Suspect filler similarity in eyewitness lineups: A literature review and a novel methodology.
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
Eyewitness lineups typically contain a suspect (guilty or innocent) and fillers (known innocents). The degree to which fillers should resemble the suspect is a complex issue that has yet to be resolved. Previously, researchers have voiced concern that eyewitnesses would be unable to identify their target from a lineup containing highly similar fillers; however, our literature review suggests highly similar fillers have only rarely been shown to have this effect. To further examine the effect of highly similar fillers on lineup responses, we used morphing software to create fillers of moderately high and very high similarity to the suspect. When the culprit was in the lineup, a higher correct identification rate was observed in moderately high similarity lineups than in very high similarity lineups. When the culprit was absent, similarity did not yield a significant effect on innocent suspect misidentification rates. However, the correct rejection rate in the moderately high similarity lineup was 20% higher than in the very high similarity lineup. When choosing rates were controlled by calculating identification probabilities for only those who made a selection from the lineup, culprit identification rates as well as innocent suspect misidentification rates were significantly higher in the moderately high similarity lineup than in the very high similarity lineup. Thus, very high similarity fillers yielded costs and benefits. Although our research suggests that selecting the most similar fillers available may adversely affect correct identification rates, we recommend additional research using fillers obtained from police databases to corroborate our findings.
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 | 0.000 |
| Meta-epidemiology (narrow) | 0.000 | 0.000 |
| Meta-epidemiology (broad) | 0.002 | 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 | 0.001 |
| 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