Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2022 Jan 1.
Published in final edited form as: Behav Ther. 2020 Feb 4;52(1):28–38. doi: 10.1016/j.beth.2020.01.009

Keeping an “I” on PRIDE: Measuring Imitation in Parent-Child Interaction Therapy

Kristen M McCabe a,b, McKenna Sakamoto a, Yessica Green Rosas a, Kelsey Kehoe a, Raymond La c, Argero Zerr a,d, May Yeh b,c,e
PMCID: PMC7826448  NIHMSID: NIHMS1568980  PMID: 33483122

Abstract

In Parent-Child Interaction Therapy (PCIT), therapists encourage parents to imitate child behaviors in order to convey approval of the child’s actions and promote the development of linguistic and social-cognitive skills. However, the Dyadic Parent-Child Interaction Coding System (DPICS-IV), used to measure skills taught during PCIT, does not include guidelines for coding parent-child imitation, making it difficult to determine how PCIT affects it. The current study addresses this problem by developing guidelines for coding imitation, which were then used to code DPICS-IV segments from 58 Mexican American families who participated in a past clinical trial. Results suggest that these coding guidelines can be used to reliably measure parent and child imitation. A series of additional analyses supported the construct validity of the codes. Specifically, there was a trend for parent imitation, but not child imitation, to increase more from pre-post treatment in PCIT relative to treatment as usual. In addition, parents who imitate their children were found to have children who imitate them more in return. Finally, improvement in parent imitation, but not child imitation, was significantly related to decrease in child behavior problems. Further study is needed to determine the optimal frequency of imitation, and findings suggest that additional attention to coaching imitation may be warranted.

Keywords: Parent-Child Interaction Therapy, Child Disruptive Behavior, Parent-Child Imitation, Dyadic Parent-Child Interaction Coding System (DPICS)

Parent-Child Interaction Therapy (PCIT) is an evidence-supported treatment for young children that has been found to significantly reduce externalizing symptoms across a number of randomized controlled trials and meta-analyses (e.g., Eyberg & Funderburk, 2011; Ward, Theule, & Cheung, 2016). In PCIT, parents learn to build warm relationships with their children while setting and consistently enforcing firm limits. PCIT therapists teach parents to use the “PRIDE” skills: (labeled) Praises, Reflective statements, Imitation, Descriptions of child behavior, and Enjoyment, to positively reinforce desirable child behavior with parental attention. At the same time, parents are instructed to eliminate questions, commands, and criticism, while ignoring minor misbehavior. When parents use PRIDE skills to reinforce positive behaviors while reducing attention to negative behaviors, it results in significantly less oppositional and more prosocial child behavior, and lays the foundation for the discipline phase of the treatment program (McNeil & Hembree-Kigin, 2010).

Parental imitation of children’s actions (P-IM), the “I” in PRIDE, is one of the primary skills parents are taught to strategically reinforce positive child behavior in PCIT. McNeil and Hembree-Kigin (2010) suggest that P-IM communicates parental approval of the child’s behavior, making it more likely that the child will repeat the behavior, while boosting child self-esteem. Furthermore, children who are imitated have been found to imitate others more frequently (Kauffman, Kneedler, Gamache, Hallahan, & Ball, 1977; Smeets & Kauffman, 1980). For example, Eckerman and Stein (1990) found that toddlers who were imitated were more likely to repeat the action that was imitated and to generate imitation games compared with toddlers who were not imitated. Thus, in addition to positively reinforcing child behavior, P-IM has the potential to increase child imitation (C-IM).

C-IM of adult actions plays an important role in children’s linguistic (Arbib, 2002) and social-cognitive development (Bauer, Wenner, Dropik, Wewerka, & Howe, 2000; Meltzoff & Decety, 2003). Infants acquire action imitation skills before spoken language (Jones, 2009), and imitation skills have been found to be positively associated with language development (Carpenter, Tomasello, & Striano, 2005; Charman et al., 2000; McEwen et al., 2007; Sundqvist, Nordqvist, Koch, & Heimann, 2016; Young et al., 2011). Infants and toddlers who often imitate the actions of others have been found to use more advanced language in both imitative and spontaneous speech (Bloom, Hood, & Lightbown, 1974). In a longitudinal study, Sundqvist et al. (2016) found that deferred C-IM in nine-month-old infants predicted productive language development at 16 months of age. Other studies have found that action imitation continues to play a role in language acquisition later in toddlerhood. For example, Gampe, Brauer, and Daum (2016) found that 30-month-olds who were randomly assigned to imitate an action were better able to learn the verbal label of the action than toddlers who only observed it. Thus, P-IM has the potential to indirectly improve child language development by increasing the child’s ability to imitate.

Reciprocal imitation between parents and children also appears to play an important role in children’s social-cognitive development. Research suggests that young children and infants rely heavily on nonverbal imitation to coordinate joint games and activities (Eckerman, Davis, & Didow, 1989). Furthermore, imitation is an integral part of parent-child interactions as it teaches the child turn-taking, helps the parent and child communicate likeness, and allows for the dyad to share meaningful experiences (Uzgiris, 1984). Cognitive processes used by infants during imitation are thought to facilitate the development of theory of mind and empathy (Meltzoff & Decety, 2003). When children recognize that they are being imitated, it heightens their awareness of others’ mental states, and increases their understanding of the internal states of others by allowing comparison to their own (Meltzoff, 2005; Meltzoff & Decety, 2003). Furthermore, being imitated is associated with more frequent prosocial behavior among infants (Carpenter, Uebel, & Tomasello, 2013). It is not surprising, then, that imitation has also been found to be important in children’s moral development (Carpenter et al., 2013; Forman, Aksan, & Kochanska, 2004). Conversely, impairment in C-IM has been associated with linguistic and social-cognitive deficits in children with Autism Spectrum Disorder (ASD; Smith & Bryson, 1994).

Despite the potential importance of P-IM in promoting crucial cognitive, social, and behavioral outcomes for young children, it is perhaps the least understood of all the PRIDE skills. This is largely because the Dyadic Parent-Child Interaction Coding System (DPICS-IV), the assessment tool used in conjunction with PCIT, does not include a code for imitation. The DPICS-IV is used to assess pre-post change in parent skills and child compliance, as well as to monitor parent skill acquisition during PCIT treatment sessions (Eyberg, Chase, Fernandez, & Nelson, 2014). The DPICS-IV provides guidelines to code parental behaviors and child responses during three standardized situations, and includes codes for the three PRIDE skills parents are required to master in the first phase of PCIT (labeled praise, reflection, and behavioral description; Eyberg et al., 2014). However, there is no formal criteria for determining when a parent has met mastery on P-IM because DPICS-IV does not include a code for its frequency. Instead, therapists are asked to make a global judgement as to whether parents’ imitation is “satisfactory” or “needs practice,” without specific scoring guidelines. Furthermore, it is difficult to assess parent progress on P-IM throughout therapy, and a lack of a monitoring tool may inadvertently lead therapists to spend less time on coaching P-IM relative to skills required for treatment graduation. Lack of a reliable code for imitation also has implications for research, as it is currently not possible to use the DPICS-IV to determine if PCIT leads to significant increases in either P-IM or C-IM, if increases in P-IM are related to increases in C-IM, and if increases in P-IM can lead to reductions in child behavior problems. Thus, a reliable measure of imitation that is compatible with the DPICS-IV has the potential to enhance both PCIT treatment and research.

Although there are a number of existing observational coding systems for imitation that have been used in child behavioral therapies, these systems have features that make them difficult to integrate with the DPICS-IV and PCIT. Systems such as the Motor Imitation Scale (Stone et al., 1997), Unstructured Imitation Assessment (Ingersoll & Lalonde, 2010), and Preschool Imitation and Praxis Scale (Vanvuchelen, Roeyers, & De Weerdt, 2011) measure children’s abilities to imitate predetermined tasks in response to scripted commands delivered by the researcher. Because these coding systems measure children’s abilities to imitate when prompted, they are a poor fit with the DPICS-IV and PCIT, which focus on increasing the frequency of P-IM during unstructured, play-based situations with parents. In an unpublished doctoral dissertation, Bhuiyan (2015) described the development of a system for coding P-IM and C-IM during DPICS-IV assessments. The primary aim of the study was to examine the ability of an enhanced version of the Child Directed Interaction phase of PCIT to improve the imitative abilities of children with ASD and their parents using a series of single case multiple baseline studies. This study provides initial evidence that it is possible to code imitation in DPICS-IV assessments with acceptable interrater reliability, and indicates that PCIT, when enhanced to boost attention paid to the coaching of imitation, can increase rates of P-IM. However, the guidelines have not been published to date, and further research conducted with larger samples of children without ASD receiving PCIT that has not been enhanced to increase attention to P-IM is still needed.

The current study addresses these needs by developing and assessing the reliability and validity of observational coding guidelines for P-IM and C-IM that can be used in conjunction with the DPICS-IV, utilizing data from a previously conducted PCIT clinical trial with Mexican American families. Because parenting practices facilitate the competencies children need in order to survive and succeed in a specific cultural context (Greenfield, Keller, Fuligni, & Maynard, 2003), and cultural groups value a range of child qualities and parenting techniques (Bornstein, 1995; Deater-Deckard & Dodge, 1997; Ripoll-Nunez & Rohner, 1997), it is possible that imitation will manifest differently in Mexican American families compared with families from other cultures. However, studies conducted to date suggest that parent-child imitation is present across all cultural groups and is a primary mode of cultural transmission across generations (Neilson, Mushin, Tomaselli & Whiten, 2016); thus, it seems likely that parent-child imitation will share common features across cultures. To our knowledge, no cross-cultural studies of parent-child imitation have included Mexican Americans to date. Conducting the study with a Mexican American sample, however, will increase our understanding of parent-child imitation among a sizable US ethnic group that has been underrepresented in behavioral parent training studies (Ortiz & Del Vecchio, 2013), and provides us with a reasonable starting point for understanding the role that imitation may play within and/or across cultures in the families of children with disruptive behavior problems. Future studies can then compare or extend these findings to other cultural groups.

The primary aim of the present study was to develop reliable and valid coding guidelines for both P-IM and C-IM. First, we sought to determine if our coding system had acceptable interrater reliability. Next, we conducted a series of analyses to examine the construct validity of the imitation code. First, we examined the hypothesis that P-IM and C-IM would show a greater increase from pre- to post- treatment in PCIT in comparison with treatment as usual (TAU). Second, based on literature suggesting that parents who imitated their children are more likely to be imitated in return, we hypothesized that P-IM frequency would be positively correlated with C-IM frequency. Finally, we examined the hypothesis that increases in PIM and CIM from pre- to post-treatment would be associated with decreases in parent-reported child behavior problems.

Method

Participants

Participants were a sample of 58 Mexican American families of young children with clinically significant behavior problems that participated in an earlier pilot clinical trial testing the effectiveness of a culturally modified version of PCIT for Mexican American families, called Guiando a Niños Activos (GANA; McCabe & Yeh, 2009). Participants were seeking treatment for their child’s problems at a community clinic and were randomly assigned to receive either standard PCIT, GANA, or TAU. In order to qualify for the pilot clinical trial, the child had to be (a) aged 3–7, (b) identified by the primary caregiver as Mexican American, and (c) in the clinical range on the Eyberg Child Behavior Inventory (ECBI). Families were excluded if they were participating in any other treatment for the child’s behavior problems, if the parent or child had a diagnosis of intellectual disability, or if the child had a diagnosis of ASD.

Respondents were female primary caregivers (92% biological mothers, 3% grandmothers, 2% aunts, and 2% other), with an average age of 32.2 years (SD = 8.1 years). The sample included more boys (n = 41, 71%) than girls. Mean child age was 52.8 months (SD = 12.4 months), ranging from 3 years 0 months, to 6 years, 8 months. Approximately 67% of the children and 81% of the primary caregivers preferred Spanish, and 76% of primary caregivers were first generation Americans. Among mothers, 69% were currently married, 12.1% were separated or divorced, and 19% were never married. Fifty percent of mothers reported a high school or lower level of education, and average total yearly income was $23,271 (SD = $14,518). The average pre-treatment ECBI intensity score was 181.12 (SD = 27.04). The demographic and clinical variables listed above did not differ significantly across the three conditions, indicating a successful randomization (see McCabe & Yeh, 2009 for details).

Measures

Demographic Questionnaire.

(Eyberg, 1999). Parents completed a demographic questionnaire about the child’s age, sex, race/ethnicity, grade level in school, health problems, medication use, family structure and income.

Eyberg Child Behavior Inventory Intensity Scale (ECBI-I).

(Eyberg & Pincus, 1999). Child behavior problems were assessed with the ECBI Intensity Scale, a 36-item parent report measure of the frequency of disruptive behaviors with established reliability and validity for both English and Spanish versions (Eyberg & Pincus, 1999; Garcia-Tornel et al., 1998). The ECBI-I scale has been found to be stable over time and sensitive to the effects of intervention. Clinical cutoffs are available for the 2–7 age range. Coefficient alpha at baseline for the current sample was .86 for the ECBI-I Scale. To qualify for the study, children had to score above the ECBI-I clinical cutoff of 131.

Dyadic Parent-Child Interaction Coding System (DPICS-III).

(DPICS 3rd ed.; Eyberg, Nelson, Duke, & Boggs, 2004). The DPICS is a behavioral observation coding system that measures the quality of parent-child social interaction during three 5-minute standard situations that vary in the degree of parental control (i.e., child directed interaction (CDI), parent directed interaction (PDI), and clean-up). For the present study, we examined CDI segments only because CDI is the situation in which P-IM is most highly encouraged and presumably most likely to occur. In this situation, parents are instructed over a bug-in-ear microphone to “tell [child’s name] s/he may play with whatever s/he chooses. Let him/her choose any activity s/he wishes. You just follow his/her lead and play along with him/her.” Convergent and discriminant validity of the DPICS categories have been extensively documented (Eyberg et al., 2004). DPICS-III observations were made once at pre- and once at post-treatment. A 4th edition of the DPICS has been released since this study was conducted (DPICS-IV; Eyberg et al., 2014); however, commonly used codes remain highly comparable and no changes were made in the way the observations themselves are conducted. Kappa coefficients ranged from .59 for Noncompliance to .85 for Unlabeled Praise, and were consistent with kappas reported from previous studies (Eyberg, Nelson, Ginn, Bhuiyan, & Boggs, 2013).

Imitation Coding Guidelines.

Imitation coding guidelines, consisting of 13 rules to identify P-IM or C-IM, were developed for the current study. After authors arrived at a working definition of imitation, pairs of research assistants trained in the DPICS-IV viewed pre- and post-treatment CDI segments from 10 families assigned to the TAU condition (20 total segments) and coded each behavior as 0 = clearly not imitation, 1 = potentially imitation, or 2 = clearly imitation. The coders also distinguished the imitative behaviors as either P-IM or C-IM. Rules to distinguish imitation from non-imitation were created through extensive discussion of coder disagreements and behaviors that were initially coded as “potential imitation.” A final version of the imitation coding guidelines can be found in Table 1.

Table 1.

Parental Imitation Coding Guidelines*

Rule Name Rule Content
Definition of Imitation Parental imitation occurs when the parent intentionally copies the child’s immediately preceding play behavior.
Five Second Rule Each time the child completes an action or gesture, imitation is coded if the parent begins to mimic the child’s immediately preceding action or gesture within 5 seconds of action completion by the child. Imitation is also coded if the parent begins to perform actions that are necessary to imitate the child’s original action during the 5 second interval.
Complete Action Rule Parental imitation is only coded for actions completed by the child. Imitation of unfinished or partially completed actions is not coded.
Compound Action Rule If the child completes an action that requires multiple steps and the steps are separated by intervals of fewer than two seconds, the set is coded as one action.
Simultaneous Action Rule If the parent and child are engaged in a behavior simultaneously and it is unclear who initiated the action, imitation is not coded. However, if it is clear who initiated the action, imitation is coded.
Sequential Imitation Rule If the child imitates a parent action or gesture and the parent then imitates the child, code imitation.
Verbalization Rule If a parent performs an imitative act, imitation is coded regardless of accompanying verbalizations.
Multiple Imitation Rule If a parent imitates the child’s action more than once, each instance of imitation that begins within the five second interval is coded as imitation.
Vagueness Rule Actions that are too vague to judge imitative intent (e.g. looking, searching, handing, selecting toys, picking up toys, putting toys down) are not coded as imitation.
Building Rule During building activities, imitation is coded when the parent imitates the child’s action on a similar structure with a similar piece and puts it in the same location that the child did.
Helping and Correcting Rule Imitation is not coded when a parent performs an action that corrects or helps the child’s completed action.
Cleanup Rule During parental imitation of child cleanup activities, imitation is coded if the child and parent are intentionally putting toys in a specific container. Imitation is not coded if the child and parent are engaging in vague cleanup actions (e.g. moving, pushing).
Game Rule Game interactions refer to actions that are performed within the scope of an activity with established, known rules. Imitation is not coded when a parent copies a child’s action in accordance with the rules of a structured game.
Open in a new tab
*

Child imitation guidelines parallel the parental guidelines.

Procedures

The GANA Study.

Families that met study eligibility criteria during a phone interview conducted by a bilingual research assistant were invited to attend a 3- to 4-hour pre-treatment assessment during which they provided written informed consent for study participation. Families were compensated $100 for their time. Families were then assigned using random permuted blocks of nine to one of three treatment conditions: GANA (n = 21), standard PCIT (n = 19), or TAU (n = 18). Research assistants conducting assessments and families were kept masked to assignment. All three programs were delivered in the community mental health clinic where the parents had sought services. Combining the standard PCIT and GANA conditions, 63% met full PCIT graduation criteria, and 37% dropped out prior to graduation. The dropout rate for TAU was 56%. The average number of sessions attended was 13.42 (SD = 8.03) for standard PCIT, 13.90 (SD = 7.99) for GANA, and 10.94 for TAU (SD = 10.01). All families were re-contacted immediately to complete a second 2- to 3-hour assessment, for which they were paid $100. All procedures were approved and conducted in accordance with the guidelines of the appropriate Institutional Review Boards.

Families assigned to standard PCIT were treated in accordance with the 1999 PCIT treatment manual (Eyberg & Members of the Child Study Laboratory, 1999). Families who were assigned to GANA also received the core features of PCIT, but delivery was modified to optimize cultural fit for Mexican American families (see McCabe et al., 2005 and McCabe & Yeh, 2009 for detailed description). All full-session videotapes were coded for integrity by independent raters, and a random sample of 25% were double-coded for reliability. Interrater agreement was 85%. In the standard PCIT condition, an average of 82% of integrity checklist items were completed. For the GANA condition, standard PCIT outlines were expanded to include cultural modifications, and an average of 77% of the expanded integrity checklist items were completed. Therapists were bilingual practicum students and staff therapists from a variety of mental health disciplines. Standard PCIT and GANA therapists were trained and received supervision/consultation from the principal investigator. TAU families were assigned to therapists without training in PCIT at the same community mental health clinic. TAU therapists were allowed complete freedom in the approaches they used and, as in both the PCIT and GANA conditions, were allowed as many sessions as they felt necessary to complete treatment. All therapists had one hour per week of face-to-face individual supervision/consultation with agency staff.

For the purposes of the current study, we have combined the standard PCIT and GANA conditions into a single group for several reasons. First, GANA only differed from standard PCIT in the tailoring of treatment rationales and engagement techniques to better fit with Mexican American culture; both groups received a program with high fidelity to the PCIT manual. Second, imitation did not emerge as a factor that was likely to influence engagement of Mexican American families in PCIT in the research that was conducted to inform modifications to increase the cultural fit of GANA; thus, imitation was implemented identically across the two conditions (see McCabe, et al, 2005). Finally, rates of imitation did not differ significantly between the standard PCIT and GANA subgroups at pre-treatment C-IM (MG = .500, MP = .790, p = .170) or P-IM (MG = .70, MP = 1.0, p = .480) or post-treatment C-IM (MG = 1.526, MP = 1.167, p = .249) or P-IM (MG = 1.63, MP = 1.67, p = .276). Furthermore, there was no significant difference between the rate of pre-to-post change in C-IM (MG = 1.00, MP = .389, p = .327) or P-IM (MG = .947, MP = .611, p = .446) between the standard PCIT and GANA subgroups. Hence, future references to the PCIT condition will include both the culturally modified and standard PCIT conditions.

Imitation Coding Study.

In the current study, the authors coded 110 (78 PCIT, 32 TAU) total CDI video segments from pre- and post-treatment for the 58 families (40 PCIT, 18 TAU) using the finalized imitation coding guidelines. A portion (20) of the TAU tapes were used to derive the imitation coding guidelines, and were later re-coded using the finalized coding system. The data analytic sample for agreement analyses included all 40 PCIT families and 18 TAU families, whereas analyses for our primary pre-to-post comparisons included only 37 PCIT and 14 TAU families (Five families were missing post assessments, one was excluded due to an outlier value of 21 parent imitations, and one family’s pre-assessment was not codable due to poor audio quality). Coders were masked to time-point. CDI segments were divided into 60 5-second intervals, and each interval was coded 0 if imitation was not present and 1 if imitation was present. P-IM and C-IM were coded separately. All segments were double coded and interrater agreement was calculated based on independent codes assigned prior to discussion. Coders then discussed any disagreements and reached a consensus on final codes. The resulting consensus codes were used for analyses involving P-IM or C-IM frequency and pre to post change.

Results

To determine interrater reliability for the newly developed imitation code, we calculated agreement for both P-IM and C-IM, as displayed in Table 2. Agreement exceeded 97% for the PCIT and TAU conditions, indicating acceptable interrater agreement (Eyberg et al., 2013). Many studies also report kappas to correct for chance agreement (e.g., Suen & Ary, 1989). Kappas for the total sample indicated moderate to substantial agreement. In the PCIT condition, kappas indicated fair to moderate agreement, whereas kappas in the TAU condition indicated moderate to substantial agreement. However, kappa has been found to underestimate true interrater reliability for infrequently occurring codes (e.g., Feinstein & Cicchetti, 1989; Viera & Garrett, 2005; Xu & Lorber, 2014), a significant concern given that both P-IM and C-IM occurred at a low frequency in this sample. On average, parents imitated their child at pre-treatment about once within the 5-minute interval (M = 1.02, SD = 1.69) and children imitated their parents less than once (M = .88, SD = 1.26). At post-treatment, parents imitated their child an average of 1.49 times (SD = 1.88) and children imitated their parents an average of 1.45 times (SD = 1.59). Thus, the most appropriate estimate of reliability in this case is likely to be Guilford’s G, which corrects for chance and provides an unbiased estimate of agreement for low base rate items (Xu & Lorber, 2014). Overall interrater agreement between coders was above G = .95 for both C-IM and P-IM, indicating very good agreement (see Table 2).

Table 2.

Interrater Agreement for Parent and Child Imitation across Conditions

% Agreement Kappa Guilford’s G
Total Sample P-IM 98.3% .618 .966
C-IM 98.2% .432 .964
PCIT P-IM 98.2% .537 .963
C-IM 98.4% .392 .968
TAU P-IM 98.6% .684 .973
C-IM 97.7% .488 .954
Open in a new tab

Note. PCIT = Parent-Child Interaction Therapy, TAU = Treatment as Usual, P-IM = Parent Imitation, C-IM = Child Imitation

In order to examine the construct validity of the imitation code, we next calculated repeated measures ANOVAs to test the hypothesis that the frequency of imitation would increase more from pre- to post- treatment in the PCIT condition than in TAU. Results of the first repeated measures ANOVA demonstrated that there was no main effect of P-IM across time (F = .738, p = .198, partial η2 = .015). However, the interaction between time and condition (F = 2.246, p = .07, partial η2 = .043) trended toward significance, indicating that pre-to-post treatment change in rates of P-IM varied by condition (see Figure 1). Post-hoc paired-samples t-tests showed that P-IM increased significantly in PCIT from pre to post treatment, but showed no significant change in TAU from pre to post (see Table 3). On average, those in PCIT increased P-IM frequency by .784 compared to those in TAU, whose rates of P-IM actually decreased by .357.

Figure 1.

Figure 1.

Open in a new tab

Pre to post change in frequency of P-IM across conditions.

Table 3.

Pre to Post Treatment Frequencies and Effect Sizes of P-IM and C-IM across conditions

Pre Post p d
Total Sample P-IM 1.02 1.49 .071 0.25
C-IM 0.88 1.45 .029 0.36
PCIT P-IM 0.86 1.65 .018 0.38
C-IM 0.65 1.35 .017 0.44
TAU P-IM 1.43 1.07 .286 −0.18
C-IM 1.50 1.71 .378 0.12
Open in a new tab

Note. PCIT = Parent-Child Interaction Therapy, TAU = Treatment as Usual, P-IM = Parent Imitation, C-IM = Child Imitation.

The second repeated measures ANOVA demonstrated that the main effect of C-IM from pre to post treatment trended toward significance (F = 2.024, p = .081) with a small effect size (partial η2 = .039). However, the interaction between time and condition was not significant (F = .583, p = .225, partial η2 = .012). Given low power to detect an interaction effect, we conducted additional exploratory analyses to further examine group differences in rates of change. When examined separately, PCIT showed significant pre to post change in C-IM, whereas TAU showed no significant change (see Table 3).

We next tested the hypothesis that P-IM frequency would be significantly correlated with C-IM frequency. This hypothesis was supported, as the one-tailed correlations between the overall frequency of child and parent imitation were positive and statistically significant when examined separately at pre-assessment (r = .517, p < .001) and post-assessment (r = .627, p < .001), and when both time points were combined (r = .580, p < .001). In the PCIT condition, rates of parent and child imitation were positive and statistically significantly correlated at pre-assessment (r = .35, p = .014), post-assessment (r = .632, p < .001), and when both time points were combined (r = .566, p < .001). Similarly, the frequency of child and parent imitation in TAU were positive and statistically significant at pre-assessment (r = .612, p <.001), post-assessment (r = .778, p <.001) and across both time points (r = .622, p <.001).

Lastly, we explored the potential contribution that increases in imitation may make to reduction in child symptoms. Regression analyses examined the relation between pre to post change in imitation and pre to post change in child behavior problems as measured by the ECBI. Change in C-IM did not significantly predict ECBI change (B = .773, one-tailed p = .796, R2 change = .001). However, change in P-IM significantly predicted improvement in ECBI scores (B = −.290, one-tailed p = .04, R2 change = .062). Specifically, for every additional increased instance of parent imitation, there was a nearly 3-point decrease in child behavior problems.

Discussion

The current study developed new codes for P-IM and C-IM that can be used in conjunction with the DPICS-IV. Further, it established the inter-rater reliability of these codes and found support for their construct validity using data from a previously conducted clinical trial comparing PCIT to usual care for Mexican American families. Specifically, we found a trend for Mexican American families receiving PCIT to increase their frequency of P-IM, but not C-IM, from pre- to post-treatment more than families who received usual care. The hypothesis that P-IM and C-IM would be significantly correlated was supported, as well as the prediction that increases in imitation would be significantly related to decreases in child behavior problems for the total sample. Taken together, these findings suggest that our measure has the potential to be a reliable and valid tool for coding both P-IM and C-IM.

Measures of interrater agreement on imitation among PCIT families provided preliminary evidence of the guidelines’ reliability. We emphasize the more conservative estimates based on the PCIT condition only, as a portion of the TAU tapes were used to derive the coding system, which could have artificially increased agreement. Inter-rater agreement for P-IM and C-IM codes in the PCIT condition was over 98%. For comparison, Bagner and Eyberg (2007) reported inter-rater agreement for other DPICS codes ranging from 67% for reflections to 90% for behavior descriptions. Most studies using the DPICS also report kappa coefficients in order to correct for chance agreement. In the current PCIT sample, kappas for overall imitation and P-IM fell in the “moderate” range, while kappa for C-IM fell in the “fair” range. Eyberg et al. (2013) reported kappas for parent and child verbalizations ranging from “moderate” to “very good,” suggesting that the P-IM code is comparable in reliability to existing DPICS codes, while the C-IM code performed less well. It could be argued that the DPICS code that is most similar to imitation is reflection, as both require the parent to mimic the child’s behavior, either verbally in the case of reflection, or behaviorally, in the case of imitation. Eyberg et al. (2013) reported a kappa of .59 for reflection, similar to our kappa for P-IM, although higher than our C-IM code. The relatively low kappa for reflections compared to other DPICS behaviors suggests that DPICS skills involving two-part parent-child interactions could be more challenging to code. Finally, according to Xu and Lorber (2014), kappas begin to underestimate the interrater agreement for codes as frequencies drop below 50%. The bias is slight at rates just below 50%, but increases substantially as the rate approaches zero (Xu & Lorber, 2014). Thus, we also calculated Guilford’s G, which is regarded as a more appropriate measure for items with base rates in the range found for both P-IM and C-IM in the current study (Holley & Guilford, 1964; Xu & Lorber, 2014). Guilford’s G for overall IM, P-IM, and C-IM with our PCIT families were in the very good range. Therefore, taking into account the low-base rates of imitation, inter-rater agreement appears to be very good and generally comparable to what has been found for other similar DPICS codes in past studies.

Next, we examined hypotheses related to the construct validity of the codes. A fundamental goal of PCIT is to increase parent imitation of their child’s behavior; therefore, we would expect to see significant pre- to post-treatment gains in P-IM in PCIT. However, less change would be expected in usual care, which is unlikely to directly target imitation. Therefore, the study tested the hypothesis that PCIT would increase average P-IM frequency more than usual care. A repeated-measures ANOVA indicated a non-significant trend for PIM to increase more in P-IM compared with treatment as usual. Post-hoc analyses further revealed that P-IM significantly increased from pre-to post treatment in PCIT, while declining in usual care. Post-treatment frequencies of P-IM (M = 1.65) suggest that while therapists do appear to be successfully coaching parents on imitation, parents are not leaving treatment near the mastery level of other “do skills” of 10 occurrences each (Eyberg & Funderburk, 2011). However, as past research on the optimal frequency for P-IM is not available to guide us, it is unclear whether therapists should increase their attention to imitation coaching or not. It is possible that coding guidelines for IM could increase therapist attention to, and parental use of imitation to levels similar to other “do skills,” or that further research using this coding system could help determine an appropriate target. Increasing P-IM has the potential to help children experience the developmental benefits of being imitated, such as increased prosocial behavior, understanding of others’ emotions, and imitation of others, which can lead to improved language skills and socio-emotional behavior (e.g., Arbib, 2002; Carpenter et al., 2013; Eckerman & Stein, 1990; Meltzoff, 2005; Meltzoff & Decety, 2003).

In contrast, comparative analyses indicated that C-IM did not increase significantly more in PCIT relative to TAU. This may be because PCIT does not explicitly coach C-IM; rather, we would expect increases in C-IM to result from parents’ increased use of imitation, which may then lead their children to imitate them more in return. Although additional exploratory analyses found a significant pre to post increase in C-IM in PCIT but no such increase in TAU, these findings must be interpreted with caution as the overall time x condition interaction was non-significant. However, it is possible that the increase in C-IM was too small to detect given the relatively small observed increase in P-IM from pre- to post-treatment and our limited sample size.

Additional support for the construct validity of the codes comes from the significant correlation observed between P-IM and C-IM. Past research suggests that reciprocal imitation teaches turn-taking, nonverbal communication, initiation of play with peers, and the sharing of meaningful experiences (Eckerman et al., 1989; Meltzoff & Gopnik, 1993; Uzgiris, 1984). Our findings are consistent with the notion that P-IM and C-IM are inter-related, which suggests that increased attention to P-IM in PCIT has the potential to enhance the social-cognitive benefits of PCIT by influencing C-IM.

Finally, PCIT promotes P-IM because it is theorized to reinforce positive child behaviors and reduce the child’s need to engage in negative attention-seeking behaviors, thereby contributing to a decrease in child behavior problems (McNeil and Hembree-Kigin, 2010). Our findings support this theory, as increases in P-IM from pre- to post-treatment were significantly related to decreases in parent-reported child behavior problems. Each additional instance of parental imitation at post-treatment was associated with a 3-point decrease in behavior problems reported on the ECBI. Increase in C-IM, however, was unrelated to changes in child behavior problems. These findings are consistent with the theory that parent imitation is an important tool for bringing about improvement in child behavior, possibly because it conveys approval of a child’s appropriate actions, thereby reinforcing prosocial behavior. Thus, the goal of increasing attention to imitation in PCIT has the potential to further improve outcomes in this form of treatment.

While the primary target of PCIT is reducing behavior problems for children with disruptive behavior disorders, increasing parent-child imitation may also have social-cognitive benefits for children on the autism spectrum. Children with ASD tend to exhibit a deficit in imitative abilities (Smith & Bryson, 1994), and interventions that target imitation have shown promise in improving those abilities (e.g. Bhuiyan, 2015; Ingersoll, 2010; Ingersoll & Schreibman, 2006). Several small trials and multiple baseline studies of adapted PCIT conducted with high-functioning children with ASD have demonstrated significant reduction in behavior problems (Masse et al., 2007; Solomon, Ono, Timmer, & Goodlin-Jones, 2008; Zlomke et al., 2017); however, they have not focused on enhancing imitation. Bhuiyan (2015) conducted a series of single case designs which demonstrated that PCIT that enhanced attention to imitation by coaching parents to imitate and reinforce children for imitating can significantly increase both P-IM and C-IM for children with ASD. Although this study did not report whether the increase in imitation explained symptom improvement for children with ASD, these findings suggest that attention to imitation may have the potential to further increase the relevance of PCIT to the specific needs of children with ASD.

This study contributes to the literature by establishing a reliable and valid set of guidelines for coding P-IM and C-IM in conjunction with the DPICS-IV. However, the study had several limitations worth noting. First, the participants were all Mexican American families. Although the effect of Mexican American culture on parent-child imitation is currently unknown, these findings cannot be generalized to other ethnic groups without further study. Second, there is a need for replication with different sets of coders. Because the authors of the guidelines also coded the parent-child interactions, it cannot be assumed that other coders would achieve the same rates of reliability. Third, coders were blind to timepoint, but not to condition. However, the rates of P-IM and C-IM for TAU at pre-test relative to PCIT suggests that a bias favoring PCIT is unlikely. Fourth, because coding was done from videotaped interactions, it is unclear whether therapists can reliably code imitation live during treatment sessions, as they do with the other PRIDE skills. It would be desirable to test these guidelines for use in live coding situations. Fifth, this study allowed coders to resolve disagreements on instances of imitation that were coded independently for the purpose of pre-post analyses. Such discussions may have resulted in greater frequencies of imitation than might be found with a single coder, as pairs of coders were able to collaboratively identify instances of P-IM and C-IM that a single coder may have overlooked. Sixth, a larger sample size would be desirable to detect what are likely to be small increases in both P-IM and C-IM, given their exclusion from mastery criteria in PCIT. Finally, future research should examine optimal levels of imitation for mastery in PCIT and determine if increased attention to imitation in PCIT treatment results in improved social-cognitive abilities and prosocial behavior for children with clinically significant behavior problems and/or ASD.

Highlights.

  • Parent and child imitation coding guidelines have acceptable interrater reliability.

  • Parent imitation, increases marginally significantly more after PCIT than treatment as usual.

  • Increases in Child Imitation do not differ between PCIT and treatment as usual.

  • Frequencies of parental and child imitation are positively correlated.

  • Increase in parent-, but not child-imitation, is related to decreased child behavior problems.

Acknowledgments

Research reported in this publication was supported by the National Institute Of Mental Health of the National Institutes of Health under Award Number K01MH01924. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institute of Mental Health or the National Institutes of Health.

Footnotes

Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

References

  1. Arbib MA (2002). The mirror system, imitation, and the evolution of language In Dautenhahn K & Nehaniv CL (Eds.), Imitation in animals and artifacts (pp. 229–280). Cambridge, MA: MIT Press. [Google Scholar]
  2. Bagner DM, & Eyberg SM (2007). Parent-child interaction therapy for disruptive behavior in children with mental retardation: A randomized controlled trial. Journal of Clinical Child and Adolescent Psychology, 36(3), 418–429. doi: 10.1080/15374410701448448 [DOI] [PubMed] [Google Scholar]
  3. Bauer PJ, Wenner JA, Dropik PL, Wewerka SS, & Howe ML (2000). Parameters of remembering and forgetting in the transition from infancy to early childhood. Monographs of the Society for Research in Child Development, 65(4), 1–213. [PubMed] [Google Scholar]
  4. Bhuiyan NN (2015). Improving imitation skills and behavior of children diagnosed with autism spectrum disorder using child-directed interaction therapy (Doctoral dissertation). Retrieved from ProQuest Dissertations Publishing; (10679082) [Google Scholar]
  5. Bloom L, Hood L, & Lightbown P (1974). Imitation in language development: If, when, and why. Cognitive Psychology, 6(3), 380–420. doi: 10.1016/0010-0285(74)90018-8 [DOI] [Google Scholar]
  6. Bornstein MH (1995). Form and function: Implications for studies of culture and human development. Culture and Psychology, 1(1), 123–137. doi: 10.1177/1354067X9511009. [DOI] [Google Scholar]
  7. Carpenter M, Tomasello M, & Striano T (2005). Role reversal imitation and language in typically developing infants and children with autism. Infancy, 8(3), 253–278. doi: 10.1207/s15327078in0803_4 [DOI] [Google Scholar]
  8. Carpenter M, Uebel J, & Tomasello M (2013). Being mimicked increases prosocial behavior in 18-month-old infants. Child Development, 84(5), 1511–1518. doi: 10.1111/cdev.12083 [DOI] [PubMed] [Google Scholar]
  9. Charman T (2006). Imitation and the development of language In Rogers SJ & Williams JHG (Eds.), Imitation and the social mind (pp. 96–117). New York, NY: The Guilford Press. [Google Scholar]
  10. Charman T, Baron-Cohen S, Swettenham J, Baird G, Cox A, & Drew A (2000). Testing joint attention, imitation, and play as infancy precursors to language and theory of mind. Cognitive Development, 15(4), 481–498. doi: 10.1016/S0885-2014(01)00037-5 [DOI] [Google Scholar]
  11. Charman T, Swettenham J, Baron-Cohen S, Cox A, Baird G, & Drew A (1997). Infants with autism: An investigation of empathy, pretend play, joint attention, and imitation. Developmental Psychology, 33(5), 781–789. doi: 10.1037/0012-1649.33.5.781 [DOI] [PubMed] [Google Scholar]
  12. Deater-Deckard K, & Dodge KA (1997). Externalizing behavior problems and discipline revisited: Nonlinear effects and variation by culture, context, and gender. Psychological Inquiry, 8(3), 161–175. doi: 10.1207/s15327965pli0803_1 [DOI] [Google Scholar]
  13. Eckerman CO, Davis CC, & Didow SM (1989). Toddlers’ emerging ways of achieving social coordinations with a peer. Child Development, 60(2), 440–453. doi: 10.2307/1130988 [DOI] [PubMed] [Google Scholar]
  14. Eckerman CO, & Stein MR (1990). How imitation begets imitation and toddlers’ generation of games. Developmental Psychology, 26(3), 370–378. doi: 10.1037/0012-1649.26.3.370 [DOI] [Google Scholar]
  15. Eyberg SM (1999). Demographic Questionnaire. Unpublished document. University of Florida, Gainesville, FL. [Google Scholar]
  16. Eyberg S, & Members of the Child Study Laboratory (1999). Parent-child interaction therapy: Integrity checklists and session materials. Gainesville, FL: University of Florida. [Google Scholar]
  17. Eyberg SM, Chase RM, Fernandez MA, & Nelson MM (2014). Dyadic parent-child \ interaction coding system (DPICS) Clinical Manual (4th ed). Gainesville, FL: PCIT \ International. [Google Scholar]
  18. Eyberg SM, & Funderburk B (2011). Parent-Child Interaction Therapy Protocol. \ Gainesville, FL: PCIT International. [Google Scholar]
  19. Eyberg SM, Nelson MM, Duke M, & Boggs SR (2004). Manual for the parent-child interaction coding system. Gainesville: FL: University of Florida. [Google Scholar]
  20. Eyberg SM Nelson MM, Ginn NC, Bhuiyan N, & Boggs SR (2013). Dyadic parent-child interaction coding system (DPICS) comprehensive manual for research and training (4th ed). Gainesville, FL: PCIT International. [Google Scholar]
  21. Eyberg SM, & Pincus D (1999). Eyberg child behavior inventory and Sutter-Eyberg student behavior inventory: Professional manual. Odessa, FL: Psychological Assessment Resources. [Google Scholar]
  22. Feinstein AR, & Cicchetti DV (1990). High agreement but low kappa: I. The problem of two paradoxes. Journal of Clinical Epidemiology, 43(6), 543–549. doi: 10.1016/0895-4356(90)90158 [DOI] [PubMed] [Google Scholar]
  23. Field T, Field T, Sanders C, & Nadel J (2001). Children with autism display more social behaviors after repeated imitation sessions. Autism, 5(3), 317–323. doi: 10.1177/1362361301005003008 [DOI] [PubMed] [Google Scholar]
  24. Forman DR, Aksan N, & Kochanska G (2004). Toddlers’ responsive imitation predicts preschool-age conscience. Psychological Science, 15(10), 699–704. doi: 10.1111/j.0956-7976.2004.00743.x [DOI] [PubMed] [Google Scholar]
  25. Gampe A, Brauer J, & Daum MM (2016). Imitation is beneficial for verb learning in toddlers. European Journal of Developmental Psychology, 13(5), 594–613. doi: 10.1080/17405629.2016.1139495 [DOI] [Google Scholar]
  26. Garcia-Tornel FS, Calzada EJ, Eyberg SM, Mas Alguacial JC, Vilamala Serra C, Baraza Mendoza C, … Trinxant Domenech A(1998). Inventario Eyberg del Comportamiento en Niños: Normalizacion de la version espanola y su utilidad para el pediatra extrahospitalario [Eyberg Inventory of Child Behavior. Standardization of the Spanish version and its usefulness in ambulatory pediatrics]. Anales Españoles de Pediatria, 48(5), 475–482. [PubMed] [Google Scholar]
  27. Ginn NC, Clionsky LN, Eyberg SM, Warner-Metzger C, & Abner JP (2017). Child-directed interaction training for young children with autism spectrum disorders: Parent and child outcomes. Journal of Clinical Child and Adolescent Psychology, 46(1), 101–109. doi: 10.1080/15374416.2015.1015135 [DOI] [PubMed] [Google Scholar]
  28. Greenfield PM, Keller H, Fuligni A, & Maynard A (2003). Cultural pathways through universal development. Annual Review of Psychology, 54, 461–490. doi: 10.1146/annurev.psych.54.101601.145221 [DOI] [PubMed] [Google Scholar]
  29. Holley JW, & Guilford JP (1964). A note on the G index of agreement. Educational and Psychological Measurement, 24(4), 749–753. doi: 10.1177/001316446402400402 [DOI] [Google Scholar]
  30. Ingersoll B (2010). Pilot randomized control trial of reciprocal imitation training for teaching elicited and spontaneous imitation to children with autism. Journal of Autism and Developmental Disorders, 40(9), 1154–1160. doi: 10.1007/s10803-010-0966-2 [DOI] [PMC free article] [PubMed] [Google Scholar]
  31. Ingersoll B, & Lalonde K (2010). The impact of object and gesture imitation training on language use in children with autism spectrum disorder. Journal of Speech, Language, and Hearing Research, 53(4), 1040–1051. doi: 10.1044/1092-4388(2009/09-0043) [DOI] [PMC free article] [PubMed] [Google Scholar]
  32. Ingersoll B, & Schreibman L (2006). Teaching reciprocal imitation skills to young children with autism using a naturalistic behavioral approach: Effects on language, pretend play, and joint attention. Journal of Autism and Developmental Disorders, 36(4), 487–505. doi: 10.1007/s10803-006-0089-y [DOI] [PubMed] [Google Scholar]
  33. Jones SS (2009). The development of imitation in infancy. Philosophical Transactions of the Royal Society B, 354(1528), 2325–2335. doi: 10.1098/rstb.2009.0045 [DOI] [PMC free article] [PubMed] [Google Scholar]
  34. Kauffman JM, Kneedler RD, Gamache R, Hallahan DP, & Ball DW (1977). Effects of imitation and nonimitation on children’s subsequent imitative behavior. The Journal of Genetic Psychology, 130(2), 285–293. doi: 10.1080/00221325.1977.10533261 [DOI] [Google Scholar]
  35. Masse JJ, McNeil CB, Wagner SM, & Chorney DB (2007). Parent-child interaction therapy and high functioning autism: A conceptual overview. Journal of Early and Intensive Behavior Intervention, 4(4), 714–735. doi: 10.1037/h0100402 [DOI] [Google Scholar]
  36. McCabe K, & Yeh M (2009). Parent-child interaction therapy for Mexican Americans: A randomized clinical trial. Journal of Clinical Child and Adolescent Psychology 38(5), 753–759. doi: 10.1080/15374410903103544 [DOI] [PubMed] [Google Scholar]
  37. McCabe KM, Yeh M, Garland AF, Lau AS, & Chavez G (2005). The GANA Program: A tailoring approach to adapting parent-child interaction therapy for Mexican Americans. Education and Treatment of Children, 28(2), 111–129. [Google Scholar]
  38. McEwen F, Happé F, Bolton P, Rijsdijk F, Ronald A, Dworzynski K, & Plomin R (2007). Origins of individual differences in imitation: Links with language, pretend play, and socially insightful behavior in two-year-old twins. Child Development, 78(2), 474–492. doi: 10.1111/j.1467-8624.2007.01010.x [DOI] [PubMed] [Google Scholar]
  39. McNeil CB, & Hembree-Kigin TL (2010). Parent-child interaction therapy (2nd ed.). New York, NY: Springer. [Google Scholar]
  40. Meltzoff AN (2005). Imitation and other minds: The “like me” hypothesis In Hurley S & Chater N (Eds.), Perspectives on Imitation: From Neuroscience to Social Science (Vol. 2, pp. 55–77). Cambridge, MA: MIT Press. [Google Scholar]
  41. Meltzoff AN, & Decety J (2003). What imitation tells us about social cognition: A rapprochement between developmental psychology and cognitive neuroscience. Philosophical Transactions of the Royal Society B: Biological Sciences, 358(1431), 491–500. doi: 10.1098/rstb.2002.1261 [DOI] [PMC free article] [PubMed] [Google Scholar]
  42. Meltzoff AN, & Gopnik A (1993). The role of imitation in understanding persons and developing a theory of mind In Baron-Cohen S, Tager-Flusberg H, & Cohen D (Eds.), Understanding Other Minds: Perspectives in Autism (pp.335–336). Oxford, England: Oxford University Press. doi: 10.1176/ajp.152.9.1392-a [DOI] [Google Scholar]
  43. Ortiz C, & Del Vecchio T (2013). Cultural diversity: Do we need a new wake-up call for parent training? Behavior Therapy, 44(3), 443–458. doi: 10.1016/j.beth.2013.03.009. [DOI] [PubMed] [Google Scholar]
  44. Pelaez M, Borroto AR, & Carrow J (2018). Infant vocalizations and imitation as a result of adult contingent imitation. Behavioral Development, 23(1), 81–88. doi: 10.1037/bdb0000074 [DOI] [Google Scholar]
  45. Penney A, & Schwartz I (2018). Effects of coaching on the fidelity of parent implementation of reciprocal imitation training. Autism. Advanced online publication doi: 10.1177/1362361318816688 [DOI] [PubMed] [Google Scholar]
  46. Ripoll-Nunez KJ, & Rohner RP (2006). Corporal punishment in cross-cultural perspective: Directions for a research agenda. Cross-Cultural Research, 40(3), 220–249. doi: 10.1177/1069397105284395 [DOI] [Google Scholar]
  47. Rogers SJ, Hepburn SL, Stackhouse T, & Wehner E (2003). Imitation performance in toddlers with autism and those with other developmental disorders. Journal of Child Psychology and Psychiatry, 44(5), 763–781. doi: 10.1111/1469-7610.00162 [DOI] [PubMed] [Google Scholar]
  48. Smeets PM, & Kauffman JM (1980). Generalized reciprocal imitation in children. Journal of Child Psychology and Psychiatry, 21(1), 67–76. doi: 10.1111/j.1469-7610.1980.tb00017.x [DOI] [PubMed] [Google Scholar]
  49. Smith IM, & Bryson SE (1994). Imitation and action in autism: A critical review. Psychology Bulletin, 116(2), 259–273. doi: 10.1037/0033-2909.116.2.259 [DOI] [PubMed] [Google Scholar]
  50. Solomon M, Ono M, Timmer S, & Goodlin-Jones B (2008). The effectiveness of parent-child interaction therapy for families of children on the autism spectrum. Journal of Autism and Developmental Disorders, 38(9), 1767–1776. doi: 10.1007/s10803-008-0567-5 [DOI] [PMC free article] [PubMed] [Google Scholar]
  51. Stone WL, Ousley OY, & Littleford CD (1997). Motor imitation in young children with autism: What’s the object?. Journal of Abnormal Child Psychology, 25(6), 475–485. doi: 10.1023/A:1022685731726 [DOI] [PubMed] [Google Scholar]
  52. Suen HK, & Ary D (1989). Analyzing quantitative behavioral observation data. Hillsdale, NJ: Lawrence Erlbaum Associates, Inc. [Google Scholar]
  53. Sundqvist A, Nordqvist E, Koch FS, & Heimann M (2016). Early declarative memory predicts productive language: A longitudinal study of deferred imitation and communication at 9 and 16 months. Journal of Experimental Child Psychology, 151, 109–119. doi: 10.1016/j.jecp.2016.01.015 [DOI] [PubMed] [Google Scholar]
  54. Uzgiris IC (1984). Imitation in infancy: Its interpersonal aspects In Perlmutter M (Ed.), Parent-child interactions and parent-child relations in child development (Vol. 17, pp. 1–32). New York, NY: Psychology Press. [Google Scholar]
  55. Vanvuchelen M, Roeyers H, & De Weerdt W (2011). Objectivity and stability of the preschool imitation and praxis scale. American Journal of Occupational Therapy, 65(5), 487–598. doi: 10.5014/ajot.2010.ajot00000414 [DOI] [PubMed] [Google Scholar]
  56. Viera AJ, & Garrett JG (2005). Understanding interobserver agreement: The kappa statistic. Family Medicine, 37(5), 360–363. Retrieved from https://fammedarchives.blob.core.windows.net/imagesandpdfs/fmhub/fm2005/May/Anthony360.pdf [PubMed] [Google Scholar]
  57. Ward M, Theule J, & Cheung K (2016). Parent-child interaction therapy for child disruptive behaviour disorders: A meta-analysis. Child Youth Care Forum, 45(5), 675–690. doi: 10.1007/s10566-016-9350-5 [DOI] [Google Scholar]
  58. Xu S, & Lorber MF (2014). Interrater agreement statistics with skewed data: Evaluation of alternatives to Cohen’s kappa. Journal of Consulting and Clinical Psychology, 82(6), 1219–1227. doi: 10.1037/a0037489 [DOI] [PubMed] [Google Scholar]
  59. Young GS, Rogers SJ, Hutman T, Rozga A, Sigman M, & Ozonoff S (2011). Imitation from 12 to 24 months in autism and typical development: A longitudinal Rasch analysis. Developmental Psychology, 47(6), 1565–1578. doi: 10.1037/a0025418 [DOI] [PMC free article] [PubMed] [Google Scholar]
  60. Zlomke KR, Jeter K, & Murphy J (2017). Open-trial pilot of parent-child interaction therapy for children with autism spectrum disorder. Child and Family Behavior Therapy, 39(1), 1–18. doi: 10.1080/07317107.2016.1267999 [DOI] [Google Scholar]

RESOURCES