Core distinction and conditions
A psychological claim can sound plausible and still exceed the evidence offered for it. Start by naming the question being asked: description, association, or a causal effect. A survey can describe a sample and reveal associations. An experiment deliberately varies an independent variable while measuring an outcome and managing plausible alternatives. Random assignment distributes enrolled participants across conditions; random sampling selects people from a population. These operations answer different questions. A well-controlled experiment using volunteers can support a causal comparison within its setting while leaving broad generalization uncertain.
Worked example: recruitment and placement
A library recruits 60 visitors for a study of how instructions affect search performance. A computer randomly assigns each visitor into a diagram-instruction group or a text-instruction group. The visitors use the same catalog for the same amount of time, and success is counted using one scoring rule. Assignment reduces systematic pre-existing differences between the groups. It does not make those visitors representative of every library user. To interpret a group difference, first check that the instruction format was the intended difference, then consider whether the scoring and other conditions were comparable. Even random assignment can produce chance imbalance; it does not force identical groups.
Contrast and common error
Suppose the diagram group instead works in a quiet room and the text group in a crowded hallway. Noise now varies with format. Calling the study an experiment does not dissolve that confound. Holding room conditions constant, or assigning them independently in a suitable design, addresses the stated alternative. Increasing sample size alone does not separate two features that always move together. Similarly, an observational association between practice time and achievement could reflect reverse direction or another influence. A sophisticated statistical result does not automatically supply the missing design conditions.
Transfer check: a stronger inference
A volunteer music club randomly assigns members to two rehearsal schedules, but only one schedule receives a new teacher. Can a later performance difference be attributed solely to scheduling? Answer before reading on. It cannot: teacher and schedule differ together. A useful redesign compares schedules under comparable teaching conditions. Also explain why random assignment of club members does not establish representativeness of all learners. You have completed this check when you can name the target inference, identify the relevant design operation, and state one limitation without dismissing everything the evidence can support.
Try it, then explain it.
Complete a short set without looking at the example. After each answer, explain which evidence or rule supports your decision and why the alternatives do not fit.
Sources & authorship
Original OpenExam explanations and examples. References support the concepts; they are not the source of copied questions. Independent educator review is pending.
OpenStax Psychology 2e, section 2-3