Items’ Difficulty in the Lower Secondary Bahasa Melayu System Diagnostic Test (LSBMSD) Instruments

1Mohd Zaharuddin Omar, *Zamri Mahamod


The main purpose of this research is to assess the suitability of the questions’ level and the items’ difficulty in the instruments used in the Lower Secondary Bahasa Melayu System Diagnostic Test (LSBMDT). The study is quantitative. Items’ detection is carried using Winstepts application (Rasch Model) to determine the overall ranking of 289 items via the analysis of items map. This analysis could provide a clear picture of the right side items’ rankings which were calibrated with the respondents on the left side of the logits scale to determine the items’ difficulty level. This analysis was validated by seven experts. This study was participated by 935 form one students in Pahang and it involved 10 schools in 10 districts. Alpha Cronbach reliability index for all items was 0.98, whereby the reliability index for each construct of the instruments was between 0.71 and 0.95. It was at the `very good’ level. The only exception was the vocabulary construct with 0.71 index (fisher, 2007). The result of the study shows the existence of item free area at the top part with the interval of 1.24 logits and at the bottom part of the variable map with the interval of 0.82 logits. The items classified at ‘knowledge’, `understanding’ and `application’ levels were at the right levels. No item was too difficult or too easy. To sum up, UDSBMR instrument contains valid and suitable items as a diagnostic tool to determine any weakness of the language system used in Form 1 Bahasa Melayu.


Question Level, Item Difficulty, Reliability Index, Item-Free Area, Items Map Analysis.

Paper Details
IssueIssue 6