Study Design Selection Task
Core summary
Choosing the right study design depends on your research question, available resources, ethical constraints, and the level of evidence needed. This lesson integrates everything from Level 4 into a practical decision framework for selecting the most appropriate design.
Detailed explanation
Detailed explanation
The study design decision begins with your question type. Treatment/intervention questions → RCT is the gold standard (if ethical and feasible); quasi-experimental if randomization is impossible. Causation/risk factor questions → Cohort (prospective or retrospective) or case-control, depending on outcome rarity and resources. Prevalence/burden questions → Cross-sectional study. Diagnostic test questions → Diagnostic accuracy study (STARD). Prognosis questions → Cohort study with adequate follow-up. Experience/meaning questions → Qualitative study. Evidence synthesis → Systematic review with or without meta-analysis. Evidence mapping → Scoping review. Process improvement → QI study with PDSA cycles. Practical constraints include: budget (RCTs are expensive), time (prospective studies take years), sample size (rare diseases may require case-control), ethics (cannot randomize to harmful exposures), and existing data availability (registry and database studies). The hierarchy of evidence is a guide, not a rule — the best design is the one that most rigorously answers your specific question within your constraints.
Clinical example
Your research question: 'Does a new physiotherapy protocol reduce falls in nursing home residents?' Ideal design: cluster RCT (randomize nursing homes, not individuals, to avoid contamination). Budget constraint: only 4 nursing homes available. Adjusted design: before-and-after study with interrupted time series in two homes, with the other two as concurrent controls — a pragmatic quasi-experimental approach.
Research example
The UK's National Institute for Health and Care Research (NIHR) recommends that researchers justify their design choice in grant applications — explaining why they selected a particular design over alternatives and how it addresses the question within practical constraints.
Knowledge check
Q1. You want to study whether a rare environmental exposure causes a rare cancer. What design is most appropriate?
Q2. When is a quasi-experimental design preferred over an RCT?
Q3. You want to understand the breadth of research on AI in radiology before planning a systematic review. What should you do first?