Research can be grouped by purpose, by method and by time. Knowing the type tells you what claims you can make.
Why it matters
A cross-sectional survey cannot prove that something causes something else, but an experiment can come much closer.
Key ideas
- Descriptive: describes what is happening
- Analytical: compares or explains
- Observational: you watch, you do not assign
- Experimental: you assign the exposure (e.g. randomised trial)
- Quantitative uses numbers, qualitative uses words and meanings
- Cross-sectional = one time point, cohort = follow over time
Real-life research example
A survey of smoking among university students is descriptive and cross-sectional. Following 1,000 smokers and 1,000 non-smokers for 10 years is an observational cohort study.
When to use
- Choose the type that matches your question and your resources
When NOT to use
- Don't use causal language ("reduces", "causes") for a simple cross-sectional survey
Quick takeaway
The study type limits the strength of your conclusion.
Try it yourself
A teacher gives one class a new method and compares exam scores with another class that she did not choose by chance. Name the study type and one limitation.
Quasi-experimental (non-random assignment). Limitation: the classes may differ in other ways (confounding), so a score difference is not clearly due to the method.