ABSTRACT
The impact of the Covid-19 pandemic has had a far-reaching effect on higher education institutions, and individual student assessments have garnered much attention during the pandemic. This study aimed to validate Science, Technology, Engineering, and Mathematics (STEM) application instruments using the Rasch analysis employing Winsteps version 3.73. A survey was conducted with 201 respondents from two provinces in Indonesia. The students were selected by convenience sampling and answered the adopted STEM application instrument. The STEM application instruments were adapted, and these were divided into seven sub-constructs derived from STEM disciplines. Rasch Modelling was employed for data analysis using Winsteps version 3.7.3 to analyse reliability, separation, item fit statistics, unidimensionality, and rating scale calibration. Each sub-construct fulfilled a minimum of 0.65 for Cronbach alpha, item, and person reliability, and most of them had more than 1.5 person and item separation. In general, each item had a good score of the mean square, Z-tolerated standard, and point measure correlation, indicating fulfilment of the Rasch measurement model. The analysis also showed unidimensionality assumption and an excellent rating scale. This study contributed to the body of STEM knowledge by using Rasch Modelling to test the validity and reliability of STEM application instruments. © 2022 Society for Research and Knowledge Management. All rights reserved.