Skripsi
Analysis of a final online teacher-made english test delivered during the covid-19 pandemic /Ismia Liainur Ifadho
Abstrak
This study analyzed the content validity construct validity test reliability item difficulty and item discrimination of a high-school teacher-made final English test for tenth graders of a private high school in Malang which was conducted online using Google Form application. The design of the study was descriptive analysis. The analysis of content validity and construct validity were done using documentation checklist instruments meanwhile the analysis of reliability used KR-20 formula item difficulty used formula from Sudijono (2009) and item discrimination used formula from Arikunto (2013) which were calculated in Microsoft Excel 2010. The result showed that the online test items had 60% content validity and construct validity (40% microskills and 100% macroskills) its reliability was 0.85 which was categorized as reliable its item difficulty had 3% difficult items 50% medium items and 47% easy items and its item discrimination had 10% poor items 27% enough items 53% good items and 10% very good items. The data revealed that in general the test fulfilled the quality of a good test yet it still needs some revisions on the test formats. Short answer or completion items can be added to minimize the possibility of cheating.