The purpose of this study is first, to investigate whether a relationship exists between self-esteem in particular academic subjects and academic achievement in the matching areas. Second, it examines the various nonacademic self-esteem dimensions and academic achievement of pupils with substantial learning difficulties (EM3) and pupils with less learning difficulties (EM2). Third, to investigate whether the self-esteem of pupils with substantial learning difficulties (EM3) and pupils with less learning difficulties (EM2). Third, to investigate whether the self-esteem of pupils with substantial learning difficulties (EM3) differs significantly from pupils with less learning difficulties (EM2). Finally, it tests whether gender differences exist in the self-esteem dimensions and academic achievement of both groups of pupils.

The sample for this study was drawn from a class of 29 EM3 pupils (boys = 12, girls = 17) and a class of 39 EM2 pupils (boys = 20, girls = 19) of an average primary school in the northern part of Singapore.

In this study, self-esteem dimensions were measured by the Self-Esteem Questionnaire (SEQ), adapted from the Self Description Questionnaire I (SDQI) which was developed by Herbert W. Marsh with high scores indicating high self-esteem. The SEQ had 6 subscales and a total of 48 items. The academic achievement of the sample was measured by their English and Mathematics school end-of-year examinations.

The data collected in the main study were analysed using the following statistical procedures : means and standard deviation, Pearson product moment correlation, t-test and One-Way Analysis of Variance.

The major findings are : (a) Academic achievement scores are correlated more highly with matching academic self-esteem than with nonmatching academic self-esteem for both EM3 and EM2 pupils. (b) Significant positive correlation between English and Mathematics achievement and a negative correlation between English and Mathematics self-esteem are observed among pupils with substantial learning difficulties (EM3). (c) No significant relationship between nonacademic self-esteem dimensions and academic self-esteem exists for the EM3 pupils. (d) No between groups differences were observed in academic self-esteem and academic achievement except for Mathematics self-esteem favouring the EM3 pupils. (e) English and Mathematics achievement scores are positively correlated for the EM3 pupils but uncorrelated for EM2 pupils. (f) Gender differences are found only within the EM3 pupils in physical abilities in favour of boys and within the EM2 pupils in Mathematics achievement in favour of boys too.

The significance and educational implications of these findings are that teachers, parents, counsellors and educational psychologists should enhance the subject-specific self-esteem of the pupils with learning difficulties through creative and varied teaching strategies. Of particular importance, is the enhancement of the English self-esteem and English achievement of these groups of pupils.