Home/Insights/Methodology reference/Types of scientific research
Types of scientific research
The useful classification is not basic versus applied. It is what kind of claim the design can support.
By purpose
Basic research seeks understanding without a specified application. Applied research addresses a defined problem. Translational work moves findings between the two. The distinction describes intent and says nothing about rigour.
By the claim the design supports
Descriptive work characterises: what exists, how often, in whom. It supports prevalence and distribution claims and no causal claim at all.
Correlational work establishes that two variables move together. It supports association and prediction, not causation.
Experimental work manipulates an exposure under the investigator’s control, with randomisation and a comparison group. It is the only design that supports a causal claim without strong additional assumptions.
Quasi-experimental work has an intervention but not random assignment. Causal claims are possible but rest on assumptions that have to be stated and defended.
This classification matters because the discussion section of a paper must not claim more than the design can carry. Most over-claiming happens here.
By data type
Quantitative work measures and counts; qualitative work interprets meaning and process; mixed methods integrates both deliberately rather than simply running them alongside each other.
Qualitative work is not "preliminary" quantitative work. It answers different questions — how and why rather than how many — and is judged by different criteria.
In practice
Fix, flag or query.
The distinction that keeps an edit defensible: what an editor may change silently, what must be shown to the author, and what only the author can answer.
- The frequent error: a correlational design with a causal discussion.
- Match: the strength of the verb in your conclusion to the design that produced it.
- Check: whether a comparison group exists before using comparative language.
Reporting standards
- EQUATOR Network, on design-specific reporting
- STROBE Statement, on observational study types
Reporting guidelines named here are hosted by the EQUATOR Network. Writing to the checklist for your design before submission is the cheapest available improvement to a methods section.
Related