Predictors of mathematics attainment in hearing impaired children
Deaf children lag behind their hearing peers in mathematical attainment. The reasons for this delay remain unclear. Two methods were used to identify the causes for this underachievement: a longitudinal investigation of predictors of mathematical attainment, and comparison with hearing children. In order for a cause of delay to be identified, both investigative strategies must produce positive results. The deaf children must lag behind the hearing children on the measures and the same measures must predict deaf children's mathematics attainment. The comparative study: The participants were: a) 42 hearing impaired (HI) children age range from 7;2 years to 9;1 years attending units and special schools located on eight different sites around London; b) 73 hearing children aged from 7;2 years to 8;11 years, classmates of some HI children attending a unit based in a mainstream school. A standardised maths test, a measure of their understanding of additive composition (the Shop Task), a memory scan task and tasks assessing understanding of time concepts were administered to all the children. The last two assessments were developed for the study. The performance by the HI children on standardised assessments was also compared to norms standardised on hearing populations. The deaf obtained significantly lower scores on nearly all of the tasks. In the maths test the mean standardised score for the hearing children was 92.68 and for the deaf children was 78.31. There were also significant differences on the memory scan task — the accuracy rates were lower, memory capacity sizes were smaller and the number processing speed was slower for the deaf children. On the time concept tasks the hearing children obtained significantly more correct responses on the tasks assessing change, ability to infer and order events. When the HI children's performance was compared to the norms of standardised assessments, a similar picture emerged. The mean Number Age was 1;1 year behind the hearing norms. The mean WISC score obtained was one standard deviation below the published mean. Raw scores obtained on the reading comprehension task were too low to be standardised. In assessments of receptive language, the HI children obtained standardised scores that were 1 standard deviation below the mean. It was concluded that all of these variables could be examined as predictor variables in the longitudinal study. The longitudinal study: The HI children participating in the comparison study were assessed twice again over the academic year. The outcome measures were scores on standardised mathematics assessments. The predictors were demographic and medical background; intelligence, language; understanding of time; memory capacity and number processing speed; numerical skills such as counting and additive composition. The only demographic variable consistently associated with mathematics scores was age. Analyses using fixed order multiple regression explored the relationships between the various cognitive, numerical and linguistic predictors and mathematics attainment. After controlling for age and non-verbal IQ, only three predictors remained significant: the language assessments, Shop Task, the Change and Inference Required time concepts tasks. When controlling for age, non-verbal IQ and language ability, only the Shop Task added a significant amount of variance in the equation. This equation explained 44% of the variance in a concurrent analysis and 66% and 64% of the variance in longitudinal predictions 4 and 7 months later, respectively. Conclusions: The present study confirms that HI children are behind their peers in mathematics achievement. Explanations for this delay were sought by identifying areas where their performance is poorer than that of hearing children and predictive of their own progress in mathematics. Although the HI children achieved lower scores in the majority of the assessments in the comparative study only the language measures and the Shop Task satisfied both criteria and added a significant amount of variance in the regression equations in the predictive study. It is concluded that these may be causally related to HI children's delay in mathematics.