Analysis items of the four-tier immune system multiple choice test instrument using rasch model

Abstract

The purpose of this study was to analyze the feasibility of the items of the four-tier multiple-choice test immune system instrument that had been developed. The development of the instrument using the Treagust (1988) model, namely defining content, collecting student misconceptions information, and developing a diagnostic test. A total of 25 items have been developed. The results of the instrument development were tested on 142 students of grade XI from several high schools in Surakarta who were selected by simple random sampling. The data analysis technique was performed using Rasch analysis in the Winstep application. The results of the construct validity test showed items number 5, 7, and 9 did not fit the validity standards. The reliability test shows that the value of Cronbach Alpha reliability is bad (n = 0.51), the value of the reliability item is special (no = 0.97), the value of person reliability is sufficient (n = 0.68), the value of person separation is weak (n = 1.44), and the item separation value is special (n = 5.38). The person discrimination test showed student 056P31 has the highest ability and student 098P51 has the lowest ability. The item discrimination test shows item number 1 is the best item and the bad item is number 14. The item difficulty analysis showed less proportionality because there were too many items in the easy and difficult categories. An expansion of the sample is needed to see a more comprehensive and diverse range of responses to instruments.