Written and maintained by CASRAI Editorial Board
Last updated
Who this is for: researchers and students choosing or describing a study design. After reading you should be able to place any design on the three axes that sources use (purpose, data type, observational versus experimental), tell a design from a method, and know what a research design plan has to state.
What research design is, and how it differs from a research method
What is the difference between research design and research method?
Research design is the overall strategy and analytical approach of a study; a research method is a specific technique used inside that strategy. The USC Libraries guide to types of research designs defines the design as “the overall strategy and analytical approach,” and says that “the research problem determines the type of design you choose, not the other way around.” The same guide notes that mixed methods “represents more of an approach to examining a research problem than a methodology.” In practice a design (for example a cohort study or a true experiment) answers how the study is organised to answer the question, while methods (a survey, an interview, a lab assay) answer how data are collected. See also the educational research methods guide for method-level detail.
Categories of research design: three ways the sources sort designs
“Research design categories”, “kinds of research design” and “research design methodologies” all point to the same family of classifications. Libraries and clinical-methods papers do not use one single scheme, so the clearest way to read them is as three overlapping axes. Classifications vary between sources, and none of the three sources read for this page gives a fixed number of “types”.
Axis 1: by data type (quantitative, qualitative, mixed)
The University of Delaware Library research methods guide describes quantitative research as “expressed in numbers and graphs” and used “to test or confirm,” and qualitative research as “expressed in words” and used “to understand.” It names mixed methods as a third option, and names experimental, correlational and descriptive designs.
Axis 2: by purpose (exploratory, descriptive, explanatory)
The USC Libraries guide lists 18 designs: action research, case study, causal, cohort, cross-sectional, descriptive, experimental, exploratory, field, historical, longitudinal, meta-analysis, mixed-method, observational, philosophical, photovoice, sequential and systematic review. Descriptive designs address who, what, when, where and how something currently exists, and cannot conclusively answer why.
Axis 3: observational versus experimental (the clinical-research split)
Chidambaram and Josephson (Pediatric Investigation 2019;3(4):245-252) describe “two major types of clinical study designs, observational and experimental.” Observational designs are descriptive (“a description of the exposure and/or the outcome”) or analytic (“a measurement of the association between the exposure and the outcome”). Experimental designs are “hypothesis testing studies.”
Cross-walk: where each design sits
The table places designs only where one of the three sources places them. A blank cell means the sources read for this page do not say.
| Design | What the source says it does | Observational or experimental (clinical-methods split) | Source |
|---|---|---|---|
| Descriptive | Describes who, what, when, where and how something currently exists; cannot conclusively ascertain why | Observational (descriptive branch) | USC Libraries; Chidambaram and Josephson 2019 |
| Cross-sectional | Examines differences among groups at a single time point | Observational; can be classified as either descriptive or analytic | USC Libraries; Chidambaram and Josephson 2019 |
| Longitudinal | Repeated measurement of the same sample | USC Libraries | |
| Case report, case series, ecologic | Describe exposure and/or outcome | Observational (descriptive) | Chidambaram and Josephson 2019 |
| Cohort, case-control (including nested case-control and case-cohort) | Measure the association between exposure and outcome | Observational (analytic) | Chidambaram and Josephson 2019 |
| Experimental / clinical trial | Tests hypotheses; a true experiment must have control, randomization and manipulation | Experimental (trial subtypes: randomized, non-randomized, cross-over, factorial; also community and field trials) | USC Libraries; Chidambaram and Josephson 2019 |
| Mixed-method | An approach combining quantitative and qualitative work rather than a single methodology | USC Libraries; UDel Library |
The cross-sectional study sits on both sides of a line: the clinical-methods paper says it can be classified as descriptive or analytic, so “descriptive design” and “cross-sectional design” are not mutually exclusive labels.
What are the key differences between descriptive and experimental research?
Descriptive research records how something is; experimental research tests a hypothesis. The USC Libraries guide says a descriptive study “cannot conclusively ascertain answers to why,” while “true experiments must have control, randomization, and manipulation.” In the clinical taxonomy, descriptive studies sit under observational designs, whereas experimental studies are hypothesis-testing studies such as clinical trials. For narrower contrasts see the CASRAI comparisons experimental vs quasi-experimental design, cross-sectional vs longitudinal study, case-control vs cohort study and within-subjects vs between-subjects designs.
What are the different types of research designs and how do they differ?
They differ in the question they serve, the kind of data they use and whether the researcher intervenes. Use the cross-walk above: pick the purpose first (the USC rule that the problem determines the design), then check whether your data are quantitative, qualitative or mixed, then, for human-health studies, whether the design is observational or experimental. The basic versus applied research distinction describes the goal of a study rather than its design. Qualitative projects have their own named designs; the qualitative descriptive design guide is one example.
What is a research design plan?
A research design plan is the written statement of the overall strategy and analytical approach a study will use, matched to the research problem. Drawing on the USC wording, it should make clear which design was chosen and why it fits the problem. Funder and journal templates add their own required headings, so follow the instructions of the specific call or journal.
- State the research problem or question first.
- Name the design (for example descriptive, cohort, randomized trial, mixed-method) using the label a reader in your field expects.
- Say whether the data are quantitative, qualitative or both.
- For human-health studies, say whether the design is observational or experimental, and for trials which subtype.
- Explain why this design can answer the question (a descriptive design, for instance, cannot answer why).
Common mistakes
- Choosing the design before the problem. The USC guide states the problem determines the design, not the reverse.
- Treating a method as a design. A survey or interview is a data-collection technique; the design is the overall strategy.
- Calling a study experimental without control, randomization and manipulation. The USC guide gives these as the requirements of a true experiment.
- Drawing causal conclusions from a descriptive study. It cannot conclusively answer why.
- Forcing cross-sectional studies into one box. The clinical-methods source treats them as either descriptive or analytic.
Checklist
- The research problem is written before the design is named.
- The design is named and distinguished from the methods used.
- Data type (quantitative, qualitative, mixed) is stated.
- Observational versus experimental status is stated where applicable.
- Any experimental label is backed by control, randomization and manipulation.
Frequently asked questions
Are there four types of research design?
Not in a single agreed scheme. The sources read for this page use different groupings: 18 designs in the USC Libraries list, two major clinical categories (observational and experimental), and quantitative, qualitative and mixed in the University of Delaware guide. Check which classification your discipline or instructor uses.
Is mixed-method a design?
The USC guide lists it among its 18 designs but describes it as more of an approach to examining a problem than a methodology. Treat it as a combination of quantitative and qualitative work that still needs a stated overall strategy.
Can one study belong to more than one category?
Yes. A study can be quantitative, descriptive and cross-sectional at once, since the three axes classify different things.
Sources
- USC Libraries, Organizing Your Social Sciences Research Paper: Types of Research Designs, libguides.usc.edu/writingguide/researchdesigns (read 2026-11-07).
- University of Delaware Library, Research Methods: Research Design, guides.lib.udel.edu/researchmethods/design (read 2026-11-07).
- Chidambaram AG, Josephson M. Clinical research study designs: The essentials. Pediatric Investigation 2019;3(4):245-252, PMC7331444 (read 2026-11-07).
