🧪 This platform is in early beta. Features may change and you might encounter bugs. We appreciate your patience!
exclusion_bias
Exclusion bias arises when the criteria used to select or filter study participants systematically remove individuals whose inclusion would change the study's results. This can happen through overly strict eligibility criteria, loss to follow-up, or removal of outliers. The remaining sample no longer represents the target population.
A clinical trial for an antidepressant excludes patients with suicidal ideation. The drug appears effective and safe, but its performance in the most severe cases — the very patients most likely to need it — remains unknown.
A vaccine efficacy trial excludes participants with autoimmune conditions. The vaccine shows strong immune responses across the study group, but the findings cannot be applied to millions of immunocompromised individuals who were never studied and may respond very differently.
A workplace productivity app study excludes employees who dropped out in the first week due to technical difficulties. The remaining users show impressive engagement metrics, but the early dropouts — likely those who struggled most with the technology — represent exactly the population the app was designed to help.
Binary (yes/no) questions an LLM must answer to identify this aspect:
Were certain participants or data points systematically excluded from the study?
Type: binaryCould the exclusion criteria disproportionately remove subjects with specific characteristics?
Type: binaryDoes the excluded group differ in meaningful ways from the included group?
Type: binaryAre the study's conclusions generalized without acknowledging the impact of exclusions?
Type: binaryExclusion bias arises when the criteria used to select or filter study participants systematically remove individuals whose inclusion would change the study's results. This can happen through overly strict eligibility criteria, loss to follow-up, or removal of outliers. The remaining sample no longer represents the target population.
Exclusion criteria often seem reasonable on the surface (safety, data quality), making it easy to overlook how they shape results. Readers rarely scrutinize who was excluded and why, focusing instead on the reported outcomes.
Scrutinize inclusion and exclusion criteria carefully. Ask who is missing from the study and how their absence might affect results. Look for intention-to-treat analyses that include all originally enrolled participants.
Drug trials routinely exclude elderly patients, pregnant women, and people with comorbidities. When these drugs are approved and prescribed broadly, their real-world effectiveness and safety may differ substantially from trial results.
Systematic difference between respondents and non-respondents distorting study results.
Occupational studies overestimate worker health because severely ill people exit the workforce.
How participants are identified or recruited systematically distorts the sample.
The statistical error of drawing conclusions from a dataset that has been filtered by a survival or success criterion, without accounting for the filtered-out cases. The surviving sample is systematically different from the full population, and conclusions drawn from it are biased.
Diagnostic test accuracy varies when evaluated across different disease severity levels.
Prevalence studies miss fatal or short-duration cases, distorting disease-exposure associations.
Participants who choose to join a study differ systematically from those who do not.
Use these tools to detect, analyze, or train this aspect.