This paper presents the efforts made to establish validity and reliability of a mathematics achievement test developed by researcher. The sample consisted of 80 students of Grade 7th, studying in two sections of an urban public sector school. Data of mathematics test scores were collected in two phases of pilot study. Two evaluation tools requiring SMEs (subject-matter experts) to rate test items on their alignment with ILOs of NCM 2006 and levels of Blooms Taxonomy, were also used to collect data for calculating content and construct validity. Item analysis was used for finding discrimination power and Cronbach Alpha used to calculate reliability value. The test is found moderate to highly reliable (0.79), content perfectly aligned (86.65 % of test items) with curriculum ILOs and well matched (K= 41.66 %, Com= 36.66 %, App= 21.68 %) with levels of Blooms Taxonomy and possesses good power of discrimination (Reasonably good=78 % test items, DI values 0.4-0.3, Marginally good=18 % test items, DI values 0.2-0.29).
Key words: Test Development, validity and reliability, ILOs, discrimination power
|
