Breaking down the question
Two tasks sit here. What are variables? asks for a precise definition and typology. Discuss their role in experimental research asks how variables function in the logic of the experiment — the manipulation of one to observe its effect on another under controlled conditions. The marks reward conceptual clarity about independent, dependent and control variables and how their interplay establishes causality.
How to approach it
Define a variable as any characteristic that varies across cases and can take different values. Distinguish the key types with an example. Then explain the experiment as the classic device for testing causal hypotheses: the researcher manipulates the independent variable, measures the dependent variable, and holds control variables constant. Close by noting the difficulty of true experiments in sociology. Ground the answer in research methods in sociology.
Model answer
A variable is any attribute, characteristic or condition that can take more than one value across individuals, groups or situations — age, income, level of education, degree of social integration. It stands in contrast to a constant, which does not vary. Before a variable can be studied it must be operationalised, that is, defined in measurable terms.
Variables are classified by their role in a causal argument. The independent variable is the presumed cause, deliberately varied by the researcher. The dependent variable is the presumed effect, whose changes are observed and measured. Control variables are other factors held constant so that they cannot confound the relationship under test. An extraneous or intervening variable is one that may distort the link if not controlled.
In experimental research these categories become the working machinery of causal inference. The classic experiment isolates a relationship by manipulating the independent variable while holding all else constant and observing the resulting change in the dependent variable. A treatment group receives the stimulus; a matched control group does not; the difference between them is attributed to the independent variable. Random assignment further ensures that unknown extraneous variables are distributed evenly across groups. This design allows the three conditions of causality to be met: covariation between cause and effect, the proper time order, and the elimination of rival explanations. Stanley Milgram's obedience experiments, varying conditions such as the proximity of the authority figure, illustrate how manipulating an independent variable reveals its effect on behaviour.
In sociology, however, true experiments are difficult. Human beings cannot always be randomly assigned or isolated in laboratories, and manipulation may be unethical. Durkheim therefore treated the comparative method as the sociologist's substitute for the experiment, using variation across groups to approximate controlled comparison. Variables nonetheless remain central: even in surveys and comparative studies, the disciplined identification of independent, dependent and control variables underpins rigorous causal analysis.
Examiner's perspective
Examiners want the typology stated cleanly and correctly — confusing the independent and dependent variable is a common and costly error. A concrete example that labels each variable demonstrates real understanding.
Strong answers explain the logic of the experiment — manipulation, control, comparison of treatment and control groups — rather than merely listing definitions. Noting the three conditions of causality, and the fact that sociology often relies on the comparative method in place of the laboratory, signals maturity. Weaker scripts define variables in the abstract but never connect them to how an experiment actually establishes cause and effect.