Developmental Changes in the Relation between Youth Disclosure and Parenting Behavior: A Cohort-sequential Analysis

Abstract

To address a gap in the literature regarding the development of youth disclosure across the transition to adolescence, the current research uses a cohort-sequential approach to study youth disclosure from middle childhood through adolescence. Longitudinal data from three cohorts of parents were utilized (N = 1359; children at T1 were in grades 2 [M = 8.00 years, SD = 0.57 years, 45% female], 4 [M = 10.12 years, SD = 0.60 years, 45% female], and 9 [M = 15.19 years, SD = 0.57 years, 48% female]). Parents were assessed annually over a 3-year time period. The focal analyses explored contemporaneous associations between characteristics of the parent-youth relationship (specifically, parental rejection and parental consistent discipline) and youth disclosure after accounting for person-specific trajectories of disclosure. Associations of gender, age, and socioeconomic status with disclosure were also assessed. Regarding trajectories of youth disclosure, results indicate that youth disclose less information to their parents about their daily lives as they get older; this trend was consistent across gender and socioeconomic status. In terms of associations with youth disclosure, when parents provided more consistent discipline or engaged in less rejection of their child, youth disclosure increased, even after accounting for their own trajectory of disclosure across time. In addition, the association of consistent discipline with youth disclosure became stronger with increased youth age. Results are discussed in terms of implications for understanding youth autonomy development, and the dyadic and developmental impact of parenting behaviors over time.

Introduction

The extent to which youth disclose aspects of their daily lives to their parents during adolescence predicts problem behavior. In general, youth who voluntarily disclose relatively high levels of information about their personal day-to-day lives to their parents are involved in lower levels of problem behavior as compared to their peers who are less open with their parents. In addition to being an important predictor of adolescent behavior, disclosure during adolescence also reflects the aspects of the parent-adolescent relationship, such that adolescents who disclose more to their parents typically believe their relationships with their parents to be relatively close and supportive. Research shows that, across adolescence, disclosure typically decreases as youth age (e.g., Laird et al., 2013; Padilla-Walker et al., 2018). However, much less is known about patterns of disclosure as youth transition into adolescence. It is also unclear how parenting during childhood affects disclosure during adolescence. Thus, the current study investigates the development of youth disclosure to their parents across the transition from childhood to adolescence and explores how it is predicted by the parenting characteristics of rejection and consistent discipline. A cohort sequential design was utilized; three cohorts of parents (with youth in Grades 2, 4, and 9, respectively) were each followed across three years.

The transition to adolescence is a time when youth seek increased autonomy and freedom (e.g., Steinberg & Morris, 2001). Parent-child relationships are in flux, with youth desiring increased jurisdiction over their day-to-day activities, relationships, and personal behaviors with increasing age. This pattern starts during childhood (Tisak, 1986) and continues through adolescence (Smetana et al., 2005). Despite this growing independence, youth continue to benefit from parental expertise, especially given the complex choices and situations that they are confronted with as they become older. Thus, characteristics of the parent-child relationship, such as the extent to which parents are knowledgeable about their children’s behavior and social experiences, are still important for understanding youth wellbeing. This information sharing about routine day-to-day aspects of life, operationalized here as disclosure, has important implications for youth outcomes (Tilton-Weaver et al., 2014). Routine disclosure is distinct from self-disclosure (where youth share aspects of their inner thoughts and feelings, often related to routine disclosure but more focused on introspection sharing). Both forms of disclosure are important to understand with regard to parent-youth relationships; however, the focus of the current investigation is on routine disclosure.

When adolescents share information about their daily lives with their parents (i.e., when they engage in routine disclosure, which will be termed disclosure for the remainder of this article), they are at a lower risk for problem behavior, delinquency, and emotional distress (Frijins et al., 2005; Goldstein, 2016; Laird & Marrero, 2010; Stattin & Kerr, 2000). Research has examined patterns of disclosure across adolescence (notably not across childhood though, or through the transition from childhood to adolescence), generally showing declines across the adolescent years (e.g., Laird et al., 2013; Padilla-Walker et al., 2018) with a few caveats (e.g., Keijsers & Poulin, 2013; declines across adolescence except for middle adolescent females; Keijsers et al., 2009; decreases based on adolescent report but curvilinear pattern based on maternal report). There is an important distinction, though, between youth disclosure and parental monitoring. Parental monitoring involves active parental information seeking; this can range from casual inquiries to persistent probing. Although parental monitoring can be an important tool in preventing problem behavior in some instances, research shows that parental monitoring’s positive implications usually occur indirectly through the acquisition of parental knowledge, or via other parental relational qualities such as warmth (Crouter & Head, 2002; Lippold et al., 2014). In fact, when taken to an extreme, parental monitoring can be perceived as intrusive by the adolescent (Dietvorst et al., 2018). Perceived intrusive parenting can serve a counterproductive role, essentially pushing adolescents to distance themselves from their parents (Hawk et al., 2009). Parental intrusiveness and perceived over-solicitation are associated with an increased likelihood of risky youth behavior, and with lower quality parent-adolescent relationship characteristics (Goldstein et al., 2005; Kapetanovic et al., 2020; Rote & Smetana, 2018).

In contrast to parental monitoring, youth disclosure occurs when the child voluntarily discusses aspects of their lives with parents; as conceptualized this occurs initially without parental inquisition or information gathering/ solicitation (Kerr & Stattin, 2000; Stattin & Kerr, 2000). In contrast to parental solicitation (where parents actively press their child for information about their activities and behavior), youth disclosure does predict parental knowledge and decreased risky behaviors over time (Kerr et al., 2010). A recent meta-analysis (Liu et al., 2020) supports this contention; across 32 studies examining predictors of parental knowledge during adolescence, youth disclosure was second only to parental warmth in terms of predicting what parents actually know about their adolescent children. It is clearly important, then, to understand what parents can do to facilitate the type of relationship with their adolescent children that supports youth disclosure. This said, with a few exceptions discussed below, the majority of research that focuses on youth disclosure has done so using it as a predictor variable.

As posited by self-determination theory (e.g., Ryan & Deci, 2017), humans are motivated by the need to achieve autonomy, relatedness, and competence. Related to parenting, these three need-supporting dimensions (as opposed to frustration) are parental involvement (as opposed to rejection), structure (as opposed to chaos), and autonomy support (as opposed to controlling parenting; Soenens et al., 2019; Vansteenkiste & Ryan, 2013). Previous research shows that when parents provide structure and warm, active acceptance and involvement, youth are encouraged to talk their parents about their daily activities (e.g., Kearney & Bussey, 2015; Soenens et al., 2006). Relatedly, research that has examined predictors of youth disclosure during adolescence are consistent in finding that warm, affirmative, and supportive parenting behavior is associated with increased youth disclosure. For example, a longitudinal study focusing on middle school students showed that when youth perceived that their parents reacted negatively to disclosure, youth felt controlled by their parents and less connected to them (Tilton-Weaver et al. 2010). In turn, these perceptions of the parent-youth relationship predicted lower levels of later disclosure. In a longitudinal study across the high school years, spending time doing fun activities with one’s family predicted increased youth disclosure over time (Willoughby & Hamza; 2011). In a longitudinal study focusing on Mexican-American adolescents, familism (operationalized as including support/closeness, reference, and obligation) predicted disclosure over time in girls, but not in boys (Son et al., 2022). In cross-sectional work focusing on college-attending emerging adults, memories of a positive childhood with one’s mother as well as satisfaction with one’s current relationship with one’s mother was associated with increased youth disclosure (to their mothers; Urry et al., 2011). In another cross-sectional study of college-attending emerging adults, parental mindfulness and family support were associated with increased youth disclosure (Goldstein et al., 2022). These studies, focusing on youth from early adolescence through emerging adulthood, are consistent in demonstrating that positive parent-youth relationship characteristics predict higher levels of youth disclosure.

Interestingly, however, none of this research has examined how parental rejection is associated with youth disclosure. Harsh and judgmental parent-child interactions are generally associated with youth adjustment concerns and lower quality relationship characteristics (Ali et al., 2019; Hale et al., 2005). For example, when youth perceive their parents to be rejecting of them, they are at a greater risk for emotional and self-esteem challenges (Ramírez-Uclés et al., 2018), and social anxiety even years later (Giaouzi & Giovazolias, 2015). In contrast, parental acceptance is associated with positive adjustment outcomes such as increased prosocial behavior (e.g., Putnick et al., 2018). Although some important research has been conducted linking parental rejection to challenges with the disclosure of sexual minority status (e.g., D’Amico & Julien, 2012) and gender minority status (e.g., Pollitt et al., 2023) among adolescents and emerging adults, no research to date has examined potential links between youth voluntary disclosure of daily whereabouts and behavior with parental rejection. It is also important to discover how this association (if apparent) might vary based on youth age. It may be that aspects of the parent-youth relationship context change in salience for understanding youth disclosure as children/adolescents age. For example, pre-adolescents may be more “okay” with their parents expressing discontent with their perspectives, whereas older adolescents may become increasingly frustrated when their parents engage in similar expressions of dismay. In addition, it is important to account for individual trajectories of youth disclosure over time when assessing the relation between parental rejection and youth disclosure, as it is possible that both behaviors are changing in tandem due to other factors (e.g., increased parent-child conflict during the transition to adolescence; Branje, 2018).

Parental warmth and support (and a lack of rejection) are not the only aspects of positive parenting, however. Conceptions of positive parenting typically include, in addition to a warmth/support dimension, an emphasis on developmentally appropriate and consistent discipline (e.g., Baumrind, 1991; 2005; Marshall & Chasin, 2000; Smetana et al., 2005). When adolescents know that parents set reasonable and consistent parameters regarding their behavior, and that they can expect reliable (but fair and non-harsh) consequences for their behavioral infractions, problem behavior is lessened and relationship outcomes are improved. Despite their increasing desire for autonomy, older children and adolescents still benefit very much from knowing that there are set, stable, and reliable boundaries placed on them by their parents. It is not yet known whether consistent discipline is associated with greater youth disclosure after accounting for youth-specific trajectories of disclosure, and if so how this relation might vary with youth age.

Additionally, regarding past research predicting youth disclosure, it is important to note that none of the above mentioned studies focused on youth during the transition from childhood to adolescence, nor did the studies follow different cohorts of youth over time. Parent-youth relationships during the pre-and-early adolescent years may be especially important for understanding the development of youth disclosure; parent-adolescent conflicts usually increase during the early adolescent years due to relationship transitions that are beginning to occur during this time (Branje, 2018; Smetana et al., 1991). Thus, research has not explored how levels of disclosure change over time from childhood through the early and middle adolescent years, and whether and how parenting characteristics during childhood and during the early adolescent years predict later youth disclosure.

Moreover, it is unclear whether disclosure trajectories during this liminal time (during the transition from childhood to adolescence) can be predicted by social and contextual factors such as child gender or socioeconomic status. Parents differentially socialize their children based on gender and this differential socialization is associated with a variety of psychosocial outcomes (e.g., McHale et al., 2003). The limited research available that has examined the role of youth gender in understanding the associations between information management and parenting shows that gender may have an impact on this aspect of the parent-child dyadic relationship as well; for example past research has found that adolescent boys may experience more rapidly increasing trajectories of secret-keeping from parents (Keijsers et al., 2010). Disclosure and secrecy are closely related (albeit distinct) constructs; thus, this may relate to the development of gender differences in trajectories of disclosure over time.

Regarding socioeconomic status, parents are less likely to engage in ideal parenting behaviors and strategies when they are stressed (Camisasca et al., 2022; Kochanova et al., 2021), and economic strain can be a critical source of stress that increases parenting challenges (e.g., Lovejoy et al., 2000; Reeb et al., 2013). As it relates to the construct of youth disclosure, this increased parental stress might have implications for relationship building opportunities or focus for socioeconomically disadvantaged families. It is also possible that socioeconomic status may have different implications for disclosure at different times during adolescence. For instance, perhaps as youth get older and more aware of their parents’ financial stress and strain (and their parents may also be less concerned about child-care/supervisory related issues pertaining to their older teen), they may be less likely to disclose information to their parents. Despite the potential association of socioeconomic status with youth disclosure, research on the relation between socioeconomic status and youth disclosure has been limited.

Current Study

To address the above discussed gap in the research on youth disclosure across the transition to adolescence, the current study explores trajectories of youth disclosure from the middle childhood period through adolescence. First, general trends of youth disclosure over time are assessed from the second grade through the eleventh grade. Next, ways in which parenting practices (specifically, parental rejection and parental consistent discipline) are related to youth disclosure are explored, after accounting for individual trajectories of youth disclosure. Previous research examining the development of disclosure during the transition from childhood to adolescence has been scant. Three cohorts were included in the current analyses; each cohort was assessed annually over a period of three years. The cohorts represent youth who were in the second grade at Wave 1 (Grade 2 cohort: followed from Grades 2 through 4), fourth grade at Wave 1 (Grade 4 cohort: followed from Grades 4 through 6), and ninth grade at Wave 1 (Grade 9 cohort: followed from Grades 9 through 11). Thus, the current study provides a cohort sequential analysis of trajectories of youth disclosure from mid-childhood through mid/late-adolescence. Several specific a priori hypotheses were forwarded. First, it was expected that youth disclosure would decrease as youth age, such that greater disclosure would occur at younger ages. This pattern was expected to occur in each of the three cohorts. Second, it was anticipated that parental rejection would be associated with lower levels of youth disclosure after accounting for person-specific trajectories of youth disclosure. Third, it was predicted that consistent parental discipline would be associated with higher levels of youth disclosure; again it was anticipated that this pattern would emerge after accounting for person-specific trajectories of youth disclosure. Fourth, it was predicted that these associations with parenting practices and youth disclosure would become stronger with increased age, due to the age-related developmental transitions of the parent-child relationship as the child transitions to adolescence. Essentially, the prediction was that disclosure would be more contingent on parent-child relational qualities with increasing youth age.

Methods

Participants

Participants were 1359 parents (M age in wave 1 = 38.6 years, SD = 7.06 years) of youth (47% female) in three cohorts (target child in Grade 2 at first assessment (n = 426), target child in Grade 4 at first assessment (n = 485), and target child in Grade 9 at first assessment (n = 448)). Parents/guardians were primarily the mothers of the target child (92%). Although race/ethnicity information is not available for parents/guardians, youth were somewhat diverse in terms of race/ethnicity (65% White, 23% Black/African-American, 12% multi-racial/ethnic, <1% Hispanic/Latino/a, 1% other race/ethnic category). Families also represented a wide range of socioeconomic status (mean parental income was 5.1 [SD = 2.8] on a 9-point scale where 1= “under $10,000/year,” 5 = “$41000-$50,000/year,” and 9 = “over $80,000/year”).

Families resided in seven communities from Iowa and Michigan. In terms of youth characteristics, the different cohorts had the following average ages and gender distributions at the first assessment (Grade 2 cohort: M = 8.00 years, SD = 0.57 years, 45% female; Grade 4 cohort: M = 10.12 years, SD = 0.60 years, 49% female, Grade 9 cohort: M = 15.09 years, SD = 0.57 years, 48% female). Families were assessed yearly over a three-year period; additional details about recruitment are provided in the procedure section below.

Measures

Youth disclosure

This construct was measured based on parents’ report of child- initiated conversations about their day-to-day lives. Two items comprise this index (Capaldi & Patterson, 1989); specifically, “how often does your child check in with you after school?” and “how often does your child talk to you about their daily plans?”. Although the alpha for this index was somewhat low (Cronbach’s α = 0.48–0.60 across waves), for a two-item index this is acceptable (Striener, 2003). The index has good face validity in that it captures the child-directed aspect of youth disclosure, rather than the parent-solicited aspect of monitoring. Response options ranged from 0 (never/almost never) through 4 (always or almost always); higher scores corresponded to greater disclosure. Items were averaged to compute the index. Identical items were asked of each cohort and at each time point.

Parental rejection

To assess the extent to which the parent is displeased with/rejecting of the child, parents answered affirmatively (1) or said no (0) to 10 questions (Eron et al., 1971; α = 0.74–75 across waves). A sample item is, “Have you thought that this child wastes too much time?” Responses were averaged across items; all participants answered these questions at each assessment point. Responses were averaged across items; all participants answered these questions at each assessment point. See Appendix A.

Consistent discipline

To assess parental consistent disciple, parents answered 9 items (Capaldi & Patterson, 1989) that focus on the extent to which they followed through with and were consistent about their child disciple practices (α = 0.79–81 across waves). Response items ranged from 0 (never or almost never) through 4 (always or almost always); items were averaged. Higher scores corresponded to more consistent discipline. The following item (reverse coded) is an example, “How often do you give up trying to get your child to do something you asked?” Participants from each cohort answered these questions at each time point. See Appendix A.

Socioeconomic status

To calculate an SES score, parent education (1 = less than 8^th grade” through 7 = “finished graduate school’), parent income (1 = “under $10,000 per year” through 9 = “over $80,000 per year) and occupational prestige (Stevens & Hoisington, 1987) were standardized and averaged to compute socioeconomic status. Higher scores on this variable correspond to greater socioeconomic status.

Procedure

All procedures were reviewed and approved by the institutional review board of the University of Michigan. Following informational presentations at participating schools by research team members, invitation letters with attached parent/guardian consent forms were sent home with students in the target starting grades (2^nd, 4^th, 9^th). That year, and in the two following years, youth whose parents/guardians consented to participation were subsequently surveyed, and then those parents/guardians were also solicited to complete their own surveys by mail and/or telephone. In this study, no significant differences were observed in responding as the function of survey type. Parents/guardians who completed surveys received $25 gift certificates at each time point.

Analysis Plan

Cohort and gender differences

To provide initial information about gender and cohort differences in study variables, a series of 2 (gender) by 3 (cohort) analysis of variance (ANOVA) tests were conducted. One ANOVA was conducted for each main study variable (adolescent disclosure, consistent discipline, parental rejection, and socioeconomic status), at each wave.

Bivariate correlations

To provide preliminary information about bivariate associations among study variables, bivariate correlations were examined among all variables (first averaged across cohorts, then individually for each cohort). Correlations were computed separately for males and females.

Unconditional growth curve models

First, unconditional latent growth curve models of youth disclosure from ages 7 to 18 were estimated, within the context of a multiple group multiple cohort (MGMC) model (Muthén & Muthén, 2017). MGMC models are latent growth curve models that take advantage of cohort-sequential or accelerated longitudinal designs by estimating developmental trajectories over the entire age range observed within the sample, with each cohort contributing to part of the overall trajectory. For the current study, youth disclosure was modeled as a function of age rather than grade, given that youth were different ages within each grade and that youth disclosure may differ more as a function of age rather than grade. Thus, youth age will be referred to in the context of the latent growth curve models, and youth grade will be referred to in the remainder of analyses. In the latent growth curve model, age at the first wave was used as a grouping factor, and growth factor loadings were constrained to be equivalent within each age across cohorts. For example, the linear slope loadings were 0, 0.1, and 0.2 at ages 7, 8, and 9, respectively, for the age 7 cohort, and the linear slope loadings were 0.1, 0.2, and 0.3 at ages 8, 9, and 10, respectively, for the age 8 cohort. Residual variances were fixed to be equivalent at the same age across cohorts, but free to vary across ages (e.g., fixed for age 8 assessments in the age 7 and age 8 cohorts, but different for age 7 and age 8 assessments).

To identify the best-fitting function of time for the trajectory of youth disclosure, the fit of a series of unconditional latent growth curve models were compared. The following models were compared: an intercept-only model (i.e., there is no change in youth disclosure from ages 7 to 18), a linear growth model (i.e., there is linear change in youth disclosure from age 7 to age 18), and a quadratic growth model (i.e., there is quadratic change in youth disclosure from age 7 to age 18, such that the rate of change itself increases or decreases). Once the best-fitting model was identified for the data, it was then tested whether relaxing equality constraints across age cohorts would improve the model fit. If model fit improves once equality constraints are relaxed, this provides evidence that a single growth curve does not adequately describe the development of youth disclosure across the age cohorts. However, if model fit does not improve meaningfully once equality constraints are relaxed, this would suggest that a single growth curve adequately captures the development of youth disclosure across age cohorts (Duncan et al., 1996; Duncan et al., 2006).

Because data were modeled based on age, 18 people whose age was unknown and 13 people who were an age with a small number of people (<10) were dropped from the original data (N = 1421), leaving 1359 in the analysis sample. Of these 1359 parents, 95% completed the first interview, 85% completed the second, and 79% completed the third. Most parents (71%) completed all three interviews. Analyses indicated that child sex was unrelated to number of completed interviews; however, parents of younger children completed more interviews on average (η² = 0.03, p < 0.001), as did parents with higher SES (η² = 0.04, p < 0.001). Wave 1 disclosure was unrelated to study retention, but parents who reported more consistent discipline (η² = 0.01, p = 0.015) and less rejection (η² = 0.02, p < 0.001) were more likely to remain in the study at waves 2 and 3. Given the observed differences between parents who remained in the study compared to those who did not, data are likely not missing completely at random (MCAR), but may instead be missing at random (MAR) such that missing data on disclosure is related to other variables observed in the dataset, but is not related to the missing values itself once these observed variables are accounted for.

The following indices are used to assess absolute model fit (i.e., how well does the hypothesized model fit the data overall): Comparative Fit Index (CFI); Tucker Lewis Index (TLI); Root Mean Squared Error of Approximation (RMSEA); and Standardized Root Mean Square Residual (SRMR). Generally, CFI and TLI values above 0.90, SRMR values below 0.08, and RMSEA values below 0.06 indicate good model fit (Hu & Bentler, 1999). The following indices are used to assess relative model fit (i.e., how well does each model fit the data compared to the other models): Aikaike Information Criterion (AIC); Bayesian Information Criterion (BIC); and sample size-adjusted BIC (aBIC). For AIC, BIC, and aBIC, values closer to 0 indicate a better fitting model (Vrieze, 2012). In addition, the Satorra-Bentler chi-squared difference test is used to compare fit of nested models, where the null hypothesis states that there is no difference in fit (Satorra & Bentler, 2001).

Conditional growth curve models

Once the best-fitting unconditional growth curve model was identified, covariates and predictors were added to the model in order to determine whether parental rejection or consistent discipline predicts youth disclosure after accounting for the growth curve of youth disclosure. Growth factors were regressed on time-stable covariates (e.g., gender). Youth disclosure at each time point was regressed on time-varying covariates and predictors (e.g., socioeconomic status, parental rejection, consistent discipline). First, bivariate analyses were conducted in which each covariate or predictor was included on its own, and then a multivariate analysis where all covariates and predictors were included simultaneously. Finally, for the time-varying predictors of parental rejection and consistent discipline, interactions with age were tested.

All models were estimated in Mplus 8 (Muthén & Muthén, 2010) with a robust maximum likelihood estimator (MLR) that adjusts the standard errors and chi-square test to account for the non-normal (negatively skewed) distribution of the outcome variable (White, 1980; Yuan & Bentler, 2000). Missing data on covariates, predictors, or the outcome were handled using full information maximum likelihood (FIML), which uses all available non-missing data to maximize each individual’s likelihood function by assuming that missing values on one variable in the data are conditionally dependent on observed values of other variables in the data. FIML generally has greater efficiency and less bias than other methods for handling missing data (e.g., pairwise or listwise deletion, multiple imputation), under various missing data mechanisms (e.g., missing completely at random, missing at random, or missing not at random; Enders, 2001; Enders & Bandalos, 2001).

Results

Gender and Cohort Differences

Table 1 presents descriptive statistics and results from ANOVAs testing for gender and cohort differences in the study variables. For youth disclosure, significant cohort main effects indicated that the 9^th grade cohort had lower disclosure than the 2^nd and 4^th grade cohorts at each time point. Further, at wave 2, the 4^th grade cohort (who were at the time in the 5^th grade) also had lower disclosure than the 2^nd grade cohort (who were at the time in the 3^rd grade). The gender main effect for waves 2 and 3 indicated that males had lower disclosure than females. There was also a gender x cohort interaction at wave 1, indicating that the difference between the 4^th and 9^th grade cohorts was greater for males than for females.

Table 1.

Gender x cohort ANOVAs

Variable	F			M (SD)					Total R2
	Gender (df = 1)	Cohort (df = 2)	Gender x cohort interaction (df = 2)	Female	Male	Grade 2	Grade 4	Grade 9
Adolescent disclosure
Wave 1 (N = 1280, df = 1274)	1.31	15.91***	3.18*	3.53 (0.75)	3.49 (0.75)	3.61 (0.68)	3.57 (0.69)	3.34 (0.84)	0.03
Wave 2 (N = 1136, df = 1130)	4.30*	21.44***	0.57	3.50 (0.77)	3.41 (0.80)	3.63 (0.62)	3.49 (0.82)	3.26 (0.86)	0.04
Wave 3 (N = 1061, df = 1055)	9.83**	29.57***	1.11	3.50 (0.80)	3.36 (0.83)	3.58 (0.73)	3.53 (0.71)	3.15 (0.95)	0.06
Socioeconomic status
Wave 1 (N = 1285, df = 1279)	1.58	0.04	1.44	−0.01 (0.82)	−0.06 (0.79)	−0.03 (0.77)	−0.03 (0.81)	−0.05 (0.83)	0.00
Wave 2 (N = 1146, df = 1140)	0.26	0.37	0.69	−0.02 (0.80)	−0.05 (0.79)	−0.03 (0.78)	−0.02 (0.83)	−0.07 (0.77)	0.00
Wave 3 (N = 1060, df = 1054)	0.40	0.92	1.06	−0.03 (0.79)	−0.06 (0.82)	−0.07 (0.79)	−0.00 (0.83)	−0.08 (0.80)	0.00
Parental rejection
Wave 1 (N = 1293, df = 1287)	20.75***	4.87**	1.25	0.34 (0.27)	0.40 (0.26)	0.35 (0.25)	0.36 (0.26)	0.40 (0.28)	0.03
Wave 2 (N = 1149, df=1143)	25.07***	0.63	1.25	0.32 (0.25)	0.40 (0.27)	0.35 (0.25)	0.37 (0.27)	0.36 (0.27)	0.02
Wave 3 (N = 1069, df=1063)	28.83***	0.48	1.06	0.32 (0.25)	0.40 (0.26)	0.36 (0.25)	0.35 (0.27)	0.37 (0.27)	0.03
Consistent discipline
Wave 1 (N = 1293, df=1287)	0.86	2.22	0.50	3.10 (0.60)	3.06 (0.62)	3.07 (0.62)	3.12 (0.58)	3.04 (0.63)	0.01
Wave 2 (N = 1149, df=1143)	0.44	3.57*	0.56	3.09 (0.63)	3.07 (0.62)	3.12 (0.60)	3.11 (0.61)	3.01 (0.65)	0.01
Wave 3 (N = 1069, df=1063)	7.38**	6.84**	2.07	3.14 (0.58)	3.05 (0.62)	3.15 (0.55)	3.13 (0.59)	3.00 (0.64)	0.02

^*p < 0.05,

^**p < 0.01,

^***p < 0.001

For socioeconomic status, there were no main effects for gender or cohort and no gender x cohort interaction. Thus, SES did not significantly differ for males and females or by cohort. Regarding parental rejection, the main effect of gender indicated that males had higher levels of parental rejection than females. The main effect of cohort at wave 1 indicated that the 9^th grade cohort had higher levels of parental rejection than the 2^nd and 4^th grade cohorts. For consistent discipline, the main effect of cohort indicated that the 9^th grade cohort had lower levels of consistent discipline than the 2^nd and 4^th grade cohorts at waves 2 and 3. Finally, the main effect of gender at wave 3 indicated that females had higher levels of consistent discipline than males during this wave (youth were in grades 4, 6, and 11 during this assessment).

Bivariate correlations

As shown in Table 2 (and also on Table 2a–c), adolescent disclosure was, in general, positively associated with consistent discipline and negatively associated with parental rejection, and each variable was positively associated with itself over time. This was the case averaged across cohorts and also when each cohort was considered individually. Moreover, socioeconomic status was generally associated with consistent discipline, such that families with higher socioeconomic status also were more likely to use consistent discipline; however, these correlations were not significant for youth in the 9^th grade cohort.

Table 2.

Bivariate correlations among study variables, averaged across cohorts, for males and females

	Adolescent disclosure			Rejection			Consistent discipline			Socioeconomic status
	Wave 1	Wave 2	Wave 3	Wave 1	Wave 2	Wave 3	Wave 1	Wave 2	Wave 3	Wave 1	Wave 2	Wave 3
Adolescent disclosure
Wave 1	--	0.44***	0.40***	−0.25***	−0.11*	−0.13**	0.31***	0.25***	0.25***	0.07	0.07	0.05
Wave 2	0.51***	--	0.52***	−0.19***	−0.20***	−0.15**	0.24***	0.33***	0.28***	0.05	0.04	0.03
Wave 3	0.45***	0.50***	--	−0.15***	−0.10*	−0.17***	0.24***	0.26***	0.31***	0.02	0.01	0.02
Rejection
Wave 1	−0.28***	−0.21***	−0.19***	--	0.70***	0.66***	−0.38***	−0.27***	−0.27***	−0.08*	−0.07	−0.03
Wave 2	−0.18***	−0.24***	−0.19***	0.67***	--	0.68***	−0.28***	−0.36***	−0.33***	−0.10*	−0.08	−0.04
Wave 3	−0.20***	−0.17***	−0.24***	0.65***	0.66***	--	−0.24***	−0.20***	−0.34***	−0.08	−0.06	−0.03
Consistent discipline
Wave 1	0.24***	0.13**	0.18***	−0.42***	−0.32***	−0.30***	--	0.70***	0.61***	0.20***	0.24***	0.22***
Wave 2	0.17***	0.17***	0.21***	−0.32***	−0.38***	−0.31***	0.66***	--	0.66***	0.21***	0.20***	0.19***
Wave 3	0.21***	0.14**	0.30***	−0.34***	−0.33***	−0.41***	0.61***	0.65***	--	0.15***	0.13**	0.13**
Socioeconomic status
Wave 1	−0.01	−0.03	−0.02	−0.08*	−0.03	0.02	0.14***	0.17***	0.07	--	0.89***	0.87***
Wave 2	0.01	−0.02	−0.06	−0.07	−0.02	0.03	0.16***	0.16***	0.05	0.84***	--	0.90***
Wave 3	−0.01	−0.04	−0.04	−0.02	0.03	0.02	0.05	0.08	0.06	0.81***	0.85***	--
(a) Bivariate correlations among study variables, for the 2nd grade cohort, for males and females
Adolescent disclosure
Wave 1	--	0.39***	0.34***	−0.22**	−0.15	−0.10	0.16*	0.16	0.12	0.07	0.02	0.05
Wave 2	0.41***	--	0.32***	−0.31***	−0.32***	−0.22**	0.17*	0.25**	0.07	0.20*	0.15	0.21*
Wave 3	0.51***	0.45***	--	−0.09	−0.12	−0.15	0.14	0.20*	0.19*	0.11	0.09	0.13
Rejection
Wave 1	−0.25***	−0.08	−0.21**	--	0.76***	0.66***	−0.25***	−0.23**	−0.15	−0.11	−0.07	−0.04
Wave 2	−0.06	−0.14	−0.19*	0.66***	--	0.66***	−0.19*	−0.33***	−0.24**	−0.08	0.01	0.00
Wave 3	−0.18*	−0.03	−0.19*	0.66***	0.65***	--	−0.14	−0.12	−0.20*	−0.09	−0.04	0.00
Consistent discipline
Wave 1	0.22**	−0.00	0.15*	−0.35***	−0.21**	−0.26***	--	0.77***	0.62***	0.27***	0.34***	0.32***
Wave 2	0.09	0.01	0.15*	−0.30***	−0.36***	−0.34***	0.70***	--	0.65***	0.24**	0.25***	0.29***
Wave 3	0.17*	−0.00	0.24***	−0.30***	−0.25***	−0.33***	0.64***	0.62***	--	0.16*	0.15	0.17*
Socioeconomic status
Wave 1	−0.05	−0.01	−0.00	−0.13	−0.09	0.03	0.19**	0.13	0.01	--	0.93***	0.91***
Wave 2	0.03	−0.02	−0.06	−0.15*	−0.06	0.07	0.26***	0.14*	−0.01	0.85***	--	0.92***
Wave 3	0.02	−0.07	0.01	−0.06	0.05	0.08	0.08	−0.02	−0.01	0.80***	0.83***	--
(b) Bivariate correlations among study variables, for the 4th grade cohort, for males and females
Adolescent disclosure
Wave 1	--	0.45***	0.53***	−0.28***	−0.14*	−0.15*	0.38***	0.32***	0.37***	0.06	0.06	0.02
Wave 2	0.46***	--	0.54***	−0.09	−0.11	−0.09	0.19*	0.34***	0.29***	0.04	0.03	0.05
Wave 3	0.31***	0.39***	--	−0.16*	−0.08	−0.12	0.29***	0.24**	0.28***	−0.04	−0.03	−0.00
Rejection
Wave 1	−19**	−0.13	−0.11	--	0.68***	0.67***	−0.40***	−0.32***	−0.39***	−0.10	−0.12	−0.07
Wave 2	−0.20**	−0.18*	−0.16*	0.65***	--	0.72***	−0.31***	−0.44***	−0.47***	−0.14	−0.14*	−0.09
Wave 3	−0.12	−0.10	−0.14	0.69***	0.64***	--	−0.25***	−0.30***	−0.43***	−0.09	−0.07	−0.04
Consistent discipline
Wave 1	0.15*	0.00	0.08	−0.43***	−0.27***	−0.27***	--	0.65***	0.62***	0.24***	0.30***	0.26***
Wave 2	0.16*	0.14	0.16*	−0.33***	−0.42***	−0.31***	0.64***	--	0.68***	0.27***	0.26***	0.23**
Wave 3	0.14	−0.01	0.20**	−0.33***	−0.39***	−0.43***	0.58***	0.66***	--	0.17*	0.17*	0.13
Socioeconomic status
Wave 1	−0.05	0.01	−0.02	−0.07	−0.13	0.02	0.19**	0.33***	0.19*	--	0.90***	0.88***
Wave 2	−0.03	0.01	−0.02	−0.04	−0.13	−0.01	0.19**	0.31***	0.18*	0.83***	--	0.91***
Wave 3	−0.03	−0.01	−0.04	−0.06	−0.13	−0.06	0.18*	0.30***	0.20**	0.80***	0.88***	--
(c) Bivariate correlations among study variables, for the 9th grade cohort, for males and females
Adolescent disclosure
Wave 1	--	0.44***	0.32***	−0.23**	−0.06	−0.15	0.37***	0.25***	0.18*	0.09	0.13	0.08
Wave 2	0.57***	--	0.57***	−0.21**	−0.25***	−0.18*	0.36***	0.36***	0.36***	−0.00	−0.00	−0.05
Wave 3	0.42***	0.58***	--	−0.18*	−0.15	−0.26***	0.29***	0.32***	0.39***	0.04	0.00	−0.02
Rejection
Wave 1	−0.33***	−0.33***	−0.16*	--	0.71***	0.64***	−0.46***	−0.27***	−0.23**	−0.05	−0.01	0.03
Wave 2	−0.26***	−0.38***	−0.21**	0.69***	--	0.66***	−0.34***	−0.34***	−0.26**	−0.07	−0.08	−0.00
Wave 3	−0.24**	−0.33***	−0.36***	0.58***	0.69***	--	−0.32***	−0.17*	−0.37***	−0.05	−0.06	−0.03
Consistent discipline
Wave 1	0.30***	0.34***	0.24**	−0.47***	−0.47***	−0.36***	--	0.68***	0.60***	0.10	0.06	0.08
Wave 2	0.20**	0.28***	0.26**	−0.29***	−0.35***	−0.26**	0.65***	--	0.65***	0.12	0.07	0.07
Wave 3	0.24**	0.31***	0.34***	−0.33***	−0.32***	−0.45***	0.63***	0.65***	--	0.12	0.06	0.09
Socioeconomic status
Wave 1	0.03	−0.11	−0.11	−0.03	0.15	0.04	0.04	0.02	−0.01	--	0.87***	0.83***
Wave 2	0.01	−0.09	−0.15	0.01	0.17*	0.06	0.00	0.01	−0.05	0.86***	--	0.87***
Wave 3	−0.05	−0.09	−0.14	0.09	0.22**	0.07	−0.11	−0.11	−0.07	0.83***	0.84***	--

Females above diagonal, males below diagonal. Ns ranged from 479 to 604 for females and 501 to 689 for males

^*p < 0.05,

^**p < 0.01,

^***p < 0.001

(a) Females above diagonal, males below diagonal. Ns ranged from 146 to 181 for females and from 169 to 228 for males

^*p < 0.05,

^**p < 0.01,

^***p < 0.001

(b) Females above diagonal, males below diagonal. Ns ranged from 178 to 225 for females and 179 to 242 for males

^*p < 0.05,

^**p < 0.01,

^***p < 0.001

(c) Females above diagonal, males below diagonal. Ns ranged from 154 to 198 for females and from 153 to 219 for males

^*p < 0.05,

^**p < 0.01,

^***p < 0.001

Unconditional Growth Curve Models

First, unconditional growth curve models for youth disclosure were conducted. The following models were tested: an intercept only model, a linear change model, and a quadratic change model. All loadings and parameters were constrained to be equivalent across age cohorts. Model fit statistics for these models are listed in Table 3 (intercept only model fixed, linear growth model fixed, and quadratic growth model fixed). Adding a linear growth factor for slope significantly improved model fit compared to the intercept only model, as indicated by smaller values for AIC, BIC, aBIC, RMSEA, and SRMR, and by larger values for CFI and TLI. In addition, the Satorra-Bentler chi-square difference test was significant (χ²(3) = 58.38, p < 0.001), indicating that the linear growth model fit the data significantly better than the intercept only model. However, adding a quadratic growth factor did not appear to significantly improve model fit compared to the linear growth model, as indicated by similar values for fit indices and a non-significant Satorra-Bentler chi-square difference test (χ²(4) = 8.34, p = 0.080).

Table 3.

Growth Factor Parameters and Model Fit Statistics for Unconditional Growth Curve Models of Adolescent Disclosure

	Intercept only model Fixed	Linear growth model Fixed	Quadratic growth model Fixed	Linear growth model Free
χ2 (df)	131.94*** (58)	66.60 (55)	58.46 (51)	29.38 (20)
AIC	7390.00	7287.66	7282.74	7295.76
BIC	7462.77	7376.03	7391.90	7566.07
aBIC	7418.302	7322.03	7325.19	7400.89
RMSEA	0.09	0.04	0.03	0.05
CFI	0.75	0.96	0.98	0.97
TLI	0.90	0.98	0.99	0.96
SRMR	0.41	0.27	0.27	0.15
I mean	3.49***	3.71***	3.62***
I variance	0.27***	0.19**	0.20*
S mean		−0.46***	0.04
S variance		0.32	2.55
Q mean			−0.48**
Q variance			2.69
I with S		−0.15	−0.28
I with Q			0.05
S with Q			−0.88*

AIC Aikaike Information Criterion, BIC Bayesian Information Criterion, aBIC sample size adjusted BIC, RMSEA root mean square error of approximation, CFI comparative fit index, TLI Tucker Lewis Index, SRMR standardized root mean squared residual, I intercept, S linear slope, Q quadratic change. N for all models is 1337

^*p < 0.05,

^**p < 0.01,

^***p < 0.001

Therefore, the best-fitting growth curve model for youth disclosure from ages 7 through 18 (corresponding to grades 2 through 11) was a linear growth model, with a high intercept and significantly decreasing slope, and significant variability around the intercept but not around the slope. To investigate whether there were age-related differences in the latent growth curve, we next tested whether relaxing equality constraints on the means and variances of growth parameters across age cohorts would improve model fit (see Table 1, linear growth model free). Relaxing the equality constraints did not appear to improve model fit, as indicated by greater values for AIC, BIC, and aBIC, similar values for RMSEA, CFI, and TLI, and a non-significant Satorra-Bentler chi-square difference test (χ²(35) = 37.40, p = 0.360). Therefore, the latent growth curve of youth disclosure appeared to be similar from ages 7 through 18 (corresponding to grades 2 through 11) regardless of youth cohort. The age-specific means of youth disclosure across cohorts, along with the growth curves fixed by age from the first three models included in Table 3 are plotted in Fig. 1. Age-specific means separated by cohort, as well as cohort-specific linear growth curves, are plotted in Fig. 2.

Fig. 1.

Age-specific sample means and estimated growth curves of adolescent disclosure from ages 7 to 18. Note that y-axis does not begin at 0

Fig. 2.

Age cohort-specific sample means and linear growth curves of adolescent disclosure from ages 7 to 18. Note that the y-axis does not begin at 0

Conditional Growth Curve Models

Next, the covariates and predictors were added to the model to see whether consistent discipline and parental rejection would predict levels of youth disclosure from ages 7 to 18, after accounting for the growth curve of youth disclosure. First, separate bivariate models were tested where we added each covariate or predictor separately, and then combined models were tested where the covariates and predictors were entered together. For the covariates, gender is a time-stable covariate predicting the intercept and slope factors of youth disclosure, while socioeconomic status is a time-varying covariate predicting youth disclosure at each age. For the predictors, consistent discipline and parental rejection were time-varying and predicted youth disclosure at each age. The coefficients from these models were originally constrained to be equivalent at each age, and are listed in Table 4.

Table 4.

Predictors of Youth Disclosure

	Bivariate (separate models)		Multivariate (same model)
	Est.	(S.E.)	Est.	(S.E.)
Time-stable covariates
Predicting intercept
Gender	0.01	(0.06)	0.01	(0.05)
Predicting slope
Gender	0.17	(0.11)	−0.10	(0.10)
Time-varying covariates
SES	0.00	(0.02)	−0.03	(0.02)
Time-varying predictors
Consistent discipline	0.29***	(0.03)	0.23***	(0.03)
Parental rejection	−0.61***	(0.06)	−0.42***	(0.06)

N for the multivariate model and bivariate gender model is 1359; N for the bivariate models with SES, consistent discipline, and parental rejection is 1341

^***p < 0.001

In both the bivariate and combined models, gender was not significantly associated with the initial level of youth disclosure or with the decreasing trend of youth disclosure, indicating that males and females have similar patterns of decreasing disclosure from childhood into adolescence. In addition, after accounting for the growth curve of youth disclosure, socioeconomic status is not associated with youth disclosure from childhood to adolescence.

For the main predictors, consistent discipline was positively and significantly associated with disclosure, and parental rejection was negatively and significantly associated with youth disclosure (p < 0.001 for both, across models). Thus, consistent discipline is associated with greater youth disclosure, and parental rejection is associated with lower youth disclosure, after accounting for the general developmental trajectory of disclosure and the influences of gender and socioeconomic status.

To assess whether the associations of consistent discipline and parental rejection with disclosure varied by age, a model was tested that included age interactions. The interaction was significant for consistent discipline (b = 0.02, p = 0.002), indicating that the association between consistent discipline and youth disclosure increased by 0.02 with each year of age, from 0.12 at age 7 (p = 0.004) to 0.39 at age 18 (p < 0.001). Thus, as youth disclosure generally decreases from childhood to adolescence, the association of parents’ consistent discipline with youth disclosure becomes stronger at later ages, although it remained significant at all included ages. The age interaction for parental rejection, however, was not significant (b =−0.02, p = 0.226), suggesting that parental rejection is similarly associated with lower youth disclosure throughout childhood and adolescence. It was also tested whether freely estimating the coefficients at each age would improve model fit, but a model with age-specific coefficients for consistent discipline and parental rejection did not fit any better than the model with age interactions.

Sensitivity Analysis

To ensure that our choice to model disclosure by age rather than grade did not substantively change results, a robustness check was conducted by additionally modeling disclosure by grade. The results from these models are substantively similar to the main results presented above, with the following exceptions: in the bivariate model, males decreased in disclosure more rapidly than females (b =−0.22, p = 0.048), but this was not the case in the multivariate model; and consistent discipline did not significantly interact with grade in predicting youth disclosure (p = 0.096).

Discussion

Youth who engage in routine disclosure with their parents (i.e., those who openly share information about their daily activities and whereabouts; Tilton-Weaver et al., 2014) are at lower risk for engaging in a variety of types of problem behavior (Frijins et al., 2005; Goldstein, 2016; Laird & Marrero, 2010; Stattin & Kerr, 2000). Although past research has explored the development of disclosure during adolescence (e.g., Laird et al., 2013), the development of disclosure in youth during the transition from childhood to adolescence has received minimal research attention. The current study addresses this gap in the research.

A few key points about disclosure during the transition to adolescence warrant highlighting. First, consistent with our expectations, disclosure declines over time from middle childhood through mid/late adolescence. Youth shared fewer aspects about their lives over time with their parents; this confirmed our first hypothesis. Second, two aspects of parenting during childhood and early adolescence predicted later youth disclosure: consistent discipline and parental rejection; greater consistency and lower rejection predicted increased youth disclosure (hypothesis two and three). Finally, consistent discipline was especially important for adolescents in predicting disclosure, as compared to younger children (it was important overall, just more important with increasing age). This last finding partially confirms hypothesis four. Each of these findings is discussed in greater detail below in a broader context.

The current results show that youth disclose less information to their parents about their daily lives as they get older. As youth age from middle childhood to late adolescence, they offer less information to their parents about their daily whereabouts and activities. This pattern was consistent across gender and socioeconomic status. These findings compliment previous research on youth autonomy development (Simões et al., 2018; Steinberg & Morris, 2001; Teuber et al., 2022). As youth get older, they tend to engage in fewer discussions with their parents about their day-to-day activities. Part of this decrease may be related to changing conceptions of parental jurisdiction and parental authority over behaviors that youth deem to be impactful to only themselves (i.e., youth believe an increasing number of behaviors to be subsumed in their personal domains; e.g., choice of appearance or friendship choice). Starting in childhood (Tisak, 1986) and moving through adolescence (Smetana et al., 2005), youth prefer increased jurisdiction over personal-domain consistent behaviors with increased age; perhaps disclosing less to their parents is a mechanism through which they achieve this goal. It would be interesting for future research to explore this link further. As a potential complimentary interpretation, it may be that developmental changes moving towards reduction in disclosure across time may be about more than autonomy – these changes may reflect developmental changes in youth time use. By age 18, many adolescents are working and engaged in intensive extra-curricular activities (e.g., varsity sports schedules that require daily after-school practices). Thus, the changes in disclosure may not be solely due to autonomy development. It also may be that adolescents feel that their parents do not need to know or would not be interested in knowing all the details of everything older adolescents are doing (see Marshall et al., 2005) and that this, coupled with the fact that youth are around parents for less time during the day as they engage their various other time commitments, explains part of the developmental patterns that are seen in the current study with regard to disclosure.

Regardless of the reasons for the developmental trajectories of disclosure, there were still variations at each time point among the sample, which in the current study were shown to be associated with variations in parenting behaviors. The current results show that if parents want to continue to engage in relatively high levels of open discussion with their children as they age into adolescence, they need to engage in accepting and consistent parenting strategies during middle childhood to maintain this contact (as demonstrated by the two younger cohorts who spanned grades 2 through 6 in the current sample). Moreover, this is also true of parenting during the early adolescent to middle adolescent years (as evidenced by the oldest cohort, spanning grades 9 through 11). This builds upon previous research on positive parent-youth relationships and youth disclosure (Tilton-Weaver et al. 2010), showing that links found previously regarding adolescents and their parents (related to disclosure and positive parent-child relationships) are also relevant for understanding parent-child relationships during middle childhood and the transition to adolescence.

A focus of the current study was to better understand ways in which parent-adolescent relationship characteristics predict trajectories of youth disclosure. Specifically, consistent discipline positively predicted youth disclosure across adolescence; notably this influence became stronger with increased adolescent age. When parents followed through with disciplinary actions, youth disclosure was more likely. Perhaps adolescents appreciated the predictability and intentionality of parenting; maybe youth of parents who consistently disciplined their children felt more safe telling their parents what they were doing and where they were going. Parents may worry that setting boundaries for their adolescent children will push them away; however, the current results suggest that this is not generally the case. This is consistent with the “structure” aspect of the self-determination theory as applied to parenting (e.g., Ryan & Deci, 2017); youth in the current study were more comfortable disclosing about their lives to their parents when their parents applied consistent and reliable (but fair) disciplinary practices.

In contrast to consistent discipline, parental rejection negatively predicted adolescent disclosure over time. In the current study, parents who accepted their children (as opposed to rejecting them) have children who engage in relatively high amounts of disclosure. When parents were disapproving and unsatisfied with adolescent behavior, adolescents were less likely to open up to their parents and talk with them about their daily experiences and plans. This pattern remained stable across adolescence. Perhaps adolescents in these families were less comfortable discussing any type of behavior with their parents, in fear of additional disapproval and rejection. It is logical to withhold information that one feels will be criticized or that will disappoint; this is consistent with recent research that shows that adolescents were less likely to disclose when they perceived that their parents had responded negatively in the past (Tilton-Weaver et al. 2010). Thus, if adolescents are already experiencing rejection as they move about their day-to-day household activities that parents can observe, it makes sense that adolescents would be hesitant to volunteer additional information to be criticized. This finding is also consistent with the parental involvement aspect of self-determination theory as applied to parenting (e.g., Ryan & Deci, 2017); youth in the current study felt more comfortable disclosing to their parents when their parents provided a non-rejecting environment for their children. The research on parental mindfulness is also relevant (e.g., Goldstein et al., 2022); these findings are consistent with previous literature showing that youth disclose more to their parents when they perceive their parents to engage in parental mindfulness (nonjudgmental acceptance). In contrast to parental mindfulness, parental rejection as operationalized in the current study as harsh, disappointed, and disapproving parental judgment as it relates to a variety of mundane daily behaviors ranging from cleaning one’s room to doing homework. Youth in the current sample opened up less to their parents when they perceived their parents to be rejecting of them.

These results suggest that parents actively create an environment that either encourages or discourages youth disclosure during the pre-adolescent and early to middle adolescent years. When parents provide a consistent, accepting family environment, they can optimize the chances that their adolescent children will continue to keep them informed about their daily whereabouts and activities. In contrast, a rejecting and inconsistent home environment makes it less likely that youth will open up to their parents. The situations that adolescents can find themselves involved in can become increasingly complex, and having parental input and guidance as adolescents even as adolescents move to autonomy is beneficial. As adolescents spend an increasing amount of time engaged in these increasingly independent peer-focused leisure activities and interaction (Montemayor, 1982; Updegraff et al., 2006; Zeijl et al., 2000), it behooves parents to be aware of and informed about what their children are doing.

Previous research shows that problem behavior is most likely to occur during unsupervised, unstructured peer socialization (Flannery et al., 1999; Goldstein et al., 2005; Hsieh et al., 2022). Thus, if parents can remain knowledgeable about where their adolescents are and who they are with, they can continue to provide advice and suggestions to help keep their children safe and out of trouble. This is also relevant for considering adolescents’ online activities, which can also be construed as unsupervised peer interaction. Thus, the best way to retain this knowledge is through adolescent voluntary youth disclosure, which as shown here is facilitated by maintaining warm, accepting, and consistent family environments. The caveat, though, as noted previously, is that this parental input would ideally be contributed in a way that does not seem intrusive (e.g., Dietvorst et al., 2018; Hawk et al., 2009; Rote & Smetana, 2018). Further, as the current results show, this parental input also cannot be something that will be judgmental and disapproving.

Gender and socioeconomic status are important variables to consider in any analyses, and are particularly important to examine when studying change over time in adolescent behavior. This said, the current data showed no evidence of gender or socioeconomic differences in trajectories of youth disclosure over time. Despite the lack of differences in trajectories, some interesting gender differences in mean levels of study variables were found. Males experienced greater rejection than females overall, and in Wave 3 (when youth were in grades 4, 6, and 11), females experienced greater consistent discipline as compared to males. Thus, females overall received parenting that is more consistent with disclosure in these regards, at least from the perspective of the parents. Thus, the main effect of gender on disclosure for the second and third waves of the study is not surprising. Given the associations of parenting with adolescent disclosure in the current study, then, parents are advised to remain warm, positive, and consistent with all of their children, regardless of child gender.

The current design had many strengths, including the cohort sequential design, the relatively large sample, and the novel focus on the impact of both parental rejection and consistent discipline on youth disclosure. Even with these strengths, though, the current study is not without its limitations. First, the current study relied entirely on parental report of youth behavior and of parenting behavior. Although parents are arguably the best reporters of whether they are disappointed in their children and whether they consistently discipline their children, it would have been interesting to have youth insight into these issues as well. One concern with this is that parental perceptions and adolescent perceptions of parent-adolescent communication and parenting may not align (e.g., Kapetanovic & Boson, 2022). Unfortunately, though, youth data were not available for the variables used in the current study. Thus, it would be advantageous for future research to use multiple informant data for studies on similar topics. Relatedly, the current study considered routine disclosure, but not active forms of concealment (e.g., keeping secrets, lying) that differ from the absence of disclosure. It would be interesting and important for future research to examine the development of these constructs as well during the transition to adolescence, from multiple perspectives. Moreover, in the current investigation, most of the parent participants were mothers. Fathers’ participation should be encouraged in future research endeavors; -the amount and type of information disclosed by adolescents may vary depending on the gender of the parent.

Another important consideration when interpreting the results is that assessment for the current study began at grade 2 for the youngest cohort. It is very plausible that some of the relevant parent-child relational patterns for predicting disclosure begin before grade 2. Therefore, future research should include earlier time points to assess whether this is the case. Relatedly, although the current study utilized cohort sequential data spanning from grade 2 through grade 11, the design of the sample was such that youth in grades 7 and 8 were not assessed. The younger two cohorts had completed their data collection by grades 4 and 6 respectively, whereas the oldest cohort entered the sample when they were in grade 9. Therefore, important changes may have occurred during grades 7 and 8; youth in these grades should be assessed in future research. Finally, the measure of adolescent disclosure utilized in the current study was relatively limited (2 items), although the items do have good face and construct validity and were identical at each wave. Thus, future research should strive to include additional items in assessment endeavors.

Despite these limitations, results from the current investigation have a number of useful implications for parents, educators, and mental health workers. First, it is important for parents to understand that the information that they receive from their adolescent will likely decrease over time. Given this developmental trend, it behooves parents to interact with their adolescents using strategies that facilitate positive parent-adolescent relationships. In the current study, two specific parenting strategies were identified (consistent discipline and low rejection) that help to promote youth disclosure. Moreover, consistent discipline seems to be especially beneficial with increasing youth age. In other words, parents of middle- and high-school aged youth should be especially careful to follow through with their (developmentally appropriate and non-harsh) disciplinary initiatives. In the moment, teens may act like they do not appreciate these gestures (e.g., following through with a punishment for violating a family rule), but in the long run this disciplinary follow-through seems to facilitate healthy communication patterns between teens and parents. Because of the clear and consistent pattern of previous research linking adolescent disclosure with lower risk of challenges with psychosocial adjustment (Frijins et al., 2005; Goldstein, 2016; Laird & Marrero, 2010; Stattin & Kerr, 2000), it is very important for parents to be mindful of interacting with their child in a warm, accepting manner that balances this affinity with disciplinary consistency. Educators and mental health workers can encourage families to work towards these goals, especially in the context of parent education classes, school-home communicate initiatives, or family counseling.

A final note about parent rejection seems important to make. Adolescence is a time for identity exploration, self-discovery, and experimentation with different ideas and goals for the future (e.g., Branje et al., 2021; Eccles, 2009; Eccles et al., 2015). Adolescents will likely make many mistakes as they age. They may also be disorganized at times, and will lack direction in the ways that their parents would like them to have direction. The current study showed that when parents reject their adolescent children (operationalized here, essentially, as feeling disappointed in their children), adolescents are less likely to open up to their parents about their daily whereabouts and activities. Thus, parents are encouraged to do their best to accept their child for the individual that they are growing up to become, realizing that their children do not have to be perfect. If children do behave in a way that is not acceptable from the parents’ perspective (which they undoubtedly will at some point), it is beneficial to engage in consistent discipline and to follow through with punishments. However, it is important to do this without expressing negative, disapproving judgment. This is a nuanced point, but based on the current results it is an important negotiation for parents to make. By doing this, parents encourage an open, honest relationship with their adolescents that encourages disclosure.

Conclusion

This research addressed a key gap in the literature; although adolescent disclosure has been explored in past studies, very little work had focused on youth disclosure changes during the pivotal transition to adolescence. The present study examined trajectories of youth disclosure across the transition from childhood to adolescence. The current work also explored ways in which parenting behaviors during childhood and early adolescence predict concurrent and later disclosure behavior. The current work used a cohort sequential design (studying 3 cohorts of families annually for 3 years each; beginning in grades 2, 4, and 9 respectively) to examine these issues. Results indicated that youth disclosure declines as youth age. Moreover, disclosure was enhanced for youth with parents who engage in consistent discipline (especially for the adolescent cohort) and was reduced for youth with rejecting parents. These results were consistent across gender and socioeconomic status. Because adolescent disclosure is associated with lower levels of problem behavior (Laird et al., 2013), it behooves parents to facilitate a warm, consistent environment for their children so that as they age. Parenting during childhood predicts adolescent behavior when it comes to the theoretically and practically important construct of youth disclosure.

Biographies

Sara E. Goldstein is a Professor of Human Development and Family Sciences at the University of Delaware. Her research focuses on social development and social cognitive development across the lifespan, with a focus on parent-adolescent relationships, peer relationships during adolescence, and the development and maintenance of aggressive behavior, harassment, and bullying.

Meagan Docherty is an Assistant Professor of Psychology at Bowling Green State University. Her research focuses on the development of aggression and related attitudes, traits, and behaviors; developmental precursors and outcomes of callous-unemotional traits in childhood and adolescence; exposure to violence and harsh parenting; handgun use in adolescence and young adulthood.

Paul Boxer is a Professor at Rutgers University. His research focuses on the development and management of aggressive behavior in children and adolescents; evidence-based practices in the juvenile justice system; impact of exposure to violence on youth development across social contexts.

Brad Bushman is a Professor of Communication and Margaret Hall and Robert Randal Rinehart Chair of Mass Communication at The Ohio State University. For over 30 years, he has studied the causes, consequences, and cures to the problem of human aggression and violence.

L. Rowell Huesmann is the Amos Tversky Collegiate Professor Emeritus of Communication Studies and Psychology and Research Professor Emeritus, Research Center for Group Dynamics, Institute for Social Research at the University of Michigan. His research focuses on the environmental experiences and the social-cognitive mechanisms that promote the learning and development of violent behaviors and the contagion of violence.

Maureen O’Brien is a Senior Survey Specialist at the University of Michigan’s Survey Research Center. Her research interests include the development of aggressive behavior across the life span and evidence based practices that address aggressive behaviors.

Craig Anderson is a Distinguished Professor of Psychology at Iowa State University. His current work focuses on the General Aggression Model, a bio-social-cognitive model integrating aggression theories from social, personality, developmental, and cognitive psychology; much of his recent empirical work involves media effects, and on climate change effects on violence.

Douglas A. Gentile is a Distinguished Professor in Liberal Arts and Sciences at Iowa State University. His research interests include understanding the complex effects of media on children, adolescents, and adults, often through the lens of the General Learning Model, focusing on topics such as prosocial media, violent media, and Gaming Disorder.

Eric F. Dubow is a Distinguished Research Professor of Psychology at Bowling Green State University and a Research Professor at the Institute of Social Research at the University of Michigan. His research focuses on the development of aggression over time and across generations; risk and resource variables in youth adjustment; effects of exposure to violence across contexts on youth and young adult outcomes; development and evaluation of school-based interventions.

Appendix A

Parental Rejection Items (Eron et al., 1971)

Has this child shown enough responsibility in doing routine chores around the house?

Have you been bothered that this child does not follow directions when they run an errand?

Have you been satisfied with how this child takes care of their things?

When this child has money to buy something for themselves, have you been displeased with what they have bought?

Have you thought that this child wastes too much time?

Have you become annoyed because this child is not ready on time or is not where they are supposed to be?

Has this child been too forgetful?

Have you been satisfied with this child’s manners?

Have you thought that this child reads as well as can be expected for a child that age?

Has the quality of this child’s schoolwork been as good as it should be?

* Parents answered affirmatively (1) or said no (0) to each question.

Consistent Discipline Items (Capaldi & Patterson, 1989)

If you ask your child to do something and they do not do it, how often do you give up trying to get them to do it?

If you warn your child that they will be punished if they do not stop doing something, how often do you actually punish them if they do not stop?

How often does your child get away with things that you feel they should have been punished for?

How often do you feel that it is more trouble than it is worth to punish your child?

How often is your child successful in getting around the rules you have set for them?

How often are you discouraged with the results of your discipline efforts?

How often do you feel that you cannot pay as much attention as you would like to discipline your child?

When you give your child a command to do something, how often do you make sure they do it?

How much of the time do you feel you are able to change or correct your child’s behavior?

* Item responses ranged from 0 (never or almost never) through 4 (always or almost always).

Data Sharing Declaration

The current dataset is not currently available to the public. However, interested readers should contact the authors of the study with any associated inquiries about the data.