It's a Piece of Cake? An evaluation of an adopter training programme.

Selwyn, Julie ; del Tufo, Sarah ; Frazer, Lesley 等

Introduction

Adoption UK, founded in 1971, is the only national charity run by and for adoptive parents, offering support before, during and after adoption. It is a registered adoption support agency and has offices in England, Scotland and Northern Ireland. In 2007, the organisation had a membership of about 5,000 adopters and prospective adopters and 204 local authorities.

Since 2000, Adoption UK has been running a training programme called It's a Piece of Cake? (shortened here to Cake). Although evaluations completed by adopters and feedback from trainers have been consistently positive, in 2005 Adoption UK decided to seek funding for a comprehensive, independent evaluation of the programme. With the help of a small grant from Lloyds TSB Foundation, Adoption UK commissioned the Hadley Centre and the Evaluation Trust to undertake this work, which involved considerable pro bono time from the two organisations. (1)

Related research

It has been established for some time that a significant minority of adopted children continue to have difficulties in their adoptive homes that threaten the stability of the placement. However, far less is known about effective methods of intervention to reduce these. Although parenting skills programmes have been found to be effective in the general population (Webster-Stratton and Hancock, 1998; Scott et al, 2001; Turner and Saunders, 2006), there has been little work to explore how effective these methods are for adoptive parents. As Rushton et al (2006) points out, the most striking difference between general parenting programmes and those designed for adopters is that adopter programmes are intended to enhance parenting skills, not to reverse poor parenting. Most research on post-adoption services has concentrated on services delivered to individual adoptive families (Barth and Miller, 2000; Rushton and Dance, 2002), rather than group-based interventions such as Cake.

There have been few systematic evaluations of any adopter or foster carer support programmes (Minnis et al, 2001; Henderson and Sargent, 2005; Herbert and Wookey, 2007; Anna Freud Centre, 2008) and only a few descriptive studies of packages provided by individual agencies (eg Swaine and Gilson, 1998). While adopters are generally very positive about such programmes, it is unclear whether they change parenting behaviours or increase skills for longer than the duration of the programme. Training has been developed based on the Webster-Stratton parenting skills programme, using a cognitive-behavioural approach (Pallett et al, 2002; Gilkes and Klimes 2003) or based on attachment theory (Schofield and Beek, 2006). In progress is Rushton and colleagues' randomised controlled trial comparing different types of intervention with adoptive families, and this will undoubtedly provide more information about effectiveness.

Attempts to evaluate foster care training (Pithouse et al, 2002; McDonald and Turner, 2005) have shown that while carers enjoyed the training and spoke highly of it, there was no evidence of sustained changes in the behaviour of the carer or any alterations in the children's behaviour. This may be because the follow-up period (often less than a year) was insufficient for changes to take effect and/or the measures were not sensitive enough to identify small developments. It is also increasingly recognised that changing behaviour is difficult, and that many children's behaviour patterns are entrenched. Consequently, in this evaluation we hypothesised that the effects of training might be seen in reduced adopter stress levels and enhanced self-esteem. In turn, these changes might have effects on parenting by reducing tension in the home and breaking the cycle of blame and guilt that arises in situations of conflict.

The Cake programme

The programme was originally developed and written by adopters drawing on the work of the adoption community and academics in the UK and US. Unusually, it is delivered by professional trainers who are themselves adopters. Ideally, parents attend the training 12 or more months after a child has been placed with them. These two elements--trainers who are adopters and delivery post placement--make it different from other adopter and parenting programmes.

Fifteen programmes are delivered per year and are purchased by local authorities or adoption agencies. Since Cake started, approximately 1,000 people have taken part. The programme consists of six modules, each lasting at least five hours, and aims to:

* enhance adopters' parenting skills and approaches;

* affirm adopters' own parenting methods and enable them to recognise that they are the key resource for their own children;

* increase understanding of attachment issues created by early trauma, separation, loss, neglect and abuse;

* increase adopters' confidence in their ability to parent damaged children;

* enable adopters to understand the importance of parental self-care.

The evaluation

The evaluation involved two separate but linked pieces of work:

* a prospective evaluation of three current Cake programmes undertaken by the Hadley Centre;

* a retrospective evaluation of past Cake programmes undertaken as a collaboration between the Evaluation Trust and Adoption UK.

The prospective evaluation

This was designed to determine whether the current programme was successful in delivering its training objectives. Within this broad aim the evaluation also set out to:

* examine whether adopters attending the programme reported improvements in their parenting skills;

* test whether any improvements could be attributed to the programme;

* understand which parts of the programme had the most impact;

* make recommendations for the improvement of the programme.

The design and sample of the prospective evaluation

Before training began, local authorities who had purchased it were asked if their adopters could be approached to request consent to participate in the evaluation. Three local authorities refused permission without giving any explanation and therefore these programmes were excluded from the evaluation.

Where agency consent was gained, adopters registered on the programme in 2005-06 were invited to take part. In the north of England six couples and five single carers had registered (total 11 families, 17 individuals). In the south there were seven couples and two single carers (total 9 families, 16 individuals). However, two couples and one single carer failed to attend any of the sessions and no reasons were given to Adoption UK by the purchasing authority. There was further drop-out from those who started the programme, with two couples and one single carer withdrawing after the first module, mainly due to family illness. Therefore, the sample comprised three couples and three single adopters from the north and six couples and two single adopters from the south--a total of 14 families. These families were caring for 26 adopted children. We were told that it was very unusual for Adoption UK to experience this rate of non-attendance. As a consequence of decreasing participant numbers, it was decided to ask a further programme in the east of England to take part. Two out of four couples who were attending the programme agreed and they were caring for three children.

A control group was sought by advertising on the Adoption UK website for help with the evaluation. The criteria for inclusion in the study were: a) that they had been an adopter for at least 12 months; and b) that they were interested in attending Cake but had not done so.

The final sample therefore comprised 16 families attending the programme (the experimental group) caring for 29 children and 19 'control group' families caring for 32 children.

Questionnaire content

Questionnaires were sent to all the adopters before the programme began, immediately after the end of training and five months after completion. The questionnaires were filled in by the adopter with the most daily contact with the child(ren). Table 1 (below) outlines the return rates.

One placement in the experimental group had disrupted before training had been completed and this questionnaire was not returned at the end. The five-month questionnaire was excluded from the evaluation because of the low number of returns.

Kraiger and colleagues' model (1993) for evaluating learning outcomes was utilised. In this model, learning outcomes are considered in three areas: cognitive, skill-based and affective outcomes. Cognitive outcomes were measured, for example, through questions about previous training and understanding of child development. Skill-based outcomes were considered through questions asking about playing with the child and strategies and skills necessary for dealing with difficult behaviour. Affective outcomes used questions asking for ratings of motivation in attending the programme, how integrated children were in the family, attitudes to contact and self-rating of confidence.

Standardised measures were also used: the Strengths and Difficulties Questionnaire (SDQ) (Goodman, 1997); the General Health Questionnaire (GHQ); and the Expression of Feelings in Relationships Questionnaire (EFR). The latter measures the ways in which children relate to parents/carers and how they show their feelings in these relationships. These measures had two purposes: first, to make sure that there were no major differences between the experimental and control groups and second, to see if there were changes in behaviours over time.

Mindful of the need to consider small changes in children's behaviour, adopters were also asked to describe two child behaviours that were currently concerning them, to explain what they thought the reasons were for these behaviours, the strategies they currently used to manage them and what happened if the strategy failed. At follow-up, the same questions were asked about the identified behaviours. Adopters' ideas and beliefs about the causes of behaviours and parental strategies could then be compared.

The data were analysed in SPSS using Fishers Exact for tests of association and Wilcoxon Signed-Rank test for changes over time.

Findings of the prospective evaluation

The vast majority of adopters in the experimental and control groups were white married couples; 69 per cent had two or more adopted children and a quarter of the families also contained birth children. The majority of adopted children had been with their families for more than three years (see Table 2). There were some differences between the groups, such as the experimental group adopters' children being older, but none of the differences reached statistical significance.

The adopters

The majority of adopters responded positively to questions about their current state of mental health. They described themselves as happy but a quarter of the experimental group was less happy than usual, most attributing their unhappiness to the present circumstances of adoption. Although 21 per cent of the control group also expressed current unhappiness, the majority attributed this to other sources of stress, such as work relationships.

Those attending the programme hoped that they would gain more knowledge about the reasons for their child's behaviour and develop improved strategies for managing it. Self-reports of motivation to attend training were high but an examination of attendance records revealed a lot of absence. Single carers in particular had great difficulty attending every session. For example, two out of three single adopters in the north only attended half the modules. Absences seemed to be due to childcare problems or sickness. Although we were not able to test this because of low numbers, we assumed that absences would have some impact on the ability of the training to deliver its objectives.

Satisfaction with the training

All those who had attended the programme reported that they were satisfied or very satisfied with the training:

Overall we found the programme to be very rewarding; it provided new insights into problems we face daily. It offered the opportunities to discuss and review similar experiences with others in the same position and to gain from their experience.

Adopters rated the group experience positively:

It was good to meet other adopters and to know that we are not alone with our problems.

The programme used a variety of teaching methods and although role play and practising skills at home were given low enjoyment ratings, they were rated as highly useful. Small group discussions were the most popular. Particular exercises had had strong impacts and had made the theoretical material more real and vivid.

Adopters thought they had learned more about child development, attachment theory and specific techniques to manage behaviour:

We used the technique 'When you've done that then you can ...' This has been extremely valuable ... makes getting things done much easier and we didn't realise how such a simple thing could be so effective.

Adopters could not identify a specific part of the programme as the most worthwhile but thought it was the whole package that had been helpful. Many commented on realising the importance of 'self-care' and 'time for oneself'. For example, when asked to identify the most important area of learning, one parent wrote: 'the importance of self-care, issues of loss and separation, understanding yourself and your emotions and reactions'. Some were surprised by what they had discovered: 'it never crossed my mind that my child would need to grieve for his past'.

When asked to be specific about alterations in family life, some reported that the biggest changes were adoptive fathers becoming more involved with the daily parenting of the child and male partners developing more understanding and sensitivity, which in turn improved marital communication. The most frequent response was that family tension and stress had reduced:

I have a greater understanding of the reasons behind my child's behaviour. This understanding means I am generally more tolerant. It means I get less stressed about things. A calmer approach often stops situations escalating.

Understanding children's behaviour

At the time of the first questionnaire, both groups of adopters reported that young people's aggression, unpredictability and hyperactivity were the most difficult behaviours to manage. The control group were slightly more confident in managing these behaviours. Some parents emphasised that children's reactions could be sudden or extreme:

[Child] rages when nervous or unsettled. This may blow up without any apparent reason. Can lead to aggression and trashing belongings.

Calm and pleasant nature but when feeling insecure [child] can have extreme rages.

Unsurprisingly, adopters reported that these children were also having problems in school, with such comments as 'lack of concentration and focus' ... 'nonsense chatter' ... 'excessive questioning'.

Before training, the experimental group of adopters often attributed difficult behaviour to their child wanting to be in control, attention-seeking, anxiety, fear or a lack of self-control. For example, one parent initially wrote that 'the child overreacts to most situations' and thought this was because 'he cannot control himself'. After training, the explanation for the behaviour was more detailed and thoughtful:

He cannot self-regulate his reactions, eg excitement, sadness. He wants to get his feelings across in as strong a way as possible so we get the message about how he is feeling.

However, adopters in the control group also showed more understanding of the child's behaviour when completing the second questionnaire. For example, initially the child was described as having daytime enuresis attributed to 'immaturity, a lack of control but mainly attention-seeking'. Six months later the reason given was 'anxious--not always aware that he needs to go or is wet'. There were improvements in both groups, with just one parent still reporting that they could not provide an explanation for their child's behaviour.

In an ideal research environment, a control group would have no access to training or other supports in order to ensure that any difference could be attributed to the impact of the programme. Real life is not like a laboratory and a third of the control group adopters had sought help from professionals such as psychologists, social workers and teachers. A quarter also reported assistance from local adoption support groups, Adoption UK message boards or from reading and watching programmes on child development. One of the experimental group's children was seeing a therapist and it is possible that other adopters also gained support elsewhere. This made it difficult to know whether improvements were due to the programme or other supports or a combination of these.

Parenting

Adopters also reported changes in the techniques used to manage behaviour. Before training, the experimental group adopters used, in order of popularity, rewards, time out, a firm telling off and restitution for damage or harm done. A third also used withdrawal of affection. Control group adopters used rewards, extra time with the child and removal of privileges/favourite toy. After training, all of the experimental group were using rewards and more (72% before training increasing to 91% post training) were spending extra time with the child. Only one family was still using withdrawal of affection. In contrast, the same percentage of control group adopters were using rewards but now a fifth were using withdrawal of affection. Replies suggested that although many of the control group had developed a greater intellectual understanding of their child's challenging behaviour, there were no improvements in their management of it.

The experimental group also reported more enjoyment of play with their child post training and this probably reflects the improving relationships and reduced stress in the households. Improvements in play did not reach statistical significance. However, there were statistically significant differences in adopter ratings of their confidence in managing children's

difficult behaviour. While the confidence of the control group adopters had remained stable over time, the experimental group's confidence had grown. (2)

There were also some surprising but non-significant responses. Adopters were asked whether the child confided in adults. Before training, slightly fewer of the experimental group adopters reported that their children talked about their pasts than those of the control group. Open communication in this area was not related to the age of the child. However, post training the experimental group reported less disclosure about the past or present concerns. The experimental group also identified more difficulties post training when asked to select on a scale of 1-4 the position that best reflected for their child the various child development dimensions such as health, education and identity.

These changes were not seen in the SDQ scores and, indeed, it would have been suspicious if parents had reported significant changes in behaviour in such a short time period. We hypothesised that the children did not have greater difficulties or communicate less, but that the experimental group adopters were now more aware of their child's problems and what might be disclosed. It will be interesting to see if this finding is replicated in other studies. As can be seen, the scores in the SDQ and EFR remained relatively stable (Tables 3 and 4 below).

Overall, the group training seemed to have given the experimental group adopters a more coherent framework for understanding children's behaviour, a wider repertoire of responses and increased confidence. However, these were measured immediately at the end of training and we were unable to test whether improvements were maintained or faded out over time. It was hoped that the retrospective evaluation would help to show whether the programme had a longer-term impact.

The retrospective evaluation

The purpose of the retrospective evaluation was to:

* identify the longer-term impact of Cake;

* identify lessons to help Adoption UK to improve and develop the programme.

The design and sample of the retrospective evaluation

The design of the retrospective evaluation of Cake attached great importance to the participatory process of the research. It was hoped that the benefits to Adoption UK of this work would be organisational growth, staff development and improved effectiveness. Participatory research of this kind provides opportunities for empowering those taking part to put forward their own realities and influence policy. It also gives rise to richer accounts, anchored in real lives and conducted in egalitarian and democratic ways (Bennett, 2004). It is essentially a learning process, which actively involves participants in reflecting critically on their organisation and the issues to which it is responding.

An indicative sample of 20 of the 70 programmes (involving approximately 200 adopters/foster carers), which had run in Scotland, England and Wales in the period 2001-05, was selected. Northern Ireland was excluded because an external evaluation had already been completed. Two local authorities refused permission for adopters to be contacted and substitute authorities were approached. It was not possible, within the very limited resources available, to trace adopters who had completed the programmes in 2000 and therefore the potential sample comprised 190 participants. Based on the profile of the prospective sample, we estimated that around 80 per cent would have attended as couples (76 families) and 20 per cent as single adopters (38 families).

Of these estimated 114 adoptive families, 94 (82%) were successfully contacted and gave consent to a parent interview and 66 parents from 64 families (56%) actually completed interviews, most often the adoptive mother. Collectively, the interviewed parents had adopted 123 children (some of whom were now adult and had left home) and fostered eight. Although the final interview sample was large, it is nonetheless important to bear in mind that these parents were not necessarily representative of all the participants and that important divergent views may therefore have been missed.

Telephone interviews were also undertaken with commissioning staff from 14 of the 20 sampled local authorities, which had themselves purchased the Cake programme, 15 individuals involved in the history of Cake or from sister agencies running other post-adoption programmes, and 13 members of Adoption UK staff. Two focus group discussions were also held with the trainers leading the programmes.

The interviews

Interview schedules were developed jointly by Adoption UK training staff and the Evaluation Trust researchers. Both organisations jointly trained a team of 12 Adoption UK volunteers (trustees and others associated with Adoption UK) and staff members whose task was to undertake the majority (75%) of the parent interviews. After interviewing, most of the interviewers took part in a conference call discussion with the Evaluation Trust research team to reflect on the interviews, as well as identifying themes and any difficulties. The Evaluation Trust researchers also undertook the remaining parent interviews (25%), completed the stakeholder interviews (with purchasers, external professionals, Adoption UK trainers, trustees and staff) and had responsibility for data analysis and the final report.

Interview questions built on the information collected from the prospective evaluation so that results could be compared. Interviewers asked for demographic details, experience of other training, views of Cake, contact with other group members and their current needs as a family. The evaluation also drew on earlier training evaluation work and discussions about Cake from the Adoption UK online message board community.

Findings of the retrospective evaluation

The retrospective evaluation mapped the development of the Cake programme and made recommendations for improvements. Space precludes this description but interested readers can find the full account of the history and theoretical basis of the programme at www.adoptionuk.org.

The Cake experience: outcomes and impact on parents

Of the 66 adopters interviewed, most were married or living with a partner, and just over half had attended the course as a couple. They had received little training before attending Cake and had been highly motivated to take part. Reasons for attending divided into two clear groups.

Fifty-six per cent were already experiencing significant difficulties with their adopted children and were seeking advice and help:

I was struggling with [my child's] behaviour ... his aggression, and his self-harming, those were the main things. I didn't feel able to manage it. I'd asked for help after he was placed and I got absolutely nothing and there just seemed to be no answers ...

The remaining 44 per cent had not yet encountered many problems and saw the offer of training as an opportunity to acquire helpful strategies for the future.

At the time of the interview, adopters were caring for children and young people aged one to 22 years (average age 10 years). Most (91%) of the children had been living with their adoptive parents for three years or more. At the time of placement, 78 per cent of the children had been under five years old and 21 per cent had been older.

In retrospect, most (73%) of the adopters felt that Cake had met or exceeded their expectations. However, just over a quarter (27%) thought they were already very experienced adopters or that their situations were so difficult that the training was not enough:

Not totally met my needs but this is because daughter's problems are so deep rooted. We did get empathy and compassion from the programme but we are still continuing to learn about attachment. There is no easy fix.

Despite the time that had elapsed--up to five years for the early programmes--adopters still had vivid memories of the programme. The most frequent recollection was of the intense relief of discovering that they were not alone in their experiences, and the support and friendship they received from the other parents and trainers who were themselves adopters:

I think everyone on Cake felt a sense of relief to be with like-minded people. When someone spoke about their child it sounded like my child. We had so much in common. People understood how bad it was and how divisive the children could be.

Many referred to the helpfulness of the group discussions and also recalled the emotional impact of particular sessions that suddenly illuminated their children's problems in new ways; some found attachment theory and its application especially memorable; others recalled and still held onto the need to care for themselves and their partners.

Just fewer than half the adopters had not known that the trainers would also be adoptive parents before they started the programme. With hindsight, 95 per cent of the parents had no reservations and were very positive about their role as professional trainers. The sense of parents' experiences being recognised, of 'not having to explain' their children to the trainers, being understood and not denied, was very important. Even those parents who were affirming of social workers did not believe they could have had the same insights or created the same learning environment. Three parents had some reservations about the practice of trainers sharing their own experiences ('their own baggage'), but not the principle of involving adoptive parents as trainers:

I think they're the best people to do it because they've been there and done it. I was exhausted with explaining [my child] to people who haven't experienced it. They don't understand; they continually say, 'but all children do that'. It would have been too overwhelming if they hadn't been adopters. I couldn't have done it. It was such a relief to find that these people knew.

When asked about the difference that taking part in Cake had made to them and their families, parents did not describe a neat linear relationship between inputs and outcomes, but rather talked about a complex interplay between personal learning, the impact of being in the group, internal processes and family changes that in many cases had worked iteratively to support and reinforce their confidence and ability to parent.

Most parents (76%) reported a new understanding of the causes of their child's behaviour. The gaining of a conceptual framework around trauma leading to attachment difficulties was explicitly linked by adopters with the development of more realistic expectations. Being in a supportive group and acquiring these new insights went hand in hand with improved feelings of well-being and confidence, and with feeling less stressed and alone, angry and guilty. This, in turn, led to new approaches and changed ways of responding to children more calmly, with less anger and more tolerance, and new strategies to manage children's behaviour:

To begin with, my relationships with the children had been very confrontational, but afterwards I saw it less negatively and worked with it. [Cake] made me understand why they were acting like that. I was better able to talk things through with them. So we could cope better--my husband too. I think otherwise we'd have gone on blaming ourselves and feeling guilty, but after Cake it felt much better. We could see where the behaviour had come from.

For just under half the parents, the growth in understanding was strongly linked to finding strategies to build attachment bonds and manage their own reactions. Others noted that Cake had helped them become better advocates for their child, and some parents reported that the programme had helped with placement stability:

The situation with the children has stabilised a lot ... There used to be a lot of conflict between us over parenting, but not any more. The kids have come on tremendously. I don't despair of their future now, whereas when I went on Cake I was pretty despondent.

Many adopters stated that they had forgotten that it was also important to care for themselves and their partner and the programme had helped them in this regard:

We knew realistically that the issues would not go away and that, at the end of the day, self-care was so important.

Cake had also resulted for many in important long-term friendships and sources of support from other group members. Linked to this was the importance for many parents that they attended Cake as a couple and were able to share the experience. Some parents described the programme as life changing:

My life improved and the children's life improved. I had lost all hope and now, beyond my wildest dreams, I am in control.

Purchasers' perceptions of the impact of Cake on parents

Like parents, the local authority purchasers thought one of the key strengths of Cake was the mutual support from the group. The purchasers' perceptions closely mirrored the perceptions of the parents themselves. They identified: the value of sharing experiences and ideas with other adoptive parents, combating isolation and encouraging self-reflection; parents' improved feelings of well-being and confidence; the benefits of adopters having improved understanding of their child's behaviour; and an increase in adopters' knowledge and skills/conceptual framework and therefore being able to move away from blaming themselves. Cake was also felt to have provided adopters with tangible strategies to manage children's behaviour. Some purchasers noted key effects on placement stability as well as supporting families during disruption:

Cake holds adoptive parents. Families who have the most benefit are the ones who are struggling. They said they wished they had had the training earlier in their placements. However, they may not have benefited as they may have thought, 'I don't have problems like others,' so the door would have been half shut and they may not have wanted to hear Cake's message ... Cake is a contribution to preventing the breakdown of placements.

Purchasers identified a variety of additional benefits, such as generating interest in other forms of support, making it acceptable to attend other seminars and events, that it indicated the authority's commitment to parents and that it opened up Adoption UK's resources to parents. Most purchasers thought that some programme participants were still in contact with one another two to three years after the programme had ended and valued this.

At a more strategic level, the purchasers saw Cake as providing comprehensive information about parenting children that the authorities could not readily offer. In this context, they also reflected on the value of the final 'feedback' session of the Cake programme, which is attended by social work staff. Most considered that this provided an important opportunity for social work staff to hear and respond to the continual support needs of adopters:

I went to the feedback session during the last module and the parents spent more time talking about the services they needed. It was very useful at the time, as we were putting contracts out to tender. We brought services in house as no one could meet the needs.

Views of other professionals

A key academic and other well-known professionals in the adoption field who had detailed knowledge of Cake identified the benefits of the programme for both parents and the purchasing agencies--for example, in enabling local authorities to fulfil their obligations to support adopters under the Adoption Act without using costly staff time.

In terms of benefits for Adoption UK itself, Cake was widely viewed as having significantly increased the organisation's profile, professionalism and credibility as an agency pioneering innovative support services for adopters. The programme was also thought to have helped Adoption UK to generate income through grants and fees, and to promote membership. Trainers themselves identified a range of personal gains from their involvement in Cake, in terms of their own parenting, their enlarged sense of community and belonging, enhanced theoretical understandings and insights, and their own confidence and access to new opportunities.

The timing and delivery of Cake: barriers to participation

Many adopters valued the variety of experiences other parents brought to the groups and discovering that others had 'got through' the different ages and stages. However, some wanted programmes to include parents at similar life stages and a few found the experiences recounted by longer-term parents negative.

Some purchasers struggled to recruit parents to the Cake programme and others were oversubscribed. Different responses were not always due to the size of the authority. The role of an active social worker in referring parents to the programme was an important factor, with 60 per cent of parents mentioning the invitation coming from their worker. Recruiting fathers to attend was a particular issue, in part due to childcare difficulties. The drop-out rate for the programmes was also variable. When the number recruited by the purchasers was already low, such dropouts began to affect the quality of the experience for the remaining parents. It was apparent, both from the interviews and attendance sheets, that having a male adopter involved in the delivery of training had a clear effect on the continued attendance of fathers.

Most, but not all, participants were white middle-class adopters. This is an issue outside the control of Adoption UK, given that the selection of participants is managed by local authorities. Black, Asian and working-class parents have attended the programme but they remain very much in the minority, even in areas of the country which have high minority ethnic populations.

Parents' perspectives on their current support needs

Almost three-quarters of the adopters spoke throughout the interviews about their current pressing needs for support and resources. Although some felt that their continuing needs were already being met, 38 per cent of parents wanted a Cake refresher support programme to remind them of previous learning and add to it. A further 15 per cent wanted a more advanced programme focusing on adolescents and their needs, contact issues and working with education. To some extent, this request has been responded to by the development by Adoption UK of one- and two-day workshops on these topics. However, the interviews reflected a low level of awareness of these modules among parents and purchasers.

Many parents wanted Adoption UK to take a strong advocacy role to ensure adoptive parents received appropriate support--a view echoed by some staff and trustees. The levels of support service offered depended upon where adopters lived. Some authorities appeared to be offering excellent post-adoption support and access to support groups, but many parents reported negative experiences of trying to access funding and support. Parents described unanswered or blocked requests for information and help, high staff turnover, a lack of practical social work and therapeutic support, and no access to respite care, support groups or ongoing training after adoption. Other parents identified negative attitudes or poor practice in their social services department that made contact and trust difficult.

Conclusions

Adoption UK is a value-based, parent-led organisation and Cake grew out of shared concerns and a commitment to respond to the needs of adoptive families. The programme is popular with a range of stakeholders. The full evaluation report identified several areas of learning and Adoption UK has begun the process of acting on the report's recommendations.

Evidencing the effectiveness of interventions is challenging and this research shared many of the problems of previous attempts to evaluate training. In the prospective evaluation, although we were able to demonstrate that the children in both experimental and control groups were similar in many ways, the children may have had unknown characteristics that influenced the way they responded to their parents. For example, we did not know whether the children in the experimental group had easier temperaments or were more resilient. It was also not possible to ensure that the control group received no interventions and the experimental group received only the support and help provided by the programme.

Reaching and interviewing a retrospective sample is also difficult, and those parents whom we interviewed, although large in number, may not have been representative of all who attended a Cake programme. Although the interviewed adopters were generally positive about their experiences of Cake and thought it had contributed to happier and more stable family lives, it is of course not possible to know what would have happened to these families without the intervention.

Bearing in mind these caveats, when we look at all the evidence from the combined evaluation results we can state that:

* Attending Cake was not harmful and was better than having no intervention at all.

* The immediate effect of the programme was an increase in adopter confidence; this seemed to be because adopters were provided with a way of understanding their child's behaviour that did not attribute blame to either the child or the adult. Adopters also developed a wider repertoire of skills to manage difficult behaviour.

* There was some evidence from adopters that the programme increased the involvement of adoptive fathers in the lives of their children.

* Most of the interviewed adopters looked back on Cake over long periods with strong and unusually vivid appreciation, and felt that participating in the programme had benefited them and their families' lives.

* The group setting for the learning, enabling parents to discuss difficult issues with other adopters, was viewed as very important. Many parents had made lasting friendships that were a continuing source of support.

* Adopters appreciated having adoptive parents as trainers.

* The effects of the programme seemed to have continued for some time for adopters who were not parenting children with multiple and overlapping difficulties. For these parents Cake was rarely enough.

The vast majority of adopters receive no post-adoption support and indeed may say they do not need any. However, we know that many adopted children (especially those placed at older ages) have scores on the borderline or in the abnormal range of the SDQ, and that the impact of early abuse and neglect continues to play out into their teenage and adult years. We are beginning to understand 'what works' in helping adoptive families but as yet we do not know enough to be able to tailor interventions to address specific needs. We do not know which adopters respond better to group-based training rather than individualised packages or whether the particular characteristics of adopters and their children influence the effectiveness of the particular methods chosen. At present, many interventions are offered without any real evidence to show whether they offer effective help or are harmful.

Access to provision continues to be patchy and is often dependent on where adopters live. Even where programmes such as Cake are available, they cannot provide the whole answer. Within Adoption UK, the programme is now seen as providing a foundation from which adopters should be able to choose from a wider menu of other services intended to meet

changing needs across the life course. The availability of such menus of support, particularly once they have been independently evaluated to assess their effectiveness, will enable adopters to make much more informed choices and to access the right help at the right time.

References

Anna Freud Centre, The Development and Evaluation of a Parenting Skills Programme for Adopters, 2008; available online at www.annafreudcentre.org/adoptive_parents_res earch.htm (last accessed 16 January 2009)

Barth R and Miller J, 'Building effective post-adoption services: what is the empirical foundation?', Family Relations 49:4, pp 447-55, 2000

Bennett F and Roberts M, From Input to Influence: Participatory approaches to research and inquiry into poverty, York: Joseph Rowntree Foundation, 2004

Gilkes L and Klimes I, 'Parenting skills for adoptive parents', Adoption & Fostering 27:1, pp 19-25, 2003

Goodman R, 'Strengths and Difficulties Questionnaire: a research note', Journal of Child Psychology and Psychiatry 38, pp 791-801, 1997

Henderson K and Sargent N, 'Developing the Incredible Years Webster-Stratton parenting skills training programme for use with adoptive families', Adoption & Fostering 29:4, pp 23-44, 2005

Herbert M and Wookey J, 'The Child Wise programme: a course to enhance the self-confidence and behaviour management skills of foster carers with challenging children', Adoption & Fostering 31:4, pp 27-37, 2007

Kraiger K, Ford K and Salas E, 'Application of cognitive, skill-based and affective theories of learning outcomes to new methods of training evaluation', Journal of Applied Psychology 78:2, pp 311-28, 1993

McDonald G and Turner W, 'An experiment in helping foster carers manage challenging behaviour', British Journal of Social Work 35:8, pp 1265-82, 2005

Minnis H, Pelosi AJ, Knapp M and Dunn J, 'Mental health and foster care training', Archives of Disease in Childhood 84:4, pp 302-06, 2001

Pallett C, Scott S, Blackeby K, Yule W and Weissman R, 'Fostering Changes: a cognitive-behavioural approach to help foster carers manage children', Adoption & Fostering 26:1, pp 39-48, 2002

Pithouse A, Hill-Tout J and Lowe K, 'Training foster carers in challenging behaviour: a case study in disappointment?', Child & Family Social Work 7:3, pp 203-14, 2002

Rushton A and Dance C, Adoption Support Services for Families in Difficulty, London: BAAF, 2002

Rushton A, Monck E, Upright H and Davidson M, 'Enhancing adoptive parenting: devising promising interventions', Child and Adolescent Mental Health 11:1, pp 25-31, 2006

Schofield G and Beek M, Attachment for Foster Care and Adoption, London: BAAF, 2006

Scott S, Spender Q, Doolan M, Jacobs B and Aspland H, 'Multi-centre controlled trial of parenting groups for child anti-social behaviour in clinical practice', British Medical Journal 323, pp 194-98, 2001

Swaine J and Gilson J, 'Meeting the needs of adoptive parents under stress: a group work approach', Adoption & Fostering 22:2, pp 44-53, 1998

Turner K and Saunders M, 'Dissemination of evidence-based parenting and family support strategies: learning from the Triple P-Positive Parenting Program system approach', Aggression and Violent Behaviour 11:2, pp 176-93, 2006

Webster-Stratton C and Hancock L, 'Training for parents of young children with conduct problems: content methods and therapeutic processes', in Schaeffer CE and Briemester JM (eds), Handbook of Parenting Training, New York: Wiley, 1998

(1) The Hadley Centre for Adoption and Foster Care Studies, based at the School for Policy Studies, University of Bristol, provided a strategic, conceptual and advisory role to the evaluation work and undertook the evaluation of the 'live' programmes. The Evaluation Trust, a registered charity and capacity-building organisation, supported and undertook the evaluation of the 20 'historic' programmes, and interviewed purchasers, trustees, staff and those involved in developing the programme. The Trust also introduced Adoption UK to the Hadley Centre.

(2) Experimental group T = 0, z = -2.82, r = .46, p<.000 Control group T = 2.5 z = -1.0, r = -0.22, p<.317. Represents a medium effect size.

Julie Selwyn is the Director of the Hadley Centre for Adoption and Foster Care Studies, University of Bristol, UK

Sarah del Tufo is the Principal Evaluation worker and Lesley Frazer is a Senior Consultant with the Evaluation Trust, Reading, UK Table 1 Questionnaire return rates Experimental group Control group Before training 100% 100% At end of training 94% 63% Five months later 54% 47% Table 2 Comparison of children's characteristics in the two groups Experimental group Control group Average age of child 8.6 years 7.2 years Average number of 68 49 months child has lived with the adoptive family Percentage of children with 70% 67% one or two previous placements Percentage of children 60% 50% with borderline or abnormal SDQ scores Table 3 The Strengths and Difficulties Questionnaire Average (mean) scores and standard deviations Experimental Pre Post Emotions 3.57 (2.31) 3.33 (2.48) Conduct 4.24 (2.79) 4.00 (2.86) Hyperactivity 5.24 (3.63) 5.14 (3.61) Peer relationships 2.76 (2.51) 2.86 (2.48) Pro-social behaviour 7.19 (2.11) 6.90 (2.39) Total score 15.81 (8.44) 15.81 (8.44) Average (mean) scores and standard deviations Control Pre Post Emotions 3.89 (3.05) 4.53 (2.61) Conduct 3.05 (2.82) 3.16 (2.87) Hyperactivity 5.16 (3.27) 5.21 (3.41) Peer relationships 2.00 (2.29) 2.42 (2.22) Pro-social behaviour 7.42 (1.87) 7.26 (2.66) Total score 14.11 (9.02) 15.32 (8.14) Table 4 The EFR Questionnaire Average (mean) scores and standard deviations Experimental Pre Post Inhibited 2.29 (1.95) 2.48 (2.38) Disinhibited 4.14 (2.57) 4.00 (2.43) Disregulated 4.86 (3.26) 4.90 (2.91) Total score 21.48 (10.98) 21.76 (10.34) Average (mean) scores and standard deviations Control Pre Post Inhibited 2.74 (2.62) 3.37 (2.75) Disinhibited 4.21 (2.42) 4.58 (2.80) Disregulated 4.21 (3.82) 4.53 (3.29) Total score 21.79 (11.69) 23.16 (12.67)