Variables

Variables make research questions testable

A variable is something that can change or take different values. In psychology, variables include behaviours, such as the number of words someone remembers, and factors that might influence behaviour, such as background noise.

Researchers need to specify exactly what they will change or measure. This is called operationalisation. ‘Memory’ is a broad idea; ‘the number of words correctly recalled from a list of 20 words within two minutes’ gives a clear measurement. Precise definitions help researchers carry out the procedure consistently and allow other researchers to repeat it.

The independent variable: what changes

In an experiment, the independent variable (IV) is the variable the researcher manipulates. Manipulation means deliberately creating different conditions so that their effects can be compared.

Suppose a psychologist investigates whether music affects word recall. Participants could learn a list of 20 words either in silence or while the same recording of rock music plays at a fixed volume. The IV is whether the words are learned in silence or with music. These are its two conditions.

The manipulation needs to be precise. Simply saying ‘some participants hear music’ leaves unanswered questions about the recording, volume and when it plays. Clear conditions make it easier to repeat the investigation accurately.

The dependent variable: what is measured

The dependent variable (DV) is the outcome measured to assess the effect of the IV. In the word-learning experiment, it could be the number of words correctly recalled from the 20-word list within two minutes.

Different questions need different measurements. A researcher investigating puzzle performance could record completion time in seconds. A researcher studying aggression could count specified aggressive acts during a 20-minute observation. The measure should produce a clear score and suit the behaviour being investigated.

A precise measure is not automatically a valid one. For example, the number of words recalled is a measure of performance on that memory task; it does not measure every aspect of a person's memory. Researchers must consider whether their chosen DV genuinely assesses what they intend to study.

Co-variables: measuring both rather than changing one

A correlation investigates whether two measured variables are related. These are called co-variables. Unlike an experiment, the researcher does not manipulate one variable to see its effect on another: both are measured.

For example, a researcher could investigate the relationship between school attendance and GCSE achievement. Attendance could be measured as the number of days present during Year 11, and achievement as the number of GCSEs achieved. Each student provides a pair of scores, one for each co-variable.

Co-variables can be measured using existing records, questionnaires or observations. Records can provide attendance totals without relying on students' memories. Questionnaires can collect information such as the number of takeaway meals eaten in a month and a numerical self-reported stress score for that month. However, people may forget information or report it inaccurately, weakening the measurements.

Both co-variables need clear definitions, including relevant units or time periods. Finding a relationship does not establish cause and effect: illness, for instance, might affect both attendance and achievement.

Extraneous variables: other influences on the outcome

An extraneous variable (EV) is any factor other than the IV that could affect the DV. In the music experiment, room temperature, time of day, tiredness and the researcher's behaviour could all influence recall.

Imagine that the music condition takes place in a hot room but the silence condition takes place in a comfortable room. A difference in recall might be caused by temperature rather than music. This weakens the validity of the conclusion because there is another possible explanation for the result.

Researchers therefore try to control relevant EVs. They could use the same room at the same temperature, test at comparable times and give everyone the same word list, learning time and recall time. These controls make the conditions more comparable while preserving the intended difference in the IV.

Standardisation: consistent procedures, with limits

Standardisation means keeping relevant instructions, materials and procedures the same. For example, the researcher could read a fixed instruction script and use the same scoring rule for every participant. A script reduces differences in explanation or encouragement; a scoring rule helps ensure that identical answers receive identical scores.

Standardisation supports reliability because the procedure can be repeated consistently. It also helps validity by reducing alternative explanations for differences between conditions.

However, not every influence can be kept identical. Participants may arrive with different moods or levels of tiredness. Highly controlled tasks can also differ from everyday behaviour, limiting how well findings apply outside the study. Good research therefore uses controls that address likely influences without assuming that standardisation removes every problem.