Individuals experience a host of stressors in their roles as parents that may contribute to high levels of distress (Creasey and Reese 1996; Haskett et al. 2006). It has been reported, for example, that parents experience heightened levels of distress if their children are temperamental, communicate negative emotions frequently, and/or display gross motor movements that are vigorous and fast (McBride et al. 2002). Research also has suggested a strong link specifically between parental distress and child aggression. Child behavior can be a very powerful determinant for how a parent responds to a child (Anderson et al. 1986; Duchovic et al. 2009). For example, children with behavior problems that are difficult to tolerate may contribute to their parents feeling distressed and critical of them.

Parental distress may partly result from experiencing a lack of parental control and the chronic level of coercive behaviors (e.g., noncompliance, whining, avoidance) parents experience from their children (Hetherington and Martin 1986; Patterson 1982). Indeed, some studies have reported that deviant child behavior is causally related to parental distress and negative mood (Pelham et al. 1997, 1998). Patterson (1980) also found that parents of socially aggressive children were especially likely to show parental depression. Further, results suggested prolonged exposure to high levels of child aggression may lead to parents’ low self-esteem, which is often associated with depression.

Other factors have been identified as contributing to a link between parental distress and child aggression, including parenting styles. In general, three types of parenting styles have been discussed in the literature based on the work of Baumrind (1971, 1991): authoritative, authoritarian, and permissive. Authoritative parents are warm, involved, and responsive to their children. They are also demanding, clear in their expectations, and hold their children accountable for adhering to the rules. They do, however, engage their children in a democratic process of developing rules and strategies of discipline, and they support their children’s autonomy. In comparison, authoritarian parents are demanding but employ this strategy with little warmth. They do not support their children’s autonomy, illustrated by their intrusive and overly restrictive behaviors, and their assertion of power. Finally, permissive parents are not demanding, do not follow through on their discipline, and ignore misbehavior. They are, however, warm, involved, and supportive of their children’s autonomy.

Parental distress has been linked with impairments in parents’ child management techniques (Middleton et al. 2009) or parenting styles. Compared to non-depressed parents, distressed parents are more inconsistent, lax, or permissive as characterized by, in general, an ineffectiveness to monitor their children’s misbehavior and administer discipline (Cunningham et al. 1988; Zahn-Waxler et al. 1990). Distressed parents also tend to utilize discipline strategies that require the least effort (Zahn-Waxler et al. 1990); these strategies have been labeled a permissive style of parenting. Moreover, research suggests distressed as compared to non-distressed parents more frequently avoid conflict (permissive parenting), are more inclined to use forceful control strategies (authoritarian parenting) when not yielding to the child’s demands (Fendrich et al. 1990), and are less likely to end disagreements in a compromise (Kochanska et al. 1987). Therefore, the stress of living with a distressed parent can be characterized by increased negative and unpredictable parental behaviors (e.g., irritability, inconsistent discipline), and decreased supportive parental behaviors or an authoritarian style of parenting (e.g., less warmth, praise, nurturance). Further, it appears parental distress leads to disruptions in parenting, including parental withdrawal (e.g., avoidant, unresponsive to their children’s needs) and features of an authoritarian style of parenting such as intrusiveness (e.g., irritable toward their children, overly involved in their children’s lives).

Dysfunctional interactions between parent and child have been frequently associated with parental distress as well. Parental depression, for instance, is often connected to more negative cognitions about the self, world, and future (Beck 1976), and a tendency toward negative attributions (Alloy et al. 2006). These negative cognitions may be expressed in more critical and intrusive parent–child interactions (Tompson et al. 2008). Critical interactions, in turn, have been linked with children’s higher self-criticism and lower self-esteem (Kuperminc et al. 1997). Thus, symptoms of parental distress may have a strong negative impact on parent–child relationships (Shelton and Harold 2008).

Although child aggression is connected to parental distress, the current study was the first to examine the contribution of specific types of perceived child aggression (reactive and proactive aggression) to parental distress. Given the earlier findings, we expected that parent’s perceptions about their children’s reactive and proactive aggression would contribute to their distress. This study was also the first investigation to examine the relationship between children’s aggression as perceived by Hong Kong parents and this group’s degree of parental distress. Previous studies on parental distress in Hong Kong have focused on children with chronic illness (e.g., cancer) and general emotional and behavioral problems. In one study, it was discovered that there was a reciprocal interaction between Hong Kong children’s emotional or behavioral problems and their parents’ psychosocial well-being, which adversely affected the children (Ma et al. 2002). In another study (Doo and Wing 2006), parents of Hong Kong children with pervasive developmental disorders and sleep problems reported greater levels of stress than parents that had children with similar disorders and no sleep problems.

This study also examined how types of parenting styles and parent–child interactions might contribute to parental distress. It was hypothesized that authoritarian and permissive parenting styles would positively contribute to parental distress, while authoritative parenting styles would negatively contribute to parental distress. In addition, both parent-to-child and child-to-parent interactions were expected to contribute to parental distress.

Method

A public seminar was initially held to introduce the project. Invitations were sent to all schools in Hong Kong (approximately 800), resulting in responses from 62 primary and 28 secondary schools. Ten of the 28 secondary schools and another 10 of 62 primary schools were selected for participation based on their representativeness to the general population. Data for this study was collected from 1,485 parents (1,057 mothers, 271 fathers, 157 other relatives) with children 8 to 17 years old drawn from 10 primary (Hong Kong P.4 to P.6; U.S. equivalent is Grades 4 to 6) and 10 secondary schools (Hong Kong Forms 1 to 3; U.S. equivalent is Grades 7 to 9). Table 1 presents demographic characteristics of the biological parents who participated in this study.

Table 1 Demographic characteristics of biological parents
Full size table

Procedure

The current study was connected to a large-scale 10-session cognitive-behavioral group intervention program in Hong Kong aimed at reducing students’ aggressive behavior in the schools. For further details about the session-by-session content of the groups, the interested reader is referred to a manual developed by Project CARE (Fung 2008).

Measures

Reactive Proactive Aggression Questionnaire (Raine et al. 2006; RPQ)

The RPQ is comprised of 23 items rated on a three-point scale ranging from 0 (never) to 2 (often) for the frequency of behaviors (e.g., “Had fights with others to show who was on top,” “Gotten angry when others threatened you”). The measure consists of two subscales. The Reactive Aggression subscale (11 items) assesses the frequency of times a person behaves aggressively under self-perceived provocative situations, while the Proactive Aggression subscale (12 items) measures the frequency of aggression-related behaviors designed to influence or take advantage of others for instrumental purposes. Internal reliabilities for the RPQ were .84 (reactive subscale), .86 (proactive subscale) and .90 (total scale). Raine et al. conducted a confirmatory factor analysis (CFA) and found support for the two-factor (reactive & proactive) model. The researchers also reported sufficient construct, convergent, criterion, and discriminant validity for a population of children.

According to Fung et al. (2009), the Chinese RPQ has sound psychometric properties as well; they reported internal reliability coefficients of .88 (proactive), .83 (reactive), and .89 (total scale). Furthermore, a CFA indicated the two-factor model was an appropriate solution. In the current study, parents were asked to share their perceptions about their children’s behaviors when completing the Chinese version of the RPQ. It should be noted that there is no evidence at this time for the validity of a parental version of the RPQ.

Kessler Psychological Distress Scale (Kessler et al. 2003; K10)

The K10 is a self-report scale of global psychological distress containing 10 questions about emotional states (e.g., “In the last 4 weeks, how often did you feel nervous?”, “During the last 30 days, how often did you feel hopeless?”) rated on a 5-point scale ranging from 1 (none of the time) to 5 (all of the time). In this study, a slightly different response format was used (1 = never, 2 = rarely, 3 = sometimes, 4 = often, 5 = always) because when translating the original response format into Traditional Chinese, the meaning of this format was difficult to comprehend. The K10 has adequate internal reliability (α = .93). A Chinese version of the K10 was found to have lower but sufficient reliability (α = .80; Zhou et al. 2008). For the current study, the K10 served as a brief assessment of parental distress.

Adjective Checklist (Friedmann and Goldstein 1993; ACL)

The ACL is a self-report measure that originally assessed expressed emotions of family members of patients with schizophrenia. More specifically, a family member is asked to describe his or her behavior toward another family member, and then to describe the family member’s behavior toward him or her. The scale includes 10 items with positive (e.g., loving, good-natured, friendly, devoted) and 10 with negative (e.g., rude, mean, lazy) values. Friedmann and Goldstein reported internal reliabilities of .88 (describing own behavior) and .92 (describing behavior toward them) for the negative adjectives, and .92 (describing own behavior) and .94 (describing behaviors toward them) for the positive adjectives.

Li and Arthur (2005) translated the ACL into Chinese and then had it back-translated by an independent expert into English to establish an accurate translation. The measure was then given to Chinese individuals living in Beijing. According to the researchers, the Chinese version of the ACL possesses acceptable content validity and also internal reliability with Cronbach’s alphas of .74 (describing own behavior) and .85 (describing behavior toward them) for negative adjectives, and .79 (describing own behavior) and .90 (describing behavior toward them) for positive adjectives.

The ACL was modified for this study. The ACL items were written in traditional Chinese and, like Arthur (2002), we used an 8-point scale (1 [never] to 8 [always]) and asked parents to rate the extent to which the adjectives described communication patterns with their children. Parents were first asked to describe their own behaviors toward their children and then asked to describe their children’s behavior toward themselves. Arthur (2002) found acceptable alpha coefficients for the adjectives “describe your behavior” toward your relative over the last 3 months (α = .82) and how well each of the adjectives “describe your relative’s behavior” (α = .75) toward you over the last 3 months for the patients in his sample, and borderline (α < .70) coefficients for the family members and nurses in his study.

Parenting Styles and Dimensions Questionnaire (Robinson et al. 2001; PSDQ)

The PSDQ is one of the most often used scales for examining parenting styles. The scale consists of three parenting scales (authoritative, authoritarian, permissive) and a total of 32 items. Sample items include “I am responsive to our child’s feelings and needs” and “I shout at him.” Items are rated on a 5-point Likert scale ranging from 1 (never) to 5 (always). The authoritative scale is comprised of subscales with items that assess warmth and involvement, reasoning/induction, democratic participation, and good natured/easy going behaviors. The authoritarian scale contains subscales for verbal hostility, corporal punishment, non-reasoning/punitive strategies, and directiveness. Finally, the permissive scale includes subscales for lack of follow through, ignoring misbehavior, and self-confidence. While no validity information is available on the PSDQ, Robinson et al. reported internal reliability coefficients of .81 (authoritarian), .83 (authoritative), and .65 (permissive). In our study, the alpha coefficients for the three scales were as follows: authoritative (α = .85), authoritarian (α = .71), and permissive (α = .66) parenting. The Chinese version of the PSDQ was used in our study.

Analyses

Pearson product correlations were performed to investigate the relationship between participants’ responses to the various measures administered in this study. Multiple regression analyses were conducted to test the proposed hypotheses.

Results

Table 2 presents descriptive statistics on all the variables investigated in this study as well as zero-order correlations between the responses to these variables. Reactive and proactive aggression responses were positively and moderately correlated. Significant correlations were found between reactive and proactive aggression responses and almost all the other variables displayed in Table 2. The results were similar for the parenting style variables. Two of the three inter-correlations between these variables were significant and responses to all three parenting styles were correlated with parental distress. Surprisingly, responses to the parenting style variables in every case but one were not correlated with the parent–child or child–parent interaction variables. Finally, there was a strong positive correlation (.77) between the parent–child and child–parent interaction variables. Given the size of this correlation, it was questionable whether these two variables were measuring unique constructs. Thus, a composite parent–child and child–parent interaction variable was created. As presented in Table 2, this new variable was minimally correlated with the two aggression variables and only the authoritarian parenting style. It was not correlated with the parental distress variable.

Table 2 Descriptive statistics and correlations for parents’ perceived child aggression, parenting styles, parent–child interactions, and parental distress
Full size table

To test the hypotheses that child aggression, parent–child interactions, and parenting styles variables would predict parental distress, three sets of standard multiple regression analyses were performed, respectively. Consistent with our prediction, when reactive and proactive aggression responses were used to predict parental distress, a significant model was obtained, F(2, 1482) = 58.36, p < .0001. Both predictors contributed to this model (see Table 3). A significant model also was discovered when parenting styles and the composite parent–child interaction patterns were used to predict parental distress, F(4, 1480) = 84.77, p < .0001. In this model, authoritative, authoritarian, and permissive parenting styles contributed significantly, while the composite parent–child interaction variable did not (see Table 3). This finding was somewhat consistent with our hypothesis in that parenting styles made a contribution to the model. One final regression was performed with the significant predictors from the two previous equations. This model was significant, F(5 1479) = 72.50, p < .0001, with all the predictors making a significant contribution to parental distress (see Table 3).

Table 3 Summary of regression analysis for predicting parental distress (N =1,485)
Full size table

Discussion

This study explored the contribution of Hong Kong parent’s perceptions about their children’s aggression, parenting styles, and parent–child interactions to their level of parental distress. In general, our hypotheses were supported by the results. First, as expected, Hong Kong parents’ perceptions about their children’s reactive and proactive aggression made a significant contribution to parental psychological distress. In fact, as anticipated, as parent’s perceptions about the severity of their children’s proactive and reactive aggression increased, so did their distress. This finding is consistent with previous research that found parental distress was linked with the aggressive behavior of the parent’s children (Anderson et al. 1986; Patterson 1980). Living with and rearing an aggressive child that tends to manipulate others, including their parents, for their own favor is not a pleasant or easy experience, especially for the parents. For instance, proactive aggressive children are deliberately aggressive with the intent of taking advantage of others for instrumental purposes (Poulin and Boivin 2000; Raine et al. 2006); this can include their parents. Further, proactive child aggressors’ experience satisfaction or pleasure by acting aggressive. Such an emotional response can be very disturbing to the proactive aggressors’ parents.

In contrast, reactive child aggressors respond to provocation or perceived threats with aggression due to an inability to regulate their emotions. Their aggressive behavior is more likely to be accompanied by anger (Poulin and Boivin 2000; Raine et al. 2006). One would suspect that parents are often the object of such anger, and again, subject to distress in the face of this form of aggressive behavior from their offspring.

As we discovered and expected, Hong Kong parents’ style of parenting also contributed to their level of distress. We hypothesized that authoritarian or permissive parenting styles would positively contribute to Hong Kong parents’ distress, while authoritative parenting styles would negatively contribute to their distress. Like others, we found that distressed parents tended to employ parenting strategies that required the least effort (Zahn-Waxler et al. 1990). Stated another way, as Hong Kong parents’ use of permissive parenting approaches increased (where they lacked follow through on their proposed actions and ignored their children’s misbehavior), so did their parental distress.

Our parents were also more likely to employ forceful control strategies (Fendrich et al. 1990) or authoritarian behaviors when managing their children’s behaviors as their distress level rose. As predicted, parental distress also increased as parents’ use of warmth, reasoning, democratic strategies, and their ability to be good natured and easy going in their parenting decreased; that is, the less inclined Hong Kong parents were to employ an authoritative parenting style with their children, the greater degree of distress they experienced.

Given these results, it is possible that Hong Kong parents experiencing psychological distress displayed less supportive parental behaviors; that is, less warmth, praise, and nurturance was demonstrated in their parenting styles. Consistent with Langrock et al. (2002), the more intrusive and overly involved our Hong Kong parents were in their children’s lives, the higher degree of parental distress they reported. Moreover, the more incompetent and lacking in efficacy with regard to parenting their children, the greater levels of parental distress shared by our Hong Kong participants.

It was expected that higher degrees of dysfunctional parent–child interactions would result in elevated levels of parents’ distress (Shelton and Harold 2008), but contrary to our prediction, our Hong Kong parents’ interactions with their children did not contribute to their level of distress. It is unclear why this finding was not discovered. Perhaps it had something to do with the fact that the parents in this study reported a very high degree of functional interactions with their children regardless of their level of distress. In fact, their mean interaction score (243.75) was almost two standard deviations above the midpoint in the potential range of scores (40 to 320), indicating their interactions with their children were extremely positive.

Limitations and Conclusion

There are several limitations worth noting. First, the data were cross-sectional, and as such, it was not possible to test for causal relationships. Next, we only investigated how a few variables contributed to the experience of parental distress; other factors might also play a role in this phenomenon. For instance, a strong social support network has been shown to decrease parents’ likelihood of developing psychological problems such as depression (Simons et al. 1993) by providing them with the confidence, patience, and energy needed to use reasoned and rational parenting approaches. Similarly, other factors that may contribute to parental distress (e.g., financial distress, community violence, family medical problems) were not included in the current study. Future research is needed to investigate the role such other factors in the prediction of parental distress and their unique contributions along with the variables investigated in this study.

Another limitation of this study is the fact that over 80 % of the respondents were mothers and all the parents were from Hong Kong. Therefore, it is unclear if the results accurately reflect the responses of fathers and if they can be generalized to parents living in locations other than Hong Kong. Future studies should investigate and also compare the responses of both mothers and fathers when predicting parental distress and persons living in Hong Kong and elsewhere. Lastly, since we only collected data on parents’ perceptions about their children's aggression and their interactions with their children, it is important to conduct a study to gather responses from children and adolescents about their own aggression and interactions with their parents.

Despite the limitations of this study, our results suggested that psychological distress among Hong Kong parents was related to their perceptions about their children’s aggression and their parenting styles. We suspect that mental health prevention and intervention programs for parents that discuss and teach specific strategies to effectively address children’s different forms of aggressive behavior and that stress overall parenting approaches intended to successfully rear children would be beneficial in not only helping to improve Hong Kong parents skills in parenting, but also lessening this group’s psychological distress.