سبک های فرزندپروری پدران و مادران: پیوندهای منحصر به فرد و ترکیبی برای بزهکاری زودهنگام نوجوانان و بزرگسالان

Abstract The present study examines the cross-sectional and longitudinal associations between fathers’ and mothers’ parenting styles and male and female delinquency using a sample of 330 Dutch families with a mid or late adolescent son or daughter (ages 14–22), followed across two measurement waves with a 5-year interval. Parenting styles of fathers and mothers were linked to delinquency. A significant parenting style by sex interaction was found: neglectful parenting was related to higher levels of delinquency in males and permissive parenting was linked to delinquency in females. A long term relationship was found between fathers’ neglectful parenting style and delinquency in males. Furthermore, results revealed that levels of delinquency were the lowest in families with at least one authoritative parent and highest in families with two neglectful parents, indicating that the level of delinquency was dependent on the combination of mother’s and father’s parenting styles.

Introduction Wide interest has been shown in the domain of the family, attempting to uncover the origins of the development of delinquent behavior. Family characteristics and, in particular, parenting has been among the strongest predictors of criminal behavior (Cottle et al., 2001 and Gendreau et al., 1996). Parental attachment, harsh parental discipline, poor relationships with parents, poor supervision, and inconsistent discipline are among the family factors that have been linked to delinquency. Despite the fact that the link between parenting and delinquency has been extensively investigated, there are several shortcomings. Although various parenting behaviors have been found to be linked to delinquency, surprisingly few studies have focused on whether combinations of parenting dimensions, that is, parenting styles are related to delinquency (Hoeve et al., 2009). In addition, the vast majority of studies has focused on parenting by the mother and has neglected the influence of paternal parenting delinquency, despite the fact that fathers’ behavior and parenting have been linked to their sons’ delinquent behavior (e.g., Simons & Conger, 2007). By examining both mother’s and father’s parenting styles it is possible to investigate whether the gender of parent and child moderates the link between parenting and delinquency and whether one parent can either compensate for or accentuate the associations found between the other parent’s style and their sons’ and daughters’ delinquency. Finally, the majority of studies have concentrated on adolescence and have analyzed concurrent links or have short time intervals. To date, it remains unclear whether or not parenting contribute to the continuation of offending after onset or for later onsets after age 20 (Farrington, 2005). Thus, the main purpose of the present paper is to test whether fathers’ and mothers’ parenting styles are linked to sons’ and daughters’ delinquency and whether combinations of parenting styles influence their children’s delinquent behavior. We concentrate on different phases of the life course from adolescence to early adulthood and analyze concurrent and longitudinal parenting-delinquency links. We first review the literature on various concepts of parenting, parenting styles and gender of parents and children and different stages in the life course in relation to parenting and delinquency. Then, we describe the current study in which we analyze the concurrent and prospective link between mother’s and father’s parenting styles and delinquency in male and female adolescents and (early) adults.

Results Concurrent links In Table 1 the delinquency means of Time 2 are presented for maternal and paternal parenting styles. Neglectful parenting by mothers and fathers is linked to the highest level of delinquency, whereas authoritative parenting is linked to the lowest level of delinquency. We next conducted multivariate cross-sectional analyses on the link between parenting styles and delinquency during mid to late adolescence (Time 2), controlling for family income. Separate analyses were conducted for maternal and paternal parenting styles. Interaction effects were analyzed in order to investigate whether the link between parenting styles were different for boys and girls and for the two age groups. Table 1. Delinquency (Adjusted) Means at Time 2 and Analyses (ANOVA) for Maternal and Paternal Parenting Styles (bivariate models). Parenting styles Maternal parenting Paternal parenting Delinquency T2 N Delinquency T2 N Authoritative 1.13 107 1.12 107 Authoritarian (supportive) 1.18 75 1.20+ 91 Permissive (poorly responsive) 1.22∗ 72 1.20+ 71 Neglectful (punishing) 1.23∗ 64 1.25∗ 28 Total 1.18 318 1.18 297 F 3.37∗ 3.52∗ η2 .03 .04 Note. Planned simple contrasts were conducted with authoritative parenting as the control category. + p < .10; ∗p < .05. Table options Maternal parenting A three-way ANCOVA with maternal parenting style, gender and age group as factors and delinquency Time 2 as dependent variable revealed that a significant main effect was found for sex of the child, F(1,286) = 50.0, p < .001, η2 = .15 ( Table 2). No significant main effects were found for parenting style and age group. Furthermore, family income was nonsignificant and no interaction effects were found. Table 2. Analyses (three-way ANCOVAs - Parenting Style by Age by Sex) for Maternal and Paternal Parenting Style. Cross-sectional model Longitudinal model Mother Father Mother Father df F-value df F-value df F-value df F-value Covariates Family income 1 .0 1 .2 1 1.2 1 1.8 Delinquency T2 1 1.6 1 .5 Factors Parenting style 3 1.0 3 3.9∗∗ 3 .3 3 2.1 Age group 1 .2 1 .1 1 5.1∗ 1 2.1 Sex of child 1 50.0∗∗∗ 1 53.0∗∗∗ 1 3.9+ 1 14.8∗∗∗ Interaction terms Parenting style ∗ Age group 3 .8 3 .3 3 .5 3 .8 Parenting style ∗ Sex 3 1.7 3 4.2∗∗ 3 .1 3 3.0∗ N 299 280 176 165 R2 .20 .23 .10 .19 + p < .10; ∗p < .05; ∗∗p < .01; ∗∗∗p < .001. Table options Paternal parenting A significant main effect of paternal parenting styles, F(3,267) = 3.9, p < .01, η2 = .04, and sex of the child, F(1,267) = 53.0, p < .001, η2 = .17, on delinquency was found ( Table 2). No significant effects were found for age group and family income. Adolescents with an authoritative father were significantly less often involved in delinquency than adolescents with permissive, t(280) = 2.15, p < .05, and neglectful, t(280) = 3.21, p < .01, fathers. The main effect of sex indicates that boys significantly engage more often in delinquency than girls. Interaction-effects were found between sex and paternal parenting type, F(3,267) = 4.2, p < .01, η2 = .05. That is, boys’ delinquency is elevated if the father is neglectful (punishing), while the level of girls’ delinquency is highest if the father is permissive ( Fig. 1, Panel a). Interaction between paternal parenting style and sex on adolescent delinquency ... Fig. 1. Interaction between paternal parenting style and sex on adolescent delinquency (Panel a) and on early adult delinquency, controlling for adolescent delinquency (Panel b). Figure options Longitudinal links In order to gain more clarity on long-term influences of parenting styles we analyzed the link between parenting styles (Time 2) and delinquency Time 3, controlling for adolescent delinquency Time 2 and family income. Again, we conducted separate analyses for maternal and paternal parenting. Maternal parenting No significant effects were found for maternal parenting style on delinquency Time 3 (see Table 2). Furthermore, effects of the covariates adolescent delinquency Time 2 and family income were nonsignificant. The nonsignificant link between delinquency at Time 2 and later delinquency at Time 3 could be due to the fact that adolescents engage more often in offending than early adults (Farrington, 1986). The majority of these adolescents will desist from delinquency when they enter adulthood. We found some evidence for this in our study: a significant main effect was found for age group, F(1, 162) = 5.1, p < .05, η2 = .03, indicating that the oldest group (ages 23–27 at Time 3) engages less often in delinquency at Time 3 than the youngest group (ages 19–22 at Time 3). A trend was found for sex of the child, F(1, 162) = 3.9, p < .10, η2 = .02. Paternal parenting The model with paternal parenting revealed that delinquency measured at Time 3 was significantly different between sons and daughters, F(1,151) = 14.8, p < .001, η2 = .09 ( Table 2). No significant main effects were found for paternal parenting styles and age group. Also, covariates, delinquency Time 2 and family income, were nonsignificant. There was an interaction effect between sex and paternal parenting style, F(3,151) = 3.0, p < .05, η2 = .06, indicating that the link between paternal parenting style and later delinquency at Time 3 was different for males and females ( Fig. 1, Panel b). Neglectful parenting by the father was related to elevated levels of later delinquency measured at Time 3 in sons. In contrast, paternal parenting style was not linked to later delinquency in daughters. Combinations of maternal and paternal parenting styles As dependency between the maternal and paternal parenting styles existed (χ2(9) = 163.33, N = 295, p < .001), it was not appropriate to conduct a factorial ANCOVA with maternal and paternal parenting styles as factors for analyzing the interaction between the parents’ styles. Therefore new constructs were created for each parenting style that represented the number of parents with a specific style. For example, one variable represented the number of authoritative parents (values 0, 1, and 2). In Table 3 the bivariate results are presented for each parenting style. The level of delinquency appeared to be dependent on the number of authoritative parents (sons and daughters with two authoritative parents had the lowest levels of delinquency) and neglectful parents (two neglectful parents are linked to the highest levels of delinquency). We also investigated whether delinquency levels differed between couples with the same parenting style and couples with different styles. A t-test revealed that delinquency levels are independent of whether parents have similar or different styles or not. Table 3. Delinquency (Adjusted) Means at Time 2 and Analyses (ANOVA) for combinations of Maternal and Paternal Parenting Styles (bivariate models). Authoritative style Authoritarian style Permissive style Neglectful style Delinquency N Delinquency N Delinquency N Delinquency N T2 T2 T2 T2 None of the parents have this style 1.22 161 1.17 170 1.17 189 1.17 220 One parent has this style 1.15+ 61 1.19 79 1.17 54 1.16 56 Both parents have this style 1.12∗∗ 69 1.19 42 1.23 35 1.39∗∗∗ 15 Total 1.18 291 1.18 291 1.18 278 1.18 291 F 4.8∗∗ .1 1.1 6.6∗∗ η2 .03 .00 .01 .04 Note. Planned simple contrasts were conducted with none of the parents have this style as the control category. + p < .10; ∗∗p < .01; ∗∗∗p < .001. Table options Furthermore, we analyzed whether concurrent links existed between combinations of maternal and paternal parenting styles and delinquency conducting multivariate analyses. The models with the number of authoritative and authoritarian parents revealed no significant main and interaction effects. Significant multivariate results were found for permissive and neglectful parenting. In Table 4 only the significant results are presented. The analysis on the number of permissive parenting styles revealed that sex of the child was significantly linked to delinquency, F(1,253) = 11.8, p = .001, η2 = .05. Furthermore, an interaction effect between the number of permissive parents and sex was found, F(2,253) = 5.4, p < .01, η2 = .04. The interaction effect indicates that levels of delinquency were higher in girls if these girls had one or two permissive parents, while the levels of delinquency in boys were independent from whether neither, one or both parents were permissive ( Fig. 2, Panel a). The model with the number of neglectful parents revealed two significant main effects ( Table 4). First, a main effect was found for the number of neglectful parents, F(2,264) = 10.2, p < .001, η2 = .07. Furthermore, we found a main effect of sex, F(1,264) = 51.2, p < .001, η2 = .16. In addition, a trend was found for an interaction effect of the number of neglectful parents and age group, F(1,264) = 2.7, p < .10, η2 = .02. It appeared that the effect of two neglectful parents was stronger for the oldest group (ages 23–27 at Time 3; see Fig. 2, Panel b). Finally, an interaction effect between the number of neglectful parents and sex was found, F(2,264) = 8.3, p < .001, η2 = .06. The interaction effect indicates that the effect of the number of neglectful parents on delinquency was mainly apparent in boys ( Fig. 2, Panel c). Thus, boys had higher levels of delinquency if both parents were neglectful compared to if their parents were non-neglectful or if one parent was neglectful. Planned Helmert contrasts revealed that delinquency was significantly higher in one or two neglectful parents compared to two non-neglectful parents, t(274) = −3.00, p < .01. Furthermore, if both parents were neglectful, delinquency was higher than if only one parent was neglectful, t(274) = −3.83, p < .001. Table 4. Analyses (three-way ANCOVAs - Parenting Style by Age by Sex) for Combinations of Maternal and Paternal Parenting Style. Cross-sectional model Permissive Neglectful df F-value df F-value Covariates Family income 1 0 1 0 Factors Combination (# parents) 2 .8 2 10.2∗∗∗ Age group 1 0 1 .9 Sex of child 1 11.8∗∗ 1 51.2∗∗∗ Interaction terms Combination ∗ Age group 2 .2 2 2.7+ Combination ∗ Sex 2 5.4∗∗ 2 8.3∗∗∗ N 263 274 R2 .21 .26 Note. Combination (# parents) refers to the number of parents with a particular parenting style (0, 1 or 2 parents). The analyses of the number of permissive parents are presented in the second and fourth columns and the analyses of the number of neglectful parents are presented in the third and fifth column. + p < .10; ∗p < .05; ∗∗p < .01; ∗∗∗p < .001. Table options Interaction between the number of permissive parents and sex (Panel a), between ... Fig. 2. Interaction between the number of permissive parents and sex (Panel a), between number of neglectful parents and sex of the child (Panel b), and between the number of neglectful parents and age group (Panel c). Figure options We next created a construct that represented all possible combinations between maternal and paternal parenting styles. Sixteen (4 × 4) combinations of parenting styles were made. As we decided to set the minimum group size at 10 families, we had to remove five combinations from further analyses (see Table 5). A three-way ANCOVA revealed again nonsignificant main and interaction effects for age group and a nonsignificant effect of family income. In order to increase the statistical power we subjected this combination-variable to two-way ANCOVA (parenting by sex) in order to analyze more specifically which combinations of parenting styles were linked to higher delinquency rates. The analysis revealed that there was a significant main effect of the combination-variable, F(10, 246) = 3.2, p = .001, η2 = .12, and sex, F(1, 246) = 31.6, p < .001, η2 = .11, and an interaction effect, F(10, 246) = 2.7, p < .01, η2 = .10, indicating that the link between combinations of parenting styles and delinquency was dependent on the sex of the adolescent. The finding that the level of delinquency in adolescents with neglectful mothers was dependent on the parenting style of the father appeared to be specifically applicable to boys. Table 5. Delinquency (Adjusted) Means and Analyses (ANCOVA) for Combinations of Maternal and Paternal Parenting Styles. Combination of parenting styles Adolescent delinquency N % Both Authoritative 1.12 71 24.1 Authoritative (M) and Authoritarian (F) 1.17 10 3.4 Authoritative (M) and Permissive (F) 1.19 13 4.4 Authoritative (M) and Neglectful (F) 1.15a 3 1.0 Both Authoritarian 1.19 42 14.2 Authoritarian (M) and Authoritative (F) 1.10 17 5.8 Authoritarian (M) and Permissive (F) 1.25 11 3.7 Authoritarian (M) and Neglectful (F) 1.10a 2 .7 Both Permissive 1.23 36 12.2 Permissive (M) and Authoritative (F) 1.16 16 5.4 Permissive (M) and Authoritarian (F) 1.24a 8 2.7 Permissive (M) and Neglectful (F) 1.09a 7 2.4 Both Neglectful 1.39a 15 5.1 Neglectful (M) and Authoritative (F) 1.27a 3 1.0 Neglectful (M) and Authoritarian (F) 1.21b 31 10.5 Neglectful (M) and Permissive (F) 1.07b 10 3.4 Total 1.18 295 100.0 F (10, 265) 2.61∗∗ η2 .09 Note: Different subscripts indicate significantly different means applying planned contrasts. ∗∗p < .01. a This combination was not included in the analyses as the number of cases was lower than 10. Table options We were particularly interested in whether fathers’ styles uniquely contributed to their children’s delinquent behavior, independent from the mothers’ style. Therefore, we tested whether maternal parenting would be dominant or complementary, that is, whether or not the levels of delinquency remained the same for the maternal parenting style regardless the style of the father. For example, if delinquency is the same for adolescents with authoritative mothers irrespective of whether their fathers are authoritative or authoritarian, then there is no unique contribution of fathers’ style beyond the authoritative style of the mother. We tested the following hypotheses. The first hypothesis considered whether adolescent delinquency was the same in families with an authoritative mother regardless the style of the father. As we hypothesized fathers to affect the development of delinquent behavior of their child, we expected to find significant differences between either authoritarian or permissive fathers and authoritative fathers who had authoritative mothers as partners. We could not use neglectful fathers to test this hypothesis as the number of families with an authoritative mother and neglectful father was less than 10 (Table 5). Thus, we tested whether the fathers’ authoritarian and permissive styles would dominate the authoritative style of the mother. The second hypothesis considered whether adolescents with neglectful mothers had different delinquency levels if their fathers were neglectful compared to if they had authoritarian or permissive fathers. We could not use authoritative fathers to test this hypothesis as the number of families with a neglectful mother and an authoritative father was also less than 10 (Table 5). In sum, we tested the following planned contrasts: (1a) authoritative mothers and authoritative fathers versus authoritative mothers and authoritarian fathers, (1b) authoritative mothers and authoritative fathers versus authoritative mothers and permissive fathers, (2a) neglectful mothers and authoritarian fathers versus neglectful mothers and neglectful fathers, and (2b) neglectful mothers and permissive fathers versus neglectful mothers and neglectful fathers. The first two planned contrasts (1a and b) revealed that families where authoritarian and permissive fathers were coupled with authoritative mothers had adolescents with relatively similar levels of delinquent behavior compared to families where both parents had authoritative parenting styles. This suggests that authoritative mothers in general reduce the chance of higher levels of delinquency, regardless of the style of the father. The remaining contrasts (2a and b) were significant. Specifically, the levels of delinquency for adolescents with two neglectful parents were significantly different from those of adolescents with neglectful mothers and authoritarian fathers, t(257) = −2.7, p < .01, and significantly different from adolescents with neglectful mothers and permissive fathers, t(257) = −3.4, p < .01. This suggests that a neglectful (punishing) style of a mother can be compensated by a permissive or authoritarian style of the father.