cross-transfer.
Although efforts were made to include targets with per-
sonal functional relevance, the aforementioned target se-
lection requirements prevented us from using personal
functional relevance as the primary selection criterion. In
addition to treated targets for each condition in each phase
(i.e., four treatment sets), we also included untreated item
sets with similar properties to examine transfer, and un-
related untreated items to establish experimental control. The
item sets for each child are briefly described in the follow-
ing paragraphs; an overview of the item sets is provided
in Table 2.
CAS001. Given the prosodic difficulties noted in her
speech, targets for CAS001 included phrases with different
stress patterns. Phase 1 targets for CAS001 consisted of
five 3-syllable phrases per condition, with a strong–weak–
strong (SwS) stress pattern (e.g., time to go) for the HFF
condition and a wSS stress pattern (e.g., before school)
for the LFF condition. Phase 2 targets consisted of five
4-syllable phrases per condition, with a wSSw stress pattern
(e.g., begin reading) in the HFF condition and an SwwS
stress pattern (e.g., tell me a joke) in the LFF condition. Five
untreated phrases with the same target stress pattern were
also generated for each of these four sets, and 48 untreated,
different control items were included as well. Finally, the
12 treated targets from the previous treatment study (Maas
reported here), for a total list of 100 items.
CAS002. Phase 1 targets for CAS002 included three
2-syllable words (e.g., circle) in the HFF condition and three
Maas et al.: Feedback Frequency in Treatment for CAS 245
TABLE 2. Target items for all study participants.

ID Condition Phase 1 targets Phase 2 targets

CAS001 HFF SwS phrases in the chair; make it quick; eat a cake; wSSw phrases without asking; the cute monkey; begin
time to go;comb your hair reading; I play soccer; escape summer
LFF wSS phrases before school; a big dog; avoid heat; SwwS phrases tell me a joke; give it away; finish the
his new bike; the small cup work; under my coat; never be late
Control SSw phrases (n = 5); wSwS phrases (n = 5); 5-syllable phrases (n = 15); 4–5-syllable words (n = 10); 1-syllable words (n = 13)
CAS002 HFF 2-syllable words tiger, balloon, circle Initial clusters broom, smoke, flame
LFF Final clusters mask, point, barn 3-syllable words bicycle, tomato, radio
Control 4-syllable words; fricative CVC words; affricate CVC words; plosive CVC words; liquid CVC words; sCC clusters (n = 10 all)
CAS005 HFF wSS phrases I don’t know; before school; a big dog; SwwS phrases give it away; never be late; thanks for
his new bike; my pink shirt your help; after the storm; tell me a joke
LFF SwS phrases this is fun; time to go; not today; wSSw phrases a hot summer; the cute monkey; without
brush your teeth; take a break asking; before Friday; begin reading
Control SSw phrases (n = 5); wSwS phrases (n = 5); 5-syllable phrases (n = 15); 4–5-syllable words (n = 10); 1-syllable words (n = 13)
CAS012 HFF Sww words bicycle, pyramid, astronaut wSwS words America, photographer, emergency
LFF wSw words flamingo, volcano, coyote SwSw words Arizona, caterpillar, helicopter
Control multisyllabic words (n = 21); multisyllabic phrases (n = 28); initial clusters (CCVC) (n = 10)

Note. HFF = high-frequency feedback, LFF = low-frequency feedback, S = strong, w = weak, C = consonant, V = vowel.

words with final consonant clusters (e.g., point) in the LFF
condition. Phase 2 targets consisted of three words with
initial clusters (e.g., broom) in the HFF condition and three 3-
syllable words (e.g., bicycle) in the LFF condition. For each
of these four treatment sets, seven untreated words
with similar structures were generated as transfer sets.
With 60 unrelated control words, the total list consisted
of 100 words.
CAS005. Given her difficulties with multisyllabic utter-
ances, targets for CAS005 included phrases with different
stress patterns. Phase 1 focused on 3-syllable phrases. The
five targets in the HFF condition had a wSS stress pattern
(e.g., my pink shirt), and the five targets in the LFF condition
had an SwS stress pattern (e.g., take a break). In Phase 2,
treatment focused on 4-syllable phrases: five targets with
an SwwS stress pattern in the HFF condition (e.g., never
be late) and five targets with a wSSw stress pattern in the
LFF condition (e.g., a hot summer). Five similar, untreated
phrases were selected for each treatment set (transfer),
and 48 different untreated items served as controls. The
12 treated items from the previous treatment study (Maas
& Farinella, 2012) were also included (but will not be
reported here), for a list of 100 items.
CAS012. Given his noted difficulties with prosody,
targets for CAS012 included multisyllabic words. Phase 1
targeted 3-syllable words. The HFF condition targets con-
sisted of three words with an Sww stress pattern (e.g.,
pyramid ), and the LFF condition targets consisted of three
words with a wSw stress pattern (e.g., coyote). In Phase 2,
six 4-syllable words were selected as targets: three HFF con-
dition targets with a wSwS stress pattern (e.g., America)
and three LFF condition targets with an SwSw stress pattern
(e.g., Arizona). In addition to three transfer items per set,
we included 59 untreated control items, for a total list of
83 items.
Procedure
Treatment and probes were administered by a trained grad-
uate student clinician under direct supervision of the third
author, who is a certified and licensed SLP with extensive
experience in the treatment of speech disorders in children.
Each child was treated by the same clinician throughout
the study.
Probe procedures. The data for this study were based on
productions on a probe task. The probe task was a repetition
task in which the clinician produced each item at a normal
rate without any other cues. No feedback was provided
during the probe task. Treatment targets, transfer items, and
control items for each child were combined into a single list
and were presented in one of four random orders. Probes
were administered three to four times before any treatment
started (baseline) and then once per week thereafter. Probes
were always administered at the beginning of a session
and were audio- and video-recorded for subsequent data
analysis and reliability scoring.
Treatment procedures. We elected to study the effects of
feedback frequency in the context of the DTTC approach
(Strand, Stoeckel, & Baas, 2006) because this treatment
has replicated efficacy support (Maas & Farinella, 2012;
Strand & Debertine, 2000; Strand, Stoeckel, & Baas, 2006)
and because it incorporates principles of motor learning,
allowing us to systematically vary one of these while
keeping the treatment otherwise the same. Furthermore, by

246 American Journal of Speech-Language Pathology • Vol. 21 • 239–257 • August 2012

selecting DTTC as the treatment, this study would contribute
yet another replication of this approach.
Briefly, DTTC involves integral stimulation (“watch me,
listen carefully, and say what I say”) combined with inten-
sive drill practice on a small set of vocabulary items. Slow
speech and tactile gestures are used as needed to highlight
the movement patterns, and the amount of time between
the clinician’s model and the child’s attempt is gradually
increased to encourage greater reliance on independent re-
trieval and planning of speech motor gestures. The number
of practice trials per session was maximized by using quick
reinforcers (e.g., stickers). Targets were practiced in ran-
dom order throughout the study, in both feedback conditions.
Children received È2 hr of treatment per week, divided over
two or three sessions.
The two conditions being compared were HFF and LFF.
Feedback was defined as explicit verbal information pro-
vided by the clinician about the accuracy (knowledge of
results; e.g., “that was almost right!”) and movement ges-
tures (knowledge of performance; e.g., “you didn’t round
your lips enough”). For each treatment target, 10 index cards
were created, some of which were marked to indicate feed-
back trials to the clinician. In the HFF condition, all cards
were marked; in the 60% condition, only six of the 10 cards
per target were marked. Before each session, the cards in
each of the two piles were shuffled to randomize trial order.
Breaks were provided at the condition switch and as needed.
All sessions were audio- and video-recorded for subse-
quent analysis and fidelity evaluation. The treatment pro-
tocol is provided in Appendix A.
Data Analysis
The dependent measure was percentage accuracy on
the probe lists, based on a 3-way perceptual scoring system
that included both segmental and suprasegmental aspects of
entire target words or phrases, rather than only segmental
accuracy (cf. Maas & Farinella, 2012; Strand & Debertine,
2000; Strand, Stoeckel, & Baas, 2006). In this system, a
score of 2 denotes a correct production, a score of 1 denotes
a close approximation, and a score of 0 denotes an incorrect
production. These scores were operationally defined (see
Appendix B). Scores were converted to a percentage based
on the total possible points for each set. Each child’s data
were analyzed by one primary analyst who was unaware of
the treated targets, and sessions were analyzed in random
order to prevent a confound of increasing familiarity with a
child’s speech. Reliability was assessed by a second blinded
analyst for a randomly selected number of sessions (see
Reliability and Fidelity section later).
Data were plotted for visual analysis and were quantified
by calculating effect sizes (ES) using Beeson and Robey’s
(2006) modification of Busk and Serlin’s (1992) d metric:
ES = (mean score post treatment – mean score pre treatment)/
pooled standard deviation.3 Following Maas and Farinella
3 In Phase 2, CAS001 demonstrated a response to treat-
We chose to use the pooled standard deviation as the denominator rather
than the pretreatment standard deviation because the latter option would
result in many instances of incalculable effect sizes due to zero variance in
baseline (cf. Beeson & Robey, 2006; Maas & Farinella, 2012).
(2012), we operationally defined an effect to be present
when ES >1 (i.e., when magnitude of change exceeded the
standard deviation). Mean level differences in percentages
(gain scores; i.e. the numerator in the ES formula) are also
provided.
Reliability and Fidelity
The proportion of randomly selected double-scored ses-
sions ranged from 18% for CAS001 to 61% for CAS005
(CAS002: 27%; CAS012: 29%). As in previous studies
using this scoring system (e.g., Maas & Farinella, 2012;
Strand et al., 2006), interrater reliability was acceptable
to high (CAS001: 86%; CAS002: 94%; CAS005: 88%;
CAS012: 71%).
An independent analyst examined several treatment as-
pects for fidelity (Kaderavek & Justice, 2010) based on
viewing a randomly selected subset of video-recorded ses-
sions. Fidelity aspects scored for each condition included
number of trials, amount of time, number of whole-word
production attempts per trial, feedback delay, and feedback
frequency. Proportion of trials adhering to protocol was also
assessed; deviations from protocol included failure to pro-
vide feedback (HFF condition), inconsistent presence of
feedback within a trial (e.g., feedback on first attempt but
not on last attempt in a trial), and feedback delays <1 s or
>5 s. An overview of these fidelity measures is provided
in Appendix C. As can be seen, treatment fidelity was
not perfect. By and large, however, conditions were well
matched.
Results
Data for each child are plotted in Figures 1–4 and sum-
marized in Table 3. In all figures, square markers (red and
pink lines) represent the HFF condition, round markers (dark
and light blue lines) represent the LFF condition, solid lines
with filled markers denote treated items, and broken lines
with open markers denote untreated transfer items. Treatment
periods are demarcated by vertical lines (see also footnote 2).
CAS001
As Figure 1 shows, both conditions were at a stable 0%
correct during baseline. After onset of treatment, improve-
ment was evident for the treated items in both conditions,
but more so for the LFF condition. After treatment, both
conditions showed performance above baseline levels, with
higher accuracy for the LFF condition. The effect size mea-
sures corroborate this visual analysis, indicating ES = 1.52
(18% gain) for the HFF condition and ES = 2.51 (35% gain)
for the LFF condition. Neither condition showed transfer
(HFF: ES = 0.15, 1% gain; LFF: ES = 0.37, 4% gain), nor did
the control items show improvement (combined weighted
ES = 0.72, 7% gain; 3-syllable SSw phrases only: ES = 1.35;
17% gain).
ment in both conditions, following a stable baseline. The
performance on probes during the treatment phase appeared
equivalent between conditions; performance on probes after
Maas et al.: Feedback Frequency in Treatment for CAS 247
 FIGURE 1. CAS001 data for Phase 1 (top), Phase 2 (middle), and control items (bottom).

Note. S = strong (stressed), w = weak (unstressed), FB = feedback; Tx = treated items. Treatment periods are delineated with
vertical lines.

treatment showed greater accuracy for the LFF condition
again. Effect size could not be calculated for the HFF con-
dition due to zero variance in both pre- and posttreatment
data; for the LFF condition, effect size calculation showed
ES = 2.28. The mean level difference was larger for the LFF
condition (24% gain) than for the HFF condition (10% gain).
Transfer was negligible for the LFF condition (ES = 0.99,
7% gain) and small for the HFF condition (ES = 1.18, 8%
gain). The combined weighted effect size for control items in
Phase 2 indicated ES = 1.69 (5% gain), suggesting improvement
on untreated items. Inspection of the bottom graph in Figure 1
reveals that this effect was driven primarily by improvements
on 3-syllable phrases with SSw stress patterns (ES = 5.93,
22% gain) and, to a lesser degree, by the 4- and 5-syllable
phrases (ES = 1.93, 3% gain).
When combining data across both phases, the LFF con-
dition shows an effect size of ES = 2.40 (30% average gain);
no combined HFF condition effect size could be com-
puted (14% average gain). No transfer is evident for either
condition, and control items show ES = 1.20 (6% gain).

248 American Journal of Speech-Language Pathology • Vol. 21 • 239–257 • August 2012

 FIGURE 2. CAS002 data for Phase 1 (top), Phase 2 (middle), and control items (bottom).
Thus, the LFF condition consistently leads to greater changes
on treated items than the HFF condition, and the effect sizes
are larger than for the control items.
CAS002
For CAS002, a stable baseline was established before
Phase 1 treatment. As can be seen in Figure 2, an increase
in accuracy for treated items emerged toward the end of the
phase, with a larger increase for HFF items than for LFF
items. However, these gains were not retained after cessation
of treatment, as evinced by the effect sizes (HFF: ES = 0.77,
6% gain; LFF: ES = 0.58, 4% gain). No transfer was evident
(HFF: ES = 0.88, 4% gain; LFF: ES = 0.38, 1% gain), nor
did control items improve (combined weighted ES = 0.70,
4% gain; liquids only ES = 1.64; 6% gain).
In Phase 2, a similar pattern emerged: a stable baseline,
followed by an increase in accuracy toward the end of the
phase, with a larger increase for the HFF condition and a
decrease in accuracy after treatment ended. Effect size cal-
culations indicated treatment effects in both conditions, with
a relative advantage for HFF (ES = 1.91, 17% gain) com-
pared to LFF (ES = 1.11, 6% gain). No transfer was evi-
dent in either condition (0% change); control items also
remained unchanged (combined weighted ES = 0.37, 0%
gain; liquids only ES = 1.48; 8% gain).
Maas et al.: Feedback Frequency in Treatment for CAS 249
 FIGURE 3. CAS005 data for Phase 1 (top), Phase 2 (middle), and control items (bottom).

Taken together, CAS002 showed a small treatment effect
for HFF items (ES = 1.32, 12% gain) but not for LFF items
(ES = 0.83, 5% gain), with no transfer (ES not computable
for both conditions; HFF 2% gain; LFF 1% gain) and no
change in control items (ES = 0.53, 2% gain; liquids only
ES = 1.56, 7% gain).
CAS005
CAS005 demonstrated a stable baseline for both con-
ditions in both phases, with no apparent change throughout
the study. Effect sizes could not be computed for treated
items, except for HFF items in Phase 2 (ES = –0.51,
–2% gain; 0% gain for other treated items in both con-
ditions and both phases). The increase in performance on
LFF transfer items following Phase 2 treatment (ES = 1.08)
was minimal (5% gain). Some improvement was noted
for untreated control items in Phase 1 (ES = 1.75, 6% gain);
this effect was driven entirely by improvement on sim-
pler one-syllable words (ES = 2.09, 22% gain). In Phase 2,
there was no change in combined control items (ES = 0.96,
5% gain), although 3-syllable SSw phrases did improve
(ES = 2.15, 11% gain) following treatment on 4-syllable
phrases.

250 American Journal of Speech-Language Pathology • Vol. 21 • 239–257 • August 2012

 FIGURE 4. CAS012 data for Phase 1 (top), Phase 2 (middle), and control items (bottom).
Combined across phases, there was no improvement or
transfer in either condition, and a slight improvement on
control items (ES = 1.36, 6% gain).
CAS012
For CAS012, the Phase 1 baseline was more variable than
for the other participants, but was relatively low and with-
out an improving trend. Following initiation of treatment, an
increase in accuracy on treated items was evident for both
conditions, with a more sustained gain for HFF. Effect sizes
indicated a small treatment effect for the HFF condition
(ES = 1.04, 24% gain) but not for the LFF condition (ES =
0.83, 20% gain). There was no transfer for either condition
(HFF: ES = 0.21, 4% gain; LFF: ES = 0.28, 7% gain), and
no change for control items (ES = 0.38, 5% gain).
The baseline for Phase 2 targets was stable, and there was
a relatively small response to treatment in the HFF condi-
tion that was not maintained after treatment (ES not com-
putable, 0% gain). For the LFF condition, there was no
change during treatment, but performance following treat-
ment was consistently above baseline levels, which was
corroborated by the effect size calculation (ES = 4.42, 22%
gain). Regarding transfer, a small improvement was noted
Maas et al.: Feedback Frequency in Treatment for CAS 251
TABLE 3. Effect sizes per condition, phase, and child. condition for treated items in both phases. These effects
exceeded changes in the control items overall, indicating that
Child Condition Phase 1 Phase 2 Combined gains were likely attributable to treatment. However, some
change was noted in some of the control items (primarily
CAS001 HFF Treated 1.52 — —
Transfer 0.15 1.18 0.67 3-syllable SSw phrases in Phase 2). Although this suggests a
LFF Treated 2.51 2.28 2.40 loss of experimental control, an alternative interpretation is
Transfer 0.37 0.99 0.68 that these control items inadvertently showed transfer from
Control 0.72 1.69 1.20 the 4-syllable phrases treatment. This interpretation is sup-
CAS002 HFF Treated 0.77 1.91 1.32 ported by the fact that other control items remained un-
Transfer 0.88 — — changed and by the fact that the improvement was coincident
LFF Treated 0.58 1.11 0.83 with the onset of Phase 2 treatment. Thus, the results from
Transfer 0.38 — —
Control 0.70 0.37 0.53 CAS001 are consistent with the motor learning literature
on adults and with findings from research with adults with
CAS005 HFF Treated — –0.51 —
Transfer 0.51 — — apraxia of speech (Austermann Hula et al., 2008).
LFF Treated — — — CAS012 showed feedback frequency effects that were
Transfer 0.66 1.08 0.87 inconsistent across phases, with a small effect for HFF only
Control 1.75 0.96 1.36 in Phase 1 and a large effect for LFF only in Phase 2. This
CAS012 HFF Treated 1.04 — — pattern of findings suggests an interaction between feedback
Transfer 0.21 1.11 0.64 frequency and target difficulty, such that LFF was more
LFF Treated 0.83 4.42 2.55 effective than HFF, and words starting with a stressed syl-
Transfer 0.28 0.12 0.20
Control 0.43 0.67 0.55 lable (Sww in Phase 1 and SwSw in Phase 2) were easier
than words starting with an unstressed syllable (wSw in
Note. Effect size = (mean post) – (mean pre) / SD(pre+post). — = Phase 1 and wSwS in Phase 2). Thus, when the “easy”
ES could not be computed due to zero variance. targets were combined with the less effective HFF condition
(Phase 1), they counteracted the potential disadvantage of
HFF to produce a slight advantage over the LFF condition.
for the HFF transfer items (ES = 1.11, 6% gain) but not In Phase 2, when the “easy” targets were paired with the
for the LFF transfer items (ES = 0.12, 1% gain). Control optimal LFF condition, there was a larger difference be-
items did not show improvement as a group (ES = 0.90, 3% tween conditions, now in favor of LFF (a “double-whammy”
gain), although multisyllabic words did show some improve- effect). Overall, combined across both phases, there was a
ment (ES = 1.02, 6% gain). clear treatment effect in the LFF condition only, and this
Combined across both phases, CAS012 showed a treat- effect exceeded changes in the control items, suggesting that
ment effect for the LFF condition (ES = 2.55, 21% gain), the effects were due to treatment. Thus, the data from CAS012
whereas no effect size could be computed for the HFF con- also indicate an advantage for LFF over HFF.
dition (12% gain). There was no transfer or change in con- In contrast, data from CAS002 indicated a slight advan-
trol items (HFF transfer: 5% gain; LFF transfer: 4% gain; tage for HFF overall in terms of the retention of treated
control items: 4% gain). targets. Treatment effects in Phase 2 exceeded changes in
combined control items, supporting the idea that improve-
ments were due to treatment. This notion is further strength-
Discussion ened by the decrease in performance following treatment
The primary purpose of the present study was to compare (i.e., the maintenance period turned out to act as a withdrawal
the relative effects of different feedback frequencies (high design feature). Inspection of Figure 2 strongly suggests that
vs. low) in treatment for CAS with respect to retention and treatment was terminated too early (based on our a priori
transfer of learning. A secondary purpose was to replicate termination criterion) because notable increases in accuracy
a version of the DTTC treatment approach (e.g., Strand, began to appear on the last probe session during each treat-
Stoeckel, & Baas, 2006). Each of these issues will be dis- ment phase. Given the severity of his CAS and the notion
cussed in turn. that CAS often requires prolonged treatment (e.g., Ballard
et al., 2010; Shriberg, Aram, & Kwiatkowski, 1997), ex-
tending our relatively short treatment phases likely would
Feedback Frequency in CAS Treatment have resulted in greater, more sustained gains, especially
Summarizing the feedback frequency effects combined in the HFF condition, although this must remain speculative
across treatment phases, it is clear that findings were mixed. at this time. Nevertheless, it is worth noting that at least
Limiting our discussion to only the three children who showed minimal gains were seen in this child, contrary to a previous
improvement on treated items in any condition or phase (i.e., study in which he showed no gains (Maas & Farinella, 2012).
excluding CAS005), it appears that two children (CAS001 One possible explanation is that the previous study used
and CAS012) demonstrated an advantage for LFF compared LFF; the present findings suggest that CAS002 might have
to HFF, whereas the other child (CAS002) showed the shown gains if HFF had been used in the previous study.
reverse. The use of two separate treatment phases raises the pos-
For CAS001, both conditions resulted in a positive re- sibility that the Phase 2 effects were influenced by Phase 1
sponse to treatment, with a consistent advantage for the LFF treatment carryover. Given that we attempted to counterbalance

252 American Journal of Speech-Language Pathology • Vol. 21 • 239–257 • August 2012

target types across phases, such carryover effects would re-
sult in a reduction of condition differences when combining
both phases. The fact that we still observed condition dif-
ferences suggests that such order effects were minimal, if
present at all. In addition, the multiple baseline aspect of the
design allowed us to demonstrate that Phase 2 targets did
not improve during Phase 1 treatment, corroborating the
notion that any potential carryover effects were negligible.
Although treatment fidelity was not perfect, the condi-
tions were generally well matched in this regard. Further-
more, to the extent that differences were noted, they tended
to favor the condition that resulted in the smallest improve-
ments (e.g., for CAS001, HFF had longer average duration
than LFF, yet the LFF condition showed the greatest change).
Thus, it is unlikely that fidelity discrepancies can account
for the pattern of findings.
Taken together, two children benefited more from LFF,
whereas the third benefited more from HFF. Based on this
small sample size, we can only speculate as to why these
different patterns emerged. At least two possibilities present
themselves; namely, age (or age-related factors) and CAS
severity. The two children who showed an LFF advantage
were older than the other child, and feedback frequency effects
may shift with age (cf. Sullivan et al., 2008). Although Sullivan
et al. (2008) studied older children (mean age = 10 years) in a
different motor task (arm movement task), it is possible that
such age-related feedback frequency reversal effects occur
earlier for speech. Perhaps a certain minimum level of error
awareness—and a reference of correctness—is required to take
advantage of LFF, and perhaps our older children met this
minimum level. Anecdotally, both children who showed an
advantage for LFF were aware of their errors, as evidenced by
attempts at self-correction. CAS002, who benefited more from
HFF, initiated very few attempts at self-correction.
Alternatively, CAS severity may play a role as well, with
more severe CAS requiring practice conditions that are con-
sidered suboptimal in the motor learning literature. Consis-
tent with this idea is the observation in the previous study
(Maas & Farinella, 2012) that to the extent that CAS002
showed gains at all, these gains appeared to be in blocked
rather than random practice. Again, the severity of his CAS
may have limited CAS002’s a priori reference of correct-
ness or his ability to detect and correct errors. HFF may have
facilitated establishing such a reference of correctness or error
detection ability. Although we did not formally test speech
perception abilities, some research suggests that speech per-
ception may also be impaired in some children with CAS
(e.g., Groenen, Maassen, Crul, & Thoonen, 1996; Maassen,
Groenen, & Crul, 2003; Nijland, 2009). Such deficits could
conceivably interfere with the ability to detect errors and with
the ability to develop a reference of correctness. Future
treatment studies might benefit from including formal, de-
tailed measures of speech perception and error detection
abilities.
Efficacy of Integral Stimulation Treatment for CAS
Regarding our secondary purpose—to provide a replica-
tion of an integral stimulation-based treatment for CAS—our
findings were consistent with previous reports in that most
but not all children showed improvements (e.g., Maas &
Farinella, 2012; Strand, Stoeckel, & Baas, 2006). Thus, this
study adds to the evidence base by providing support for
a modified version of the DTTC approach. The magnitude
of treatment effects was generally comparable to the treat-
ment effects obtained in a previous study using this modified
DTTC approach (Maas & Farinella, 2012), but smaller than in
the original DTTC studies (Strand & Debertine, 2000; Strand,
Stoeckel, & Baas, 2006).
As discussed in Maas and Farinella (2012), the differ-
ence in magnitude of effects between the original and this
modified DTTC treatment may be related to one or more
factors, including treatment distribution or intensity (ten
30-min sessions per week for 6 weeks in Strand, Stoeckel, &
Baas, 2006; three 1-hr sessions per week for 4 weeks per
phase in the present study), the nature of treatment targets
(relatively simple, functionally relevant items in Strand &
Debertine, 2000; relatively complex items that were not
selected strictly based on personal functional relevance in
the present study), and the probe measurement procedures
(multiple opportunities per target in Strand, Stoeckel, &
Baas, 2006, and probes administered at the end of treatment
sessions in Strand & Debertine, 2000; one opportunity per
target, obtained at the beginning of a treatment session in
the present study).
It is important to note that three of the four children in this
study also participated in a previous treatment study (Maas
& Farinella, 2012). Although we selected new targets
(neither treated nor included on the probe lists in the previous
study) and demonstrated stable, low baseline performance
for these items, these children were already familiar with
the general treatment procedures, which may have enhanced
the gains seen in the present study. However, this appears
unlikely for two main reasons. First, the treatment effects
were comparable in magnitude to those observed in the prior
study (if anything, smaller than in the previous study). Sec-
ond, one of these children (CAS005) showed considerable
treatment effects in the previous study but showed no effects
in the current study. In the case of CAS005, her previous
participation may have reduced her motivation; anecdotal
observations suggested greater noncompliance during treat-
ment compared to the other children, as well as compared
to herself in the prior study. This may explain her lack of
improvement in the present study. Thus, although participa-
tion in the previous study may have influenced the present
findings, such influences do not necessarily enhance treat-
ment effects. Future studies are needed to replicate this treat-
ment with additional children (who may or may not have
received prior treatment).
With respect to transfer, the general conclusion is that
transfer was quite limited for all of the children studied. This
may be due to the nature of the relation between treated
and untreated targets. We defined our target sets in terms of
word/phrase length and stress pattern for three out of four
children, and it is possible that this type of similarity was not
sufficient to allow transfer. Perhaps similarity in terms of
place or manner of articulation classes is a more appropriate
way to assess transfer; however, our materials were not de-
signed or controlled in this way. We did note changes in
some control items for some children that seemed to coincide
Maas et al.: Feedback Frequency in Treatment for CAS 253
with onset of treatment on similar but more complex tar-
gets. For example, untreated words with singleton liquids
showed some improvement for CAS002 during Phase 2
treatment, in which the HFF condition targeted word-initial
clusters with liquids. For CAS001, untreated 3-syllable SSw
phrases improved during Phase 2, which targeted 4-syllable
phrases. Nevertheless, limited transfer is not uncommon in
integral stimulation approaches, and in fact appears to be the
rule in CAS treatment more than the exception (e.g., Maas
& Farinella, 2012; Strand, Stoeckel, & Baas, 2006). This
general finding argues for targeting words and phrases in
CAS treatment that are personally relevant and functional.
Future research with integral stimulation (and other) approaches
is warranted to identify ways to increase the degree of trans-
fer to untrained items.
Overall, the present study adds to a growing literature
that supports the efficacy of integral stimulation-based
treatments for CAS (e.g., Edeal & Gildersleeve-Neumann,
2011; Maas & Farinella, 2012; Strand & Debertine, 2000;
Strand, Stoeckel, & Baas, 2006). Further research is needed
to replicate and extend these findings to determine the op-
timal amount, distribution, and child characteristics to max-
imize the utility of this type of treatment.
Conclusion and Future Directions
The present study shows that reducing the frequency of
augmented clinician-provided feedback enhances learning in
some but not all children with CAS, compared to a high-
frequency feedback regimen. It is possible that younger
children, or those with more severe CAS, benefit from high-
frequency feedback, although further research is needed to
determine the child-related factors that predict the optimal
feedback frequency. Future studies should also extend this
line of research to examine other feedback-related variables
in this population, including feedback delay (immediate vs.
delayed), feedback control (self-controlled vs. clinician-
controlled), feedback schedule (fading schedule vs. random-
ized schedule), and feedback type (verbal vs. biofeedback),
as well as potential interactions with other practice conditions
(Maas et al., 2008).
Finally, the findings from this study add to the small
but growing literature supporting the efficacy of integral
stimulation-based treatment approaches such as DTTC
for CAS. Future studies should replicate and extend these
findings to determine optimal treatment candidacy and
treatment dosage and distribution.
Acknowledgments
This research was supported in part by a generous grant from
the Childhood Apraxia of Speech Association of North America
(CASANA). The authors also wish to thank Cass Faux, Lydia
Garcia, and Becky Vance for their assistance in identifying potential
participants; Jenna Beltrami for assistance in treatment delivery;
Lawrence Andrade, Ashley Davis, Krista Durr, Kelly Gallardo, and
Amanda Lopez for their assistance with data analysis; and of course,
the children and their parents who participated in this study. Por-
tions of these data were presented at the American Speech-
Language-Hearing Association Convention (Philadelphia, PA,
2010), the National Conference on Childhood Apraxia of Speech
254 American Journal of Speech-Language Pathology • Vol. 21 • 239–257 • August 2012
(Pittsburgh, PA, 2010), and the Arizona Speech-Language-Hearing
Association Convention (Phoenix, AZ, 2011); portions of these
data were also part of Christine Butalla’s master’s thesis at the
University of Arizona.
References
American Speech-Language-Hearing Association. (2007).
Childhood apraxia of speech [Technical Report]. Retrieved
from http://www.asha.org/policy.
Austermann Hula, S. N., Robin, D. A., Maas, E., Ballard, K. J.,
& Schmidt, R. A. (2008). Effects of feedback frequency and
timing on acquisition, retention, and transfer of speech skills
in acquired apraxia of speech. Journal of Speech, Language, and
Hearing Research, 51, 1088–1113.
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.
Beeson, P. M., & Robey, R. R. (2006). Evaluating single-subject
treatment research: Lessons learned from the aphasia literature.
Neuropsychology Review, 16, 161–169.
Bohland, J. W., & Guenther, F. H. (2006). An fMRI investigation
of syllable sequence production. NeuroImage, 32, 821–841.
Bruechert, L., Lai, Q., & Shea, C. H. (2003). Reduced knowl-
edge of results frequency enhances error detection. Research
Quarterly for Exercise and Sport, 74(4), 467–472.
Bullock, D., Grossberg, S., & Guenther, F. H. (1993). A self-
organizing neural model of motor equivalent reaching and tool
use by a multijoint arm. Journal of Cognitive Neuroscience,
5(4), 408–435.
Busk, P. L., & Serlin, R. C. (1992). Meta-analysis for single case
research. In T. R. Kratochwill & J. R. Levin (Eds.), Single-case
research design and analysis: New directions for psychology
and education (pp. 187–212). Hillsdale, NJ: Erlbaum.
Campbell, T. F. (1999). Functional treatment outcomes in young
children with motor speech disorders. In A. J. Caruso & E. A.
Strand (Eds.), Clinical management of motor speech disorders
in children (pp. 385–396). New York, NY: Thieme.
Chiviacowsky, S., Wulf, G., Laroque de Medeiros, F., Kaefer, A.,
& Wally, R. (2008). Self-controlled feedback in 10-year-old
children: Higher feedback frequencies enhance learning.
Research Quarterly for Exercise and Sport, 79(1), 122–127.
Duffy, J. R. (2005). Motor speech disorders: Substrates, differ-
ential diagnosis, and management (2nd ed.). St. Louis, MO:
Mosby-Year Book.
Dunn, L. M., & Dunn, D. M. (2007). Peabody Picture Vocabulary
Test, Fourth Edition. Bloomington, MN: Pearson Assessments.
Edeal, D. M., & Gildersleeve-Neumann, C. E. (2011). The
importance of production frequency in therapy for childhood
apraxia of speech. American Journal of Speech-Language
Pathology, 20, 95–110.
Goldman, R., & Fristoe, M. (2000). The Goldman–Fristoe Test
of Articulation 2. Austin, TX: Pro-Ed.
Grigos, M. I., & Kolenda, N. (2010). The relationship between
articulatory control and improved phonemic accuracy in child-
hood apraxia of speech: A longitudinal case study. Clinical
Linguistics & Phonetics, 24(1), 17–40.
Groenen, P., Maassen, B., Crul, T., & Thoonen, G. (1996).
The specific relation between perception and production errors
for place of articulation in developmental apraxia of speech.
Journal of Speech and Hearing Research, 39, 468–482.
Hall, P. K. (2000). A letter to the parent(s) of a child with
developmental apraxia of speech. Part IV: Treatment of DAS.
Language, Speech, and Hearing Services in Schools, 31, 179–181.
Iuzzini, J., & Forrest, K. (2010). Evaluation of a combined
treatment approach for childhood apraxia of speech. Clinical
Linguistics & Phonetics, 24(4–5), 335–345.
Kaderavek, J. N., & Justice, L. M. (2010). Fidelity: An essential
component of evidence-based practice in speech-language
pathology. American Journal of Speech-Language Pathology,
19, 369–379.
Kearns, K. P. (1986). Flexibility of single-subject experimental
designs. Part II: Design selection and arrangement of experimental
phases. Journal of Speech and Hearing Disorders, 51, 204–214.
Lai, Q., & Shea, C. H. (1998). Generalized motor program (GMP)
learning: Effects of reduced frequency of knowledge of results
and practice variability. Journal of Motor Behavior, 30(1), 51–59.
Maas, E., & Farinella, K. A. (2012). Random versus blocked
practice in treatment for childhood apraxia of speech. Journal of
Speech, Language, and Hearing Research, 55, 561–578.
Maas, E., Robin, D. A., Austermann Hula, S. N., Freedman,
S. E., Wulf, G., Ballard, K. J., & Schmidt, R. A. (2008).
Principles of motor learning in treatment of motor speech disorders.
American Journal of Speech-Language Pathology, 17, 277–298.
Maassen, B., Groenen, P., & Crul, T. (2003). Auditory and
phonetic perception of vowels in children with apraxic speech
disorders. Clinical Linguistics & Phonetics, 17(6), 447–467.
McCauley, R. J., & Strand, E. A. (1999). Treatment of children
exhibiting phonological disorder with motor speech involvement.
In E. A. Strand & A. J. Caruso (Eds.), Clinical management of
motor speech disorders in children (pp. 187–208). New York, NY:
Thieme.
McCauley, R. J., & Strand, E. A. (2008). A review of standard-
ized tests of nonverbal oral and speech motor performance in
children. American Journal of Speech-Language Pathology, 17,
81–91.
Nicholson, D. E., & Schmidt, R. A. (1991). Scheduling infor-
mation feedback to enhance training effectiveness. Proceedings of
the Human Factors Society 35th Annual Meeting (pp. 1400–1403).
Santa Monica, CA: Human Factors Society.
Nijland, L. (2009). Speech perception in children with speech output
disorders. Clinical Linguistics & Phonetics, 23(3), 222–239.
Nijland, L., Maassen, B., Van der Meulen, S., Gabreëls, F.,
Kraaimaat, F. W., & Schreuder, R. (2002). Coarticulation
patterns in children with developmental apraxia of speech.
Clinical Linguistics & Phonetics, 16(6), 461–483.
Robin, D. A., Maas, E., Sandberg, Y., & Schmidt, R. A.
(2007). Motor control and learning and childhood apraxia of
speech. In P. Hall, L. Jordan, & D. Robin (Eds.), Developmental
apraxia of speech: Theory and clinical practice (2nd ed.,
pp. 67–86). Austin, TX: Pro-Ed.
Rosenbek, J., Lemme, M., Ahern, M., Harris, E., & Wertz, R.
(1973). A treatment for apraxia of speech in adults. Journal of
Speech and Hearing Disorders, 43, 462–472.
Sakai, K., Kitaguchi, K., & Hikosaka, O. (2003). Chunking
during human visuomotor sequence learning. Experimental
Brain Research, 152, 229–242.
Salmoni, A. W., Schmidt, R. A., & Walter, C. B. (1984).
Knowledge of results and motor learning: A review and critical
appraisal. Psychological Bulletin, 95, 355–386.
Schmidt, R. A. (1975). A schema theory of discrete motor skill
learning. Psychological Review, 82, 225–260.
Schmidt, R. A., & Bjork, R. A. (1992). New conceptualizations
of practice: Common principles in three paradigms suggest new
concepts for training. Psychological Science, 3(4), 207–217.
Schmidt, R. A., & Lee, T. D. (2005). Motor control and learning: A
behavioral emphasis (4th ed.). Champaign, IL: Human Kinetics.
Semel, E., Wiig, E. H., & Secord, W. A. (2003). Clinical
Evaluation of Language Fundamentals, Fourth Edition.
San Antonio, TX: Pearson.
Shea, C. H., Kohl, R. M., & Indermill, C. (1990). Contextual
interference: Contributions of practice. Acta Psychologica, 73,
145–157.
Shea, C. H., & Wulf, G. (2005). Schema theory: A critical ap-
praisal and reevaluation. Journal of Motor Behavior, 37, 85–101.
Shriberg, L. D., Aram, D. M., & Kwiatkowski, J. (1997).
Developmental apraxia of speech: I. Descriptive and theoretical
perspectives. Journal of Speech, Language, and Hearing Research,
40, 273–285.
Shriberg, L. D., Austin, D., Lewis, B. A., McSweeny, J. L., &
Wilson, D. L. (1997). The percentage of consonants correct
(PCC) metric: Extension and reliability data. Journal of Speech,
Language, and Hearing Research, 40, 708–722.
Skinder, A., Connaghan, K., Strand, E., & Betz, S. (2000).
Acoustic correlates of perceived lexical stress errors in chil-
dren with developmental apraxia of speech. Journal of Medical
Speech-Language Pathology, 8, 279–284.
Skinder, A., Strand, E. A., & Mignerey, M. (1999). Perceptual
and acoustic analysis of lexical and sentential stress in children
with developmental apraxia of speech. Journal of Medical
Speech-Language Pathology, 7(2), 133–144.
Steinhauer, K., & Grayhack, J. P. (2000). The role of knowledge
of results in performance and learning of a voice motor task.
Journal of Voice, 14, 137–145.
Strand, E. A., & Debertine, P. (2000). The efficacy of integral
stimulation intervention with developmental apraxia of speech.
Journal of Medical Speech-Language Pathology, 8, 295–300.
Strand, E. A., McCauley, R. J., & Stoeckel, R. (2006, March).
Dynamic assessment of acquired and developmental apraxia
of speech in children. Paper presented at the Conference on
Motor Speech: Motor Speech Disorders, Austin, TX.
Strand, E. A., & Skinder, A. (1999). Treatment of developmental
apraxia of speech: Integral stimulation methods. In E. A. Strand
& A. J. Caruso (Eds.), Clinical management of motor speech
disorders in children (pp. 109–148). New York, NY: Thieme.
Strand, E. A., Stoeckel, R., & Baas, B. (2006). Treatment of
severe childhood apraxia of speech: A treatment efficacy study.
Journal of Medical Speech-Language Pathology, 14, 297–307.
Sullivan, K. J., Kantak, S. S., & Burtner, P. A. (2008). Motor
learning in children: Feedback effects on skill acquisition.
Physical Therapy, 88(6), 720–732.
Terband, H., & Maassen, B. (2010). Speech motor development
in childhood apraxia of speech (CAS): Generating testable
hypotheses by neurocomputational modeling. Folia Phoniatrica
et Logopaedica, 62, 134–142.
Terband, H., Maassen, B., van Lieshout, P., & Nijland, L.
(2011). Stability and composition of functional synergies for
speech movements in children with developmental speech
disorders. Journal of Communication Disorders, 44, 59–74.
Wilde, H., & Shea, C. H. (2006). Proportional and nonpropor-
tional transfer of movement sequences. Quarterly Journal of
Experimental Psychology, 59(9), 1626–1647.
Willingham, D. B. (1998). A neuropsychological theory of motor
skill learning. Psychological Review, 105(3), 558–584.
Winstein, C. J., & Schmidt, R. A. (1990). Reduced frequency
of knowledge of results enhances motor skill learning. Journal
of Experimental Psychology: Learning, Memory, and Cognition,
16, 677–691.
Wolpert, D. M., Ghahramani, Z., & Flanagan, J. R. (2001).
Perspectives and problems in motor learning. Trends in Cogni-
tive Sciences, 5(11), 487–494.
Wulf, G., Schmidt, R. A., & Deubel, H. (1993). Reduced
frequency feedback enhances generalized motor program
learning but not parameterization learning. Journal of Exper-
imental Psychology: Learning, Memory, and Cognition, 19,
1134–1150.
Maas et al.: Feedback Frequency in Treatment for CAS 255
Wulf, G., & Shea, C. H. (2002). Principles derived from the study
of simple motor skills do not generalize to complex skill
learning. Psychonomic Bulletin and Review, 9, 185–211.
Wulf, G., & Shea, C. H. (2004). Understanding the role of aug-
mented feedback: The good, the bad and the ugly. In A. M. Williams
& N. J. Hodges (Eds.), Skill acquisition in sport: Research,
theory and practice (pp. 121–144). London, England: Routledge.
Wulf, G., Shea, C. H., & Matschiner, S. (1998). Frequent
feedback enhances complex motor skill learning. Journal of
Motor Behavior, 30(2), 180–192.
Zimmerman, I. L., Steiner, V. G., & Pond, R. E. (2002).
Preschool Language Scale, Fourth Edition. San Antonio,
TX: Pearson.

Appendix A
Treatment protocol

Clinician: “Watch me, listen carefully, and repeat after me,” and models target
Child responds
Clinician waits 2–3 seconds
If child’s response is correct:
: FB trial: Provide FB – Knowledge of Results (KR) and Knowledge of Performance (KP)
: No FB trial: Present next target
If child’s response is incorrect:
: FB trial:
: Provide FB (KR and KP)
: Slow, simultaneous production (up to 2 times); tactile cues as needed
: Fade simultaneous cue (mouth the words)
: Immediate repetition
: Wait 2–3 seconds and provide FB (KR and KP)
: No FB trial:
: “Let’s do it slowly together”
: Slow simultaneous production (up to 2 times); tactile cues as needed
: Fade simultaneous cue (mouth the words)
: Immediate repetition
: Wait 2–3 seconds, then “Now let’s do another one”

As child improves, increase delay between clinician model and child’s response (i.e., delayed repetition instead of
immediate repetition) and eventually elicit targets without clinician model.

Note. FB = feedback.

256 American Journal of Speech-Language Pathology • Vol. 21 • 239–257 • August 2012

Appendix B
Scoring guidelines

Score = 0 Example Score = 1 Examples

More than one sound in error kæt Y gæd

(distorted, substituted,
omitted, added)
Omission of sound/syllable
Addition of sound/syllable
Vowel “substitution” (sounds
like a good exemplar
of a different vowel)
Gross consonant error
(more than one
feature off target)
Sound error plus prosodic
error (e.g., syllable
segregation, incorrect
stress pattern)
Any other errors that do
not qualify for a 1
tbÃtɚflaI Y tbætɚfaI
kæt Y kæ
tbÃtɚflaI Y tbÃflaI
tbÃtɚflaI Y tbÃtɚfaI
kæt Y kætəl
tbÃtɚflaI Y tbÃtɚflaIp
kæt Y k I t
tbÃtɚflaI Y tbætɚflaI
kæt Y kæf, bæt
tbÃtɚflaI Y tbÃzɚflaI
tbÃtɚflaI Y bætɚtflaI
Unintelligible
Two minor prosodic errors
Excessive plosive release and /or kæt Y kæth
aspiration (unless also included
in the clinician’s model)
Vowel distortion: kæt Y kæt æ/ɛ?
(sounds like a bad exemplar of the
target vowel, or between target and
other vowel)
Also includes excessive lengthening. tbÃtɚflaI Y tbÃtɚflaI Ã /I?
No other errors: kæt Y kæ:t
Rest of word should be entirely correct.
Minor consonant errors: kæt Y gæt, kæp
No more than one feature off target tbÃtɚflaI Y tdÃtɚflaI, tmÃtɚflaI
(may result in perceived substitution) tbÃtɚflaI Y tbÃtɚf:laI
for one consonant.
Also includes excessive lengthening.
No other errors: Rest of word should
be correct.
Minor prosodic error: tbÃtɚflaI Y bÃtɚtflaI
Incorrect stress pattern or syllable tbÃtɚflaI Y tbÃ(.)tɚ(.)flaI
segregation (only for multisyllabic items).
No other errors: Rest of word should
be correct.

Appendix C
Treatment fidelity measures

Child Sessions checked Condition Duration (min.) # Trials # Attempts per trial FB delay (sec.) %FB % Adherence

CAS001 3/24 HFF 26 16 4.2 2.4 83% 76%

LFF 18 15 4.5 2.4 48% 77%
CAS002 3/24 HFF 26 19 7.8 2.1 81% 68%
LFF 28 21 6.8 2.0 45% 66%
CAS005 3/24 HFF 17 11 6.2 2.5 91% 64%
LFF 28 14 6.2 2.4 66% 83%
CAS012 3/12 HFF 22 14 4.5 2.4 91% 86%
LFF 19 16 4.6 2.6 63% 92%

Note. HFF= high-frequency feedback, LFF = low-frequency feedback.

Maas et al.: Feedback Frequency in Treatment for CAS 257

Copyright of American Journal of Speech-Language Pathology is the property of American Speech-Language-
Hearing Association and its content may not be copied or emailed to multiple sites or posted to a listserv
without the copyright holder's express written permission. However, users may print, download, or email
articles for individual use.