Detecting Atypical Assessment Perceptions via Rasch Person Fit
Location
Hamilton A
Proposal Track
Research Project
Session Format
Presentation
Abstract
The primary curricular aim of a classroom assessment course is to train educators to develop a balanced assessment system that facilitates student learning. Since the age of high-stakes accountability, student motivation and readiness to participate has been clouded by misinformation and the unintended consequences of policy. An understanding of students’ beliefs facilitate dialogue that can help instructors address student misconceptions head-on, and subsequently guide students to an appreciation that sound classroom assessment practice is paramount.
To make accurate inferences about student perceptions and readiness, instructors need an avenue to separate atypical student response from typical response. Such an analysis requires a psychometric model that separates items and persons and detects misfit for each. This study used Rasch modeling to calibrate The Terse Self-Test on Testing (Popham, 2013). The model placed items and persons onto the same scale. Subsequently, variable mapping permitted a thorough understanding of the match between examinees and the ease of endorsement of each item. Person misfit indices were examined to detect and evaluate each atypical response pattern (Rupp, 2013). We illustrate how the procedure facilitated an appreciation of perceptions and readiness that directed course conversations. We recommend using person fit to provide a deep understanding of survey response.
Keywords
Assessment Perception, Rasch model, Person fit
Recommended Citation
Ferster, Amanda E.; Walker, A Adrienne; and Purcell, Sam, "Detecting Atypical Assessment Perceptions via Rasch Person Fit" (2017). Georgia Educational Research Association Conference. 18.
https://digitalcommons.georgiasouthern.edu/gera/2017/2017/18
Detecting Atypical Assessment Perceptions via Rasch Person Fit
Hamilton A
The primary curricular aim of a classroom assessment course is to train educators to develop a balanced assessment system that facilitates student learning. Since the age of high-stakes accountability, student motivation and readiness to participate has been clouded by misinformation and the unintended consequences of policy. An understanding of students’ beliefs facilitate dialogue that can help instructors address student misconceptions head-on, and subsequently guide students to an appreciation that sound classroom assessment practice is paramount.
To make accurate inferences about student perceptions and readiness, instructors need an avenue to separate atypical student response from typical response. Such an analysis requires a psychometric model that separates items and persons and detects misfit for each. This study used Rasch modeling to calibrate The Terse Self-Test on Testing (Popham, 2013). The model placed items and persons onto the same scale. Subsequently, variable mapping permitted a thorough understanding of the match between examinees and the ease of endorsement of each item. Person misfit indices were examined to detect and evaluate each atypical response pattern (Rupp, 2013). We illustrate how the procedure facilitated an appreciation of perceptions and readiness that directed course conversations. We recommend using person fit to provide a deep understanding of survey response.
