Bell 1968 (ibid.) proposed a different way of investigating disturbed behaviour. He re-examined data from a classic study carried out by Sears et al 1957 and argued for a reversal of their interpretation, that social environments were causal, by proposing a different direction of effect in which the child's temperament determined its aggressive behaviours and parent’s subsequent disciplinary responses were an attempt to alter the child's disturbing behaviour. In this he was arguing that children are active in the process and therefore a transaction between child and environments had occurred. His ideas were later taken up by Sameroff and Chandler. Sameroff 1991(in George et al 2005) proposes a transactional model which emphasised two-way interaction between active-child and active-environments across time. He argues for continuous, dynamic systems of change where neither child nor environment is solely responsible for resultant behaviours. The model does not seek simple cause/effect explanations. It looks at associations (co-variances) between multiple risk factors, protective factors and behaviours acknowledging nature and nurture as being complicit in the origins of disturbed behaviours in young children. It sees behavioural outcomes as resulting from multiple, complex transactions which may encompass several or all of the factor that linear models each individually argue for as causes.
Social learning theorists cite imitation of others negative behaviours as causal for children's disturbing behaviour. Bandura’s 1965 (in Ding et al 2005) “Bobo doll” study assessed the effects of children experiencing adult behaviours. Children were shown videos of adults exhibiting aggressive or non-aggressive behaviours towards the doll and were then themselves exposed to the doll. Children's subsequent behaviours towards the doll were observed. The study found that the significant number of children who saw aggressive videos acted aggressively towards the doll. Further they found that when children see the adult rewarded for aggression the frequency of child’s imitation increased significantly as opposed to no reward or punishment scenarios. Bandura argued that this study showed how a child takes up behaviours by imitating what it experiences. However Bandura recognised the possible impacts of other factors such as child's initial temperament and child's attitudes. These other factors mediated the level of imitation in the child therefore a simple cause/effect explanation does not fully explain the outcome.
Individual differences in children, especially temperament, are a major area of research into the causes of disturbing behaviour. Eysenck (nativist view) proposed that an infant’s temperament is a product of genetics and fixed through life. Most temperament theorists argue for temperament stability throughout life for most people. Using data from the Dunedin multidiscipline longitudinal study Henry et al 1996 (ibid.) showed that only temperament discriminated between participants who had committed a violent crime and participants who had committed a non-violent crime thus ostensibly providing strong evidence for child's initial temperament’s major influence in later disturbed behaviour however the small sample and very specific extreme behaviours examined raises questions about claims for temperament as the sole cause for the specific outcomes. Moreover Flouri and Buchanan 2002 (ibid.) provided evidence for the involvement of IQ as a protective factor against later criminality in boys thus indicating the need to consider multiple individual difference and gender factors as influencing outcomes. Jaffee et al 2003 (ibid.) provided evidence for environments’ impact on children's antisocial behaviour. Their findings suggested that quality of father/child transactions is highly influential in children's development.
The idea of a stable temperament, across time and situations, underpins individual differences cause/effect arguments and at the same time serves to limit the possibility for change by transactions as proposed by transactional models. Buss and Plomin 1984 (in Oates et al 2005) carried out meta-analysis of four Twin studies and produced evidence for heritability of temperament (MZ twins showed correlations r > 0.5, DZ twins r ≈ 0). Plomin et al 1988 (ibid.) provided evidence for stability, via their Colorado adoption project study, in some dimensions of temperament (e.g. age 1-3 years, emotionality r > 0.6). However it is obvious, by definition of correlation coefficients, that neither of these results provides a full account for their particular focus of research in that inherent in any value of r < 1 is an acknowledgement of other factors’ involvement in the outcome. Dunn and Kendrick 1982a (ibid.) provided evidence for environments’ affect on the child's temperament by examining the effect of a new baby’s arrival on firstborn children. They found that most firstborn children's behaviours changed, predominantly towards negative, and they also suggested that the mother/firstborn interactions (post new baby) were influential in the firstborn behavioural changes (Dunn and Kendrick, 1982b ibid.). When Dunn and Kendrick's findings are considered in tandem with Johnson et al’s 2002 (in Ding et al 2005) findings, which highlighted how children's negative behaviours affect mothers’ reciprocal responsiveness and this in turn increased the children's negative behaviours, a positive argument is generated for using transactional models to understand these complex interrelated issues.
Lerner and Galambos 1985 (in Oates et al 2005) found that children's difficult behaviours influenced mother’s work outcomes and was influential in the mother’s mental states supporting transactional model’s ideas for children's active involvement in its psychological outcomes. Gallagher 2002 (in Ding et al 2005) argues that the weight of evidence indicates a bidirectional relationship between parenting and children's temperament as opposed to a linear cause/effect direction. Kochanska 1995, 1997 (ibid.) proposes that the progression from difficult child to later disturbed behaviours is intricately linked with complex maternal/child interactions where different maternal environmental adaptations to children's temperaments are crucial to the outcomes. Although Kochanska’s study focused on maternal interactions it follows that any significant individual in the child's social environment can have influence on behavioural outcomes.
Attachment theorists believe that the child's later behaviours are strongly influenced by early relationships to primary care givers. They argue that poor early relationships can produce negative outcomes in later life via maladjusted internal working models (e.g. Murray and Stein 1991; Garver 1997; ibid.). Murray 1992 (ibid.) linked mother’s post natal depression to children being assessed using the strange situation method as insecure. His findings are supported by subsequent studies which also link attachment problems and disturbed behaviours in children (e.g. Greenberg et al 1993 ibid.). However Cox et al 1987 (ibid.) showed that relationship difficulties were not guaranteed by a mother’s psychopathology alone. In fact Zahn-Waxler et al 1992 (ibid.) found that some depressed mothers were able to maintain positive responsive caring styles towards their offspring. These findings indicate the mediation of outcomes by other environmental factors. It has been suggested (e.g. Crockenburg 1981 ibid.) that positive external social support (e.g. health visitor, family help, parenting skills education) for depressed mothers has positive influences on outcomes. Crockenburg also highlighted that irritable infants were more likely to show insecure attachment behaviours as a result of evoking less responsive maternal care. Crockenburg’s study indicated that a range of personal and environmental influences could be associated with outcomes. Tiet et al 2001 (ibid.) suggested a high IQ in children protects against maternal psychopathology. Ritalin’s effects, on ADHD diagnosed children, studies by Johnson et al 2000 (ibid.) also looked at parent’s attitudes towards and perceptions of their children's behaviours. They found that Ritalin reduced disturbed behaviours in children and in addition, knowing that the child was medicated, the parents’ perceptions of child's behaviours changed. Positive changes in the child's behaviours evoked more positive responses from the parents. Significantly parents also reacted less negatively to children's negative behaviours viewing these as a less attributable to the child. These findings clearly indicate dynamic transactions taking place.
The weight of evidence presented so far points to the need to consider multiple factors when looking at the origins and development of disturbed behaviours in children. Arguments for single causes do not fully address the complex multiple pathways to disturbed behaviour. Many co-variables (e.g. child’s or parents’ temperament, parental behaviours, attitudes, mental states, social environments and cultural pressures) impinge on behavioural outcomes. Acknowledgement of the dynamic reciprocal activities of child and environments promotes transactional models as best suited to understanding the complex interactions involved. It can be argued that transactional models offer a meta-theory which allows consideration of all variables from an integrative viewpoint. Transactional models can synthesise the causal explanations of other approaches into a network of transactions which act upon each other to produce outcomes thus illustrating how a change in peripheral transactions can have ramifications in the primary transaction being considered (see diagram 1). It is important to note that no transaction is fixed in time rather it involves a continuous process across time.
Diag.1
By allowing for these changes transactional models can account for the influence of protective factors in mediating outcomes. Transactional models can also be used to highlight appropriate intervention strategies by illuminating the whole picture as opposed to a single facet of it. Interventions can be designed to redress the multiple negative variables using positive variables in support (e.g. as practised by Dr Tanya Byron 2006).
In summary; traditional research into understanding the origins and development of disturbed behaviour in young children is conducted at a linear cause/effect level of analysis. Each approach offers good evidence in support of their particular claim for the “cause” however all of them, to differing degrees, recognise other factors as being involved. Traditional approaches focus on the passive child’s behaviours being determined solely by either Nature or Nurture. This reductionist discontinuous view of development does not allow for dynamic change across time. The weight of evidence indicates multiple complex interactions across time are involved in the development of disturbed behaviour in children. Transactional models recognised multiple pathways, active child-active environment interactions and continuous processes operating in a bidirectional level of effect. Therefore transactional models provide a more useful and integrative way of understanding the full spectrum of factors involved in the origins and development of disturbed behaviour in young children.
2175 words.
References
Byron, T. (2006) The House of Tiny Tearaways. London, BBC3 British Broadcasting Company. Viewed 14th-18th May 2006.
Ding, S. & Littleton, K. (eds) (2005) Children’s Personal and Social Development. Oxford, Blackwell Publishing Ltd.
George, R., Oates, J. & Wood, C. (2006), Methods and Skills Handbook, ED209, Milton Keynes, The Open University.
Oates, J. Wood, C. & Grayson, A. (2005) Psychological Development and Early Childhood. Oxford, Blackwell Publishing Ltd.
