External Validity Bias of Purposive and Random Site Selection When Sites Can Opt Out: Evidence from the Head Start Program
Bibliographic record
Abstract
Randomized controlled trials (RCTs) produce impact evidence with high internal validity but uncertain external validity. External validity bias has been defined as the expected difference between the average impact in the sample and the average impact in the population. This study estimated the external validity bias from several site selection methods by simulating hypothetical RCTs of the Head Start program in which some selected sites decline to participate. Three main findings emerged from the analysis. First, purposive site selection consistently produced biased impact estimates that varied in magnitude based on the outcome examined, the factors used to select sites, and the factors that influenced site decisions to participate. Second, simple random site selection yielded less external validity bias than purposive site selection under most tested conditions. Third, stratified random site selection yielded virtually no external validity bias, but the results likely overstate the method’s performance when data on impact moderators are unavailable before sites are selected. These findings offer lessons on how to select sites in future RCTs to minimize external validity bias.
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 | 0.029 | 0.029 |
| Meta-epidemiology (narrow) | 0.000 | 0.000 |
| Meta-epidemiology (broad) | 0.001 | 0.000 |
| Bibliometrics | 0.001 | 0.001 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.001 | 0.000 |
| Open science | 0.002 | 0.001 |
| Research integrity | 0.000 | 0.002 |
| 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 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".