Siti Qomariyah
Universitas Islam Negeri Sunan Kudus, Indonesia

Published : 1 Documents Claim Missing Document
Claim Missing Document
Check
Articles

Found 1 Documents
Search

Psychometric evaluation of a statistical literacy instrument for pre-service mathematics teachers: Rasch analysis using R Putri Nur Malasari; Siti Qomariyah; Savira Najwa Meilani
Journal of Advanced Sciences and Mathematics Education Vol. 6 No. 3 (2026): Journal of Advanced Sciences and Mathematics Education
Publisher : CV. FOUNDAE

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.58524/jasme.v6i3.1240

Abstract

Background: Statistical literacy is an essential competency for pre-service mathematics teachers, characterized by persistent challenges in interpreting and critically evaluating statistical information. Despite the development of several statistical literacy instruments, comprehensive Rasch-based psychometric evidence for instruments targeting pre-service mathematics teachers remains limited. Purpose: This study aimed to evaluate the psychometric quality of a statistical literacy instrument for pre-service mathematics teachers using the Rasch measurement model implemented in R. The evaluation focused on item functioning, reliability, validity, and overall measurement characteristics. Methods: Data were collected from 132 pre-service mathematics teachers enrolled in an education program. Rasch analysis was conducted to examine item fit and difficulty, reliability, dimensionality, local independence, and person-item targeting. Results: The entire 15 items met the Rasch fit criteria (Infit and Outfit MNSQ = 0.5–1.5), thereby supporting the construct validity and internal consistency, with no misfitting items identified. Most items ranged from easy to moderate (−1.47 to +0.35 logits), with only a particular difficult item. Wright Map analysis showed that respondents' abilities generally exceeded the highest item difficulty, suggesting suboptimal targeting and reduced measurement precision for high-ability respondents. Conclusions: The results showed a mild ceiling effect, with 40.2% of respondents exceeding the hardest item and 3.0% achieving perfect scores. Generally, the instrument exhibited adequate psychometric quality but required more difficult items to improve measurement precision for high-ability respondents. The script-based Rasch workflow in R also enhanced the transparency and reproducibility of psychometric evaluation.