Answers are grounded in this paper and cite their sources.
Mokken Scale Analysis (MSA) can offer a more flexible approach to common Item Response Theories (IRT) as a non-parametric method for evaluating educational tests with ordinal data. This study applied MSA on a Vocabulary Levels Test (VLT) originally developed with IRT methods among 234 English as a Foreign Language (EFL) students to test the assumptions of MSA, including unidimensionality, monotonicity, local independence, and invariant item ordering (IIO) with mokken package in R. MSA assumptions successfully met with the VLT validated with validating the VLT being a unidimensional test with Automated Item Selection Procedures (AISP) and Genetic Algorithm (GA) methods utilized. Scalability coefficients indicated a medium scale, and monotonicity was upheld, which ensured consistent participant rankings. IIO showed that item difficulty remained consistent across respondents. MSA highlights its potential for exploratory analysis, particularly with smaller sample sizes and ordinal data, offering a less restrictive alternative to parametric IRT methods.
No takes yet. Share an insight, caveat, or question.
Jahanaray et al. (2023) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: