Internal Validity vs External Validity
Internal validity is how well a study's results reflect the truth for the people studied; external validity (generalisability) is how well they apply beyond those participants and setting. A study needs internal validity first, because a biased result cannot be meaningfully generalised.
Threats
- Internal validity: selection bias, confounding, measurement error, attrition, performance and detection bias, and flawed analysis.
- External validity: narrow eligibility criteria, unusual settings, highly selected participants, short follow-up, surrogate outcomes, and differences between trial conditions and real practice.
The trade-off
Tight control (strict eligibility, closely supervised treatment) improves internal validity but may limit generalisability. Pragmatic studies that reflect usual practice improve external validity but may be harder to keep free of bias. Good studies state their aims and limit their claims accordingly.
Judging each when you read
| Question | Validity type |
|---|---|
| Were groups comparable and outcomes measured consistently? | Internal |
| Was loss to follow-up low and handled properly? | Internal |
| Were participants similar to my patients? | External |
| Is the intervention available and deliverable in my setting? | External |
| Do outcomes matter to patients? | External |
Illustrative example
Invented for teaching. A trial of a blood-pressure drug enrols healthy adults aged 40–60 without comorbidity, with intensive follow-up. The result may be internally valid but not directly applicable to frail 85-year-olds with kidney disease. A clinician should consider whether benefits and harms might differ.
Pragmatic and explanatory trials
Trials sit on a spectrum. Explanatory trials test whether an intervention can work under ideal conditions, so they prioritise internal validity. Pragmatic trials test whether it works in routine practice, so they include broader participants and flexible delivery. When you read a trial, ask where on that spectrum it sits, because the answer shapes how you apply the result.
Common pitfalls
- Assuming results apply to every patient.
- Dismissing a study because it is "not generalisable" without checking whether the mechanism is likely to differ.
Key takeaways
- Internal validity comes first; external validity follows.
- Compare participants and setting with your own.
References and further reading
- Hulley SB, Cummings SR, Browner WS, Grady D, Newman TB. Designing Clinical Research. 4th ed. Lippincott Williams & Wilkins; 2013. Link
- Schulz KF, Altman DG, Moher D; CONSORT Group. CONSORT 2010 statement: updated guidelines for reporting parallel group randomised trials. BMJ. 2010;340:c332. doi:10.1136/bmj.c332
- Higgins JPT, Thomas J, Chandler J, et al., editors. Cochrane Handbook for Systematic Reviews of Interventions. Cochrane. Link
Related reading
This page is general educational information about research methods. It is not medical, legal or regulatory advice; follow your institution's ethics board and your target journal's instructions.