Relatively speaking: contrast effects influence assessors’ scores and narrative feedback
Bibliographic record
Abstract
CONTEXT: In prior research, the scores assessors assign can be biased away from the standard of preceding performances (i.e. 'contrast effects' occur). OBJECTIVES: This study examines the mechanism and robustness of these findings to advance understanding of assessor cognition. We test the influence of the immediately preceding performance relative to that of a series of prior performances. Further, we examine whether assessors' narrative comments are similarly influenced by contrast effects. METHODS: Clinicians (n = 61) were randomised to three groups in a blinded, Internet-based experiment. Participants viewed identical videos of good, borderline and poor performances by first-year doctors in varied orders. They provided scores and written feedback after each video. Narrative comments were blindly content-analysed to generate measures of valence and content. Variability of narrative comments and scores was compared between groups. RESULTS: Comparisons indicated contrast effects after a single performance. When a good performance was preceded by a poor performance, ratings were higher (mean 5.01, 95% confidence interval [CI] 4.79-5.24) than when observation of the good performance was unbiased (mean 4.36, 95% CI 4.14-4.60; p < 0.05, d = 1.3). Similarly, borderline performance was rated lower when preceded by good performance (mean 2.96, 95% CI 2.56-3.37) than when viewed without preceding bias (mean 3.55, 95% CI 3.17-3.92; p < 0.05, d = 0.7). The series of ratings participants assigned suggested that the magnitude of contrast effects is determined by an averaging of recent experiences. The valence (but not content) of narrative comments showed contrast effects similar to those found in numerical scores. CONCLUSIONS: These findings are consistent with research from behavioural economics and psychology that suggests judgement tends to be relative in nature. Observing that the valence of narrative comments is similarly influenced suggests these effects represent more than difficulty in translating impressions into a number. The extent to which such factors impact upon assessment in practice remains to be determined as the influence is likely to depend on context.
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.
How this classification was reachedexpand
Full frame machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.035 | 0.320 |
| Meta-epidemiology (narrow) | 0.000 | 0.000 |
| Meta-epidemiology (broad) | 0.001 | 0.001 |
| Bibliometrics | 0.001 | 0.000 |
| Science and technology studies | 0.001 | 0.002 |
| Scholarly communication | 0.002 | 0.002 |
| Open science | 0.001 | 0.002 |
| Research integrity | 0.001 | 0.001 |
| Insufficient payload (model declined to judge) | 0.005 | 0.001 |
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 itClassification
machine, unvalidatedMachine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".