Performance of methods for analyzing continuous data from stratified cluster randomized trials – A simulation study
Bibliographic record
Abstract
Background: The adoption of cluster randomized trials (CRTs) with the stratified design is currently gaining widespread interest. In the stratified design, clusters are first grouped into two or more strata and then randomized into treatment groups within each stratum. In this study, we evaluated the performance of several commonly used methods for analyzing continuous data from stratified CRTs. Methods: This is a simulation study where we compared four methods: mixed-effects, generalized estimating equation (GEE), cluster-level (CL) linear regression and meta-regression methods to analyze the continuous data from stratified CRTs using a simulation study with varying numbers of clusters, cluster sizes, intra-cluster correlation coefficients (ICCs) and effect sizes. This study was based on a stratified CRT with one stratification variable with two strata. The performance of the methods was evaluated in terms of the type I error rate, empirical power, root mean square error (RMSE), and width and coverage of the 95% confidence interval (CI). Results: GEE and meta-regression methods had high type I error rates, higher than 10%, for the small number of clusters. All methods had similar accuracy, measured through RMSE, except meta-regression. Similarly, all methods but meta-regression had similar widths of 95% CIs for the small number of clusters. For the same sample size, the empirical power for all methods decreased as the value of the ICC increased. Conclusion: In this study, we evaluated the performance of several methods for analyzing continuous data from stratified CRTs. Meta-regression was the least efficient method compared to other methods.
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 | 0.260 | 0.485 |
| Meta-epidemiology (narrow) | 0.003 | 0.001 |
| Meta-epidemiology (broad) | 0.004 | 0.010 |
| Bibliometrics | 0.003 | 0.003 |
| Science and technology studies | 0.001 | 0.002 |
| Scholarly communication | 0.003 | 0.003 |
| Open science | 0.003 | 0.002 |
| Research integrity | 0.003 | 0.004 |
| 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; the direct Gemma label and the distilled Codex classifier 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".