Evaluation of a Training Programme Designed to Assist Foster Parents in the Management of Disruptive Behaviour Funder: Centre for Evidence Based Social Services, ExeterGeraldine Macdonald and Ioannis Kakavelakis, School for Policy Studies, University of BristolTel +44 (0) 117 9546755E: April 2000-September 2002EXECUTIVE SUMMARY
1. Introduction (Chapter1)
1.1 Children looked after have consistently fared less well than others on a range of indicators, including health, education, and social adjustment. One of the factors that increase children’s risks of adverse outcomes is lack of placement stability. Placement breakdowns and frequent changes of carers can undermine children’s capacity for developing meaningful attachments, disrupt friendships, and contribute to discontinuities in education and health.
1.2 Children looked after have long been recognised as experiencing high rates of behavioural and emotional problems. Among the many reasons why behaviour problems are important is that they are strongly correlated with placement breakdown.
1.3 Previous research has also highlighted the need for foster carers to be better equipped to manage the range of demands placed on them by children with emotional and behavioural problems.
1.4 This study was designed to test whether training foster carers in methods designed to help them manage challenging behaviour would have benefits for looked-after children and foster carers. Whilst primarily concerned to test whether it enabled carers better to manage difficult behaviour, we were also interested in whether it would enhance carers’ confidence in their capacity to care for challenging children and young people, and, ultimately, whether improved skills and/or confidence would enhance placement stability.
2. Managing Challenging Behaviour (Chapter 2)
2.1 Cognitive-behavioural training was chosen on the basis of its track record of effectiveness in dealing with a wide range of emotional and behavioural programmes; its effectiveness in training birth parents (i.e. ‘parent training’), and evidence from some previous studies of training foster carers.
2.2 Typically, a cognitive-behavioural approach aims to help participants to develop i) an understanding of the ways in which behaviour can be shaped by the environment, ii) skills in analysing behaviour – commonly referred to as the ABC of behaviour[1], iii) selecting strategies to change the relationship between behaviour and its consequences and/or changing the antecedents (e.g. asking a child to get ready for bed after making it clear what his/her bedtime is and turning the television off before asking), and iv) monitoring and reviewing progress.
2.3 The study had six hypotheses. These were as follows:
(i) Participants in the training programme condition would score significantly higher on the Knowledge of Behavioural Principles as Applied to Children (KBPAC) scale than foster carers in the wait-list control condition.
(ii) Participants in the training programme would be significantly more likely than foster carers in the wait-list control group to use behavioural techniques in managing children’s behaviour.
(iii) For children in the experimental group[2] there would be fewer unplanned terminations of placement in which behaviour problems were implicated.
(iv) Participants in the training programme would report a significant reduction in the range of problems that they found particularly difficult or challenging, compared with foster carers in the wait-list control.
(v) Foster carers in the training programme would report success in dealing with behaviour problems.
(vi) Foster carers would feel more confident in their abilities to manage difficult behaviour.
3. The Training Programme (Chapter 3)
3.1 The content of the training programme developed for this study mirrored that of programmes that have proved effective when provided to groups of parents facing difficulties with their children.
3.2 The training sought to familiarise carers with an understanding of social learning theory, both in terms of how behaviour develops and how it can be influenced using interventions derived from social learning theory. There was an emphasis throughout on developing the skills to observe, describe and analyse behaviour in behavioural terms – the so-called ‘ABC’ analysis.
3.3 The programme gave due consideration to the fostering context, and to the particular constraints and pressures under which carers.
3.4 The programme was designed to promote a sense of confidence or ‘self-efficacy’. This has been shown to be correlated with effectiveness.
3.5 The programme was originally designed as five weekly, three-hour sessions plus a follow-up session and this format was used for the first two groups. Because these groups felt rather pressured, we decided to move to four weekly, five-hour sessions plus a follow up for the remaining four groups. The content of the programme remained unchanged.
3.6 Each participant foster carer was provided with a Course Handbook. This contained details of each of the sessions and information relevant to each. It was written and presented in what we hoped would be an easily accessible format. Participants were asked to read certain sections between sessions as part of their ‘homework’, and had an opportunity to discuss the material at the beginning of each session.
3.7 The training was organised to reflect a collaborative approach to problem-solving, such as that described by Webster-Stratton (1998). The training was designed to marry the expertise and experience of foster carers with the specialist knowledge of the trainers and their experience of tackling child behaviour problems.
3.8 Two trainers ran the training programme: The first (MH) was a clinical psychologist who has had a great deal of experience in working with troubled and troublesome children and young people. The other (GM) was a professor of social work who has a special interest in behaviour problems and looked after children.
3.9 One of the trainers was the grant holder. To minimise bias, all data were collected and analysed independently by a research recruited specifically to the project.
4. Methodology (Chapter 4)
4.1 The most secure ways of answering questions about the effectiveness of a particular intervention is by means of an experimental research design. The most common design is a randomised controlled experiment, in which research participants are randomly allocated either to receive the intervention or to a group that does not. The latter is referred to as a control group.
4.2 This was the design used in this study. Carers who agreed to participate in the study were randomly assigned to either the training group (the experimental group) or a ‘wait-list’ control. Those in the control group continued to receive standard services and were told that should the training prove helpful, it would be made available to them in the future. The Directors of the South West Consortium gave this undertaking.
4.3 The participants in this study were foster carers in five local authorities in the South West.
4.4 Although we initially received statements of interest from a large number of carers, this number withered for a variety of reasons. In addition to those who did not meet our inclusion criteria, some withdrew from the study once the dates and/or location of the training were confirmed. Others withdrew once they realised they had been allocated to the control group.
4.5 Final numbers of willing participants allowed us to randomly allocate them to the two conditions, but we had to do so on a 1 in 2 basis (rather than the 1 in 4 that we had intended) and we had to do so within geographical regions. This effectively means that we ran three small trials, rather than one. This resulted in a reduction of statistical power, and may be a contributory factor to the failure of this study to suggest positive outcomes for the experimental (training) group.
4.6 The outcome measures used were:
(i) A measure of participants’ knowledge of behavioural principles as applied to children (O’Dell et al. 1979).
(ii) The number of unplanned breakdown of placements. These data were obtained from interview data, which covered the month prior the onset of training, the five weeks of training, and the six months after training.
(iii) The Child Behavioural Checklist (Achenbach 1993)
(iv) A Foster Carer Satisfaction Questionnaire.
4.7 Participants were interviewed before and after training, and at six months follow up. Each interview schedule was developed specifically for the purpose of the study. All interviews were tape-recorded after consent was obtained.
5. Results (Chapter 5)
5.1 A preliminary analysis indicated no significant effects and/or interaction of the Local Authority variable with any other variables. Since no significant effect of this variable was observed, this factor was excluded from all the following analyses.
5.2 117 participants comprised the final sample, of whom all but two were white. Of these, 67 foster carers (female foster carers=52, male foster carers=6, couple foster cares=9) participated in the training group and 50 in the control group (female foster carers=37, male foster carers=6, couple foster carers=7). The ages of these participants ranged from 32 to 65 years, with a mean age of 45 years. Experience of fostering (in years) ranged from one to fifty years, with a mode of one year (i.e. carers with one year’s experience occurred more frequently than carers with other profiles). The overall mean was 8.68 years.
5.3 Two of the six hypotheses were clearly confirmed. These were hypotheses (1) that participants in the training programme condition will score significantly higher on the Knowledge of Behavioural Principles as Applied to Children Scale (KBPAC), and (6) enhanced confidence in their abilities to manage difficult behaviour. The qualitative data collected provides ample evidence to support this hypothesis. Carers generally said that the course had made them more confident to deal with difficult situations and difficult behaviour.
5.4 The hypotheses concerning the use of behavioural techniques (Hypothesis 2) received support in relation to the use of ABC analysis, tokens and ‘grounding’. With regard to ABC analysis, those who had participated in training were significantly more likely to be using this at all time points measures (i.e. post-test and follow-up) compared with those in the control group. They were also more likely to use tokens and less likely to be using ‘grounding’ after training. At six months follow-up, this last difference (i.e. use of tokens) disappeared, and a rather unexpected result occurred, namely that those in the control group were more likely to report themselves as using strategies which are best categorised as ‘response.
5.5 The hypothesis that children in the experimental group would have fewer unplanned terminations of placements in which behaviour problems were implicated was not proven. If anything, carers in the training group showed a slight increase in the number of unplanned terminations of placement from post-training to follow up.
5.6 Generally, carers in the training group did not use behavioural terminology when describing the strategies they used to manage behaviour, despite their increased knowledge of behavioural principles.
5.7 Participants in the training programme did not report a significant reduction in the range of problems that they found particularly difficult or challenging, compared with foster carers in the wait-list control (Hypothesis 4). There were no differences between the two groups at any time point.
5.8 Foster carers in the training programme did not report more success than those in the control group in dealing with behaviour problems (Hypothesis 5). However, particular caution is required when interpreting these data as we secured so few completed Child Behaviour Checklists, on which these data primarily depend. Data from the Course Task and from qualitative data provided during the interviews, suggest that carers in the training programme are both using these techniques and having success where they themselves would otherwise not have expected success.
6. Foster Carers’ Use of Behavioural Methods (Chapter 6)
6.1 Evidence from the records kept by foster carers indicates a secure basis in behavioural analysis, but weakness in the use of cognitive-behavioural interventions.
6.2 The conclusions drawn from these data and the trainers’ experience of running the training programme, are (i) that the programme needs to ensure that participants understand the principles underpinning behavioural interventions, (ii) that they have a firm understanding of how to use them appropriately and effectively, and (iii) that the programme provides more opportunities to practice their implementation, both in the training situation and at home.
6.3 These aspects were essential components of the programme, but the size of the group, poor attendance, and the challenges facing carers meant that this aspect of training was often ‘squeezed’. Carers were also not reliable in doing the homework assignments, a key component of the course.
7. Foster Carer Satisfaction (Chapter 7)
7.1 With the exception of one or two participants, satisfaction with the programme was generally very high. All but seven of the respondents said they were either satisfied (2) or very satisfied (16) with the training they had received.
7.2 Most respondents reported improvements in the behaviour problems that had prompted them to participate in training.
7.3 The majority of carers thought that their child’s behaviour had improved in general since participating in the course, and most were satisfied with the degree of progress. All but nine foster carers saw the training programme as having helped bring about these improvements.
7.4 Over 82% indicated that they felt very positively about the future, and all but five said they would ‘recommend’ (n=20) or ‘strongly recommend’ (n=23) the programme to other carers.
8. Discussion (Chapter 8)
8.1 We should be rightly suspicious of attempts explain away results, but it is also negligent to disregard lessons learned and plausible explanations for why a particular pattern of results have fallen as they have.
8.2 Certain lessons learned in the course of the training, and in the conduct of the study, urge caution in concluding that training in cognitive-behavioural methods is not an effective means of helping carers to manage difficult behaviour or to prevent unplanned terminations of placement. There is much here to support the view that training foster carers in cognitive behavioural skills is an appropriate undertaking, but that it may require more than a short course with some of the problems that this programme encountered.
8.3 Essentially, the current study points to the need to make certain changes in the training provided in order to maximise (or perhaps to ensure) its impact. These are essentially to provide a longer training programme, to limit the size of the groups, and to find a way of improving / securing full attendance and completing homework. This should enable an appropriate emphasis to be maintained on developing the skills of cognitive-behavioural interventions, as well as cognitive-behavioural analysis. Some carers might do better on a one-to-one basis, at least initially, as some appeared to struggle in the group context.