Enhancing Consistency of Maximal Responding in Behavior Description Interviews: An Exploration of Priming and Response Length
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
In a Behavior Description Interview (BDI), candidates are asked to describe past experiences that demonstrate skills and abilities important for the position (Janz, 1982). A recent study by Huffcutt et al. (2020) found that only around half of participants (48.1 percent) describe an experience reflecting maximal performance capability. Random mixing of maximal capability with day-to-day typical performance tendencies is problematic psychometrically because candidates are not all providing comparable information and top candidates could be overlooked. Given notable methodological concerns with Huffcutt et al.’s approach, our first purpose was to provide empirical confirmation that maximal responding in BDIs is, in fact, inconsistent. Our estimate of the proportion of maximal responding was even lower (41.3 percent), further amplifying concerns when assessment of maximal performance capability is desired (e.g., for many professional positions). The second purpose was to investigate two factors that could increase the consistency of maximal responding: rewording the main BDI question to focus directly on absolute top-end experiences (i.e., priming) and longer response length. Both were found to have significant effects. A number of directions for future research were identified, which, along with these findings, could help researchers move closer to the long-term goal of uniform description of experiences that reflect each candidate’s maximal capability (or typical tendencies if so desired).
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 | 0.002 | 0.000 |
| Meta-epidemiology (narrow) | 0.000 | 0.000 |
| Meta-epidemiology (broad) | 0.000 | 0.000 |
| Bibliometrics | 0.001 | 0.000 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.000 | 0.001 |
| Open science | 0.000 | 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