Method This study explored the effect of parental support, school and the pupil characteristics of cognitive style and gender on attendance, conduct behaviour and attainment.
Sample. The sample comprised 222 sixteen-year-old Year 1.1 pupils (112 boys and 110 girls) from two comprehensive secondary schools, A and B.
Materials and procedure. The Cognitive Styles Analysis (Riding, 1991) was administered during Year 10 to determine pupil style on two fundamental dimensions: the Wholist-Independent-Analytic and the Verbaliser-Bimodal-Imager.
The school records gave the percentage attendance of each pupil for the school Year 11. During Year 11 the form teachers rated the pupil's behaviour on a scale of: 1, very poor; 2, poor; 3, moderate; 4, good; 5, very good.
Attainment was the mean General Certificate of Secondary Education (GCSE) performance in the three subjects English language, mathematics and science. The grades of U, G to A were coded as 0 to 7 respectively. Pupils who did not have the necessary academic attainment to be entered for the examination were coded as 0.
The quality of the schools was based on the Office for Standards in Education (OfStEd) inspections, which indicated that school B was better than school A. During Year 11 the Head of Year teacher in each school rated the parental support received by each pupil on a five-point scale: 1, very poor; 2, poor; 3, moderate; 4, good; 5, very good. Taken together, the degree of parental support and the quality of the school represent a continuum of support.
Results and discussion Each cognitive style dimension was divided into three according to a secondary school standardisation sample (Riding, 2000), Wholist-Analytic: Wholist-Intermediate-Analytic; Verbal-Imagery: Verbaliser-Bimodal-Imager. For each measure of school performance, analyses of variance took all combinations of the variables of school (two), parental support (two), gender (two), Wholist-Analytic style (three), Verbal-Imagery style (three) with the variables in groups of four, with the performance measure at Year 11 as the dependent variable. The results will be reported in the four groupings.
1 Parental support and Wholist-Analytic style. The interaction significances between Wholist-Analytic style and parental support were: attendance (p = 0.070), behaviour (p = 0.043) and attainment (p = 0.033). The interactions are shown in Figure 1. For all performance indicators when parental support is low, Wholists are superior to Analytics, but as support increases Analytics improve more than Wholists. Parental support will reflect a loving environment and the encouragement to learn, and the effects on the pupil may be on stability and reasoning ability.
2 Gender, parental support and Verbal-Imagery style. The interactions between parental support, gender and Verbaliser-Imager style were: for attendance (p = 0.002) and attainment (p = 0.005). For behaviour there was no significant interaction (p = 0.315), perhaps because of the relatively greater effect of gender on behaviour masking the other effects. The interactions are shown in Figure 2.
For both attendance and attainment, with males, Verbalisers are most affected by the level of parental support, while for the females Imagers are most influenced. …