POSTER: Detecting Differential Item Functioning: A Comparison of Two Effect Size Measures in Logistic Regression Analysis
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
The purpose of the current project was to compare two different effect size measures used in assessment of differential item functioning (DIF). Specifically, in the context of logistic regression analysis, which is a common methodology for assessing DIF, we compared the DIF decisions based on the use of ∆R 2 effect size measure and decisions based on use of log odds ratios ( ∆ LR ). This problem was addressed within a study concerned with DIF of a delinquency scale commonly used in antisocial/delinquent behavior research. We examined the data collected from 3290 students in the city of Toronto (Canadian portion of international study concerned with behaviour and misbehaviour of students in grades 7 to 9; the International Self-report Delinquency Study, Enzmann et al., 2010). We evaluated DIF of the utilized delinquency scale in relation to four grouping variables relevant for delinquent behaviour: gender, age, socio-economic status, and neighbourhood context (i.e., crime in neighbourhood). According to the results, conclusions about DIF were related to the choice of effect size measure; that is, different conclusions resulted from the use of the different effect size measures. Our results, obtained by utilizing real data, were in line with the recent simulation studies that pointed towards low power of the ∆R 2 effect size measure in detecting DIF (Hidalgo & Lopez-Pina, 2004; Hidalgo et al., 2014). The results emphasize a need for further examination of the logistic regression effect sizes and their optimal cutoffs in regard to DIF. As DIF is of importance in development of psychological tests and measures as well as in interpretations/conclusions based on tests and measures, we discuss the results in the context of methodological choices in detecting DIF and practical consequences of such choices.
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 | 0.001 | 0.001 |
| Meta-epidemiology (narrow) | 0.000 | 0.000 |
| Meta-epidemiology (broad) | 0.001 | 0.000 |
| Bibliometrics | 0.000 | 0.000 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.000 | 0.000 |
| Open science | 0.000 | 0.000 |
| Research integrity | 0.000 | 0.000 |
| Insufficient payload (model declined to judge) | 0.003 | 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