Research Methodology · Short notes

Study Design: Short Notes & Cheat Sheet

Observational and experimental designs, the hierarchy of evidence, randomisation and blinding, bias and confounding, causation criteria and research ethics.

Attempt Study Design MCQs

Short notes

Read these before you attempt the topic's MCQs.

The hierarchy of evidence

Designs further right control bias better. A well-done cohort study can still beat a badly done RCT.

Designs

Case reports and case series describe individual patients and generate hypotheses.
Cross-sectional studies measure exposure and outcome at one time point and estimate prevalence.
Case-control studies start with the outcome and look back for exposure. They suit rare diseases, give odds ratios and are prone to recall bias.
Cohort studies start with exposure and follow people forward (or use past records). They give incidence and relative risk and suit rare exposures, but suffer loss to follow-up.
Randomised controlled trials randomly allocate treatment, which balances known and unknown confounders.
Systematic reviews and meta-analyses combine several studies and sit at the top of the evidence hierarchy.

Trial methods and bias

Blinding can be single (participants), double (participants and investigators) or triple (including analysts). Allocation concealment hides the next assignment.
Intention-to-treat analysis keeps everyone in their randomised group, preserving the benefit of randomisation.
Selection bias affects who enters or stays in a study. Information bias includes recall and observer bias.
A confounder is linked to both exposure and outcome. It is controlled by randomisation, restriction, matching, stratification or multivariable analysis.
Internal validity is whether results are correct for the study; external validity is whether they apply to other people.

Causation and ethics

Bradford Hill criteria: strength, consistency, specificity, temporality, dose–response, plausibility, coherence, experiment and analogy. Temporality (cause before effect) is the only essential one.
Ethics milestones: the Nuremberg Code (1947), the Declaration of Helsinki (1964) and the Belmont Report (respect for persons, beneficence, justice).
All studies with humans need ethics committee approval and informed consent.
Reporting guidelines: CONSORT (trials), STROBE (observational studies), PRISMA (systematic reviews).

Cheat sheet

High-yield facts in one place. Cover the right-hand column and test yourself.

Design comparison

Design	Starts with · measure
Cross-sectional	Snapshot · prevalence
Case-control	Outcome · odds ratio
Cohort	Exposure · relative risk, incidence
RCT	Random allocation · effect of treatment
Meta-analysis	Many studies · pooled estimate

Reporting guidelines

Guideline	For
CONSORT	Randomised trials
STROBE	Observational studies
PRISMA	Systematic reviews, meta-analyses
STARD	Diagnostic accuracy studies

Memory aids

Case-control looks back at Causes; Cohort goes forward: Case-control starts from the outcome; cohort starts from the exposure.

Exam traps

Case-control studies cannot measure incidence or relative risk directly.
Randomisation controls unknown confounders; matching controls only the matched factors.
Temporality is the only Bradford Hill criterion that must always be met.
Intention-to-treat analysis includes people who did not take their assigned treatment.

Revised? Now test yourself.

Attempt the Study Design MCQs while the notes are fresh. Anything you get wrong is saved to your incorrect list for the next revision round.