DEVELOPMENT AND VALIDATION OF AN INDIGENOUS INTELLIGENCE TEST

Abstract

The aim of the study was to construct and standardize an Indigenous Intelligence Test (IIT) to measure general intelligence ‘g’ and a screening tool for youth intending to apply in Armed Forces, Public Service Commission and similar organizations where English carries a status of official language. The items of IIT were developed with the help of material obtained from own culture and were constructed in simple English language. The IIT comprises of six sub-scales: Vocabulary, Arithmetic, Analogy, Information, Comprehension, and Similarity. The test was planned to measure various aspects of intelligence verbal ability, to understand concepts, general reasoning factor, element of discovery, acquired knowledge, social intelligence, eduction of relations and correlates. The usefulness of the items was judged through item analysis by administering the test to a sample of 200 subjects with minimum Intermediate qualification. Items were evaluated by considering two main aspects; discrimination power and difficulty level. Items with discriminatory power less than .30 and level of difficulty below .30 and above .70 were discarded. The final draft of the test comprised of 71 items; Vocabulary and Arithmetic sub-scales consists of 12 items each, Analogy 14, Information 11, Comprehension 12 and Similarity 10. In the main study the test was administered to 1669 students both male and female, belonging to various Government Colleges & universities of four provinces of the country including AJK and FATA. Candidates both from urban and rural areas were included in the study. Two methods were used to establish reliability of the test i.e., KR 20 and Split-half methods. The estimated indices of reliability were, .79 and .69 respectively. To determine validity of indigenous intelligence test different validity criteria were used including urban/rural differences, province wise differences, correlation of the total test and sub-scales (construct validity) and marks in last qualified examination to determine concurrent validity Significant differences in the mean scores of the male and female candidates on vocabulary, arithmetic, analogy and information were found whereas non significant difference was observed on comprehension and similarity subscales. Significant mean differences on vocabulary, arithmetic, analogy and information and non significant difference on comprehension and similarity were also found. Inter-correlation among the subscales and IIT were calculated to determine construct validity of the test. The concurrent validity was determined by correlating scores on IIT with the college marks obtained by each student in the last annual examination. Both indices established the evidence of high validity of the test. Significant differences in urban/rural mean scores on vocabulary, analogy, information, comprehension subscales and IIT were observed. 12Significant differences were also observed in Province wise mean scores on vocabulary, arithmetic, analogy, information and comprehension subscales. Whereas, non significant mean scores were found on similarity subscale. Income group wise comparison revealed that higher the income groups better the performance. Percentile norms were developed for IIT and subscales while administering the test on a sample of 1669 subjects representative of Pakistani youth. In addition, province wise norms were also developed to gauge true performance of candidates hailing from different provinces.