The relationship between personality and attainment in 16-19-year-old students in a sixth form college. I: Construction of the Student Self-Perception Scale.

Abstract

Of the research that has been undertaken into the relationship between personality and attainment, relatively little exists relating to the 16-19 age range. In a substantive study examining the relationship between academic self-concept, attainment and personality in sixth form students, a first requirement was to design a self-perception instrument. The psychometric element of the study aimed to construct a Student Self-Perception Scale (SSPS) that would be effective for students in the FE (further education) context. The samples comprised a pilot sample of 152 students (aged 16-17 years from two sixth from colleges) and a main sample of 364 students (mean age, 16yrs 10mths, range 16:0 to 18:6 years, from one sixth form college). The main sample included similar numbers of male and female students (46% male, 54% female) and ethnic minority students comprised 14% of this sample. An initial item pool of 88 four-point Likert type statements was compiled from comparable existing scales and from responses to a Student Induction Questionnaire. Item analysis was based on oblique factor analysis of the pilot sample responses, followed by cross-validation on the main sample to refine the scale structures. Construct validity was established from the substantive study, especially the Nowicki & Strickland (1973) locus of control results. Exploration of the four- and five-factor structures led to a final specification based on 52 items from five oblique factors. The constituent scales were Passivity (12 items, alpha = .81). Mastery (15 items, alpha = .79), Work Related Inadequacy (11 items, alpha = .72), Extraversion (4 items, alpha = .70) and Social Dependence (10 items, alpha = .66), all statistics compiled from the cross-validation sample. Correlations with Locus of Control ranged from 0.52 for Mastery to -.34 for Work Related Inadequacy. Distribution statistics for Locus of Control matched a comparable American sample. The five-scale structure exhibits good cross-validation characteristics and supports revealing analyses of relationships within the substantive study. Its 52-item format is suitable for research or exploratory use within its intended FE context.