Analysis of Teacher-Made Tests Used in Summative Evaluation at SMP Negeri 1 Tompaso

Main Author: Nihta V.F. Liando, Eunike Serhalawan, Ceisy Wuntu
Format: Article Journal
Terbitan: , 2021
Subjects:
Online Access: https://zenodo.org/record/5775342
Daftar Isi:
  • The purpose of the present study is to describe the quality of English tests for use in summative evaluation in sixth semester in 2019-2020 and in fifth semester 2020-2021, The study was descriptive-and evaluative in nature in that it tried to determine the advantages (strengths) and disadvantages (weaknesses) ofthe summative tests used. in terms of their validity, reliability and item analysis. The data were the fifth and sixth semester students’ responses to the two test in semester examinations. The data obtained were statistically analyzed using Point-Biserial for validity, KR-20 for reliability and item analysis for level of difficulty and discrimination power. Result of the statistical and item analysis indicated that (1)concerning the validity and reliability of the fifth semester 2020/2021 summative test, the data analysis shows that all the items are valid with validity coefficientranges between 0.53 and 0.98, whereas reliability coefficient is 0.82, meaning that the test is highly reliable. It can be concluded that the test is valid and highly reliable, (2) concerning validity and reliability of the sixth semester summative test in 2019/2020, the analysis indicates that all the items are valid with validity coefficient ranges between 0.53 and 0.85, whereas reliability coefficient is 0.88, meaning that the test is highly reliable. It can be concluded that the test is good in its validity and highly reliable, (3) Item analysis item facility of the fifth semester 2020/2021 summative test shows that item facility index of the 50 items in the test ranges between 0.55 and 0.67, meaning that the test is good in terms of its item facility index; and of 40 multiple-choice items in the sixth semester 22019/2020 summative test, 27 items are considered very good items, and 13 items are reasonably good in their discrimination power. It can be concluded that the test is good in terms of its item facility and discrimination power, and (4) Item analysis of the sixth semester 2019/2020 summative test shows that in terms of item facility, all the 50 items are recommended for use. In terms of item discrimination, the analysis shows that 30 are very good or acceptable, and, therefore, are recommended for use, 19 are reasonably good, and only 1 is marginal and subject to improvement.It can be concluded that the test is good in its item facility and discrimination power. In general, the two summative tests, one used in the fifth semester 2020 and the other one in the sixth semester 2019 are good in terms of the characteristics of a good objective type test in multiple-choice format.