Considerable research in recent decades has shown that externalising and internalising behaviours in children consistently predict maladjustment in adulthood (Bennett et al., 1999; Huesmann, Eron, & Dubow, 2002; Loeber, 1982). Consequently, a vast amount of literature has examined the myriad of variables contributing to behavioural difficulties. For the purposes of this review, however, the focus will be on childhood temperament, parenting behaviours and parenting stress. These constructs are emphasised because they are considered unique and interactive contributors to the development of behavioural difficulties (Smeekens, Riksen-Walraven, & van Bakel, 2007). Thus, the aim of the present paper is to examine research studies regarding the independent effect of child temperament, parenting and parenting stress on child behaviour, and how these factors interact to produce and sustain child behavioural problems. Gender differences will be addressed throughout the review, along with differences between clinical and community samples where possible.
The present review begins with a description of child behavioural difficulties by referring to normal child development, and explaining internalising and externalising problems. This is followed by a discussion of how child temperament can act either as protective or risk factor on children's development. Next, the individual effect of parenting on internalising and externalising behaviours is reviewed, utilising social learning theory and attachment theory. Lastly, the literature on parenting stress, particularly stress arising from daily hassles and socio-economic status (SES) is discussed. In concluding, the cumulative impact of parenting stress, parenting behaviours and temperament on behavioural difficulties are assessed along with the current research gaps.
Behavioural difficulties present one of the most challenging aspects of childhood development. Many researchers have established that early childhood behavioural difficulties impair functioning in social, emotional and occupational domains in adulthood (Bennett, et al., 1999; Keenan, Shaw, Delliquadri, Giovannelli, & Walsh, 1998; Loeber, 1982; Mash & Wolfe, 2007; Moffitt, 1993; White, Moffitt, Earls, Robins, & Silva, 1993). The more we understand about why behavioural problems develop, the more clinicians and community professionals will be able to divert children away from a detrimental path. In order to achieve this understanding, typical children's behavioural patterns must be addressed. Mash and Wolfe (1997) note that it is expected that children will learn language skills, self control and compliance during the preschool period. Through middle childhood and adolescence, children are expected to develop social relationships with peers, behave within a regulated society, learn academic skills and form a cohesive self identity. Children who do not achieve these developmental milestones may be at risk of future maladjustment. Furthermore, the authors indicated that deviancy from established norms of development must be examined in terms of same-age peers. For example, a temper tantrum by a preschool child is considered a normal response to conflict as children have underdeveloped communication and social skills at this age (Tremblay, 2000). However, a temper tantrum by an adolescent is considered abnormal as this is not an appropriate behaviour at this age.
Reference to gender differences must also be made as community samples of girls are more compliant, prosocial, easier to manage and engage in less physical and verbal bullying than non-risk boys (Prior, Sanson, Smart, & Oberklaid, 2000; Zahn-Waxler, 1993). Boys generally display more aggressive and oppositional tendencies in preschool than girls (Fagot, 1984). An understanding of these different behavioural trends by gender allows clinicians to assess whether the behaviour significantly differs from the norm. Consequently, the achievement of developmentally appropriate skills and the comparison to same age and gender peers provides criteria for determining normal and abnormal behaviour.
Child Behaviour Difficulties
While the overall literature regarding abnormal behavioural difficulties is extensive (Campbell, Shaw, & Gilliom, 2000; Dishion & Patterson, 2006; Rhee et al., 2007), internalising difficulties have received less academic attention than externalising problems. The presenting symptoms are often less disruptive, noticeable and less consistently documented throughout childhood (Mesman, Bongers, & Koot, 2001). According to Rubin and Mills (1991), internalising behaviours are manifestations of "over-control", and include social withdrawal, anxiety and depressed mood. The social internalising symptoms include an unwillingness to engage in play with peers, poor social skills and social anxiety (Rubin & Mills, 1991). The emotional symptoms include low self esteem, insecurity (Pettit, Laird, Dodge, Bates, & Criss, 2001), distress, anxiety and fear (Eisenberg et al., 2001). These behaviours present as mood and anxiety disorders in clinical populations with 12.8 percent of children being diagnosed with internalising disorders in Australia (Sawyer et al., 2001). However, Sawyer and her colleagues also reported comparable emotional and behavioural problems in non clinical populations. This data suggests that further investigation in the general community is needed. The understanding of social withdrawal, anxiety and depression in childhood is constantly evolving as research uncovers more about the causes of internalising problems and their subsequent prevalence (Bell, Foster, & Mash, 2005; Prior, et al., 2000).
In order to further understand developmental pathways to internalising behaviours, it is useful to explore their presentation by gender in both community and clinical samples. Stacks and Goff's (2006) study of 63 children aged between three and five years explored gender differences in psychological symptoms, stress, social support, parenting attitudes toward discipline and child behaviour. Findings revealed no significant differences between girls and boys' internalising problems. Only 4.5% of males and 2.4% of females were considered to have clinically significant internalising difficulties at home. These findings are contrary to the intuitively held belief that boys have fewer internalising difficulties than girls. Previous research may help explain Stacks and Goff's unexpected result, as internalising problems arising from negative emotionality can be co-morbid with externalising symptoms (Rhee, et al., 2007). Therefore, boys may display aggression and frustration due to difficulties in social interactions. As such, their externalising behaviours are the result of internalising problems. Alternatively, boys may experience emotions such as sadness if they are rejected by peers, teachers and parents for their disruptive externalising tendencies (Eisenberg, et al., 2001).
Additionally, Stacks and Goff's finding suggest that preschool internalising symptoms are similar in boys and girls. Perhaps this can be attributed to the fact that only 20 percent of the sample reached the threshold of clinically significant symptoms (Stacks & Goff, 2006). The remaining 80 percent of the sample who did not present with clinical behavioural difficulties were not analysed, yet this population of participants could yield interesting results about these concerns in the general community. If the remaining 80 percent of the sample were examined, a difference between subclinical internalising behaviours in boys and girls may be found. Thus, the continuum of behavioural difficulties from typical to clinical symptoms in boys and girls is worth considering for future research to address the shortcoming of Stacks and Goff's study.
The strength of the study lies in the authors' discussion of significant methodological issues in assessing internalising problems, which have been noted by other researchers in the field. In particular, Stacks and Goff made reference to the high level of discord between teachers and parents in ratings of problem behaviours. Webster-Stratton (1996) also noted this issue, that teachers were less likely to report internalising behaviours in girls while fathers often reported internalising symptoms in their daughters. In addition, Stacks and Goff's research referred to the accepted understanding that girls present with internalising difficulties at a much higher rate during childhood and adolescence. Indeed, this notion was also supported by Webster- Stratton who noted that by adolescence, girls were twice as likely to experience internalising problems as boys. These researchers have implied that preschool may be a turning point whereby girls and boys follow different pathways that lead to different behavioural outcomes. As it is less acceptable for girls to display disruptive behaviour beyond preschool, girls' behaviour may be channelled into internalizing problems as a result of gender specific socialization processes (Keenan & Shaw, 1997). In saying so, future research may explore the underlying features of social and emotional development in preschool and childhood to further understand how internalising difficulties in both genders follow diverging paths later in life (Bell, et al., 2005).
In comparison to internalising problems, research on externalising problems in young children is extensive and consistently documented. Defined as behaviours of "under control" (Rubin & Mills, 1991), the externalising problems in community samples include aggression, non compliance, poor concentration and impulsivity (Kroneman, Loeber, Hipwell, & Koot, 2009). The clinical manifestations of these behaviours are Conduct Disorder (CD), Oppositional Defiance Disorder (ODD) and Attention Deficit Hyperactive Disorder (ADHD) once the behaviours reach the formal diagnostic criteria in terms of frequency, severity and impairment of function (American Psychiatric Association, 1994). Boys have typically been more represented than girls in community and clinical samples of externalising behaviours (Mesman, et al., 2001).
An influential study by Campbell and colleagues (1994) examined these externalising symptoms of aggression, noncompliance and inattention / impulsivity in a sample of 112 preschool boys drawn from the community. Campbell's et al. (1994) findings were not unexpected as problem boys were more active and disruptive; more likely to be noncompliant with the teacher; and more likely to engage in rough play at age four and age six. The data suggests that the underlying causes of these externalising symptoms are due to a diminished capacity to self-regulate behaviour. In addition, Campbell's study showed that these behaviours are not transient stages but persistent across a two year period (Campbell, Pierce, March, Ewing, & et al., 1994). A particular strength of this study was its ability to compare preschool boys across different contexts (home, laboratory and preschool), against a control group of normally developing boys and across two time periods. Furthermore, the authors' data provided the groundwork for future studies to explore the stability of early externalising behaviours in boys (Keenan, et al., 1998). While these behaviours are not male specific, the presentation of externalising behaviours in girls has a significantly different profile to boys.
At the outset, there is little doubt that boys are more physically aggressive than girls (Loeber & Hay, 1997). This is not to say girls never engage in violence as young girls are increasingly at risk for involvement in violent criminal activities and drug abuse (Cote, Zoccolillo, Tremblay, Nagin, & Vitaro, 2001). Theorists have concluded that girls are simply less aggressive than boys however; research indicates girls' alternative favour relational aggression and callous behaviour, such as spreading rumours and excluding peers. A pioneering study, known as the Pittsburgh Girls Study (Hipwell et al., 2002) investigated the increase in female specific externalising behaviours. The large sample comprised 2451 girls aged between five and eight years. Their parents and/or teachers completed questionnaires on the global functioning, antisocial behaviour and peer relations of the child. Hipwell et al. (2002) found that parent ratings of callous, unemotional behaviour were more common among five-year-old girls while teacher ratings of relational aggression were more common in the seven to eight-year-old age group. This data suggests that girls may outgrow callous behaviour in favour of relational aggression. Hipwell et al. (2002) speculated whether these age trends in girls' externalising behaviours would persist and rightly concluded that a follow-up study was necessary. On this point, Maccoby's (1990) work on gender and relationships may provide context for Hipwell's data. He theorised that girls are more likely to abstain from using physical violence in interpersonal conflicts because they are socialised to value their relationships. Therefore, relational aggression is a successful tool to exert control in social conflicts without being overtly aggressive.
In conclusion, boys frequently demonstrate externalising behaviours such as physical aggression, noncompliance and impulsivity while girls tend to display relational aggression. It has been incorrectly assumed that girls are not aggressive though their use of relational aggression shows that they display a different form of externalising behaviour. Taken with the research on internalising problems, behavioural difficulties have a range of presentations in community and clinical populations of children. Existing research does favour clinical populations, especially male samples, though some strides have been made to address these concerns in the community. Equipped with the knowledge of the forms behavioural difficulties take, the child and family specific conditions under which these behaviours develop can be explored.
Childhood Temperament and Child Behaviour
An empirically validated determinant of child behaviour is childhood temperament (Berden & Calkins, 2008). Conceptually, a myriad of interpretations of temperament exist, although theorists agree that temperamental dimensions have an innate basis which also includes activity level and emotionality (Goldsmith et al., 1987). Specifically, Thomas and Chess proposed that temperament is interactive with the social environment (Thomas & Chess, 1977). Rothbart et al. (2000) supplements this definition by emphasising self regulation. These definitions will be utilised to explain how 'easy' and 'difficult' temperament influences behaviour and interacts with protective and risk environments.
Childhood Temperament and Protective Factors
The 'easy' temperamental traits of positive emotionality, high self regulation and effortful control have often been cited as determinants of adaptive behaviour and functioning (Rothbart, Ahadi, & Evans, 2000). These traits allow children to adjust effectively to their social environments. A study by Blair, Denham, Kochanoff and Whipple (2004) attempted to assess the temperamental traits that lead to positive adjustment in childhood and adolescence. The study of 153 preschool children examined the impact of self regulation, temperament and social competence on future development. Blair et al. (2004) found that children who were optimally self-regulated demonstrated the most adaptive behaviour. These children had the ability to abstain from hedonistic urges. As such, they were more popular and more likely to experience and exhibit positive emotions. Thus, the authors concluded that positive emotionality and moderate self-regulation are protective factors in the development of positive adjustment in childhood. Blair et al.'s study is a departure from the tradition in child psychopathology of focusing solely on diminishing risk factors. Few studies have taken a collaborative approach that includes both risk and protective factors and Blair's study adequately demonstrates how both factors cumulatively effect child development.
Childhood Temperament and Risk Factors
In contrast to easy temperament and its subsequent pathways to adaptive behaviour, difficult temperamental dimensions such as negative emotionality, low self-regulation and poor effortful control have been consistently linked with behavioural difficulties (Caspi, Henry, McGee, Moffitt, & Silva, 1995; Eisenberg, et al., 2001). Views about temperament dimensions are fragmented, however, as there is little agreement amongst theorists about the characteristics that comprise each of these dimensions (Rothbart, et al., 2000). For clarification purposes, the present paper will utilise Eisenberg and colleagues' (2001) summary of the dimensions and their subordinate traits. The temperamental traits of sadness, fear and anger are coded as emotionality; inhibitory control falls under self-regulation; and effortful control comprises attentional focus and attentional shifts.
Eisenberg et al.'s (2001) study of 214 children aged between four and eight years seamlessly integrates the research on temperament and provides a simple and coherent way of explaining which temperamental characteristics are linked with certain behavioural outcomes. The study allowed multiple comparisons between; clinical and community populations of boys and girls, informants and types of behavioural difficulties. This comparative feature of different populations and contexts is severely lacking in other research. Eisenberg et al. (2001) found that boys typically possessed negative temperamental emotionality, with a propensity for anger; low in effortful control and high in impulsivity. Therefore, these children were less able to restrain hedonistic desires and displayed externalising symptoms. In contrast, girls generally possessed negative emotionality with a propensity for sadness, low attentional control and low impulsivity. They were more likely to display internalising symptoms (Eisenberg, et al., 2001) as they lack the spontaneity necessary to adapt to their environment (Caspi, et al., 1995).
In fact, Caspi and his colleagues (1995) also provided interesting results about frustration, a temperamental trait of emotionality. This is akin to Eisenberg's proneness to anger. Caspi et al. (1995) found that children who are prone to frustration were more likely to respond aversively to obstacles that prevented them from doing what they want. Thus, increased frustration is likely result in aggression and non compliance to achieve their goals (Caspi, et al., 1995). Furthermore, Caspi's study explored a temperamental trait specific to girls that was neglected in Eisenberg's research; that is, sluggishness. This trait manifests as wariness, shyness and fearfulness (Coplan, Bowker, & Cooper, 2003; Cote, Zoccolillo, Tremblay, Nagin, & Vitaro, 2002) and not only were girls more likely to display high levels of sluggishness; they were also more likely to experience anxiety and social withdrawal throughout childhood and adolescence (Caspi, et al., 1995; Prior, et al., 2000).
These temperamental profiles are considered risk factors in the development of internalising and externalising behaviours. However, it is important to remember Thomas and Chess's definition; that is, temperament alone is not the sole cause of behavioural problems (Bates, Pettit, Dodge, & Ridge, 1998). In saying so, temperament not only influences the development of behavioural difficulties but also affects the way a child is parented (Rothbart and Bates, 1998). As such, the effect of parenting on child behavioural difficulties (Hollenstein, Granic, Stoolmiller, & Snyder, 2004; Zubrick et al., 2008) will be reviewed.
Parenting and Behavioural Difficulties
Parenting is undeniably one of the principle influences on child behaviour (Dishion & Patterson, 2006). It has been consistently documented that suboptimal levels of parenting has a direct impact on the development of internalising (Bayer, Sanson, & Hemphill, 2006; Hipwell et al., 2008) and externalising problems in children (Baumrind, 1971; Hiramura et al., 2010; Hollenstein, et al., 2004; Schaffer, Clark, & Jeglic, 2009; Verhoeven, Junger, van Aken, DekoviÄ‡, & van Aken, 2010). Social learning theory and attachment theory provide perspectives to examine how ineffective parenting styles such as harsh authoritarian parenting and permissive parenting are associated with behavioural difficulties (Campbell, 1997; Hipwell, et al., 2008; Patterson, 1986). Social learning theory postulates that children absorb the behaviours of significant role models, such as parents, and apply those behaviours in other situations (Bandura, Ross, & Ross, 1961). In contrast, attachment theory emphasises the quality of the parent-child relationship in predicting future adjustment (Ainsworth, Blehar, Waters, & Wall, 1978).
Social Learning Theory
With reference to social learning theory, authoritarian parents who issue harsh, physical punishments, inadvertently teach their children that aggression is an effective conflict resolution strategy (Bandura, et al., 1961; Baumrind, 1971). This is particularly relevant to young boys as a study by Lytton and Romney (1991) found that boys are dealt physical punishment more often than girls. Acknowledging this finding, the literature on aggression in boys is extensive as boys are more likely to have physically aggressive models to emulate (Campbell, Spieker, Vandergrift, Belsky, & Burchinal, 2010; Loeber & Hay, 1997). Bandura's social learning theory provides a context to explore how aggression develops in boys though the principles of the theory are applicable to female populations as well.
Wilson, Parry, Nettelbeck, and Bell (2003) relied on social learning theory to investigate whether a sample of 333 children aged 10 to 11 years modelled conflict resolution tactics on parents, television characters or peers. Teacher and children identified the children as either a bully or victim and the children reported on what type of behaviour their parents, peers and characters on television would exhibit in an argument. Although the surprising outcome of the study was that children model peers' behaviour in an argument (Wilson, Parry, Nettelbeck, & Bell, 2003), it was also found that bullies model their aggressive behaviours on their parents' conflict resolution tactics. However, Wilson et al. (2003) acknowledged that the finding of social learning from parents was unclear. The authors referred to prior research that established aggressive children were likely to have learned their behaviours from hostile family treatment and exposure to adult aggression (Schwartz, Dodge, Pettit, & Bates, 1997). Overall, the study demonstrated the salience of social learning theory in the development of aggression though in an unexpected direction.
A study by Sanders, Pidgeon, Gravestock, Connors and Brown (2004) clarified the impact of parenting behaviour on children. Participants were 98 families with a child aged two to seven years. The study aimed to test the efficacy of an enhanced behavioural family intervention (EBFI) program, based on social learning principles. The EBFI aimed to decrease coercive parenting strategies with attributional retraining, anger management and self-regulation training (Sanders et al., 2004). The program sought to teach parents how to control and manage their own behaviour as they act as a model for their child's behaviour (Sanders, 2008). The EBFI showed long term benefits in families with preschool girls and boys which is a departure from existing research that favours males as girls do not display as much overt aggression (Gorman-Smith & Loeber, 2005; Loeber & Hay, 1997). However, the surge in studies on relational aggression in girls (Loeber et al., 2009) may utilise social learning theory to explain girls' typical socialization focussing on interpersonal relationships and curtailing physical aggression (Kroneman, et al., 2009). Therefore, girls use relational aggression as they have learnt through observation that it is more socially acceptable (Kroneman, et al., 2009). In summary, children internalise the values of their parents and imitate their behaviour though this may depend on the strength of the attachment relationship (Wiggins, Sofronoff, & Sanders, 2009). Social learning theory offers an insightful analysis into behavioural difficulties and provides an appropriate segue to discuss the role of attachment as a complementary theoretical perspective.
Attachment theory, as the study of parent-child relationships is particularly relevant to discuss the impact of parenting on behavioural difficulties. This theory presents two main attachment styles: secure and insecure attachment. Secure attachment is typically associated with positive adjustment in children, while insecure attachments have been consistently linked with poor future relationships and maladjustment in adulthood (Ainsworth, et al., 1978; Stacks, 2007). A recent study by Vando et al. (2008) examined the impact of infant attachment on conduct problems at age six and the mediating roles of hostile parenting and maternal depression. Findings revealed that insecurely attached infants were directly affected by hostile parenting but not maternal depression. In this instance, hostile parenting is similar to parenting that is low in sensitivity and high in responsivity (Vondra & Barnett, 1999). This type of parent responds to their child's needs with aggression and little warmth, (Rubin & Mills, 1991) leading to social withdrawal and anxiety. In addition, authoritarian and overcontrolling parents do not allow their child to explore their environment independently and take over for their socially reticent child or they are overly assertive in instructing their child to perform tasks (Coplan, et al., 2003).
The contrasting parenting style of high sensitivity and high responsivity in the form of anxious, overprotective parenting can lead to internalising problems as well (Mantymaaa et al., 2009; Pettit, et al., 2001). Generally, anxious parents are so preoccupied with the negative situations that may befall their child that they do not allow their child to experience their environment independently (Coplan, et al., 2003). In this way, the concepts of sensitivity and responsivity of parents has significant implications in the study of behavioural difficulties in children. However, internalising difficulties are not the sole outcome of insecure and disorganised attachment; externalising problems are also likely to develop if the parent-child relationship is difficult.
Attachment theory provides a lens to understand externalising problems in children (Braungart-Rieker, Garwood, Powers, & Wang, 2001; Vando, Rhule-Louie, McMahon, & Spieker, 2008). Disorganised attachment, characterised by inconsistent caregiver response, is frequently cited as a contributor to externalising difficulties (Lyons-Ruth, Easterbrooks, & Cibelli, 1997). Lyons-Ruth et al. (1997) study of attachment strategies in 50 seven-year-old children found that disorganised attached infants' needs are not met which may contribute to aggressive and maladaptive responses towards others (Lyons-Ruth, et al., 1997). Therefore, parenting behaviour that is intrusive and dismissive of infant initiated communication sets the scene for later parent-child interactions. In addition, disorganised attachment typically precedes the onset of coercive cycles (Lyons-Ruth, et al., 1997). This notion was first introduced by Patterson (1986) in his study of performance models in boys. The coercive cycle model also has implications in the study of girls externalising behaviours and disruptions in the parent-child attachment relationship.
Miller, Loeber and Hipwell (2009) study of externalising behaviours in girls can be viewed within the framework of Patterson's coercive model. Miller et al. (2009) found in a sample of 2451 parent-daughter dyads aged five, six, seven and eight years that harsh discipline strategies and low parental warmth were strongly associated with girls' disruptive behaviours. These experimental findings were established from measures of association with delinquent peers, parenting practices and conduct and oppositional problems. The authors acknowledged that parents may employ punitive discipline strategies in an effort to control increasingly difficult child behaviour. The externalising behaviours intensify as daughters become frustrated at the restrictions placed on them, noted in Patterson's coercive model. Reference to attachment theory can also be made in the development of this coercive cycle as Buist et al. (2004) theorises that an aggressive child may place a strain on the relationship between child and parent. The child may perceive the parent as less sensitive and responsive when he or she is punished for a transgression, thereby increasing the externalising symptoms as a reaction to the punishment.
A significant shortcoming of the Miller et al. (2009) is the neglect of internalising concerns in girls which is a conceptual weakness, especially as it is understood that girls are prone to internalising symptoms (Webster- Stratton, 1996). This does not, however, diminish the value of the study as it provided an adequate explanation of how low parental warmth causes emotional dysregulation in girls which leads to externalising concerns.
Based on current knowledge, social learning theory and attachment theory provide a framework for understanding parenting effects on internalising and externalising behaviours in girls and boys. That is not to say one process solely affects girls and one solely affects boys; it is the cumulative effect of a multitude of factors that influences child behaviour (Webster-Stratton, 2001). In fact, the literature reviewed so far has focused on how temperament and parenting factors influence child behaviour. The importance of parenting in the prediction of behavioural difficulties is clear therefore factors that affect parents' behaviour should be discussed. These factors may be personal and environmental. For example, parental stress is likely to interfere with their ability to effectively manage their child's behaviour.
Parenting Stress and Behavioural Difficulties
Parental stress, as a family environment variable, has been significantly correlated with child behavioural problems (Crnic & Greenberg, 1990). Crnic and Greenberg defined minor daily hassles as the annoying and distressing interactions with the environment that occur on a regular basis (Crnic & Greenberg, 1990). Their study of 74 mother-child dyads found that stress from daily hassles was a significant predictor of poor parental, child and family functioning (Crnic & Greenberg, 1990). However, Crnic and Greenberg downplayed the significance of major stressful life events despite persistent data suggesting that SES, unemployment and poverty significantly impact on parenting practices and subsequent child behaviour (Flouri, 2008; Klebanov, Brooks-Gunn, & Duncan, 1994). However, Crnic and Greenberg's findings are noteworthy as the daily hassles included in the study can be generalised to 'regular' families in the community. In summary, Crnic and Greenberg made strides towards identifying how parenting stress relating to child externalising behaviours significantly impact childhood development.
A study conducted by Klebanov et al. (1994), acknowledges the pitfalls in neglecting SES, as Crnic and Greenberg did, in a discussion of parental stress. The authors found that families living on the poverty line experience more daily stresses which in turn weakens their capacity to deal with stress effectively over a long period of time. This diminished capacity and psychological distress negatively impacts on parenting behaviours, placing the child at risk for behavioural problems (Klebanov, et al., 1994). SES not only has an indirect impact on child behaviour through suboptimal parenting; it is also a risk factor for children who are temperamentally prone to frustration and low effortful control (Veenstra, Lindenberg, Oldehinkel, De Winter, & Ormel, 2006). As low SES restricts access to opportunities, children may experience increased frustration and react negatively without considering the consequences of their actions. As such, the role of SES in internalising and externalising difficulties should not be understated.
These two studies demonstrate how everyday hassles and low SES are significant contributors to parental stress. Parental stress therefore, acts as the lynchpin in the parent-child relationship. The stressors, such as low socio-economic status and daily hassles, impact on the quality of parenting which further impacts on child behaviour.
Parenting Stress, Parenting Style and Temperament: Cumulative Impact on Childhood Behavioural Difficulties
In the family environment, temperamentally difficult children pose a challenge to parents' child management skills. For example, children who display high levels of emotional reactivity, or a predisposition towards anger and aggression are likely to elicit a negative response from their parent (Coplan, et al., 2003). This situation was explained in Coplan, Bowker and Cooper (2003). The study of daily parenting hassles, child temperament and social adjustment in preschoolers comprised a sample of 122 preschool children and their parents. Coplan and colleagues found that child temperament interacted with the daily hassles experienced by parents. Coplan et al. (2003) described the cycle of negative parent-child relations, similarly noted in Patterson's coercive model, as the parent is likely to respond with harsh punishments to control temperamentally difficult behaviour (Hipwell, et al., 2008). This, in turn, leads to escalating behavioural problems as children who are temperamentally prone to anger will respond negatively when they are thwarted in their attempts to get what they want (Caspi, et al., 1995). Therefore, the behavioural problems of aggression and negative emotionality increase, leading to further parenting stress. This cycle of difficult behaviour, parental stress, impaired parental response and further behavioural difficulties presents a valid model for how negative behaviours become entrenched over time (Coplan, et al., 2003). Coplan et al. (2003) attempted to assess the interactive effect of parenting stress and temperament on behavioural difficulties though future research may focus more strongly on the impact of parenting stress on parenting behaviour.
A study by Paulussen- Hoogeboom, Stams, Hermanns and Peetsma (2008) addressed the shortcoming of Coplan et al.'s (2003) study by investigating a parent's capacity to respond to their temperamentally difficult child's needs. The longitudinal study of 59 three-year-olds and their mothers included observational measures and questionnaires to assess parental stress and child negative emotionality. The authors hypothesised that women from middle to high SES backgrounds will have greater resources than women from low SES background to respond suitably to their child's negative emotionality (Flouri, 2008). This was prediction was supported but their ability depended on their own parenting stress and the sex of the child (Paulussen-Hoogeboom, Stams, Hermanns, & Peetsma, 2008). These two studies are recent attempts to address the cumulative nature of factors affecting child behavioural difficulties.
In contrast to coercive parent-child relations that become entrenched over time, over-protective style parenting and internalising symptoms are likely to co-occur in a similarly damaging cycle. For example, as children develop more anxiety, over-involved parents are likely to engage in further controlling parenting to reduce their child's distress (Coplan, et al., 2003). The over-involvement then causes further distress which theoretically would lead to parental stress in not being able to soothe their child. These two models of coercive and over-involved parent-child relations provides a clearer picture of how the interaction of temperament, parenting style and parenting stress can influence child adjustment. On the basis of this model, it is likely that high risk boys will follow the coercive cycle in developing externalising difficulties. This assumption is based on their increased temperamental likelihood of anger and the subsequent harsh punishments that it elicits. In contrast, high risk girls are likely to follow the over-involved cycle in developing internalising difficulties as they are more likely to be temperamentally shy and prone to anxiety which in turn may elicit over-involved parenting.
Parental stress therefore, has significant implications in the family environment and child development. Everyday hassles and SES affect parenting practices and child temperamental profile. The interactive relationship between child and family variables has not been directly studied in previous research, which highlights a gap in the knowledge of child behaviour difficulties. It is clear, however, that the effect of these variables is intertwined and cumulative. As such, the impact of stress should be considered in any research examining internalising and externalising symptoms in childhood.
In sum, the study of child behavioural difficulties has come a long way in the past thirty years. Armed with the knowledge that behavioural difficulties have a significant impact in adulthood, researchers have investigated a wide range of causes of internalising and externalising behaviours. Current research has identified temperamental traits of low effortful control, propensity towards anger and shyness as potential predictors of internalising and externalising difficulties. In addition, the extensive research on parenting has alerted academics and community professionals to the detrimental effects of permissive, authoritarian and over-involved parenting. Lastly, the literature regarding parenting stress demonstrates how everyday hassles and SES can influence the development of behavioural problems. Moreover, it is expected that the cumulative effect of difficult child temperament, suboptimal parenting and high levels of parenting stress places a child at risk for behavioural problems, though this requires further investigation. The individual impact of these variables is significant though the combination of risk factors will have a greater effect on child behavioural difficulties.
While the literature regarding the causes of behavioural difficulties in young children is expansive, there are notable research gaps that must be addressed. First, the majority of the research conducted has focused on male populations despite the growing prevalence of girls' conduct, oppositional and internalising problems (Cote, et al., 2001). While studies of male populations have significantly contributed to current knowledge of childhood behaviour, understandings of behavioural difficulties in girls has suffered. Studies have assumed girls develop in a similar way to boys or simply make very little reference to girls' development. Fortunately, this dearth in the research is slowly being addressed with enlightening studies by Hipwell et al. (2002), Cote et al. (2002) and Kroneman et al. (2009). With the advent of new empirical studies on girls' development, more appropriate gender comparisons can be made.
Second, the majority of the research has followed clinical populations of children who display these behaviours so the knowledge of the frequency and intensity in community samples is limited. As such, future research may examine how temperament, parenting and parenting stress affect children who do not present with a diagnosable behavioural disorder but still exhibit behavioural difficulties.
Third, the old adage 'the whole is greater than the sum of its parts' explains the additive nature of childhood development. Each variable discussed in the present paper has a significant impact on behavioural difficulties but the variables should not be only considered in isolation. The three factors form part of a cyclic process that is inexorably linked. The relationship between child temperament, parenting and parenting stress is not linear therefore future research must investigate the compounding effect on child behavioural difficulties.
In conclusion, the research on child behavioural difficulties is extensive and has provided a robust groundwork for understanding the predictors of internalising and externalising behaviours. However, there are significant research gaps, particularly in the consolidation of factors that contribute to child behavioural difficulties.