The study revalidated the “Renolds Adaptable Intelligence Test (RAIT-NV)” for use in Nigeria. Instrumentation design was used. The researcher purchased and administered four hundred (469) copies of Raynolds Adaptable Intelligence Test-Non-Verbal (RAIT-NV) to a sample drawn across educational institutions, hospitals, corporate organizations and religious bodies. Before then, the researcher had quickly carried out a pilot test using 100 respondents to determine the validity and reliability of RAIT-NV in its original form. Discriminant validation process was used in determining the validity while test retest was used in determining its reliability. For the earlier, the Leibowitz social anxiety scale was administered alongside RAIT-NV. After the scores the researcher correlated the scores and an index of 0.13 was realized testifying the discrepancy in these scales. The test retest also showed a correlation coefficient of 0.86 indicating a high reliability. For determination of reliability, the researcher used Cronbach Alpha, split half method and KR20. In determining the mean differences in the groups, the researcher used ANOVA and t-test in the process. The result showed that thirty (30) and thirty-three (33) items were either too difficult or easy for respondents here in Nigeria. Hence, thirty-two (32) items were considered normal with overall difficulty index for RAIT-NV is 0.55 (55%) in Nigeria. Nineteen (19) items discriminated better. Construct validity showed that there is a significant difference in age while there were non in terms of educational level, ethnic groups and gender. Internal consistency reliability reveals that overall Cronbach of 0.66, KR20 of 0.62 for NVA and 0.58 for SEQ and Split-Half reliability index of 0.07 for NVA and 0.18 for SEQ.
Keywords: Reliability, Renolds adaptable intelligence test, revalidation, validity
