DEVELOPMENT AND STANDARDIZATION OF AN INDIGENOUS NONVERBAL TEST OF INTELLIGENCE

Abstract

The study was aimed at developing and standardizing an indigenous nonverbal test of intelligence for the assessment of general intelligence of Pakistani youth. The items of the test were developed using figurative material taken from our own cultural and folk heritage and were constructed in the framework of English script writing (from left to right). The test comprised of five subtests viz. , Series, Matrices, Analogies, Odd One Out and Similarities. It was planned in a manner so as to measure various aspects of intelligence including abstract reasoning, spatial relations, conceptual ability, accuracy of discrimination, eduction of relations and correlates. The effectiveness of the items was judged through item analysis by administering the test to a sample of 200 subjects of grade 12. In evaluating the items, three major aspects; internal consistency, discrimination index and level of difficulty were considered. Only those items were retained that were internally consistent (p < .001). Items having discrimination index less than. 3 and level of difficulty below . 3 and above . 7 were discarded. The .final draft of the test comprised 90 items; the first three subtests consist of 20 items each and the last two 15 items each. In the main study, the test was administered to 200 subjects; 100 urban and 100 rural. Three methods were used to establish reliability of the test viz., KR-20, split half and test-retest. The estimated indices of reliability were, .89, .85 and .90 respectively. To determine validity of indigenous nonverbal test of intelligence, different validity criteria were used including grade/age differentiation, correlation of the test with other measures of general ability (construct validity) and marks in HSSC examination (criterion-related validity). Significant differences in the mean scores of three grade/age groups on the subtests and the full test, except similarities subtest, were found. To determine the construct validity of the test, factor analysis and the convergent and discriminant validation approaches were adopted. To find out convergent vahdity, the test was correlated with Intelligence Test Bat/elY (ITB) and adapted version of Raven's Standard Progressive Matrices (RSPM). The results demonstrate sign(ficant correlations between all these tests (p < .001). The discriminant validity was determined by correlating the test with Individual Obsticles Test, which was insign!ficant. Criterion-related vahdity of the test is also high ( 74: p < .001). No sign(ficant d(fferences emerged between urban and rural groups. Similarly, medium of 11 instruction (English/Urdu) do es not seem to influence the test scores. Nevertheless. sign[ficant d[fferences were found between Science and Arts groups. Three types of norms viz .. Percentiles. T scores and the Deviation IQ were developed by administering the test on a sample of 1 000 su~jects representative of Pakistani youth.