Successful Therapist–Parent Coaching: How In Vivo Feedback Relates to Parent Engagement in Parent–Child Interaction Therapy

Abstract

Although behavioral parent training is considered efficacious treatment for childhood conduct problems, not all families benefit equally from treatment. Some parents take longer to change their behaviors and others ultimately drop out. Understanding how therapist behaviors impact parental engagement is necessary to improve treatment utilization. This study investigated how different techniques of therapist in vivo feedback (i.e., coaching) influenced parent attrition and skill acquisition in parent–child interaction therapy (PCIT). Participants included 51 parent–child dyads who participated in PCIT. Children (age: M = 5.03, SD = 1.65) were predominately minorities (63% White Hispanic, 16% African American or Black). Eight families discontinued treatment prematurely. Therapist coaching techniques during the first session of treatment were coded using the Therapist–Parent Interaction Coding System, and parent behaviors were coded with the Dyadic Parent–Child Interaction Coding System, Third Edition. Parents who received more responsive coaching acquired child-centered parenting skills more quickly. Therapists used fewer responsive techniques and more drills with families who dropped out of treatment. A composite of therapist behaviors accurately predicted treatment completion for 86% of families. Although group membership was correctly classified for the treatment completers, only 1 dropout was accurately predicted. Findings suggest that therapist in vivo feedback techniques may impact parents’ success in PCIT and that responsive coaching may be particularly relevant.

Full text

Full Terms & Conditions of access and use can be found at
http://www.tandfonline.com/action/journalInformation?journalCode=hcap20
Download by: [University of California, Los Angeles (UCLA)] Date: 23 October 2015, At: 12:31
Journal of Clinical Child & Adolescent Psychology
ISSN: 1537-4416 (Print) 1537-4424 (Online) Journal homepage: http://www.tandfonline.com/loi/hcap20
Successful Therapist–Parent Coaching: How In
Vivo Feedback Relates to Parent Engagement in
Parent–Child Interaction Therapy
Miya L. Barnett, Larissa N. Niec, Samuel O. Peer, Jason F. Jent, Allison
Weinstein, Patricia Gisbert & Gregory Simpson
To cite this article: Miya L. Barnett, Larissa N. Niec, Samuel O. Peer, Jason F. Jent, Allison
Weinstein, Patricia Gisbert & Gregory Simpson (2015): Successful Therapist–Parent Coaching:
How In Vivo Feedback Relates to Parent Engagement in Parent–Child Interaction Therapy,
Journal of Clinical Child & Adolescent Psychology, DOI: 10.1080/15374416.2015.1063428
To link to this article: http://dx.doi.org/10.1080/15374416.2015.1063428
Published online: 14 Oct 2015.
Submit your article to this journal
Article views: 30
View related articles
View Crossmark data

Successful Therapist–Parent Coaching: How In Vivo
Feedback Relates to Parent Engagement in
Parent–Child Interaction Therapy
Miya L. Barnett, Larissa N. Niec, and Samuel O. Peer
Center for Children, Families and Communities, Central Michigan University
Jason F. Jent, Allison Weinstein, Patricia Gisbert, and Gregory Simpson
University of Miami, Miller School of Medicine
Although behavioral parent training is considered efficacious treatment for childhood conduct
problems, not all families benefit equally from treatment. Some parents take longer to change
their behaviors and others ultimately drop out. Understanding how therapist behaviors impact
parental engagement is necessary to improve treatment utilization. This study investigated
how different techniques of therapist in vivo feedback (i.e., coaching) influenced parent
attrition and skill acquisition in parent–child interaction therapy (PCIT). Participants included
51 parent–child dyads who participated in PCIT. Children (age: M=5.03, SD = 1.65) were
predominately minorities (63% White Hispanic, 16% African American or Black). Eight
families discontinued treatment prematurely. Therapist coaching techniques during the first
session of treatment were coded using the Therapist–Parent Interaction Coding System, and
parent behaviors were coded with the Dyadic Parent–Child Interaction Coding System, Third
Edition. Parents who received more responsive coaching acquired child-centered parenting
skills more quickly. Therapists used fewer responsive techniques and more drills with families
who dropped out of treatment. A composite of therapist behaviors accurately predicted
treatment completion for 86% of families. Although group membership was correctly classi-
fied for the treatment completers, only 1 dropout was accurately predicted. Findings suggest
that therapist in vivo feedback techniques may impact parents’success in PCIT and that
responsive coaching may be particularly relevant.
Young children with high levels of disruptive behaviors are
more likely to develop serious conduct and emotional dis-
orders than their peers (Reef, Diamantopoulou, van Meurs,
Verhulst, & van der Ende, 2011). Although behavioral par-
ent training (BPT) programs, such as parent–child interac-
tion therapy (PCIT), can disrupt this negative trajectory
(Eyberg, Nelson, & Boggs, 2008), not all families engage
in and benefit equally; some take longer to complete treat-
ment and others ultimately drop out (Holden, Lavigne, &
Cameron, 1990). Attrition rates for general outpatient child
mental health services range from 40% to 60% (Kazdin,
2008), whereas rates for PCIT range from 27% to 67%
(Boggs et al., 2005; Pearl et al., 2012). Successfully com-
pleting treatment by mastering the targeted parenting skills
is necessary for families to fully benefit from PCIT (Boggs
et al., 2005). Thus, to increase the number of families that
benefit from the treatment, it is critical to identify factors
that lead to improved engagement.
Most research on treatment engagement in BPT focuses
on parent characteristics. Low socioeconomic status, ethnic
minority status, high maternal stress, and maternal criticisms
of their children have been linked to attrition (Fernandez &
Eyberg, 2009; Holden et al., 1990; Werba, Eyberg, Boggs,
& Algina, 2006). Other factors have been related to the
parents’rate of skill acquisition. For example, parents
from ethnic minority backgrounds typically require more
sessions and skill practice to complete treatment success-
fully (Lau, 2012; McCabe & Yeh, 2009). Exploring client
Correspondence should be addressed to Larissa N. Niec, Center for
Children, Families and Communities, 2480 W. Campus Drive, B100, Central
Michigan University, Mount Pleasant, MI 48858. E-mail: niec1L@cmich.edu
Journal of Clinical Child & Adolescent Psychology,1–8, 2015
Copyright © Taylor & Francis Group, LLC
ISSN: 1537-4416 print/1537-4424 online
DOI: 10.1080/15374416.2015.1063428
Downloaded by [University of California, Los Angeles (UCLA)] at 12:31 23 October 2015

characteristics is insufficient to improve treatment engage-
ment, however. It is also necessary to understand how
therapist behaviors influence parents. For example, thera-
pists’use of directive techniques has been associated with
greater parent resistance (Patterson & Forgatch, 1985), and
therapist communication style in PCIT has predicted treat-
ment completion, with lower attrition linked to therapists
using more facilitative statements and fewer closed-ended
questions and supportive statements (Harwood & Eyberg,
2004).
PCIT has demonstrated efficacy in treating young children
with conduct problems (e.g., Boggs et al., 2005). The treat-
ment model emphasizes the parent–child relationship and
uses in vivo coaching to develop parent skills, which are
features associated with larger effect sizes in BPT
(Kaminski, Valle, Boyle, & Filene, 2008). PCIT is conducted
in two phases, a parent–child relationship-enhancement
phase (Child-Directed Interaction [CDI]) and a discipline-
focused phase (Parent-Directed Interaction). During CDI,
parents are taught to increase their positive, child-centered
verbalizations (praises, reflections, and descriptions of child
behavior) while decreasing their demanding and leading ver-
balizations (questions, commands, and criticisms) as a way to
reinforce their children’s positive behaviors and improve the
parent–child relationship. Parents complete the CDI phase
when they demonstrate mastery of the child-centered skills as
measured during a weekly 5-min behavior observation,
which typically occurs after five to six sessions (Harwood
& Eyberg, 2006). Parents then learn developmentally appro-
priate discipline during Parent-Directed Interaction, includ-
ing how to give effective commands and provide consistent
follow-through for child compliance and noncompliance.
Reductions in parenting stress and dysfunctional parenting
practices often occur during CDI (Harwood & Eyberg, 2006),
but families need to complete both phases of treatment for
clinically significant changes in children’s conduct problems
to occur and persist (Boggs et al., 2005).
Therapists’coaching of parents, a core component of
PCIT, has been identified as a strategy that leads to efficient
acquisition of parenting skills (Shanley & Niec, 2010) and
improved treatment outcomes (Kaminski, Valle, Filene, &
Boyle, 2008). Coaching techniques have been classified as
being directive or responsive (Borrego & Urquiza, 1998).
Directive techniques tell the parent what to do (e.g., ‘‘ Praise
her for sharing’’ ), whereas responsive techniques reinforce
the parent’s use of a skill (e.g., ‘‘Excellent labeled praise’’).
A third technique, constructive criticism, corrects the par-
ent’s behavior (e.g., ‘‘Don’t ask questions’’ ;Herschell,
Capage, Bahl, & McNeil, 2008). PCIT experts recommend
that therapists utilize frequent positive and responsive tech-
niques in order to shape parents’behaviors. In a clinical
sample, responsive coaching was associated with parents’
behavior change between two subsequent PCIT sessions,
whereas directive coaching was not related (Barnett, Niec,
& Acevedo-Polakovich, 2013).
Investigations of the relation between therapists’in vivo
feedback techniques and treatment outcomes lag behind other
areas of PCIT research; the need for this research is ‘‘ impera-
tive’’ (McNeil & Hembree-Kigin, 2010, p. 20). Therapist
behaviors during the noncoaching portions of initial PCIT
sessions predict dropout (Harwood & Eyberg, 2004), but
limited research has investigated how therapist behaviors
during the core component of PCIT—coaching—impact par-
ents’engagement. To address this gap in the empirical litera-
ture, we used a measure of therapist coaching, the Therapist–
Parent Interaction Coding System (TPICS; Barnett, Niec, &
Peer, 2013), to explore an important question: How does
coaching influence parental engagement in PCIT?
Specifically, we investigated how coaching techniques influ-
ence (a) attrition from PCIT and (b) parents’speed of skill
acquisition during the first phase of PCIT.
Given the relation between directive therapist behavior
and parental resistance (Patterson & Forgatch, 1985) and
the relation between responsive coaching and parent behavior
change (Barnett, Niec, & Acevedo-Polakovich, 2013), we
predicted that therapist coaching techniques during the first
CDI coaching session (CDI 1) would discriminate between
parents who successfully completed treatment and those who
discontinued prematurely. Specifically, we predicted that less
directive coaching and more responsive coaching would sig-
nificantly predict treatment completion, over and above par-
ent stress and parental criticisms of their children at intake,
which have been associated with PCIT attrition (Werba et al.,
2006). We further hypothesized that responsive coaching in
CDI 1 would predict parents’speed of acquisition of the
child-centered skills, over and above the contribution of
parent stress and child conduct problems at intake.
METHOD
Participants
Parent–child dyads. Data for this study were
archival from a clinical evaluation of PCIT. Participants
included 51 parent–child dyads (41 mothers, 10 fathers).
To avoid problems with nesting, only one caregiver was
included in the analyses if both caregivers participated in
treatment. Fathers were included if they were the primary
caregiver or if the mother was not coached for at least 10
min in CDI 1. Families were referred for treatment if their
primary presenting concern was distress regarding their
children’s conduct and/or if the family was designated by
a child welfare agency as being at risk for physical abuse.
For study inclusion, parent and child had to speak and
understand English, and children could not have
significant impairments in receptive language (Standard
Score ≤70) on the Peabody Picture Vocabulary Test,
Fourth Edition (Dunn & Dunn, 2007). Seventy-three
2BARNETT ET AL.
Downloaded by [University of California, Los Angeles (UCLA)] at 12:31 23 October 2015

percent of families reported that they were bilingual
(English/Spanish).
Children included in this study were 2 to 9 years old (M=
5.03, SD =1.65). Children older than 7 (n=4) were included if
the family was referred for risk of child physical abuse, as
PCIT has been identified as an efficacious intervention for
abusive parents with children in this age range (Chaffinetal.,
2004). On average, child behavior problems were rated by
parents as clinically significant at intake (Eyberg Child
Behavior Inventory [ECBI]: M=150.20, SD = 29.85). The
majority of children were male (77%) and parent-identified as
White Hispanic (63%). The sample came from an urban center
with a heterogeneous Hispanic population primarily from the
Caribbean, South America, and Central America.
Therapists
Therapists in this study included three postdoctoral fellows,
seven predoctoral interns, and six practicum students who
were enrolled in doctoral clinical psychology programs.
Therapists were predominately female (n=15) and non-
Hispanic White (69%; Asian: n= 2, Hispanic: n= 4). A
licensed clinical psychologist and certified PCIT trainer
provided training and supervision. All predoctoral interns
and practicum students conducted cotherapy with a post-
doctoral fellow. If cotherapy occurred, the therapist who
provided coaching to the parent during CDI 1 was included
in the analyses.
Measures
Demographics form. Parents completed demographic
information (e.g., race, ethnicity, and ages) about themselves
and their children.
Treatment attrition. Attrition was defined as occurring
when participants discontinued treatment after completing at
least one session of CDI coaching and before meeting criteria
for successful treatment graduation as defined by the PCIT
protocol (Eyberg & Funderburk, 2011).
Speed of parenting skill acquisition. The speed
with which parents acquired the CDI skills was defined as
the number of sessions it took for caregivers to meet the
mastery criteria as defined by the current PCIT protocol
(i.e., at least 10 Labeled Praises, 10 Behavior Descriptions,
and 10 Reflections, and fewer than three combined
Questions, Commands, and Negative Talks during the
weekly 5-min behavior observation of parent–child
interactions; Eyberg & Funderburk, 2011).
Therapist–Parent Interaction Coding System. The
TPICS (Barnett, Niec, & Peer, 2013) is a behavioral
observation coding system that assesses therapists’coaching
statements and categorizes them as directive, responsive, or
critical (Tabl e 1 ). The TPICS has been found to have excellent
TABLE 1
Therapist–Parent Interaction Coding System Therapist Categories
Coaching Technique Definition Example
Directive Techniques
Modeling Therapist verbalization of a parent–child interaction therapy parenting skill. “Thank you for sharing with me.”
Prompting Therapist verbalization of the beginning of an appropriate skill intended to
allow a parent to finish the statement.
“Thank you for …”
Direct Command Therapist declarative statement that contains a direction for a behavior to be
performed by the parent.
“Praise him for sharing.”
Indirect Command Therapist suggestion for a behavior to be performed by the parent. “Can you think of something to praise him
for?”
Drill An exercise during which the therapist tells the parent to focus on one targeted
parenting behavior for a specified duration or frequency.
“We are going to see how many labeled praises
you can give him in one minute.”
Child Observation Any therapist observation about the child that is used to draw the parent’s
attention to the child.
“He just shared with you.”
Responsive Techniques
Labeled Praise Therapist verbalization that provides a positive evaluation of a specific
behavior, activity, product, or verbalization of the parent.
“That was a great labeled praise you gave him.”
Process Comment Therapist statement that ties a child’s behavior to the parent’s treatment-related
behavior.
“She shared again because you praised her for
that.”
Reflective Description Therapist non-evaluative, declarative sentence or phrase about the parent’s
most recent verbalization or behavior.
“That was an unlabeled praise.”
Unlabeled Praise Therapist verbalization that provides nonspecific positive evaluation of the
parent or parent’s behavior.
“Excellent job!”
Corrective Criticism Therapist verbalization that is negatively stated or gently critical of a parent’s
behaviors.
“Stop paying attention to that behavior.”
Note: From Barnett, Niec, and Peer (2013).
SUCCESSFUL THERAPIST–PARENT COACHING 3
Downloaded by [University of California, Los Angeles (UCLA)] at 12:31 23 October 2015

reliability and can predict parents’skill acquisition from one
session to the next (Barnett, Niec, & Acevedo-Polakovich,
2013). In this study, approximately 25% of video-recorded
coaching sessions were coded for interrater reliability, which
was good to excellent for the majority of behavioral codes
(intraclass correlation coefficient [ICC] =.74–1.00), and fair
for two codes (Tabl e 2).
Dyadic Parent–Child Interaction Coding System-
III. The Dyadic Parent–Child Interaction Coding System-
III (DPICS-III; Eyberg, Nelson, Duke, & Boggs, 2005)isa
behavioral observation coding system that was designed to
assess the quality of parent–child interactions. The measure
has good interrater reliability (Eyberg et al., 2005)and
treatment sensitivity (e.g., McCabe & Yeh, 2009). Consistent
with the PCIT protocol, this study used the following DPICS-
III categories—Behavior Description, Labeled Praise,
Unlabeled Praise, Reflection, Question, Negative Talk,
Indirect Command, and Direct Command—to measure
parents’skill use in CDI 1 and determine when parents met
the CDI mastery criteria. We used parents’frequency of
Negative Talk (i.e., criticism) at intake to predict dropout,
based on previous findings that it relates to attrition. Of the
51 video-recorded intakes and CDI 1 sessions, 13 (≈25%) of
each session type were randomly selected for reliability
coding. Interrater reliability was excellent (ICC =.88) for
Negative Talk at intake, good to excellent for the majority of
codes in CDI 1 (ICC =.70–.93), and poor for Negative Talk in
CDI 1 (ICC =.26; Tabl e 3).
Eyberg Child Behavior Inventory. The ECBI
(Eyberg & Pincus, 1999) is a 36-item parent-rating scale
of conduct problems for children between the ages of 2 to
16. Parents rate the frequency of each disruptive behavior
on a 7-point Likert scale from 1 (never)to7(always), which
are summed to yield the Intensity Scale. The ECBI is
sensitive to treatment effects (McCabe & Yeh, 2009) and
has excellent internal reliability (α=.92–.95; Burns &
Patterson, 2001). The ECBI was used to control for
children’s conduct problems.
Parenting Stress Index–Short Form. The PSI-SF
(Abidin, 1995) is a 36-item self-report measure of
parenting stress, with acceptable internal consistency (α
= .74–.88) and test–retest reliability (Haskett, Ahern, Ward,
& Allaire, 2006). The PSI-SF was used to control for
parenting stress.
Procedure
Treatment. Parents first completed a phone call with
clinic staff and a screening packet. If they did not rate their
child’s behaviors as clinically significant on the ECBI, they
were offered less intensive services but were allowed to
receive PCIT if they felt it was the best match for their
family. Treatment was conducted in one clinic on a medical
campus and was grant-funded. Prior to participation, adult
participants signed a written informed consent, and children
7 years and older signed a written informed assent, both of
which had been approved by the University of Miami
Miller School of Medicine Institutional Review Board.
Treatment followed the evidence-based PCIT protocol
(Eyberg & Funderburk, 2011). Therapists promoted family
engagement through the use of motivational interviewing
strategies, support in addressing treatment barriers,
flexibility in rescheduling missed sessions, and phone
consultations to promote skill acquisition.
Coder training. Coders included graduate and
undergraduate research assistants. Training included
tutorial and discussion of the DPICS-III and TPICS codes
and practice coding transcripts and video-recorded sessions.
Prior to beginning coding for the study, the coders exceeded
the criterion of k≥.80 with the expert rating tapes for both
the DPICS-III and TPICS. Biweekly coding meetings were
completed to reduce rater drift.
Video-recorded session samples. All sessions
were video-recorded. Of the 113 families who were
enrolled in treatment, 51 families were included in the
current study. The additional 62 cases had problems with
video-recordings (e.g., sessions lost due to a corrupted
computer drive). Institutional Review Board approval was
TABLE 2
Interrater Reliability of the Therapist–Parent Interaction Coding
System
Coaching Techniques
Directive Coaching Responsive Coaching
MO PR DC IC DR CO LP RD PC UP CC
ICC 1.00 .96 .97 .98 .93 .65 .98 .92 .77 .74 .51
Note: MO = Modeling; PR = Prompt; DC = Direct Command; IC =
Indirect Command; DR = Drill; CO = Child Observation; LP = Labeled
Praise; RD = Reflective Description; PC = Process Comment; UP =
Unlabeled Praise; CC = Constructive Criticism; ICC = intraclass correlation
coefficient.
TABLE 3
Interrater Reliability of the Dyadic Parent–Child Interaction Coding
System for CDI 1
BD LP UP RF QU CM NTA
ICC .91 .93 .70 .84 .73 .82 .26
Note: CDI 1 = Child-Directed Interaction first coaching session; BD =
Behavior Description; LP = Labeled Praise; UP = Unlabeled Praise; RF =
Reflection; QU = Question; CM = Command; NTA = Negative Talk; ICC =
intraclass correlation coefficient.
4BARNETT ET AL.
Downloaded by [University of California, Los Angeles (UCLA)] at 12:31 23 October 2015

obtained prior to accessing the video-recorded sessions for
the current study. The first 10 min of coaching of the first
CDI coaching session were coded.
Data Analysis
Treatment attrition. A general linear model that
controlled for nesting at the therapist level was used to
determine if significant differences existed between
therapist coaching techniques in CDI 1 for treatment
completers and dropouts. Predictors of attrition in earlier
PCIT studies (parenting stress and the number of parent-to-
child criticisms at intake) were examined for differences
between groups to determine if these variables would be
included in the model (Werba et al., 2006). Discriminant
function analysis (DFA), which can detect group differences
in small samples, was used to determine whether group
membership (i.e., successful completer or dropout) could be
predicted reliably from the set of hypothesized variables. The
eight dropouts were distributed among seven therapists;
therefore, therapist was not controlled in the DFA.
Speed of parents’skill acquisition. A mixed-effects
model was used to evaluate the relation between child
conduct problems, parental stress, therapist use of
responsive coaching in CDI 1 and the length of the CDI
phase of treatment. Parental stress (PSI-SF Total Stress),
child conduct problems (ECBI Intensity Scale), and
therapist responsive coaching techniques were included as
fixed effects. To control for nesting, the therapist was
entered as a random effect. The random effect (i.e.,
therapist) estimate was not significant (Estimate =1.26,
SE = 1.33, p=34), suggesting that nesting was not of
concern; therefore, a linear regression was conducted to
evaluate the influence of coaching techniques on parents’
rate of skill acquisition.
RESULTS
Who Completed Treatment?
Parents completed treatment in an average of 13 sessions (M=
13.38, SD = 4.14). Of the 51 parent–child dyads, 43 success-
fully completed treatment and eight dropped out. No signifi-
cant differences existed between dropouts and completers on
the demographic variables, parent stress, use of criticisms at
intake, or level of children’s conduct problems (Tabl e 4).
Parents who completed treatment had somewhat higher skills
at CDI 1 than those who dropped out, using significantly more
Behavior Descriptions and fewer Questions (Table 5).
Does Therapist Coaching Influence Family Retention?
As predicted, therapists provided significantly more respon-
sive coaching statements for treatment completers than
dropouts. The overall level of directive coaching did not
differ across groups, but the technique drill approached
significance in the hypothesized direction, and thus was
also included in the DFA (Table 6). Treatment completers
received fewer drills than dropouts.
There was a significant violation of the homogeneity of
variance–covariance assumption for responsive coaching and
drills (Box’sM=78.27, F=23.30, p< .001), which was
likely influenced by the discrepancy between the two groups’
TABLE 4
Parent and Child Variables for Treatment Completers and Dropouts
Completers Dropouts
% or M n or SD % or M n or SD t or χ
2
p
Child Age 5.07 1.69 4.83 1.47 t(49) =.37 .63
Child Gender χ
2
(1) = 2.31 .13
Boys 77% 33 100% 8
Girls 23% 10 0% 0
Child Ethnicity (Hispanic) 65% 28 75% 6 χ
2
(1) =.30 .59
Child Race χ
2
(3) = 1.27 .74
White 77% 33 75% 6
Black 14% 6 25% 2
Other 9% 4 0% 0
Caregiver Age 38.08 7.13 38.84 8.07 t(49) =−.27 .79
Caregiver Gender χ
2
(1) = 1.77 .18
Female 81% 35 100% 8
Male 19% 8 0% 0
ECBI Intensity Scale 148.81 31.29 157.63 20.39 t(49) = −.76 .45
PSI-SF Total Stress 94.90 29.56 91.38 28.24 t(45) =.31 .76
Criticisms 9.97 9.36 13.25 15.27 t(45) =−.80 .43
Note: Completers N= 43; Dropouts N= 8. ECBI = Eyberg Child Behavior Inventory; PSI-SF = Parent Stress Inventory Short Form.
SUCCESSFUL THERAPIST–PARENT COACHING 5
Downloaded by [University of California, Los Angeles (UCLA)] at 12:31 23 October 2015

sizes. Results should be interpreted with some caution due to
this violation. The linear composite of these therapist beha-
viors predicted group membership, Wilks’sλ= .84, χ
2
=8.64,
p< .05. Analysis of the structure matrix revealed that drills
(.93) were a stronger predictor of attrition than responsive
coaching (−.49). These therapist and parent behaviors com-
bined correctly classified actual group membership for 86%
of families. All 43 families were correctly classified as treat-
ment completers; however, only one out of eight families was
correctly classified as treatment dropouts.
Does Therapist Coaching Influence Parents’Speed of
Skill Acquisition?
Parents met CDI mastery criteria in an average of five to six
coaching sessions (M= 5.67, SD =2.51). Child conduct
problems and parenting stress were not included in the
linear regression, as they did not significantly correlate to
the length of the CDI phase. As predicted, therapists’
responsive coaching during the first 10 min of CDI 1 was
found to significantly predict the speed with which parents’
acquired the child-centered skills, β=−.30, t(43) =−2.04, p
< .05, with higher levels of responsive coaching related to
quicker mastery of the skills. Directive, β=−.10, t(43) =
−.66, p> .05, and critical coaching techniques, β=−.05, t
(43) =−.31, p> .05, did not significantly predict parents’
speed of mastery.
DISCUSSION
To improve families’engagement in PCIT, it is impor-
tant to understand the role of therapists’coaching. This
study sought to investigate how coaching techniques
influence parents’treatment retention and rate of skill
acquisition, which are important markers of treatment
engagement (Holden et al., 1990). Findings suggest
that coaching, specifically more responsive techniques
and fewer drills, was associated with treatment comple-
tion. However, although all treatment completers were
accurately predicted, only one family who dropped out
was correctly predicted. As hypothesized, higher rates of
responsive coaching predicted quicker mastery of the
child-centered interaction skills, whereas directive and
critical coaching techniques did not. It is notable that
the first 10 min of therapists’coaching significantly
influenced parents’mastery of the child-directed interac-
tion skills, which typically occurred more than 1 month
later.
Parents in this sample demonstrated high levels of treatment
engagement. Treatment attrition was only 16%, lower than has
previously been reported (e.g., Fernandez & Eyberg, 2009),
and on average, child-centered skill mastery occurred after five
sessions, similar to efficacy trials of PCIT (Harwood &
Eyberg, 2006). This is especially significant because families
were predominately from ethnic minority backgrounds, which
has been associated with higher rates of premature termination
(Holden et al., 1990) and slower skill acquisition (McCabe &
Yeh, 2009). In addition, parents and therapists rarely matched
in race or ethnicity, which can negatively impact treatment
outcomes (Halliday-Boykins, Schoenwald, & Letourneau,
2005). It is still unclear what cultural adaptations may need
to be made to improve BPT outcomes and engagement for
minority families, but behavioral rehearsal seems especially
important (Lau, 2012). An important step to increase minority
parents’engagement in BPT may be identifying therapist
behaviors, such as responsive coaching, that promote skill
acquisition during behavioral rehearsal.
Although the high retention rate in this study is a strength
of the implementation of the intervention, it limited our ability
to predict attrition. Therefore, replication is warranted, and
TABLE 5
Parenting Behaviors in CDI 1 for Treatment Completers and
Dropouts
Completers Dropouts
M SD M SD t(49) p d
Behavior Description 3.90 4.19 .71 1.11 2.17 .04 1.04
Reflection 6.51 6.69 5.14 5.82 .56 .58 .22
Labeled Praise 3.76 4.14 .86 2.29 1.97 .06 .87
Unlabeled Praise 4.73 3.98 2.86 2.27 1.32 .19 .58
Question 9.88 7.14 17.43 12.61 −2.51 .02 −.74
Command 4.41 4.46 4.71 3.99 −.18 .86 −.07
Negative Talk .47 1.14 .99 1.68 −1.69 .10 −.33
Note: Completers N= 43; Dropouts N= 8. CDI 1 = Child-Directed
Interaction first coaching session; d= Cohen’sd.
TABLE 6
Coaching Techniques in CDI 1 for Treatment Completers and
Dropouts
Completers Dropouts
M SD M SD F(1, 38) p η2
p
Total Directive 51.56 26.98 68.86 24.41 2.73 .11 .07
Modeling 21.98 17.94 36.43 20.20 1.84 .18 .05
Prompting 2.70 3.86 1.57 1.62 0.15 .70 .00
Indirect Command 13.00 10.04 14.43 11.77 0.12 .73 .00
Direct Command 9.18 6.52 11.14 11.14 3.33 .08 .08
Child Observation 4.68 4.58 4.71 3.77 0.01 .94 .00
Drills 0.02 0.15 1.57 1.13 3.82 .06 .09
Total Responsive 65.58 19.18 54.38 17.47 4.57 .04 .11
Labeled Praises 44.36 16.17 37.86 14.92 3.36 .08 .08
Process Comments 2.91 2.43 2.00 2.00 0.51 .48 .01
Reflective Descriptions 3.64 3.58 5.29 3.95 0.27 .61 .01
Unlabeled Praises 14.32 8.03 9.86 6.36 1.73 .20 .04
Corrective Criticism 1.39 2.41 2.43 3.95 0.57 .46 .02
Note: Completers N= 43. Dropouts N= 8. Estimated marginal means
used in analyses, with unadjusted means reported for clarity. CDI 1 = Child-
Directed Interaction first coaching session; η2
p= partial = partial eta-squared.
6BARNETT ET AL.
Downloaded by [University of California, Los Angeles (UCLA)] at 12:31 23 October 2015

recommendations based on our findings regarding optimal
coaching techniques and the training of PCIT therapists should
be made with caution. Although the aim of this study was not
to investigate system-level factors that supported family reten-
tion, it may be that the engagement strategies of the clinical
team (e.g., flexible scheduling, motivation enhancement state-
ments) positively impacted retention. These strategies were
similar to those implemented in other evaluations of PCIT
with higher attrition (e.g., Niec, Barnett, Prewett, & Shanley,
2015) but deserve additional attention in the future.
This study has begun to address how therapist coaching
impacts parental engagement, but further work is needed to
better understand what is likely to be a bidirectional influ-
ence between therapists’and parents’behaviors. Previous
research suggests that parental resistance might lead thera-
pists to become more directive or confrontational, which in
turn can lead to more resistance from the parent and even-
tual dropout (Patterson & Forgatch, 1985). In PCIT, parents
who independently generate fewer skills or who are resistant
to therapist coaching may elicit more directive statements
and corrective criticisms and fewer responsive statements
from therapists. In our sample, treatment completers began
CDI 1 with significantly higher skills in two parenting
behaviors (behavior descriptions, questions). It is possible
that this difference impacted therapist behaviors, which in
turn may have impacted the parents’engagement. Finally,
coaching in CDI 1 is somewhat different from later coach-
ing sessions in that coaches are recommended to avoid
correcting parents to keep the experience positive (Eyberg
& Funderburk, 2011). Sequential coding of therapist and
parent behaviors in PCIT coaching across sessions would
help to illuminate how different coaching techniques impact
parental engagement in the moment and over time.
Implications
Within the first 10 min of the first PCIT coaching session,
significant differences in therapists’in vivo feedback techni-
ques existed between families who completed treatment and
those who did not, and also predicted the speed with which
parents acquired the child-centered interaction skills. Although
it is striking that such a short segment of early therapist–parent
interactions can have a significant association with engage-
ment, these findings are consistent with past research of non-
coaching interactions. Therapist verbalizations in 30 min of
early PCIT sessions correctly predicted family attrition
(Harwood & Eyberg, 2004). It appears that the therapist–parent
interactions arising in the earliest moments of treatment may be
important indicators of the ultimate success of treatment.
This study further supports the TPICS as a valuable psy-
chometric tool that can be used both to investigate important
therapeutic processes and to guide PCIT training. To provide
the most effective training for therapists, it is valuable to have
the ability to (a) understand the empirical links between
coaching techniques and treatment outcome, (b) reliably
identify the coaching techniques used by PCIT trainees, and
(c) evaluate changes in trainees’coaching over time.
Consistent with previous research (Barnett, Niec, &
Acevedo-Polakovich, 2013) and expert recommendations
(Eyberg & Funderburk, 2011), our findings suggest that
responsive coaching is a critical skill for therapists to master.
Although directive techniques are likely also valuable teach-
ing tools for parents, our study suggests that these techniques
may need to be balanced with a high level of responsiveness
to maintain parent engagement. The identification of therapist
behaviors that promote parents’skill acquisition is important,
as it may help to move families more quickly through treat-
ment. Future studies investigating the utility of the TPICS as a
therapist training tool may also play a part in improving
treatment outcomes for families in community settings.
FUNDING
Funding for this project was provided by The Children’s
Trust.
REFERENCES
Abidin, R. R. (1995). Parenting stress index: Professional manual (3rd
ed.). Odessa, FL: Psychological Assessment Resources.
Barnett, M. L., Niec, L. N., & Acevedo-Polakovich, I. D. (2013). Assessing
the key to effective coaching in parent–child interaction therapy: The
therapist–parent interaction coding system. Journal of Psychopathology
and Behavioral Assessment,36,211–223. doi:10.1007/s10862-013-
9396-8
Barnett, M., Niec, L., & Peer, S. (2013). The therapist-parent interaction
coding system: Child directed interaction component (Unpublished man-
ual). Mount Pleasant, MI: Central Michigan University.
Boggs, S. R., Eyberg, S. M., Edwards, D. L., Rayfield, A., Jacobs, J.,
Bagner, D., & Hood, K. K. (2005). Outcomes of parent–child interaction
therapy: A comparison of treatment completers and study dropouts one
to three years later. Child & Family Behavior Therapy,26,1–22.
doi:10.1300/j019v26n04_01
Borrego, J., & Urquiza, A. J. (1998). Importance of therapist use of social
reinforcement with parents as a model for parent–child relationships: An
example with parent–child interaction therapy. Child & Family Behavior
Therapy,20,27–54. doi:10.1300/j019v20n04_03
Burns, G. L., & Patterson, D. R. (2001). Normative data on the Eyberg
child behavior inventory and Sutter-Eyberg student behavior inventory:
Parent and teacher rating scales of disruptive behavior problems in
children and adolescents. Child & Family Behavior Therapy,23,15–
28. doi:10.1300/J019v23n01_02
Chaffin, M., Silovsky, J. F., Funderburk, B., Valle, L. A., Brestan, E. V.,
Balachova, T., …Bonner, B. L. (2004). Parent-child interaction therapy
with physically abusive parents: Efficacy for reducing future abuse
reports. Journal of Consulting and Clinical Psychology,72, 500–510.
doi:10.1037/0022-006X.72.3.500
Dunn, L. M., & Dunn, L. M. (2007). Examiner’s manual for the Peabody
picture vocabulary test fourth edition. Circle Pines, MN: American
Guidance Service.
Eyberg, S. M., & Funderburk, B. (2011). Parent–child interaction therapy:
Treatment manual. Gainesville, FL: PCIT International.
SUCCESSFUL THERAPIST–PARENT COACHING 7
Downloaded by [University of California, Los Angeles (UCLA)] at 12:31 23 October 2015

Eyberg, S. M., Nelson, M. M., & Boggs, S. R. (2008). Evidence-based
psychosocial treatments for children and adolescents with disruptive
behavior. Journal of Clinical Child and Adolescent Psychology,37,
215–237. doi:10.1080/15374410701820117
Eyberg, S. M., Nelson, M. M., Duke, M., & Boggs, S. R. (2005). Manual
for the dyadic parent-child interaction coding system (3rd ed.).
(Unpublished manuscript). Gainesville, FL: University of Florida.
Eyberg, S. M., & Pincus, D. (1999). Eyberg child behavior inventory and
Sutter-Eyberg student behavior inventory-Revised professional manual.
Odessa, FL: Psychological Assessment Resources.
Fernandez, M. A., & Eyberg, S. M. (2009). Predicting treatment and
follow-up attrition in parent–child interaction therapy. Journal of
Abnormal Child Psychology,37, 431–441. doi:10.1007/s10802-008-
9281-1
Halliday-Boykins, C., Schoenwald, S. K., & Letourneau, E. J. (2005).
Caregiver-therapist ethnic similarity predicts youth outcomes from an
empirically based treatment. Journal of Consulting and Clinical
Psychology,73, 808–818. doi:10.1037/0022-006X.73.5.808
Harwood, M. D., & Eyberg, S. M. (2004). Therapist verbal behavior early
in treatment: Relation to successful completion of parent–child interac-
tion therapy. Journal of Clinical Child & Adolescent Psychology,33,
601–612. doi:10.1207/s15374424jccp3303_17
Harwood, M. D., & Eyberg, S. M. (2006). Child-directed interaction:
Prediction of change in impaired mother-child functioning. Journal of
Abnormal Child Psychology,34, 335–347. doi:10.1007/s10802-006-
9025-z
Haskett, M. E., Ahern, L. S., Ward, C. S., & Allaire, J. C. (2006). Factor
structure and validity of the parenting stress index-short form. Journal of
Clinical Child & Adolescent Psychology,35, 302–312. doi:10.1207/
s15374424jccp3502_14
Herschell, A. D., Capage, L. C., Bahl, A. B., & McNeil, C.B. (2008). The role
of therapistcommunication style in parent-child interaction therapy. Child &
Family Behavior Therapy,30,13–35. doi:10.1300/j019v30n01_02
Holden, G. W., Lavigne, V. V., & Cameron, A. M. (1990). Probing the
continuum of effectiveness in parent training: Characteristics of parents
and preschoolers. Journal of Clinical Child Psychology,19,2–8.
doi:10.1207/s15374424jccp1901_1
Kaminski, J. W., Valle, L. A., Filene, J. H., & Boyle, C. L. (2008). A meta-
analytic review of components associated with parent training program
effectiveness. Journal of Abnormal Child Psychology,36, 567–589.
doi:10.1007/s10802-007-9201-9
Kazdin, A. E. (2008). Evidence-based treatments and delivery of psycho-
logical services: Shifting our emphases to increase impact. Psychological
Services,5, 201–215. doi:10.1037/a0012573
Lau, A. S. (2012). Reflections on adapting parent training for Chinese
immigrants: Blind alleys, thoroughfares, and test drives. In G. Bernal
& M. M. Domenech-Rodriguez (Eds.), Cultural adaptations: Tools for
evidence-based practice with diverse populations (pp. 133–156).
Washington, DC: American Psychological Association.
McCabe, K., & Yeh, M. (2009). Parent–child interaction therapy for
Mexican Americans: A randomized clinical trial. Journal of Clinical
Child & Adolescent Psychology,38, 753–759. doi:10.1080/
15374410903103544
McNeil, C. B., & Hembree-Kigin, T. (2010). Parent-child interaction
therapy (2nd ed.). New York, NY: Springer Science & Business Media.
Niec, L. N., Barnett, M., Prewett, M., & Shanley, J. (2015). Group parent-
child interaction therapy: A randomized control trial for the treatment of
conduct problems in young children. Unpublished manuscript.
Patterson, G. R., & Forgatch, M. S. (1985). Therapist behavior as a
determinant for client noncompliance: A paradox for the behavior modi-
fier. Journal of Consulting and Clinical Psychology,53, 846–851.
doi:10.1037//0022-006x.53.6.846
Pearl, E., Thieken, L., Olafson, E., Boat, B., Connelly, L., Barnes, J., &
Putnam, F. (2012). Effectiveness of community dissemination of parent–
child interaction therapy. Psychological Trauma: Theory, Research,
Practice, and Policy,4, 204–213. doi:10.1037/a0022948
Reef, J., Diamantopoulou, S., van Meurs, I., Verhulst, F. C., & van der
Ende, J. (2011). Developmental trajectories of child to adolescent exter-
nalizing behavior and adult DSM-IV disorder: Results of a 24-year
longitudinal study. Social Psychiatry and Psychiatric Epidemiology,46,
1233–1241. doi:10.1007/s00127-010-0297-9
Shanley, J. R., & Niec, L. N. (2010). Coaching parents to change: The
impact of in vivo feedback on parent’s acquisition of skills. Journal of
Clinical Child & Adolescent Psychology,39, 282–287. doi:10.1080/
15374410903532627
Werba, B. E., Eyberg, S. M., Boggs, S. R., & Algina, J. (2006). Predicting
outcome in parent-child interaction therapy: Success and attrition.
Behavior Modification,30, 618–646. doi:10.1177/0145445504272977
8BARNETT ET AL.
Downloaded by [University of California, Los Angeles (UCLA)] at 12:31 23 October 2015