What Is a Research Design?

The blueprint linking question to evidence

A research design is the overarching plan that coordinates research questions, data collection, and analysis strategies. It specifies what will be observed, on whom, when, and how comparisons are structured, while simultaneously controlling threats to internal and external validity. Broadly classified as experimental, quasi-experimental, or non-experimental (observational/descriptive), the choice of design directly shapes the credibility and scope of a study's conclusions.

Defining the Research Design

A research design is the structural framework that determines how a study will be conducted. It forms the logical bridge between the research question and the evidence needed to answer it. Choosing a design goes beyond selecting a data-collection technique; it encompasses decisions about who the participants are, how comparisons will be structured, which variables will be controlled, and how broadly findings can be generalized. In Shadish, Cook, and Campbell's (2002) framework, a design represents the totality of safeguards taken against threats to validity.

Major Design Types

Research designs fall into three broad categories. Experimental designs randomly assign participants to conditions, enabling the strongest causal inferences. Quasi-experimental designs investigate an intervention without random assignment, approximating causality through methods such as time-series analyses or comparison groups. Non-experimental designs involve no manipulation; descriptive studies, cross-sectional surveys, and longitudinal observations belong here. Each design type offers a different balance between internal validity (strength of causal inference) and external validity (generalizability of findings).

How to Select a Research Design

Design selection is driven by the structure of the research question. If the question is causal ("Does X affect Y?"), an experimental or quasi-experimental design is preferred, though ethical and practical constraints often push researchers toward quasi-experimental alternatives. If the question aims to describe a situation or understand associations, descriptive or correlational designs are appropriate. Researchers must also consider the time horizon (cross-sectional vs. longitudinal), data type (quantitative, qualitative, mixed-methods), and available resources. A sound design maximizes question-method fit while minimizing validity threats.

Common Pitfalls and Good Practice

One of the most common misconceptions is equating a research design with a single analysis technique — for instance, concluding "I used regression, therefore my design is experimental" is incorrect. A design denotes a logical structure independent of the chosen technique. Another pitfall is assuming that internal and external validity can be simultaneously maximized; in practice, an unavoidable tension exists between the two. Good practice requires researchers to document design decisions and their rationale explicitly in the methods section, and to address components such as measurement validity, statistical power, and sample representativeness during the planning stage.

Key terms

Internal Validity: Confidence that the observed effect is genuinely caused by the independent variable.
External Validity: The extent to which findings can be generalized across persons, settings, and times.
Random Assignment: Assigning participants to conditions by chance to control for confounding variables.
Quasi-Experimental Design: A research plan that attempts causal inference without random assignment to conditions.
Validity Threat: A plausible alternative explanation that undermines the accuracy or generalizability of findings.