Validity of Items and Spurious Overlap (ESO) Effects Unidimensionally and Multidimensionally (A Study of the Effects of Spurious Overlap on the Test of Logical Thinking Ability

Abstract

At the junior high school level, there are still very few standardized tests of logical thinking abilities that can be used to measure students' logical thinking abilities. This research aims to analyze the validity of items and the effect of spurious overlap (ESO) unidimensionally and multidimensionally (a study of the effect of spurious overlap on tests of logical thinking abilities. This research is a type of instrument calibration research. The population of this research is all students of class VIII SMP, with the sample in the form of answers from 100 students selected using purposive sampling. Data collection method using tests. The results of data analysis obtained findings from testing the validity of unidimensional logical thinking ability test items before being corrected for the spurious overlap effect; 49 items were suitable for use as part of the test, and testing the validity of the multidimensional logical thinking ability test items before being corrected for the spurious overlap effect, 49 items are suitable for use as test constructors, in the unidimensional test of item validity after being corrected for the spurious overlap effect, 47 items are suitable for use as test constructors; and testing the validity of the multidimensional logical thinking ability test items before being corrected for the spurious overlap effect, 45 items are suitable to be used as test components, and the unidimensional ESO coefficient is smaller when compared to the multidimensional ESO coefficient of the items that make up the logical thinking ability test.