P2–308: Common rater errors when scoring the ADAS‐Cog
Bibliographic record
Abstract
There is a growing concern that rater administration and scoring errors on the Alzheimer's Disease Assessment Scale-Cognitive subscale (ADAS-Cog) may be masking positive drug effect in some Alzheimer disease trials (Kobak, 2010; Schafer et al., 2011). Identification of key problem areas for raters may inform development of training programs so that raters can more skillfully perform the ADAS-Cog and thus minimize the risk of masking positive drug effects. This study examined rater performance on an ADAS-Cog training exercise and identified common rater errors when scoring the ADAS-Cog. A total of 326 raters from 4 AD clinical trials were selected to participate in a standard ADAS-Cog training program based on their reported educational status and experience with the measure. This training program included a 1 hour didactic session that reviewed the standard ADAS-Cog administration and scoring guidelines. The didactic session was followed by an evaluation exercise where raters were required to score a video-recorded ADAS-Cog administration. Raters' scores were compared to the criterion scores established for the video-recorded administration. Preliminary analyses indicated that 95% of the raters made at least one error rating the video-recorded ADAS-Cog; 17% (1 error) 29% (2 errors), 24% (3 errors), 12.5% (4 errors), and 12.5% (>5 errors). On average the group made 2.7 errors in scoring the video-recorded administration, which impacted the ADAS-Cog total score by ±1.2 points. Errors were identified in all ADAS-Cog subtests, with more frequent errors made on Remembering Test Instructions, Word Finding Difficulty, Comprehension, and Spoken Language Ability. Although raters were experienced administering the ADAS-Cog and were provided with training, most made at least 1 error on the ADAS-Cog training exercise. The errors committed by raters impacted the ADAS-Cog total score to varying degrees, which reaffirms the concern that rater errors may be masking true trial outcomes. Raters had greater difficulty rating items that involved clinical judgment, such as the language-evaluation items of the ADAS-Cog. Raters are likely to benefit from more substantial training and ongoing monitoring programs for these items.
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 | 0.100 | 0.231 |
| Meta-epidemiology (narrow) | 0.001 | 0.001 |
| Meta-epidemiology (broad) | 0.001 | 0.001 |
| Bibliometrics | 0.002 | 0.002 |
| Science and technology studies | 0.001 | 0.001 |
| Scholarly communication | 0.001 | 0.001 |
| Open science | 0.001 | 0.002 |
| Research integrity | 0.001 | 0.001 |
| Insufficient payload (model declined to judge) | 0.002 | 0.002 |
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".