For several decades, research from most Western countries has consistently reported that children in out-of-home care tend to be low achievers in school and are at high risk of entering adulthood with a low level of education (eg Bohman and Sigvardsson, 1980a, b; Festinger, 1983; Dumaret, 1985; Runyan and Gould, 1985; Stein and Carey, 1987; Barth, 1990; Weiner and Weiner, 1990; Christoffersen, 1993;Veland, 1993; Cheung and Heath, 1994; Cook, 1994; Jackson, 1994; Cashmore andPaxman, 1996; Vinnerljung, 1996; Blome, 1997; Courtney et al, 2001; Social Exclusion Unit, 2003; Pecora et al, 2006; Egelund et al, 2008; Clausen and Kristofersen, 2008).
Vinnerljung and colleagues (2005) used Swedish register data for eight national birth cohorts to examine the educational attainments of over 31,000 former child welfare clients and almost 750,000 majority population peers. Compared to majority population peers with low-educated mothers (only compulsory schooling), young people who had been in long-term stable foster care had a two-to three-fold elevated relative risk of reaching adulthood with only a compulsory education--after controlling for influence of the birth mother's education. Majority population peers with low-educated mothers were, at the age of 25, between two and four times more likely to have a post-secondary education degree when compared to former foster children who had been in long-term care. A decade earlier, the British researcher David Berridge had concluded that the compensatory long-term effects of care on education seemed, at best, to be neutral (Berridge, 1994, 1997). In a US doctoral dissertation, Deborah Matthews (1997) examined 293 children using various standardised measurements and related the results to the length of time spent in out-of-home care. She found that reading achievement, mental development and overall behaviour problems were negatively correlated with the duration of care, but found no association between time in care and intelligence scores or achievements in maths.
This large body of research is based on cross-sectional data. We know less about foster children's progress--or its absence--in school during their time in care. The UK Social Exclusion Unit (2003) found that the gap between looked after children's school achievements and that of their peers tended to widen with age. However, it measured educational outcomes for different age groups at the same point in time. Studies based on longitudinal designs show less uniform results. Fanshel and Shinn's classic Children in Foster Care (1978) found that foster children's school achievements actually deteriorated during the first two-and-a-half years in care, but improved during the subsequent 30 months. After Fanshel and Shinn, several US longitudinal studies employing matched comparison designs concluded that out-of-home care does not seem to facilitate children's cognitive or academic development (Berzin, 2008; Berger et al, 2009; Stahmer et al, 2009; cp. Doyle, 2007). In the UK, Heath and associates (1994) followed 49 foster children through three years of care and reported a lack of educational progress compared with national age standardised norms, even for those who were in stable long-term placements.
Poor academic performance in primary school seems to be a robust predictor for future psychosocial problems for all children and adolescents (Jablonska et al, 2009; Vinnerljung et al, 2010). A series of recent national cohort studies by Vinnerljung et al (2010) showed that Swedish children who grow up in foster care had substantially lower performance in primary school than their peers with similar cognitive ability. Foster children in long-term care also displayed higher risks (RR = 6-10) for future suicide attempts, serious criminality, substance misuse, long-term dependence on public welfare and several other negative outcomes. This increase in risk was reduced by roughly half after adjustment for school failure (cp. Zingraff et al, 1994). Sonia Jackson in the UK has for decades argued that poor school performance and low education are the strongest risk factors for looked after children's futures (Jackson, 1994). The Swedish cohort studies strongly support her hypothesis.
The dismal educational performance of looked after children has been explored in numerous UK studies during the last two decades (see, for example, Jackson, 2001; Jackson, 2007). But, in spite of all these research efforts, there are few examples of evaluated attempts to do something about the problem, even when using a wide definition of the concept 'evaluated'. The by-and-large successful US 'emancipation programmes' include educational support, but are limited to foster children in their mid- to late teens (eg Montgomery et al, 2006).
We have come across only three examples of interventions aimed at primary school age foster children that have been evaluated with regard to their effects. In Olisa and colleagues' pre-post intervention study from London (undated report), ten foster children were given extra literacy and numeracy training over a 20-week period outside of the curriculum. Five children did not receive any such interventions. All participants were tested with standardised instruments for cognitive capacity and literacy/ numeracy skills at the start and at the end of the project. The results suggested that the training sessions had some effects: the children who had received extra training had made progress and were catching up with their peers in reading, spelling and maths.
A literacy-focused intervention in Kent, UK, was evaluated by Wolfendale and Bryans (2004). In this project, 58 foster children were provided with books, a hand-held computer and other tools to stimulate their interest in reading. A comparison of pre- and postscores on a standardised literacy test showed significant gains in reading accuracy, spelling and comprehension. The intervention seemed most beneficial for children with low pre-test scores. In an ongoing Canadian randomised field trial, foster carers were trained to be adult tutors to their 77 foster children. Recently reported results from the first year follow-up are promising. Children who had received foster carer tutoring had significantly better results on several measures of academic performance (eg numeracy skills) than peers in the control group (Flynn et al, 2010).
In this article, we report on the results of a Swedish intensive project aimed at improving foster children's school achievements. By employing standardised instruments for baseline and follow-up measurements, it shares some common ground with Olisa et al and the Kent studies. However, in this project baseline test results were also used for assessing individual potential and educational service needs, and for tailoring interventions to meet the needs of individual children. After a two-year period, the same tests were used for assessing post-intervention outcomes.
The project was staged in Helsingborg, a town in southern Sweden with about 125,000 inhabitants. We included all 30 children aged 7-11 in foster family care whose placements were perceived by case workers as likely to last another two years. From this sample, we excluded five children who either had been diagnosed with a neuropsychiatric disorder or were placed in special education schools due to very disruptive behaviour. This left us with 25 children. Their median time in care since birth was 3.5 years with a median of more than two years in their present foster family. Subsequently, the large majority of the children can be characterised as placed in long-term foster care. At the start of the project, the children's median age was ten years. The majority were already in care when they started primary school (age 7). Compared to all children in Swedish out-of-home care, the children in the project were younger, as two out of three children entering foster- or residential care are teenagers (Vinnerljung et al, 2007).
Most children (13 of 23; data were unavailable for two children) had experienced more than one placement in out-of-home care since birth and ten out of 23 had been in three placements or more. But most had experienced reasonably stable schooling. About two-thirds had not changed school at all since they started primary education or had changed school only once; a third had changed two or three times.
At the start of the project (baseline/T1), a psychologist assessed each child's cognitive ability. The results of the cognitive tests were compared to results on standardised tests for reading, spelling and numeracy, administered by a special education teacher, and also to achievements in school, as perceived and reported by the teachers. Standardised tests were also used to assess baseline psychological well-being and behaviour, as well as child-teacher relations.
The psychologist and the special education teacher were employed in the project and henceforth became external resource people for the schools. Both had long experience in their fields and were well suited to pursue the intentions of the project. The psychologist worked part time in the project but had a flexible schedule and could adapt her working hours to the needs of the children at...