4.2.3.1 - Variables

What variables do

Psychological research is built from variables. A researcher usually begins with an idea, such as "sleep affects memory" or "stress is related to social media use". To investigate that idea scientifically, the researcher has to decide which factors can change and what job each factor has in the study.

Variables can play different roles depending on the research question. Sleep could be the thing a researcher changes, the outcome a researcher measures, or a factor the researcher tries to keep under control. The role is not fixed by the word itself. It is fixed by the design of the study.

In an experiment, the two central variables are the independent variable and the dependent variable.

The outcome that may respond to this manipulation is the dependent variable.

For example, a researcher investigates whether background noise affects recall of psychology terms. The IV is the noise condition, such as silence or background noise. The DV is recall, such as the number of terms correctly recalled.

There are also variables the researcher is not mainly investigating, but that could still affect the result.

Researchers must also state exactly how the variables will be handled in the particular investigation.

In the noise and recall study, extraneous variables might include the difficulty of the word list, the time allowed to learn it, participants' prior knowledge, tiredness, room temperature or interruptions. These are not the focus of the study, but they could still influence recall.

A variable’s role comes from the procedure. If students are assigned to learn in silence or noise, sound condition is the IV. If students instead report their usual noise exposure and complete a recall task, both quantities are measured co-variables in a correlation. An association would not show whether noise caused poor recall, whether people with memory difficulties chose quieter places, or whether another factor affected both.

Manipulating the IV

Manipulating a variable means deliberately varying it. In an experiment, the researcher manipulates the IV by creating two or more conditions or levels. The researcher then compares the DV across those conditions.

The manipulation must be clear enough that another researcher would know exactly what happened in each condition. "Music" is too vague as an IV because it does not say what kind of music, how loud it is, when it is played, or what the comparison condition is.

A good manipulation changes the intended variable while keeping the rest of the situation as similar as possible. In the music example, the silence condition still uses headphones. This matters because otherwise headphone use itself might affect concentration. The researcher wants the difference between conditions to be the sound, not a bundle of other differences.

This is the heart of experimental control. A clean manipulation helps the researcher argue that any difference in the DV is more likely to be due to the IV. A messy manipulation makes the result harder to interpret. If the music group studies in a warm room for five minutes and the silence group studies in a cold room for two minutes, a recall difference could be due to room temperature, learning time or the music condition.

The IV also needs to fit the aim. If the aim is to investigate whether "background noise affects concentration", a manipulation involving loud unpredictable alarms may not represent ordinary background noise. If the aim is to investigate "acute noise stress", the same manipulation might be more suitable. The variable must match the concept being tested.

In AQA answers, strong IV wording often includes:

• the conditions being compared
• the amount, duration or intensity of the manipulation where relevant
• the materials or setting needed to make the conditions clear
• wording that shows only the intended IV changes systematically between conditions

Not every study manipulates an IV. In a natural experiment the relevant difference occurs naturally; in a quasi-experiment it is an existing characteristic. The researcher compares its levels rather than creating it. A correlation measures co-variables without manipulating either one, so naming an IV and DV is inappropriate unless the design establishes those roles.

Measuring the DV

The DV is the evidence the researcher uses to judge whether the IV made a difference. If the DV is poorly chosen, the study can look controlled but still fail to answer the question.

A good DV measure should be:

• directly linked to the aim
• specific enough to score consistently
• sensitive enough to detect meaningful differences
• practical for the participants and setting
• described so clearly that another researcher could repeat the measurement

For example, "memory" is not a DV measure. It is a broad psychological concept. The researcher needs to decide how memory will be measured in this study.

Sometimes the DV is a score, time, frequency or rating. Sometimes it can be a written explanation or interview response. For AQA research-methods scenarios, you should usually make the DV measurable in a clear, practical way. If the question asks for an operationalised DV, vague phrases such as "how well they did", "their behaviour" or "their level of stress" are not enough.

There is also an AO3 judgement. A very precise DV is not automatically a valid one. Suppose a researcher investigates whether a mindfulness exercise reduces stress, but measures stress only by counting how often participants tap a pencil. Pencil tapping may be easy to count, but it may not capture the psychological idea of stress very well. A standardised self-report scale might address the experience more directly. Heart rate can add physiological evidence, but it also varies with activity or caffeine. Combining measures can help investigate a construct; it does not automatically validate either measure.