Assessing the Effect Size and Homogeneity of Selected Standardized Tests in Education

Abstract

The study was conducted to analyse the empirical differences in the studies on procedures for validating standardised tests in Education. It examined the difference in the findings of the selected studies in terms of the sample size and also determined the homogeneity of the effect size of the selected studies. The study adopted an ex-post factor research design. The study population comprised empirical studies on validating standardised tests in Education (1988 to 2017). The sample size consisted of 130 empirical studies on the validation of standardised tests in education (1988-2017), which were selected using a purpose sampling technique. The research instrument comprised both published and unpublished articles, PhD and Masters theses. Data were analysed using counts and percentages, study variance, and statistical transformations to convert t-test and f-test to corresponding ‘r’ statistics. Also, Q statistics, which is the sum of weighted squares based on Chi-square, I2, Tau2 and Fishers Z, were used for the analysis. Results showed an inverse relationship between sample and effect sizes in the selected studies. Also, there is no significant difference in the effect sizes of studies (Zr=1.14403). Furthermore, the test for homogeneity showed that the studies significantly differ in their effect sizes; Q=24940.437, I2 = 99.48%, and Tau2 = 0.209>0 Tau=0.457 do not belong to the same population.