Analysis of Validity and Reliability of Economic Achievement Test Based on Rasch Measurement Model

Abstract

Teachers regularly attend assessments on students to identify students’ levels of mastery. Achievement tests as a quality measurement tool are quintessential so that the conclusions obtained are reliable and significant. Therefore, high-quality achievement tests need to satisfy specific criteria by going through standard procedures. Nonetheless, time and competency constraints lead teachers to utilise economic test questions that do not reach specific standards. Therefore, this research intended to develop and test the validity and reliability of economic achievement tests. An economic achievement test instrument was developed, consisting of 30 objective questions based on Bloom’s taxonomy. The testing of the instrument involved 40 respondents of Form Six economics students. The researchers appointed five experts to evaluate the validity of the content of the achievement test questions. At the same time, the construct validity and instrument reliability test analysis involved item-respondent reliability analysis, item-respondent separation index, Cronbach’s alpha, item polarity, item fit, standardised residual item correlation and respondent item-ability difficulty level distribution using Rasch measurement approach through Winsteps software 3.72.3. The data of the tests conducted determined that the achievement test confirmed good content validity and reliability values. The analysis also established that six questions needed to be modified. The development of the economic achievement test offers an alternative measurement design over future performance test testing. The researchers proposed that the implementation of this measurement on other subjects too.