PAPER: Examining Position Effects in Large-Scale Assessments Using an SEM Approach
Why this work is in the frame
A frame that forgets how it found something cannot be audited. These are the routes that admitted this work.
Bibliographic record
Abstract
Introduction Item position effects have been an important concern in educational and psychological measurement. This type of effect may occur in both paper-pencil tests and computer-based assessments when examinees receive the same test items at different positions. If there is a significant item position effect for the items, this may lead to an unfair assessment. Objectives Previously, item position effects were often examined using Hierarchical Generalized Linear Mixed Model (HGLM), which is computationally burdensome for large data, and is mathematically limited to Rasch model. In this study, we aim to introduce a Structural Equation Modeling (SEM) approach to overcome the disadvantages of HGLM, and to demonstrate the proposed method using data from an operational large-scale reading assessment. Methodology The data come from a statewide reading assessment in the US. The sample consisted of 11734 3 rd grade students who responded to 45 reading items related to 7 reading passages. Because the test was given in computers, item positions were scrambled across students, which resulted in four different patterns of item positions (referred to as test forms in this study). We examined the overall form effects, passage position effects, and item position effects using the SEM approach. Results The results showed that one of the 7 passages and 10 of the 45 items showed significant position effects in the test, although there was no overall form effect detected across the four forms. All SEM models converged in less than 2 minutes, whereas their HGLM counterparts either took more than 25 minutes or failed to converge. Conclusion This study contributes to the literature by introducing a flexible SEM approach to estimate position effects. Compared to HGLM, the SEM approach is computationally more efficient, easier to interpret, and allows the examination of position effects not only for Rasch model but also 2PL model.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
Full frame distilled prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. Learned from the 10,348 direct Codex labels and 10,348 direct Gemma labels. Candidate is the union of thresholded teacher heads; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels or direct frontier model labels.
Codex and Gemma teacher scores by category
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.006 | 0.009 |
| Meta-epidemiology (narrow) | 0.000 | 0.000 |
| Meta-epidemiology (broad) | 0.000 | 0.000 |
| Bibliometrics | 0.000 | 0.001 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.000 | 0.001 |
| Open science | 0.001 | 0.000 |
| Research integrity | 0.000 | 0.000 |
| Insufficient payload (model declined to judge) | 0.000 | 0.000 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from it