Relationships between family variables and children's mathematical achievement
This study aims to assess the relationship between family variables and specific mathematical achievement in secondary school children. It employs both quantitative and qualitative methods. In the early stages, re-analysis of data from the National Child Development Study (NCDS), (e. g. Fogelman, K. 1983), suggested that previous findings may have confounded mathematical ability/ achievement with general educational achievement and/or measures of cognitive ability. In addition these studies tended to use only family variables which could be measured quantitatively and in ways which facilitated statistical analysis. However, the re-analysis of NCDS data showed a significant association between teachers' subjective assessment of parental interest in their children's education when they were aged 7 years, and the children's mathematical achievement at age 11 years. This significant relationship led to the focus in the second, qualitative phase on the involvement of parents in their children's education (Steinberg and Silverberg, 1986; Lareau, A. 1987,1989), on intra-family interactions (Wentzel, K. R. 1994), and the possible relations between them and the children's mathematical achievement. In the later stages, therefore, following a methodological model developed by Goetz and Lecompte (1995), qualitative techniques were utilised to elicit from parents details of intra-family interactions, with particular emphasis on their involvement with their children's educational activities. The results of the qualitative phase of this study suggested that high mathematics achievers were more likely to be found in families where mothers and fathers had both obtained 0 Levels, where a similar history of educational achievement in the wider families was reported, and where there was a high level of parent/child/sibling interaction. In contrast, low mathematics achievers were more likely to be found in families where only one or neither parent had obtained 0 Levels, where there was little evidence offered of educational achievements in the wider families.