Learn / Epidemiology / Study designs
The landscape of designs, ecological and cross-sectional studies
13 minute read · 6-question quiz
What you will learn
Classify studies as descriptive or analytic, observational or experimental
Describe ecological and cross-sectional studies, with their uses and weaknesses
Calculate and interpret a prevalence ratio from a cross-sectional survey
A study design is the plan for collecting and comparing data so that a question can be answered with as little bias as possible. The right design depends on the question, the frequency of the disease and the exposure, time and money, and ethics.
Ecological studies
An ecological study compares groups (countries, districts, years) using aggregate data on exposure and outcome, with no individual-level information.
Example: a plot of national salt consumption against stroke mortality for 30 countries.
- Strengths: quick and cheap using existing data; good for hypotheses and for exposures that vary mainly between populations (air pollution, policy).
- Weakness: the ecological fallacy. An association between groups does not necessarily hold for individuals. Districts with more bed-net coverage may have fewer malaria deaths because of something else common to those districts.
- Confounding is hard to control, and misclassification of the group exposure is common.
Cross-sectional studies (surveys)
A cross-sectional study measures exposure and outcome in a defined population at one point in time. The Demographic and Health Surveys, STEPS surveys of risk factors and HIV population-based impact assessments are examples.
- Measures: prevalence of the outcome, and association measured by the prevalence ratio or odds ratio.
- Strengths: one visit; gives the burden; can assess many exposures and outcomes at once; no loss to follow-up.
- Weaknesses: it cannot tell which came first (temporality): does low income cause depression or the reverse? It captures prevalent (existing, usually longer-lasting) cases, which can bias toward long-duration disease (prevalence-incidence bias). Not suited to rare diseases or short-lived conditions.
Where a cross-sectional study fits in a research career
Because it is quick, a cross-sectional survey is often a first study in a new setting and a common design for MPH theses. Do it well: a clearly defined population, a probability sample (chapter 3 of biostatistics), a validated questionnaire or measurement, a response rate above about 70%, sample size calculated for the main prevalence, and analysis that accounts for survey design (weights, clustering). Report it with the STROBE checklist (cross-sectional version).
Summary
Designs are experimental or observational, descriptive or analytic.
Ecological studies use group-level data and risk the ecological fallacy.
Cross-sectional studies measure prevalence and associations at one time, with no clear temporality.
The prevalence ratio compares prevalence between exposed and unexposed.
Cross-sectional surveys favour long-lasting cases (prevalence-incidence bias).
Check your understanding
Draft study guide. Spotted an error or something unclear? Tell us at hello@scintria.com. This is educational material, not clinical or regulatory advice.