Parent-adolescent relationship and externalizing behavior : the role of sensory sensitivity

Parent-adolescent relationship and externalizing behavior.

The role of sensory sensitivity.

Masterthesis Forensische Orthopedagogiek Graduate School of Child Development and Education Universiteit van Amsterdam Student: J. van Wijk Number: 11861606 Mentor: mw. K. Fisher MSc Second assessor: dhr. prof. dr. G.J. Overbeek Amsterdam April 2019

2 Abstract

Externalizing behavior in adolescents is a major risk factor for later juvenile delinquency, adult crime and violence. Over the years, different scientists agreed that a dysfunctional parent-adolescent relationship influences the development of externalizing behavior. Some adolescents are more susceptible to a dysfunctional parent-adolescent

relationship than others. The present research investigated one of the traits that could make an adolescent more susceptible; sensory sensitivity.

Based on literature, we expected that (1) a negative parent-adolescent relationship is associated with later increases in child externalizing behavior and (2) adolescents with higher levels of sensory sensitivity and a negative parent-adolescent relationship will show more externalizing behavior compared to adolescents with lower levels of sensory sensitivity. Adolescents at High schools in Amsterdam were asked to fill in three self-report measures about externalizing behavior (YSR, delinquency and aggression), parent-adolescent relationship (NRI, parental support, negative interaction and power) and sensory sensitivity (HSC Scale). After a 12-month interval, 176 participants were included a correlation and moderation analysis en 166 participants were included in a two-way mixed ANOVA analysis.

This results showed that adolescents who experienced less support show more delinquent and aggressive behavior. Also adolescents who experienced a lot of negative interaction showed more delinquent and aggressive behavior. There was no difference in delinquent or aggressive behavior for the amount of parental control. Most of the associations did not change over a 12-month period. Only adolescents who experience little negative interaction seem to develop more aggressive behavior over a 12-month period. The present study showed that sensory sensitivity did not moderate the association between the parent-child relationship at time 1 and externalizing behavior at time 2. Suggestions about future research were given.

3 Introduction

It is important to investigate externalizing behavior in adolescents since it is a major risk factor for later juvenile delinquency, adult crime and violence (Betz, 1995; Moffitt, 1993; Okado & Bierman, 2015). Aggression and delinquency are two characteristic features of externalizing behavior, related to conduct problems (Stanger, Achenbach & Verhulst, 1997) and were found to be a strong indicator of adult crime and violence (Moffitt, 1993). Aggression is characterized by physical or verbal behaviors that harm or threaten others, including children, adults and animals (American Psychiatric Association, 1994) whereas delinquency is characterized by violent behavior not directly to others (Achenbach, 1991).

Examples of aggressive behavior are attacking others, bullying, cruelty and fighting, examples of delinquent behavior are stealing, swearing, substance abuse, truancy and vandalism (Fergusson, Horwood & Lynsky, 1994; Gillmore et al., 1991; Lahey & Loeber, 1994; Rey & Morris-Yates, 1993). Some children show both aggressive and delinquent behavior, howevermost children only show one of these behaviors (Stanger et al., 1997). For example, McConaughy and Achenbach (1994) measured only a comorbidity rate of 28% in the USA. This small amount of comorbidity is the reason why aggression and delinquency are considered two different types of externalizing behavior.

Research shows that there is a strong association between a dysfunctional parent-adolescent relationship and parent-adolescent’s externalizing behaviors such as aggression and delinquency (e.g., Chassin, Pillow, Curran, Molina & Barrera., 1993; Lamborn, Mounts, Steinberg, & Dornbusch, 1991; Pinquart, 2017; Reuben et al., 2016). Parental support seems to be one of the primary predictors of adolescents’ psychological and/or behavioral control (Costigan, 1996; Forehand & Nousiainen, 1993). Parent-child conflict seems to act as a vulnerability that increases risk for multiple externalizing disorders, like attention-deficit/hyperactivity disorder (ADHD), oppositional defiant disorder (ODD) and conduct

4 disorder (CD) (Burt, Krueger, McGue & Iacono, 2003). Research shows that parent’s rejective and unresponsive behavior towards their children seems to encourage their children to behave in socially unacceptable ways, resulting in the development of externalizing behavior (Rothbaum & Weisz, 1994). Last, Pinquart (2017) showed with a meta-analysis that psychological and harsh control showed the strongest associations with externalizing problems. There are certain factors that are hypothesized to make some children more susceptible to a negative parenting environment than others (Bakermans-Kranenburg & van IJzendoorn, 2011). For example, there is scientific support that certain genes can buffer or enhance the effects of a negative parenting environment on children’s development (Belsky & Pluess, 2009). This susceptibility to a stressful environment could be explained by the diathesis-stress model (Monroe & Simons, 1991; Zuckerman, 1999); some individuals carry “risk” genes that make them disproportionately susceptible to adverse environmental conditions (Jolicoeur-Martineau et al., 2017). Whereas genes have been more commonly investigated as a susceptibility factor moderating the effects of dysfunctional parenting environments on child outcomes, there are other moderating factors that have been investigated to a lesser extent. One of those factors is the so-called sensory sensitivity. Research shows that sensory sensitivity is able to function as a moderator (Booth, Standage & Fox, 2015; Karam et al., 2019). For example, sensory sensitivity seems to be a moderator between negative childhood experiences and adult life satisfaction (Booth et al., 2015). There was a difference observed by those who reported negative childhood experiences; adults with higher levels of sensory sensitivity reported lower life satisfaction, which supports the diathesis-stress model.

Some psychology studies have considered sensory sensitivity as a personality trait (Aron & Aron, 1997; Aron, Aron & Jagiellowicz, 2012). Sensory sensitivity refers to a tendency to more strongly process a variety of external stimuli (Aron & Aron, 1997). Highly sensory sensitive adolescents seem to be more aware of details in their physical environment, recognize

5 body language faster, experience touches more strongly and smell, taste and hearing are well developed. Research suggests that this leads to an easier recognition of positive and negative face expressions (Jagiellowicz, Aron & Aron, 2016), a more severe pain sensation (Wyller, Wyller, Crane & Gjelsvik, 2017) and being more aware of other people’s moods (Aron & Aron, 1997). Approximately 25% of the population is hypothesized to be highly sensitive (Aron & Aron, 1997).

One aspect of being highly sensory sensitive is being more aware of physical signals and emotional cues from other people (Aron & Aron, 1997; Aron et al., 2012). Literature supports that this awareness results in the ability to learn more from social interactions (Aron, Aron & Davies, 2005; Baumeister, Vohs, DeWall & Zhang, 2007). Learning from social interactions leads to more successful responses in future similar situations (Aron et al., 2005; Baumeister et al., 2007). For example, if an adolescent uses dysfunctional behavior in this interaction with someone else, he can recognize this behavior by a negative response (Aron et al., 2005; Baumeister, et al., 2007). Next time, he knows it was dysfunctional behavior, so he will try a different behavior in the hope that this provokes a positive reaction. Highly sensory sensitive adolescents are more aware of positive and negative reactions and will feel more negative after a negative interaction. This leads to a faster adjustment of behavior.

Although highly sensory sensitive adolescents learn more from social experiences, they also get more easily overwhelmed by stimuli because of their higher sensitivity (lower threshold) (Aron et al., 2012). Freedman et al. (1987) suggests that disruption of pre-attentional processes may interfere with filtering of inappropriate stimuli and may lead to overstimulation or insufficient modulation of behavior responses. The sensory disturbance may be described as the inability to filter out environmental stimuli with a subsequent impaired ability to focus on attention (Serafini et al., 2017); These adolescents can experience those stimuli as too intense, frequent or novel (Aron et al., 2012). According to Dunn (1997, 2001) there are two

6 patterns that refer to a lower threshold: (1) individuals who are sensation avoiders and are usually described as introspective or reclusive, since they actively limit exposure to sensory information (sensation avoiders); (2) individuals who are sensory sensitive and feel discomfort with regular sensations but do not actively limit their exposure to the uncomfortable stimuli (sensory sensitive). So sensory sensitive adolescents get more easily overwhelmed, but do not actively avoid sensation.

Serafini and colleagues (2017) were one of the first researchers who focused on sensory sensitivity and impulsivity. They found that high sensory sensitivity correlates with greater attentional and motor impulsivity. Attentional control may be especially important for reducing anger that is implicated in reactive (emotionally driven) aggression and defiance (Eisenberg et al., 2009). So it might be that highly sensory sensitive adolescents have a greater attentional impulsivity, which might lead to showing more externalizing behavior. In the present research it is hypothesized that the overwhelming of sensory sensitive adolescents will result in externalizing behavior.

The present study

The present study uses longitudinal data from the Development and Parent Peer Transactions (ADAPT) study (N = 177) to determine whether sensory sensitivity is a moderator for the association between parent-adolescent relationship and externalizing behavior.

This leads to the following hypothesis: (1) We expect that a negative parent-adolescent relationship is associated with later increases in child externalizing behavior. (2) We expect that adolescents with higher levels of sensory sensitivity and a negative parent-adolescent relationship will show more externalizing behavior compared to adolescents with lower levels of sensory sensitivity.

7 Method

Sample

To answer the aforementioned questions, the data of the longitudinal ADAPT study was used. This was a self-report study among adolescents about learning and motivation, personality, temperament, sensory sensitivity, bullying, relation with their best friend and parents, externalizing- and internalizing behavior, social phobia, life events and alcohol and smoking. The data was collected in 2 waves, with a 12-month interval. 177 adolescents participated in both waves. The participants were going to three different schools in Amsterdam, the Netherlands, and approximately half of them where boys (54%). The age of the participants at the start of the study was between 11.30 and 16.53 years (M = 13.40, SD = 1.06). Most of the participants attended the highest level of Dutch High schools (79.5%).

Procedure

Letters were sent to schools in Amsterdam and followed up by a phone call to ask about participating in the study. When the schools agreed to participate, master students went to the participating schools to complete data collection, which consisted of distributing the questionnaire and supervising the classroom while they were completed. After one year, the same adolescents were approached to fill in the questionnaire again. Parents gave passive consent. The study has been approved by the Education and Child Studies Ethics Committee of the University of Amsterdam.

Measures

Sensory processing sensitivity. Sensory sensitivity was measured by the High Sensitive Child Questionnaire (HSC Scale; Pluess et al., 2018). This version had twelve questions with three factors: ease of excitation, aesthetic sensitivity and low sensory threshold.

8 Items like ‘I love nice smells’ and ‘When someone is sad, that makes me feel sad too’ had to be rated by the participants on a scale from 1 (not at all) to 7 (extremely). Higher total scores indicate higher levels of sensitivity. Pluess and collegues (2018) tested the reliability of this twelve item version on a 8 to 11 year old English sample (M = 9.82, SD = .45), and found an acceptable (Barker, Pistrang & Elliot, 1994) internal consistency (α = .71 and α = .74). The test-retest reliability was questionable (r = .68). Pluess and colleagues (2018) also tested the twelve item version on adolescents from 15 to 19 years old (M = 17.06, SD = .88). This study showed a good internal consistency for the total scale (α = .82). In the present study, the internal consistency of the first (α = .74) and second (α = .78) wave were acceptable for the total scores. Externalizing behavior. Externalizing behavior was measured by the Youth Self-report (YSR; Achenbach, 1991), Dutch version. The two factors measuring the externalizing behavior are aggression (17 items) and delinquency (15 items). An example item for aggression is: ‘I argue a lot’ and for delinquency: ‘I drink a lot of alcohol without approval from my parents’. Items are assessed on a 3-point Likert scale ( 0 = ‘not true’, 1 = ‘somewhat or sometimes true’, and 3 = ‘very true or often true’) (de Groot, Koot & Verhulst, 1996). The reliability and validity data of both scales provided by Achenbach (1991) had an acceptable internal consistency (α > .75) and test–retest reliability (p > .60) (Song, Singh & Singer, 1994). In the present study, the internal consistency of the first wave was good for aggression (α = .86) and delinquency (α = .83). In the second wave, the internal consistency was good for aggression (α = .83) and for delinquency (α = .81) as well.

Parent-adolescent relationship quality. The quality of adolescents’ relationship to parents was measured by the Network of Relationships Inventory – Short form (NRI; Furman & Buhrmester, 1985), Dutch version. There are three subscales: ‘Support’ (8 items), ‘negative interaction’ (6 items) and ‘power’ (6 items). In this research, negative interaction between the parents and adolescent are characterized by conflict, criticism and antagonism. Power was

9 defined by how much power the parents had in the life of their child, like making decisions for them or telling what to do. Example items include: ‘Do your parents admire you?’ (‘support’) ‘Do you and your parents annoy each other?’ (‘negative interaction’) and ‘How often do your parents get what they want when you disagree?’ (‘power’). Participants rate on a 5-point Likert scale how much such an interaction occurs, with the scale ranging from 1 (‘little or none’) to 5 (‘could not be more’). Furman and Buhrmester (2009) investigated the reliability of the NRI subscales support and negative interaction. The subscales ‘support’ and ‘negative interaction’ were minimally related (r = -.30, p < .001 to r = .07, p = .86). The test-retest reliability for both subscales was good (r = .67 to . 70, p < .01). The test-retest reliability of the scale ‘Power’ has not yet been tested (Furman & Buhrmester, 2009).In de present study, the internal consistency of the first wave was acceptable for support (α = .76), excellent for negative interaction (α = .91) and acceptable for power (α = .76). The internal consistency of the second wave was acceptable for support (α = .79), excellent for negative interaction (α = .94) and good for power (α = .83).

Analysis

Statistical analyses were performed using version 22 of the Statistical Package of Social Sciences (SPSS), with the significance level set at α < .05. Descriptive statistics, including boxplots were conducted and used for the purposes of checking the data distribution. Internal consistencies were calculated using Cronbach’s alpha. Externalizing behavior was divided into delinquency and aggression. The parent-adolescent relationship was divided into three variables: Parental support, negative interaction and power. Pearson correlational analyses were computed between all variables, to investigate associations among variables for time 1 (T1) and time 2 (T2).

10 To investigate whether a negative parent-adolescent relationship was associated with later increases in child externalizing behavior, a two-way mixed ANOVA (GLM Repeated Measures) was performed (Leard Statistics, 2015). Only participants who filled in the YSR both times were included in the analysis. Because SPSS does not have the option to control for an earlier time measure with only continues variables, the independent variables (between-subject variables) had to be divided into groups for appropriate analyses. The parent-adolescent relationship variables (support, negative interaction and power) were used as between-subject variables and were thus divided into three groups to create categorical variables, needed for the two-way mixed ANOVA. When the variable was normally distributed, the groups were divided based on the standard deviation (SD). A low score was defined as a score smaller than the mean score – 1SD, an average score was defined between mean – 1SD and mean + 1SD and a high score was defined as a score higher than the mean + 1SD. When the variable was not normally distributed, the groups were created based on the percentiles, namely low scores (<.25 percentile), average scores (<.25 score < .75 percentile) and high scores (>.75 percentile).

After inspecting the data for outliers and normal distributions, a two-way mixed ANOVA was performed for every combination of dependent variables (within-subject factor, delinquency and aggression), and between-subject factor (support T1, negative interaction T1 and power T1). The within-subject factors delinquency and aggression were defined by T1 and T2. When the two-way mixed ANOVA revealed a significant difference between the groups of the between-subject factor, a Bonferroni post-hoc test was performed. The Bonferroni post-hoc test is a recommended (Maxwell & Delaney, 2004) test for the purpose of testing all possible pairwise combinations of levels of the within-subjects factor. It provides the statistical significance level (p – value) for each pairwise comparison. The two-way mixed ANOVA and Bonferroni post-hoc test are not suitable for examining an increase or decrease over time within groups. Therefore an ANOVA profile plot was created to examine if an increase or decrease of

11 the dependent variable over time seemed present within the groups of the between-subject factor. When a difference is expected based on visual inspection, a paired-samples t-test was performed for the groups in which a difference was expected to provide a p-value.

Moderation analysis.

In order to test the moderating effect of sensory sensitivity (T1) on the relationship between the adolescent-parent relationship (T1) and externalizing behavior (T2), multiple regression analyses were performed (Cohen, Cohen, West & Aiken, 2003). Before testing for moderating effect, the independent variables and moderator were centralized (Methodologiewinkel, n.d.). To do so, the means of sensory sensitivity, support, negative interaction and power were calculated and used to center the variables around zero. Centering was used to ease the interpretation of regression models (Lohse, Bland & Lang, 2016). Next, an interaction variable was created between the centered independent variables (support T1, negative interaction T1 and power T1) and the centered moderator (sensory sensitivity T1), for instance support centered * sensory sensitivity centered. Last, the multiple regression analysis was executed for every combination of variables separately. For instance; delinquency T2, support centered T1, sensory sensitivity centered T1 and the interaction between support centered and sensory sensitivity centered T1. The coefficients table was interpreted to examine the interaction (moderation) effect (Howitt & Cramer, 2011).

Results Descriptive statistics

The sample consisted of 177 participants. We examined all variables to determine if the variables were normal distributed (z-skewness < 3.29; West, Finch & Curran, 1995). One extreme value was deleted, because this participant used hard drugs which resulted 176 participants. Delinquency T1 and T2, aggression T1 and T2, support T1 and T2 and negative

12 interaction T2 were not normal distributed (See Table 1). This corresponds to the normal population, because there were less adolescent who showed aggressive and delinquent behavior compared to adolescents who did not. Also parenting styles with more support and less negative interaction are normal in a non-clinical population. Therefore other extreme values were not considered as abnormal for a non-clinical population.

For the Two-way Mixed ANOVA, only the participants who filled in the YSR both times were included. This resulted in a population of 166 participants. 84 boys (50.6%) and 81 girls (48.8%), 1 unknown (0.6%) were divided into three groups based on the levels of support they received from their parents, the negative interaction they experience and the amount of power their parents try to exert (See Table 2). Support was divided into groups based on the SD and negative interaction and power based on the .25 and .75 percentiles. The groups were labeled low, average and high. Boys and girls were not equally divided over the three groups of support (X2 (2) = 7.07, p < .05), but were equally divided over the three groups of negative interaction (X2 (2) = .30, p = .86) and power (X2 (2) = .71, p = .70). The participants were 13 years old at measure time one (T1) in the total population (M = 13.43, SD = 1.07). The age differences were equal between the groups for support (F (2, 164) = 1.51, p = .22), negative interaction (F (2, 164) = 1.09, p = .34) and power (F (2, 164) = .35, p = .71).

For the moderation analysis, a regression analysis was performed in which all 176 participants were included. As a prerequisite to conduct a regression analysis, linearity had to be checked. Based on visual inspection of boxplots, all combinations of dependent and independent variables showed a linear relationship. Furthermore, the assumption of homoscedasticity was checked, and was found for delinquency T2 and power T1 (F = 1.39, p = .14) and aggression T2 and power T1 (F = 1.29, p = .20). Lastly, data was checked for multicollinearity, none was found between any of the independent variables (support, negative interaction and power).

13 Parent-Adolescent Relationship and Externalizing Behavior

For all different variables at T1 and T2, a correlation analysis was performed (See Table 3). Delinquency and aggression, representing externalizing behavior, were strongly correlated at T1 and T2. Also correlations were found between some of the variables representing the parent-adolescent relationship. Support T1 had a moderate negative correlation with negative interaction at T1. Negative interaction T1 was weakly correlated with power at T1. Support T1 and power T1 were not correlated. Comparing the different variables, the correlations showed that support T1 was weakly correlated delinquency T1, delinquency T2 and aggression T2 (See Table 3). Support T1 was moderate correlated with aggression T1. Negative interaction T1 showed a moderate correlation with delinquency T1 and T2 and aggression T1 and T2. Power T1 was only weakly correlated with aggression T1.

Delinquency – Support.

The two-way mixed ANOVA showed that there was no significant interaction between support and time on delinquent behavior in the total population, F (2, 163) = 1.01, p = .37, partial η2 _{= .01. The main effect of delinquency showed no statistically significant difference} in mean delinquency at the different time points, F (1, 163) = 1.16, p = .28, partial η2 = .01, even though the main effect of support showed that there was a statistically significant difference in mean of delinquency between the different levels of support, F (2, 163) = 8.59, p < .001, partial η2 _{= .10. Post hoc Bonferonni correction revealed that adolescents with less} support (M = .35, SD = .03) showed more delinquency compared to adolescents with an average amount of support (M = .20, SD = .02) and a high amount of support (M = .21, SD = .03). The profile plot of the ANOVA suggested an increase in delinquent behavior from T1 to T2. However, a paired T-test showed that this suggestion was non-significant (t = -1.44, p = .08).

14

No interaction between negative interaction and time was found on delinquent behavior, F (2, 163) = .13, p = .88, partial η2 = .00. Also no significant main effect was found for differences in mean delinquency at the different time points, F = (1, 163) = .81, p = .37, partial η2 _{= .01.} However, between groups a main effect was found between differences in the subgroups of negative interaction (F (2, 163) = 19.62.59, p < .001, partial η2 = .20). Post hoc analysis with Bonferroni correction revealed that adolescents with more negative interaction (M = .41, SD = .03) showed an increase of delinquency compared to both adolescents with an average or low amount of negative interaction (M = .20, SD = .02 and M = .18, SD = .04 resp.). The ANOVA profile plot suggested an increase of delinquent behavior from T1 to T2 in the low negative interaction group. However, this suggestion was found non-significant after conducting a paired T-test (t = -1.52, p = .07).

Delinquency – Power.

There was no significant interaction between power and time on delinquent behavior, F (2, 163) = 1.23, p = .30, partial η2 = .02. There was also no main effect in difference from T1 to T2 in the amount of delinquent behavior, F = (1, 163) = .51, p = .47, partial η2 _{= .00. Neither a} main effect for group differences was found, F (2, 63) = 1.62, p = .20, partial η2 = .20.

Aggression – Support.

The results of the ANOVA showed no significant interaction effect between support and time on aggressive behavior, F (2, 163) = .43, p = .65, partial η2 = .01. There was also no main effect found in the difference in mean aggression at the different time points, F (1, 163) = 1.07,

p = .30, partial η2 = .01. However, a significant main effect for differences between groups of support was found, F (2, 163) = 11.14, p < .001, partial η2 = .12. Post hoc analysis with Bonferroni correction revealed that adolescent with less support (M = .49, SD = .04) showed more aggression compared to both adolescents with an average or high amount of support (M = .28, SD = .03 and M = .30, SD = .03 resp.). The ANOVA profile plot suggested a decrease

15

of aggressive behavior in the group with low and average levels of support over a 12-month period. However, paired T-tests showed that this suggestion was not significant for both low (t = .77, p = .22) and average (t = 1,42, p = .08) aggression, so all the levels of aggression stayed the same in the different groups of support.

Aggression – Negative Interaction.

No significant interaction between negative interaction and time on aggressive behavior was found, F (2, 163) = 2.42, p = .09, partial η2 = .03. Also no main effect in difference in the amount of aggressive behavior between T1 and T2 was found, F (1, 163) = .43, p = .51, partial η2 _{= .00. However, we did find a significant main effect for between-group differences, F (2,} 163) = 24,95, p < .001, partial η2 _{= .23. Post hoc analysis with Bonferroni correction revealed} that adolescent with a high amount of negative interaction (M = .57, SD = .04) showed more aggression compared to both adolescents with an average or low amount of negative interaction (M = .30, SD = .02, p < .001 and M = .21, SD = .04, p < .001 resp.). No differences were found between average and low negative interaction (p = .22). The ANOVA profile plot indicated a decrease of aggressive behavior in the group with high levels of negative interaction. To test this visual hypothesis, a paired t-test was conducted and found no significant decrease in aggressive behavior (t = 1.25, p = .11). The profile plot also indicated an increase of aggressive behavior in the group with less negative interaction. The paired t-test did confirm this expectation (t = -2.24, p < .05).

Aggression – Power.

No significant interaction effect between power and time on aggression was found, F (2, 163) = 2.58, p = .08, partial η2 = .03. Also no significant main effect was found on the difference in mean in aggression at T1 and T2, F (1, 163) = 1.77, p = .19, partial η2 = .01. However, a main effect for between-group differences with respect to the association between power and the amount of aggressive behavior was found, F (2, 163) = 3.15, p < .05, partial η2

16

= .04. In contrast, post hoc analysis with Bonferroni correction revealed no significant difference between the groups. The ANOVA profile plot indicated a decrease in the high power groups between T1 (M = .49, SD = .05) en T2 (M = .37, SD = .37), but paired t-test showed that this suggestion was not significant (t = 1.68, p = .05).

Moderation by Sensory Sensitivity.

Last the moderator analysis as described in the analysis section was executed with sensory sensitivity as moderator on the relationship between the parent-adolescent relation aspects (T1) and externalizing behavior (T2). The results (See Table 4) indicate that sensory sensitivity did not moderate any of the relationships (p < .05). For delinquency T2 and aggression T2, no interaction effects between sensory sensitivity and support T1, negative interaction T1 or power T1 were found.

Discussion

In the present study, we examined whether a negative parent-adolescent relationship was associated with later increases in child externalizing behavior. Additionally, we examined if sensory sensitivity moderates the associating between parent-adolescent relationship at T1 and externalizing behavior at T2. Our results showed that adolescents who experienced less support showed more delinquent and aggressive behavior. Also adolescents who experienced a lot of negative interaction showed more delinquent and aggressive behavior. There was no difference in delinquent or aggressive behavior for the amount of parental control. Most of the associations did not change over a 12-month period, which might suggest that the level of delinquent or aggressive behavior is a predictor for the same level over a 12-month period. Only adolescents who experienced little negative interaction seemed to develop more aggressive behavior over a 12-month period. The present study showed that sensory sensitivity did not

17 moderate the association between the parent-child relationship at T1 and externalizing behavior at T2.

The results indicate that adolescents who experience a small amount of support from their parents show more delinquent and aggressive behavior. This supports previous research about parental support being a predictor of behavioral control (Costigan, 1996; Forehand & Nousiainen, 1993). The meta-analysis of Vazsonyi, Mikuška and Kelley (2017) showed that there was a strong link between low self-control and deviance and crime, based on 100 studies. Independent of the type of study design (cross-sectional versus or longitudinal) or culture. This suggested that adolescents with less support develop less self-control, which leads to more delinquent and aggressive behavior.

Another expected finding was that adolescents who experience a lot of negative interaction with their parents show more delinquent and aggressive behavior. This corresponds to the finding of Burt and colleagues (2003) who suggested that parent-child conflict seems to act as vulnerability that increases risk for multiple externalizing disorders. Also the results of a meta-analysis (Weymouth, Buehler, Zhou & Henson, 2016) based on 52 studies confirmed that parent-adolescent conflict is associated positively with youth maladjustment. An unexpected finding was that adolescents who experience a small amount of negative interaction seem to develop more aggressive behavior in a 12-month period. A possible explanation for this finding was that parent-adolescent conflict facilitates transformational processes during puberty in which adolescents attempts to adjust parent-adolescent boundaries, renegotiate parental authority, and increase their autonomy and independence (Steinberg, 2001; Smetana, 2005; Weymouth et al., 2016). We suggest that adolescents who experience a small amount of negative interaction try to provoke a little negative interaction by searching for boundaries.

Another unexpected finding was that there was no difference between the amount of parental control (power) on the amount of delinquent or aggressive behavior. We did expect

18 that more control would lead to more externalizing behavior, based on the meta-analysis of Pinquart (2017). A possible explanation was that there is no difference between groups, because every adolescent responses different to parental control. This hypothesis was supported by the research of Blossom, Fite, Frazer, Colley and Evans (2016). They found that the emotional (dys)regulation of children from 8 to 12 years moderates the effect of parental psychological control on aggression. Blossom and colleagues (2016) suggested that psychologically controlling parenting strategies contribute to increased relational and decreased psychical aggression among emotionally well-regulated children. They found the opposite pattern among emotionally dysregulated children.

Also no moderation effect was found, which was expected since Serafini and colleagues (2017) found that high sensory sensitivity correlates with greater attentional, which leads to problems with reducing anger (Eisenberg et al., 2009). We expected that this might lead to more externalizing behavior. An interesting finding is that we did not find a association between sensory sensitivity and delinquency or aggression. We expect that this unexpected findings might be explained by the difference between two coping strategies for high sensory sensitivity. In the introduction, two coping strategies were mentioned: sensation avoiders and sensory sensitive. Sensation avoiders seem to be especially associated with depression (Serafini et al., 2017), negative emotionality (Ben-Sasson et al., 2008) and anxiety (Engel Yeger & Dunn, 2011a; 2011b). Sensation sensitivity on the other hand seems to be more associated with behavioral problems (Reynolds, Lane & Thacker, 2012; Shochat, Tzischinsky & Engel-Yeger, 2009) and impulsivity (Serafini et al., 2017). A recommendation for future research is to use a measurement which makes a distinction between sensory sensitivity and sensory avoiders. The Adolescent/Adult Sensory Profile (AASP, Brown & Dunn, 2002) could be used to measure the sensory processing style (Jerome & Liss, 2005). When using the AASP in a similar kind of research as the present one, a moderation effect might be found.

19 One of the strengths of this study was the longitudinal design, which started with a large sample (N = 419). This way, the sample was still large after wave 2, despite the large amount of drop outs and the including criteria (N = 166 and N = 176). Another strength of this study was the internal consistency of the measuring instruments; these were all at least acceptable. Using self-report data was also a strength, because participants know most information about themselves (Paulhus & Vazire, 2007). Besides, self-report has proven to be equally, if not more valid, for questions about externalizing behavior, especially when the behaviors assessed rarely lead to contacts with authorities or court appearances (Gold, 1970; Jensen & Rojek, 1980). Also, respondents are more likely motivated to answer questions about themselves then others (Paulhus & Vazire, 2007).

Despite these strengths, the self-report measures were also a limitation. Self-report measures are vulnerable for ‘response biases’ (Moskowitz, 1986; McGee Ng et al., 2016). For example, participants tend to give answers which presents them in a better way (Paulhus, 1991). Another reason for more positive answers was explained by having a distorted outlook about who we are (John & Robins, 1994; Yüksel, 2017). Another limitation was the implementation of the moderation analysis, which did not control for T1. Due to this limitation, the results were not able to tell us anything about moderation a difference in delinquency or aggression between T1 and T2. For future research, controlling for measuring T1 is recommended. Another recommended future research is investigating the change in support, negative interaction and power over time, and the effect of this change on different levels of delinquent or aggressive behavior.

The present study did find that most of the levels of delinquency and aggression stayed the same. This might indicate that training parents to become more supportive for example, could result in a decrease of delinquent or aggressive behavior. Further research is recommended to evaluate this hypothesis. This further research could help clinical practitioners

20 to adapt parental advice and trainings. The last recommendation for future research is including more different levels of education. In the present research, most of the participants attended the highest level of the Dutch High schools, which makes it not a representative sample for all 13-year old adolescents in the Netherlands.

Despite the mentioned shortcomings, this was the first study exploring the relationships between support, negative interaction, power, delinquency, aggression and sensory sensitivity. In addition to earlier research we suggested that a negative parent-adolescent relationship would increase later externalizing behavior. Interestingly, most levels of delinquency or aggression stayed the same, but adolescents who experience a small amount of negative interaction seemed to develop more aggressive behavior in a 12-month period. We did not find any support for sensory sensitivity moderating these found relationships, but gave some recommendations for future researches. In the future, a moderation effect might be found.

21 References

Achenbach, T. M. (1991). Manual for the youth self-report and 1991 profile. Burlington: Department of Psychiatry, University of Vermont.

American Psychiatric Association. (1994). Diagnostic and statistical manual of mental disorders (4th ed.). Washington, DC: Author.

Aron, E. N. & Aron, A. (1997). Sensory-processing sensitivity and its relation to introversion and emotionality. Journal of Personality and Social Psychology, 73, 345-368.

Aron, E. N., Aron, A., & Davies, K. M. (2005). Adult shyness: The interaction of

temperamental sensitivity and an adverse childhood environment. Personality and

Social Psychology Bulletin, 31, 181-197.

Aron, E. N., Aron, A., & Jagiellowicz, J. (2012). Sensory processing sensitivity: A review in the light of the evolution of biological responsivity. Personality and Social Psychology

Review, 16, 262-282. doi: 10.1177/1088868311434213

Baumeister, F. R., Vohs, D. K., DeWall, N. C., & Zhang, L. (2007). How emotion shapes behavior: Feedback, anticipation, and reflection, rather than direct causation. Personality and Social Psychology Review, 11, 167-203.

Bakermans-Kranenburg, M. J., & van IJzendoorn, M. H. (2011). Differential susceptibility to rearing environment depending on dopamine-related genes: New evidence and a meta analysis. Development and psychopathology, 23, 39-52.

Barker, C., Pistrang, N., & Elliot, R. (1994). Research methods in clinical and counselling

psychology. John Wiley & Sons.

Belsky, J., & Pluess, M. (2009). Beyond diathesis stress: Differential susceptibility to environmental influences. Psychological bulletin, 135, 885. doi: 10.1037/a0017376 Ben‐Sasson, A., Cermak, S. A., Orsmond, G. I., Tager‐Flusberg, H., Kadlec, M. B., & Carter,

22 A. S. (2008). Sensory clusters of toddlers with autism spectrum disorders: Differences in affective symptoms. Journal of Child Psychology and Psychiatry, 49, 817-825. Betz, C.L. (1995). Childhood violence: A nursing concern. Comprehensive Pediatric Nursing,

18, 149–161.

Blossom, J. B., Fite, P. J., Frazer, A. L., Cooley, J. L., & Evans, S. C. (2016). Parental psychological control and aggression in youth: Moderating effect of emotion dysregulation. Journal of Applied Developmental Psychology, 44, 12-20.

Booth, C., Standage, H., & Fox, E. (2015). Sensory-processing sensitivity moderates the association between childhood experiences and adult life satisfaction. Personality and

individual differences, 87, 24-29.

Brown, C. & Dunn, W. (2002) Adult/Adolescent Sensory Profile: User's Manual . San Antonio, TX: Psychological Corporation.

Burt, S. A., Krueger, R. F., McGue, M., & Iacono, W. (2003). Parent-child conflict and the comorbidity among childhood externalizing disorders. Archives of general

psychiatry, 60, 505-513.

Chassin, L., Pillow, D. R., Curran, P. J., Molina, B. S. G., & Barrera, M. (1993). Relation of parental alcoholism to early adolescent substance use: A test of three mediating mechanisms. Journal of Abnormal Psychology, 102, 3-19.

Cohen, J., Cohen, P., West, S. G., & Aiken, L. S. (2003). Applied multiple

regression/correlation analysis for the behavioral sciences (3rd ed.). Mahwah, NJ: Lawrence Erlbaum Associates.

Costigan, C. L. (1996). Parental values related to control in families with children with special needs. Unpublished doctoral dissertation, Michigan State University.

Dunn, W. (1997). The impact of sensory processing abilities on the daily lives of young children and their families: a conceptual model. Infants Young Child. 9, 23–35.

23 Dunn, W. (2001). The sensations of everyday life: Empirical, theoretical, and pragmatic considerations. The American Journal of Occupational Therapy, 55, 608-620. Eisenberg, N., Valiente, C., Spinrad, T. L., Cumberland, A., Liew, J., Reiser, M., ...

Losoya, S. H. (2009). Longitudinal relations of children’s effortful control, impulsivity, and negative emotionality to their externalizing, internalizing, and co-occurring behavior problems. Developmental psychology, 45, 988.

Engel-Yeger, B., & Dunn, W. (2011a). Exploring the relationship between affect and sensory processing patterns in adults. British Journal of Occupational Therapy, 74, 456-464. Engel-Yeger, B., & Dunn, W. (2011). The relationship between sensory processing

difficulties and anxiety level of healthy adults. British Journal of Occupational

Therapy, 74, 210-216.

Fergusson, D. M., Horwood, L. J., & Lynskey, M. (1994). The childhoods of multiple problem adolescents: a 15‐year longitudinal study. Journal of child psychology and

psychiatry, 35, 1123-1140.

Forehand, R., & Nousiainen, S. (1993). Maternal and paternal parenting: Critical dimensions in adolescent functioning. Journal of Family Psychology, 7, 213–221.

Freedman, R., Adler, L.E., Gerhardt, G.A., Waldo, M., Baker, N., Rose, G.M., … Franks, R. (1987). Neurobiological studies of sensory gating in schizophrenia. Schizophr. Bull. 13, 669e78

Furman, W., & Buhrmester, D. (2009). Methods and measures: The network of relationships inventory: Behavioral systems version. International journal of behavioral

development, 33, 470-478.

Furman, W., & Buhrmester, D. (1985). Children's perceptions of the personal relationships in their social networks. Developmental psychology, 21, 1016.

24 Structure of problem behaviors in preadolescence. Journal of Consulting and Clinical

Psychology, 59, 499.

Gold, M. (1970). Delinquent behavior in an American city.

de Groot, A., Koot, H. M., & Verhulst, F. C. (1996). Cross-cultural generalizability of the Youth Self-Report and Teacher's Report Form cross-informant syndromes. Journal of

abnormal child psychology, 24, 651-664.

Howitt, D. & Cramer, D. (2011). Introduction to SPSS Statistic in Psychology. For version 19

and earlier (5th ed.). Pearson.

Jagiellowicz, J., Aron, A., & Aron, E. N. (2016). Relationship between the temperament trait of sensory processing sensitivity and emotional reactivity. Social Behavior and Personality: an international journal, 44, 185-199.

Jensen, G. F., & Rojek, D. G. (1980). Delinquency, a sociological view. Lexington, Mass.: DC Heath.

Jerome, E. M., & Liss, M. (2005). Relationships between sensory processing style, adult attachment, and coping. Personality and individual differences, 38, 1341-1352. John, O. P., & Robins, R. W. (1994). Accuracy and bias in self-perception: Individual

differences in self-enhancement and the role of narcissism. Journal of Personality and

Social Psychology, 66, 206-219.

Jolicoeur-Martineau, A., Belsky, J., Szekely, E., Widaman, K. F., Pluess, M., Greenwood, C., & Wazana, A. (2017). Distinguishing differential susceptibility, diathesis-stress, and vantage sensitivity: Beyond the single gene and environment model. Development and

psychopathology, 1-11.

Karam, E. G., Fayyad, J. A., Farhat, C., Pluess, M., Haddad, Y. C., Tabet, C. C., ... Kessler, R. C. (2019). Role of childhood adversities and environmental sensitivity in the development of post-traumatic stress disorder in war-exposed Syrian refugee children

25 and adolescents. The British Journal of Psychiatry, 1-7.

Lahey, B.B., & Loeber, R. (1994). Framework of developmental model of Oppositional Defiant Disorder and Conduct Disorder. In D. Routh (Ed.), Disruptive behavior

disorders in childhood. New York: Plenum.

Lamborn, S. D., Mounts, N. S., Steinberg, L., & Dornbusch, S. M. (1991). Patterns of

competence and adjustment from authoritative, authoritarian, indulgent and neglectful families. Child Development, 62, 1049-1065.

Laerd Statistics (2015). Two-way mixed ANOVA using SPSS Statistics. Statistical tutorials and software guides. Retrieved from https://statistics.laerd.com/

Lohse, K., Bland, M. D., & Lang, C. E. (2016). Quantifying change during outpatient stroke rehabilitation: a retrospective regression analysis. Archives of physical medicine and rehabilitation, 97, 1423-1430.

McConaughy, S. H., & Achenbach, T. M. (1994). Comorbidity of empirically based syndromes in matched general population and clinical samples. Journal of Child

Psychology and Psychiatry, 35, 1141-1157.

McGee Ng, S. A., Bagby, R. M., Goodwin, B. E., Burchett, D., Sellbom, M., Ayearst, L. E., ... Baker, S. (2016). The effect of response bias on the Personality Inventory for DSM 5 (PID–5). Journal of Personality Assessment, 98, 51-61.

Methodologiewinkel (n.d.). Handleidingen. Moderator analyse. Netherlands, Twente: Universiteit Twente. Retrieved from https://www.utwente.nl/nl/bms/m-winkel/ moderator%20handleiding.pdf

Moffitt, T.E. (1993). Adolescence-limited and life-course-persistent anti-social behavior - A developmental taxonomy. Psychological Review, 100, 674–701.

26 research: Implications for the depressive disorders. Psychological Bulletin, 110, 406– 425.

Moskowitz, D. S. (1986). Comparison of self-reports, reports by knowledgeable informants, and behavioral observation data. Journal of Personality, 54, 294-317.

Okado, Y., & Bierman, K. L. (2015). Differential risk for late adolescent conduct problems and mood dysregulation among children with early externalizing behavior problems. Journal of abnormal child psychology, 43, 735-747.

Paulhus, D. L., & Vazire, S. (2007). The self-report method. Handbook of research methods

in personality psychology, 1, 224-239.

Pinquart, M. (2017). Associations of parenting dimensions and styles with externalizing problems of children and adolescents: An updated meta-analysis. Developmental psychology, 53, 873.

Pluess, M., Assary, E., Lionetti, F., Lester, K. J., Krapohl, E., Aron, E. N., & Aron, A. (2018). Environmental sensitivity in children: Development of the Highly Sensitive Child Scale and identification of sensitivity groups. Developmental psychology, 54, 51. Reuben, J. D., Shaw, D. S., Neiderhiser, J. M., Natsuaki, M. N., Reiss, D., & Leve, L. D.

(2016). Warm parenting and effortful control in toddlerhood: independent and

interactive predictors of school-age externalizing behavior. Journal of abnormal child psychology, 44, 1083-1096.

Rey, J. M., & Morris-Yates, A. (1993). Are oppositional and conduct disorders of adolescents separate conditions?. Australian and New Zealand journal of psychiatry, 27, 281-287. Reynolds, S., Lane, S. J., & Thacker, L. (2012). Sensory processing, physiological stress, and sleep behaviors in children with and without autism spectrum disorders. OTJR:

Occupation, Participation and Health, 32, 246-257.

27 nonclinical samples: A meta-analysis. Psychological bulletin, 116, 55.

Shochat, T., Tzischinsky, O., & Engel-Yeger, B. (2009). Sensory hypersensitivity as a contributing factor in the relation between sleep and behavioral disorders in normal schoolchildren. Behavioral Sleep Medicine, 7, 53-62.

Serafini, G., Gonda, X., Canepa, G., Pompili, M., Rihmer, Z., Amore, M., & Engel-Yeger, B. (2017). Extreme sensory processing patterns show a complex association with

depression, and impulsivity, alexithymia, and hopelessness. Journal of affective

disorders, 210, 249-257.

Smetana, J. G. (2005). Adolescent–parent conflict: Resistance and subversion as

developmental process. In L. Nucci (Ed.), Conflict, contradiction, and contrarian elements in moral development and education (pp. 69–91). Mahwah, NJ: Erlbaum. Song, L. Y., Singh, J., & Singer, M. (1994). The Youth Self-Report inventory: A study of its measurements fidelity. Psychological Assessment, 6, 236.

Stanger, C., Achenbach, T. M., & Verhulst, F. C. (1997). Accelerated longitudinal comparisons of aggressive versus delinquent syndromes. Development and

Psychopathology, 9, 43-58.

Steinberg, L. (2001). We know some things: Parent–adolescent relationships in retrospect and prospect. Journal of Research on Adolescence, 11, 1–19. doi:

10.1111/1532-7795.00001

Vazsonyi, A. T., Mikuška, J., & Kelley, E. L. (2017). It's time: A meta-analysis on the self control-deviance link. Journal of Criminal Justice, 48, 48-63.

West, S. G., Finch, J. F., & Curran, P. J. (1995). Structural equation models with nonnormal variables: Problems and remedies.

28 adolescent conflict: Disagreement, hostility, and youth maladjustment. Journal of

Family Theory & Review, 8, 95-112.

Wyller, H. B., Wyller, V. B., Crane, C., & Gjelsvik, B. (2017). The relationship between sensory processing sensitivity and psychological distress: A model of underpinning mechanisms and an analysis of therapeutic possibilities. Scandinavian Psychologist, 4. Yüksel, A. (2017). A critique of “Response Bias” in the tourism, travel and hospitality

research. Tourism Management, 59, 376-384.

Zuckerman, M. (1999). Vulnerability to psychopathology: A biosocial model. Washington, DC: American Psychological Association.

29 Appendix

Table 1

Descriptive Statistics variables

N M SD Z-Skewness Z-Kurtosis Sensory Sensitivity T1 174 3.49 .89 .52 -1.38 Sensory Sensitivity T2 176 3.49 .90 .98 -.51 Externalizing behavior T1 Delinquency T1 167 .23 .24 16.81 46.03 Aggression T1 167 .35 .31 12.89 27.73 Externalizing behavior T2 Delinquency T2 175 .25 .24 9.43 13.29 Aggression T2 175 .33 .27 7.78 7.28 Parent-adolescent relationship T1 Support T1 166 3.29 .52 -6.71 5.39 Negative Interaction T1 166 1.22 .90 3.02 -.93 Power T1 165 2.81 1.93 .47 -.29 Parent-adolescent relationship T2 Support T2 175 4.28 .55 -5.12 1.17 Negative Interaction T2 175 2.23 .98 4.00 -.82 Power T2 175 3.67 .69 -2.21 -.81

Table 2

Descriptive statistics groups

Support Negative interaction Power

Low Average High Low Average High Low Average High

N 41 66 59 26 106 34 22 112 32 Age M 13.68 13.34 13.36 13.24 13.41 13.64 13.48 13.38 13.55 SD 1.15 .99 1.07 .71 1.06 1.06 1.22 1.04 1.08 Sekse Man 27 31 36 12 54 18 12 54 18 % 65.9% 47% 61% 46.2% 50.9% 52.9% 54.5% 48.2% 56.2% Vrouw 14 34 23 14 51 16 10 57 14 % 34.1% 51.1% 39% 53.8% 48.1% 47.1% 45.5% 50.9% 43.8% Missing 1 1 1 Table 3

Correlations for externalizing behavior and parent-adolescent relationship

Variable 1 2 3 4 5 6 7 8 9 10 11 1. Delinquency T1 2. Delinquency T2 .45** 3. Aggression T1 .82** .34** 4. Aggression T2 .41** .73** .53** 5. Support T1 -.18* -.25** -.31** -.26** 6. Support T2 -.15 -.29** -.31** -.34** .63** 7. Negative Interaction T1 .40** .32** .49** .36** -.46** -.47** 8. Negative Interaction T2 .32** .30** .40** .43** -.27** -.43** .60** 9. Power T1 .14 -.02 .24** .08 -.01 -.15* .22** .17* 10. Power T2 .09 .06 .24** .21** -.12 -.21** .22** .27** .42** 11. Sensory Sensitivity T1 .00 -.11 .12 .07 -.10 -.13 .13 .14 .16* .05 12. Sensory Sensitivity T2 -.07 -.14 .00 -.04 -.02 .01 .02 .03 .08 .01 .65** Note. *** p < .001, ** p < .01, * p < .05

Table 4

Moderation analysis coefficients

Model Variable B SE β t

1 Support T1 -.14 .04 -.28 -3.57***

Sensory Sensitivity T1 -.04 .02 -.13 -1.75 Interaction Suport Sensory Sensitivity .02 .04 .04 .50

2 Negative Interaction T1 .09 .02 .35 4.59***

Sensory Sensitivity T1 -.04 .02 -.14 -1.89 Interaction Negative Interaction SS -.02 .02 -.07 -.99

3 Power T1 -.02 .03 -.04 -.46

Sensory Sensitivity T1 -.03 .02 -.12 -1.58 Interaction Power Sensory Sensitivity .06 .04 .13 1.57

4 Support T1 -.13 .04 -.23 -2.93**

Sensory Sensitivity T1 .02 .02 .06 .76

Interaction Suport Sensory Sensitivity -.06 .05 -.09 -1.17

5 Negative Interaction T1 .11 .02 .36 4.81***

Sensory Sensitivity T1 .01 .02 .03 .41

Interaction Negative Interaction SS -.02 .02 -.06 -.75

6 Power T1 .02 .04 .04 .50

Sensory Sensitivity T1 .01 .03 .04 .53

Interaction Power Sensory Sensitivity .08 .05 .14 1.75

Note. The Moderation analysis is executed for every independent variable separately.

Delinquency (T2) was the dependent variable at Model 1,2 and 3. Aggression was the dependent variable at Model 4, 5 and 6.