NORMATIVE DATA ON THE RAVEN’S STANDARD PROGRESSIVE MATRICES AND THE SLOSSON INTELLIGENCE TEST AMONG GHANAIAN CHILDREN

Abstract

Intelligence as a construct is difficult to define and measure. In non-Western populations, it is even more difficult to do so. The lack of locally derived normative data on most available intelligence testing instruments results in inaccurate assessments due to unfair comparisons between Western and non-Western populations. In addition, various factors that affect cognitive ability have scarcely been elucidated in most sub-Saharan African countries. This has impeded the understanding of intelligence, why it appears to differ among various segments of the population, and how it can be improved. The link between risk factors such as poverty and low educational attainment in a caregiver and the lowering of cognitive ability has been well established. Furthermore, mental health problems are fairly common in the extremes of the intelligence spectrum: among the intellectually gifted and those with intellectual disabilities. Data from a recent meta-analysis showed an unusually high prevalence (2-3%) of Intellectual Disability in developing countries as compared to high-income countries, a situation likely due to poor obstetric and antenatal care. There is therefore the need to develop accurate and locally acceptable means of assessing intelligence among the local population of children of developing countries such as Ghana. There is also the need to examine factors associated with performance on intelligence testing to enhance our understanding of intelligence and introduce interventions as much as is possible. Objective: This study aimed to provide normative data for the Raven’s Standard Progressive Matrices (RSPM) and the Slosson Intelligence Test (SIT) among Ghanaian children and adolescents aged 6-19 years to allow for more accurate local assessment of intelligence, and to assess sociodemographic factors most predictive of intelligence scores on these tests. Methodology: In a cross-sectional survey, 619 children and adolescents aged 6-19 years were recruited into the study from 11 randomly selected schools (7 public schools and 4 private schools) located in rural and urban Kumasi in the Ashanti region of Ghana. Participants with known cognitive deficits were excluded from the study. Sampling was by multi-stage random sampling. Once informed consent was obtained and socio-demographic data was collected, screening for depression and anxiety symptoms was done using the Patient Health Questionnaire-9. Finally, the intelligence instruments were administered to the participants. Instruments used to establish intelligence as a construct is difficult to define and measure. In non-Western populations, it is even more difficult to do so. The lack of locally derived normative data on most available intelligence testing instruments results in inaccurate assessments due to unfair comparisons between Western and non-Western populations. In addition, various factors that affect cognitive ability have scarcely been elucidated in most sub-Saharan African countries. This has impeded the understanding of intelligence, why it appears to differ among various segments of the population, and how it can be improved. The link between risk factors such as poverty and low educational attainment in a caregiver and the lowering of cognitive ability has been well established. Furthermore, mental health problems are fairly common in the extremes of the intelligence spectrum: among the intellectually gifted and those with intellectual disabilities. Data from a recent meta-analysis showed an unusually high prevalence (2-3%) of Intellectual Disability in developing countries as compared to high-income countries, a situation likely due to poor obstetric and antenatal care. There is therefore the need to develop accurate and locally acceptable means of assessing intelligence among the local population of children of developing countries such as Ghana. There is also the need to examine factors associated with performance on intelligence testing to enhance our understanding of intelligence and introduce interventions as much as is possible. Objective: This study aimed to provide normative data for the Raven’s Standard Progressive Matrices (RSPM) and the Slosson Intelligence Test (SIT) among Ghanaian children and adolescents aged 6-19 years to allow for more accurate local assessment of intelligence, and to assess sociodemographic factors most predictive of intelligence scores on these tests. Methodology: In a cross-sectional survey, 619 children and adolescents aged 6-19 years were recruited into the study from 11 randomly selected schools (7 public schools and 4 private schools) located in rural and urban Kumasi in the Ashanti region of Ghana. Participants with known cognitive deficits were excluded from the study. Sampling was by multi-stage random sampling. Once informed consent was obtained and socio-demographic data was collected, screening for depression and anxiety symptoms was done using the Patient Health Questionnaire-9. Finally, the intelligence instruments were administered to the participants. Instruments used to establish Multiple linear regression models revealed that children in private schools had an over 5 point increase in scores compared to public schools independent of all other variables (95% CI 1.58 to 9.36). Children with a reading habit had a 3.39 point increase in scores (95% CI 1.55 to 5.24), and teachers with higher levels of training were associated with 4.44 point increase in scores (95% CI 1.15 to 7.73). Conclusions: This study reports the first attempt to standardize the RSPM and SIT in Ghana across the wide age range of 6-19 years. The different normative values obtained from western standards underscore the need for local norming of psychometric instruments to be able to use them meaningfully. The vastly different scores obtained for different sub-groups within the population appeared to be largely connected to standard of living. These significant differences appear to justify the calls for establishing different normative data for different subsets of the population within the national population of developing countries for the same reason local normative data are recommended rather than importing western norms. This will help to avoid unfair comparisons across widely disparate socio-demographic sub-groups. Also, the strong case for the effect of environmental factors on intelligence scores over and above genetic factors has been enhanced by this study. Interventions for improving the quality of education and the standard of living of developing countries are thus desirable for the improvement of the cognitive ability of children