Internal and External Validity

Validity is the ultimate measure of study quality. A study can produce a result, but if the result is not valid, it is meaningless — or worse, misleading. Internal validity refers to whether the observed effect is truly due to the intervention/exposure and not to bias, confounding, or chance. A study with high internal validity gives you confidence that the results accurately reflect what happened in the study population. Threats to internal validity include selection bias, information bias, confounding, attrition, and chance. External validity (generalizability) refers to whether the study results can be applied to other populations, settings, or conditions beyond the study itself. A study with high external validity produces results that are broadly applicable. Here is the tension: studies designed for maximum internal validity often sacrifice external validity, and vice versa. A tightly controlled RCT with strict eligibility criteria (only patients aged 40-65, no comorbidities, one specific drug dose) has high internal validity but may not generalize to real-world patients who are older, sicker, and on multiple medications. Conversely, a large observational study in a diverse real-world population has broad generalizability but may suffer from confounding and bias — lower internal validity. The pragmatic trial movement attempts to bridge this gap by designing RCTs with broad eligibility criteria, real-world settings, and clinically relevant outcomes — prioritizing external validity while maintaining reasonable internal validity. As a reader, always consider both: 'Is this result true?' (internal validity) AND 'Does it apply to my patients?' (external validity).