Abstract
The current study assessed the extent to which parental validating and invalidating behaviors (a) could be reliably measured in parent–adolescent relationships, (b) differed significantly between clinic and nonclinic families, and (c) were associated with measures of adolescent emotion dysregulation, behavior problems, and parent–adolescent relationship satisfaction. Adolescents (N = 29; age range = 12–18; 62% female) and their parents completed a variety of self-report and parent-report measures of adolescent functioning. Ratings of parents’ validating and invalidating responses during video-recorded social support and problem-solving interactions were obtained. Results indicated that parental validating and invalidating behaviors (a) were measured with a high degree of reliability, (b) differed significantly between clinic and nonclinic families, and (c) were correlated, in expected directions, with adolescent emotion dysregulation, externalizing problem behaviors, and adolescent relationship satisfaction. The implications of these findings are discussed in terms of both research and potentially improved family interventions.
Emotion dysregulation, or the difficulty in deploying behavioral strategies that effectively modulate the form, frequency, or magnitude of an emotional response (Diamond & Aspinwall, 2003; Gross, 1998), is a process variable associated with a variety of psychological outcomes in children and adolescents, including higher incidences of externalizing behaviors (Eisenberg et al., 2001), anxiety (Suveg & Zeman, 2004), hyperactivity (Walcott & Landau, 2004), depression (Kobak & Ferenz-Gillies, 1995; Silk, Steinberg, & Morris, 2003), and suicidal behaviors (Tamas et al., 2007). The relationship between emotion dysregulation and a variety of child and adolescent clinical outcomes has led to a growing body of research identifying key determinants of emotion regulation that inform models of emotional development and clinical intervention. Much of this research has focused on person-level variables that affect a child’s ability to regulate emotions, such as physiological reactivity (Beauchaine, Gatzke-Kopp, & Mead, 2007; Cole, Zahn-Waxler, Fox, Usher, & Welsh, 1996; Shipman et al., 2007), temperament (Eisenberg et al., 2005; Ellis, Rothbart, & Posner, 2004), and specific behavioral strategies acquired to regulate emotions (Blair, Denham, Kochanoff, & Whipple, 2004; Hannesdottir & Ollendick, 2007; Zeman, Shipman, & Suveg, 2002). In addition to person-level variables, specific aspects of the family environment are linked to varying degrees of child and adolescent emotion dysregulation. When parents respond negatively to a child’s emotional expression, the child is more likely to react negatively (Eisenberg, Cumberland, & Spinrad, 1998). More specifically, when a parent responds with emotional invalidation and minimization of an emotional expression, the child has more difficulty regulating his or her emotions (Gottman & Katz, 2002) and is more likely to learn problematic means of regulating emotions that are linked to clinical outcomes (Aldao, Nolen-Hoeksema, & Schweizer, 2010; Berlin & Cassidy, 2003; Krause, Mendelson, & Lynch, 2003). In contrast, when parents demonstrate warm, understanding, and accepting responses to an expressed emotion, children are more likely to develop understanding of their emotional experience, accurately express emotions, regulate emotional reactions, and comply with parental directions (Calkins, Smith, Gill, & Johnson, 1998; Eisenberg & Fabes, 1994; Shipman, Zeman, Penza, & Champion, 2000; Spinrad, Stifter, Donelan-McCall, & Turner, 2004). The family context then, and in particular parent–child interactions, can play an important role in regulating a child’s emotions, shaping a child’s ability to learn emotion regulation strategies, while serving as an important target in family interventions.
Linehan (1993) and colleagues (Fruzzetti, Shenk, & Hoffman, 2005) assert that emotion dysregulation results from an ongoing transaction between parents and children, including vulnerabilities from a child’s temperament, prior learning, and problematic interactions with parents, which leads to the development and maintenance of various forms of psychopathology. This model proposes that specific aspects of family communication, validating and invalidating behaviors, are key determinants of a child’s ability to regulate emotions. A validating behavior occurs when a child or adolescent expresses his or her private experience to a parent and this expression is met with understanding, legitimacy, and acceptance of this experience (Linehan, 1997). A validating behavior does not directly seek to change or alter a child’s emotional experience; instead, it seeks to highlight the emotional experience in order to facilitate an individual’s acceptance and experiencing of the emotion. Validating responses can influence individual emotion regulation in several ways. First, validating behaviors can promote the learning of skills for regulating emotions because they promote more disclosures of emotional states which facilitate the experiencing of an emotion and consequently its expression and regulation (Fruzzetti & Shenk, 2008; Fruzzetti & Worrall, 2010). Second, validating behaviors minimize the frequency, intensity, and duration of an emotional reaction, making regulation more likely. Conversely, an invalidating behavior is “one in which communication of private experiences is met by erratic, inappropriate, and extreme responses. In other words, the expression of private experiences is not validated; instead it is often punished or trivialized” (Linehan, 1993, p. 49). Such a response conveys to a child or adolescent that his or her emotional experience in a given situation is incorrect and attributes that experience to socially unacceptable or undesirable standards. Parental invalidating behaviors have a significant impact on emotion dysregulation by worsening a child’s emotional reactivity and by impeding his or her ability to learn and use skills for regulating emotions. Validating and invalidating behaviors have demonstrated moderate to large effect size differences (δ = 0.73–1.10) on emotion regulation outcomes, such as heart rate, skin conductance, and negative affect (Shenk & Fruzzetti, 2011).
However, there is no research directly examining the relationship between parental validating and invalidating behaviors and child and adolescent outcomes. The current study is a preliminary test of parental validating and invalidating behaviors and their relationship to broad domains of adolescent functioning. There were several aims of the current research: (a) determine whether validating and invalidating behaviors can be reliably measured in parent–adolescent relationships, (b) test whether validating and invalidating behaviors discriminate between clinic and nonclinic families, and (c) examine whether validating and invalidating behaviors are related to adolescent emotion dysregulation, externalizing and internalizing behaviors, and relationship satisfaction as predicted by biosocial models of psychopathology (e.g. Linehan, 1993).
Method
Sample
Clinic (n = 14) and nonclinic (n = 15) families were recruited for participation. Clinic families, defined as an adolescent currently participating in family-based psychological treatment, were recruited from local behavioral health clinics. Nonclinic families, where no family member was currently receiving psychological treatment, were recruited through public service announcements and advertisements in local newspapers. Families responding to recruitment efforts contacted the program coordinator for the study and scheduled the research assessment. Inclusion criteria were (a) families with at least one caregiver with custodial rights, (b) an adolescent child between the ages of 12 and 18, and (c) a willingness to participate in two videotaped interaction tasks including one parent and his or her adolescent child. In the case of two parent homes, each parent was required to participate in the study in order to be eligible. The mean age of children in the sample was 14.86 (SD = 1.55), the median family income was $40,000–$49,000, 62% of the children were female, with 93% identifying themselves as Caucasian. See Table 1 for detailed demographic information by clinic status membership.
Demographic Characteristics.
Note. AFLSI = Adolescent Family Life Satisfaction Index–Parental subscale; CBCL = Child Behavior Checklist; LPI = Life Problems Inventory.
*p < .05. **p < .01. ***p < .001.
Measures
Validating and Invalidating Behaviors Coding Scale (VIBCS; Fruzzetti, 2001)
The VIBCS is an observational rating scale used to measure levels of validating and invalidating behaviors within families (Fruzzetti, 2001). The VIBCS uses an ordinal rating scale ranging from 1 to 7 where family members are given a global rating on validating and invalidating behaviors observed in each interaction (the coding manual is available upon request from the second author). The VIBCS has demonstrated good interrater reliability when rating couples’ interactions in previous research with an intraclass correlation coefficient (ICC) of .77. The concurrent validity of the VIBCS was also examined in prior research with couples where validating behaviors were associated with greater relationship satisfaction (r = .37, p < .001), invalidating behaviors were associated with greater interpartner aggression (r = .39, p < .001), and where moderate to large effect size differences (η2 = .13 to .20) were observed between distressed and nondistressed couples (Lowry, Mosco, Shenk, & Fruzzetti, 2002). The VIBCS was used in this study to establish the initial reliability of measuring validating and invalidating behaviors in parent–adolescent relationships. Only parents were assigned ratings of validating and invalidating behaviors in the current study. When a two-parent home completed the study, the highest level of validating and invalidating behaviors on the VIBCS across the two parents was used for analysis.
Life Problems Inventory (LPI)
The LPI is a 60-item self-report questionnaire measuring adolescent emotion regulation consistent with the biosocial theory of emotion dysregulation (Rathus & Miller, 1995). Example items include “When I don’t get my way, I quickly lose my temper,” “Once I get upset, it takes me a long time to calm down,” and “Relationships with people I care about have a lot of ups and downs.” Items are rated on a 5-point scale ranging from 1 (not at all like me) to 5 (extremely like me). Reliability of the LPI in the current study is α = .96. The total score on the LPI was used in this study and is derived by summing all 60 items with higher scores indicating greater emotion dysregulation.
Child Behavior Checklist/4-18 (CBCL)
Parents completed the CBCL, a well-established, comprehensive multiaxial parent report measure of children’s behavioral functioning with reliabilities ranging from α = .72 to .96 and stability coefficients ranging from r = .70 to .74 in prior research (Achenbach, 1991). The CBCL generates standardized scores for broadband scales of internalizing and externalizing behavior problems. T-scores derived from the internalizing and externalizing scales of the CBCL were used in the current study as indicators of global adolescent problem behaviors. An average score was used in cases where two parents each provided a score on externalizing and internalizing behaviors for a single adolescent.
Adolescent Family Life Satisfaction Index—Parental subscale (AFLSI)
The AFLSI is a self-report questionnaire assessing global family satisfaction as reported by the adolescent (Henry, Ostrander, & Lovelace, 1992; Henry & Plunkett, 1995). In prior research, the AFLSI has demonstrated reliability (α = .90) and concurrent validity (r = .72) with other measures of family satisfaction (Henry et al., 1992). The AFLSI has a 7-item, Parental subscale measuring the degree to which an adolescent agrees with an item assessing how satisfied he or she is with the parental relationship. Items are ranked on a scale ranging from 1 (strongly disagree) to 5 (strongly agree) with higher scores indicating greater satisfaction. Example items include “How much my parents approve of me and the things I do” and “The amount of freedom my parents give me to make my own choices.” The Parental subscale of the AFLSI was used in this study as an index of parent–adolescent relationship satisfaction. Reliability and concurrent validity of the Parental subscale in prior research is α = .88 and r = .78, respectively (Henry et al., 1992). The reliability of the Parental subscale in the current sample is α = .82.
Procedure
All procedures were approved by the local institutional review board prior to beginning the study. Upon the family’s arrival for the research assessment, informed consent and child assent was reviewed with the parent/parents and adolescent. Following consent and assent, each family member was given the appropriate questionnaires to complete in private. Once the questionnaires were completed, each parent and adolescent participated in two, 10-min videotaped interactions. In two-parent homes, the adolescents participated in two 10-min videotaped interactions with each of their parents. Each family was prompted to discuss two topics. The first topic involved a discussion of an issue that promotes closeness between the adolescent and parent. The second topic involved a discussion of an issue that the adolescent and parent agreed was a mild to moderate conflict in their relationship. Families were asked to discuss each of these topics for 10 min while being videotaped. Once the family had completed both the questionnaires and the videotaped portions of the assessment, they were financially compensated for participating in the study.
Graduate and undergraduate students were trained in the VIBCS prior to coding the videotaped interactions. Training consisted of 10 weekly meetings with each meeting lasting 90 min. The first five meetings involved an overview of observational rating systems with families and detailed instruction on the VIBCS, including the theoretical background, coding structure, content of each level of validating and invalidating behaviors, and decision rules for promoting reliability. The final five meetings involved consensus coding of specified training sessions. After the 10th meeting, coders rated a new set of training sessions to determine whether their ratings met a sufficient criterion of reliability. An ICC of .75 was adopted as the lower bound criterion of reliability as coefficients of .75 and above indicate excellent reliability (Fleiss, 1986; Shrout & Fleiss, 1979). Coders meeting the ICC = .75 criterion were permitted to code the interactions. Coders not meeting this criterion were provided with additional training. Coder drift was minimized via weekly checks of adherence on a videotape rated by each coder. All coders were blind to the family’s clinic status and all other data.
Data Analytic Strategy
The data analytic strategy involved several planned analyses to establish the preliminary evidence of parental validating and invalidating behaviors. First, ICCs were obtained to estimate the interrater reliability of parental validating and invalidating behaviors measured in parent–adolescent interactions. Second, ratings obtained from the VIBCS were analyzed using multivariate analysis of variance (MANOVA) to determine whether parental validating and invalidating behaviors differed significantly between clinic and nonclinic families. Finally, multiple regression determined whether validating and invalidating behaviors were related to proposed processes of change (emotion dysregulation) and adolescent outcomes (externalizing behaviors, internalizing behaviors, and parent–adolescent relationship satisfaction). Validating and invalidating behaviors were entered simultaneously as predictors in the regression models. Regression models were then examined for outliers and model assumptions.
Results
Preliminary Data Analysis
Demographic and study-related variables were assessed using chi-square and analysis of variance (ANOVA) to detect significant differences between clinic and nonclinic families. Results from chi-square tests indicated that clinic and nonclinic families did not differ significantly on race, sex of the adolescent, whether the family was a single-parent or dual parent home, or family income. The ANOVA revealed significant mean differences between clinic and nonclinic families on the LPI, CBCL externalizing, CBCL internalizing, and AFLSI scores (see Table 1). There were no significant age differences between clinic and nonclinic families.
Reliability of the VIBCS in Parent–Adolescent Dyadic Interactions
Four coders provided ratings of validating and invalidating behaviors observed during the parent–adolescent interactions. Seventeen percent of the families in the sample were randomly selected as a means to assess interrater reliability. Interrater reliability on this subset of families was determined via ICC using a two-way random effects model with absolute agreement among coders (McGraw & Wong, 1996). Interrater reliability was estimated using a single measure ICC, which is based on each individual rating across raters, a conservative estimate of interrater reliability. The resulting estimate was ICC = .86, indicating excellent interrater reliability. The correlation between parental validating and invalidating behaviors was r = −.57, p = .001.
Validating and Invalidating Behaviors in Clinic and Nonclinic Families
A MANOVA compared ratings of validating and invalidating behaviors obtained using the VIBCS between clinic and nonclinic families. The results of the MANOVA demonstrated significant between-group differences on levels of validating behaviors, F(1, 27) = 4.23, p = .05, δ = .80, and invalidating behaviors, F(1, 27) = 16.27, p < .001, δ = 1.55. Specifically, clinic families had significantly lower levels of validating behaviors and significantly higher levels of invalidating behaviors when compared to nonclinic families (see Table 1) with large effect size differences observed between the groups.
Validating and Invalidating Behaviors and Global Adolescent Functioning
LPI
The multiple regression model with validating and invalidating behaviors as predictors of LPI scores provided a good fit to the data, F(2, 26) = 4.77, p = .02, with validating and invalidating behaviors accounting for 27% of the variance in LPI scores. Validating behaviors significantly predicted LPI scores, b = −11.88, p = .01, indicating that a one-level increase in ratings of validating behaviors was associated with an approximate 12-point decrease in LPI scores. Invalidating behaviors were not significantly related to LPI scores.
CBCL
Validating and invalidating behaviors were simultaneously estimated as predictors of CBCL externalizing behavior scores. This model produced a good fit to the data, F(2, 26) = 6.62, p = .01, that accounted for 34% of the variance in externalizing scores. Invalidating behaviors significantly predicted CBCL externalizing scores, b = 3.83, p = .01, with a one-level increase in invalidating behaviors associated with an almost 4-point increase in externalizing T-scores. Validating behaviors did not significantly predict externalizing scores after accounting for invalidating behaviors. This same model was used to fit CBCL internalizing behavior scores. Results indicated a poor fit to the model, F(2, 26) = 1.06, p = ns, with validating and invalidating behaviors accounting for only 8% of the variance in internalizing scores. Neither validating nor invalidating behaviors significantly predicted internalizing scores.
AFLSI
A final model was fit where validating and invalidating behaviors were entered as predictors of AFLSI scores. Results demonstrated a good fit to the model, F(2, 26) = 8.42, p < .01, that accounted for 39% of the variance in AFLSI scores. Both validating behaviors, b = 1.51, p = .03, and invalidating behaviors, b = −1.19, p = .04, predicted AFLSI scores, indicating the mutual importance of both variables when assessing adolescent relationship satisfaction.
Discussion
Results provide preliminary support to theoretical models (Diamond & Aspinwall, 2003; Fruzzetti et al., 2005) and prior research (Shenk & Fruzzetti, 2011) examining the role of family-level determinants, specifically validating and invalidating behaviors, of emotion dysregulation and corresponding behavioral outcomes. Parental validating and invalidating behaviors can be measured with a high degree of reliability using a global observational rating scale. These behaviors differed significantly between clinic and nonclinic families where families receiving treatment had significantly lower levels of validating behaviors and significantly higher levels of invalidating behaviors when compared to families who were not in treatment. Even with a modest sample size, analyses were sufficiently powered to detect the large effect size differences between clinic and nonclinic families on ratings of validating and invalidating behaviors, highlighting the potential importance of evaluating these behaviors in parent–adolescent relationships. Validating and invalidating behaviors were also differentially related to process variables and adolescent outcomes as predicted. Validating parent behaviors were associated with more effective emotion regulation and greater satisfaction in parent–child relationships. Invalidating parent behaviors, on the other hand, were associated with higher externalizing behavior problems and lower relationship satisfaction. These results support previous research indicating that how parents react to emotional responses and disclosures is related to the psychological functioning of their children currently and later in development (Krause et al., 2003; Spinrad et al., 2004). It is also important to note that neither validating behaviors nor invalidating behaviors were associated with internalizing behavior problems. Thus, validating and invalidating behaviors may be more useful when understanding the development of externalizing behaviors as opposed to internalizing behaviors, although future research will be needed to support this claim. Future longitudinal research will also need to examine the relationship between validating and invalidating responses and other ratings of interests (e.g., parent social support or criticality) to evaluate the predictive utility of various theoretical models. Overall, the results support existing research while extending the literature through the identification of specific parenting behaviors linked to both processes of change and clinical outcomes.
There are several potential clinical implications based on findings from this study. The VIBCS is a readily applicable assessment tool that could be used when conceptualizing clinical cases and developing a family treatment plan. For instance, if invalidating behaviors are related to clinical outcomes, then focusing on decreasing invalidating behaviors, and potentially increasing validating behaviors, may serve as important treatment targets. Based on present outcomes, the extent to which invalidating behaviors are contributing to individual or relationship outcomes can be assessed reliably and efficiently before starting therapy. By rating interaction samples from clients at various times throughout therapy, clinicians can track changes in invalidating behaviors as a result of implementing a treatment plan that includes these behaviors as targets. Also, because brief interactions can be coded in real time, immediate in-session feedback can be provided to families in therapy. Reducing the level of invalidating behaviors in parent–adolescent interactions may help in reducing problem behavior while removing important barriers to relationship satisfaction throughout adolescence. In turn, increasing the use of parental validating behaviors, both in their frequency and intensity, can be used to facilitate children labeling their emotional experiences, the accurate expression of their emotional states, as well as their abilities to regulate their emotional reactivity. In this context, validating behaviors can help improve adolescent functioning during the course of treatment while promoting relationship enhancement with parents.
There are also important considerations that limit broad conclusions about parental validating and invalidating behaviors from this study. The sample size is modest (N = 29) and, although representative in terms of sex and family constellation, not racially or ethnically diverse. The implication of having a modest sample size with primarily Caucasian families raises the possibility that findings may not generalize to the larger population of families, although previous research with the VIBCS has included ethnically and racially diverse samples (Lowry et al., 2002; Shipman et al., 2007). The research design is cross-sectional and causal inferences are not appropriate despite significant relationships among validating and invalidating behaviors, adolescent emotion dysregulation, and clinical outcomes. Only longitudinal research examining the temporal relations between these variables can tease out whether adolescent problem behavior develops as stated in biosocial models of psychopathology. Overall, this study provides a base from which to launch further research on parental validating and invalidating responses, advancing prior research by identifying specific parent behaviors contributing to the well-established connection between parenting and adolescent outcomes (Eisenberg et al., 1998). This study also offers theoretically informed targets for intervention to aid clinicians treating emotion dysregulation concerns with adolescents and their families.
Footnotes
Declaration of Conflicting Interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author(s) received no financial support for the research, authorship, and/or publication of this article.
