Characteristics of Math national-standardized school exam test items in junior high school: What must be considered?

Abstract

National Standardized School Exam (USBN) is used to determine Student’s graduation. This research aims to determine the characteristics of items for Math USBN in SMP on grade 9. This kind of research is a descriptive-explorative quantitative research. The data collected is a USBN test instrument at SMP Negeri 3 Pati and participants' answers, which are collected by documentation. The USBN instrument was validated by experts and the characteristic items of USBN instrument were analyzed using the classical test theory approach. The question items of math USBN test at SMP Negeri 3 Pati is generally moderately good. Based on the classical theory approach, the result of the validity is 0.924. 56.7% of items are very valid. The reliability is 0.78 categorized as high reliability. Generally, Math USBN items are in the easy category with a percentage of 83.3%. The results of discrimination index indicate that in general, USBN items are in a moderate category with a percentage of 60%. The distraction effectiveness shows that USBN items are in the functional category with a percentage of 50%.

What might be frequently overlooked is actually still beneficial: Learning from post national-standardized school examination

“…Such type of error that makes student's test score unpredictable can arise due to students, test-specific, or scoring specific factors (Wells & Wollack, 2003). In a study conducted by Argianti and Retnawati (2020), the students' scores obtained from a test used in national-standardized school examination was reliable, meaning that this result is consistent with our result. In another study (Sampouw & Retnawati, 2020), however, showed a contrast result, in which it was found that the scores obtained from the test used in the examination was unreliable.…”

“…With respect to this focus, the results of our study revealed that more than half of all multiple-choice items performed well in distinguishing students who took the test based on their score assumed to reflect their competence. The previous study focusing on investigating the characteristics of a mathematics test and its items used in the national-standardized school examination (Argianti & Retnawati, 2020) has also shown similar results in terms of the number of items that have good discriminating power. Furthermore, a study conducted by Sampouw and Retnawati (2020) which has a similar focus to our study obtained a similar result with what we have found that of 30 multiple-choice items, one item has a negative discrimination coefficient.…”

“…Many previous studies have used the classical test theory (CTT) (e.g., Argianti & Retnawati, 2020;Awopeju & Afolabi, 2016;Maharani & Putro, 2020;Muna et al, 2019;Quaigrain & Arhin, 2017;Retnawati et al, 2017;Sampouw & Retnawati, 2020) or item response theory (IRT) (e.g., Awopeju & Afolabi, 2016;Gleason, 2008) approach in performing item analysis. CTT is based on a linear relationship between an observable variable, namely the observable test score (X), and the sum of two unobservable (or latent) variables, namely the true score (T) and error score (E).…”

“…In general, there are three views regarding statistical methods used by researchers in their studies to distinguish between functioning and non-functioning distractors on a multiple-choice item of a test. The first view is only focused on the frequency of students choosing each distractor, in which a distractor is said to be functional when at least 5% of the total number of students taking the test choose that distractor (e.g., Argianti & Retnawati, 2020;Haladyna & Downing, 1993;Hingorjo & Jaleel, 2012;Sampouw & Retnawati, 2020). The second view is emphasized on the frequency of students choosing each distractor and the index or coefficient of discrimination.…”

“…Therefore, based on the willingness to obtain things that can be learned from the nationalstandardized school examination instrument, the present study seeks to describe characteristics of the instrument that reflect the quality or characteristics of the test and its items on the examination by performing post-examination item analysis. Although several studies focusing on item analysis of examination instruments have been carried out (e.g., Argianti & Retnawati, 2020;Sampouw & Retnawati, 2020), our study was still relevant to be conducted considering that the national-standardized school examination instrument could be regarded as a teacher-developed test. Consequently, each test instrument in a particular subject used by schools in a regency or city would provide opportunities for learning that are different from the instruments used by schools in other regencies or cities.…”