Ongoing monitoring of data clustering in multicenter studies
Bibliographic record
Abstract
BACKGROUND: Multicenter study designs have several advantages, but the possibility of non-random measurement error resulting from procedural differences between the centers is a special concern. While it is possible to address and correct for some measurement error through statistical analysis, proactive data monitoring is essential to ensure high-quality data collection. METHODS: In this article, we describe quality assurance efforts aimed at reducing the effect of measurement error in a recent follow-up of a large cluster-randomized controlled trial through periodic evaluation of intraclass correlation coefficients (ICCs) for continuous measurements. An ICC of 0 indicates the variance in the data is not due to variation between the centers, and thus the data are not clustered by center. RESULTS: Through our review of early data downloads, we identified several outcomes (including sitting height, waist circumference, and systolic blood pressure) with higher than expected ICC values. Further investigation revealed variations in the procedures used by pediatricians to measure these outcomes. We addressed these procedural inconsistencies through written clarification of the protocol and refresher training workshops with the pediatricians. Further data monitoring at subsequent downloads showed that these efforts had a beneficial effect on data quality (sitting height ICC decreased from 0.92 to 0.03, waist circumference from 0.10 to 0.07, and systolic blood pressure from 0.16 to 0.12). CONCLUSIONS: We describe a simple but formal mechanism for identifying ongoing problems during data collection. The calculation of the ICC can easily be programmed and the mechanism has wide applicability, not just to cluster randomized controlled trials but to any study with multiple centers or with multiple observers.
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 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 | 0.801 | 0.884 |
| Meta-epidemiology (narrow) | 0.000 | 0.000 |
| Meta-epidemiology (broad) | 0.003 | 0.000 |
| Bibliometrics | 0.001 | 0.002 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.000 | 0.000 |
| Open science | 0.005 | 0.004 |
| Research integrity | 0.000 | 0.001 |
| Insufficient payload (model declined to judge) | 0.004 | 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; both teacher heads agree on what is shown here.
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".