Aims, Hypotheses and Variables

An aim is a general statement of the purpose of a study — what the researcher intends to investigate. It sets the direction of the whole investigation but does not, by itself, make a specific prediction.

An aim usually begins "to investigate…", "to find out whether…" or "to examine…". For example: "to investigate whether caffeine affects memory." That tells you what the researcher is interested in, but not what they expect to happen or how "caffeine" and "memory" will be measured.

Aims come from observations, from gaps in existing theory, or from earlier research findings. Once a researcher has an aim, they turn it into one or more precise predictions — the hypotheses — which are what the study actually tests.

An aim states the purpose of a study in general terms. It is not a prediction and it does not need operationalised variables. A hypothesis does both.

Getting this distinction right matters because AQA questions frequently ask for an aim in one part and a hypothesis in the next, and marks are lost when the two are confused.

A hypothesis is a precise, testable statement that predicts the relationship between the variables in a study. Where the aim is broad, the hypothesis is specific: it must be clear enough that the results of the study can either support it or fail to support it.

Two features separate a good hypothesis from a restated aim:

• It predicts a difference or relationship, rather than just naming a topic of interest.
• Its variables are operationalised — defined so they can be measured or manipulated (covered on a later slide).

Compare the two for the same study:

Statement	Type	Why
To investigate whether caffeine affects memory	Aim	States the purpose; makes no prediction
Participants who drink caffeine will recall more words from a 20-word list than participants who drink water	Hypothesis	Predicts a difference; variables are measurable

A hypothesis must be testable and falsifiable: it has to be possible, in principle, for the data to disagree with it. "Caffeine is interesting" is not a hypothesis; "caffeine drinkers recall more words" is.

There are two forms of the experimental (alternative) hypothesis, and choosing between them is a common exam skill.

A directional (one-tailed) hypothesis predicts the direction of the difference or relationship. It says not just that the groups will differ, but which way.

Example — directional: "Participants who drink caffeine will recall more words than participants who drink water."

A non-directional (two-tailed) hypothesis predicts that there will be a difference or relationship, but does not state its direction.

Example — non-directional: "There will be a difference in the number of words recalled by participants who drink caffeine and those who drink water."

Which one you choose depends on the existing evidence:

Choose	When
Directional (one-tailed)	Previous research suggests a particular direction for the effect
Non-directional (two-tailed)	There is no previous research, or previous findings are conflicting

The key phrase to look for in an exam scenario is whether prior studies are described. If the scenario says "previous research has found that…", that justifies a directional hypothesis. If it gives no clear prior direction, a non-directional hypothesis is the safe choice.

The hypotheses above are versions of the alternative hypothesis — the prediction that there is an effect. Alongside it sits the null hypothesis, which predicts no difference or relationship between the variables (any difference seen is due to chance).

For the caffeine study, the null hypothesis would be: "There will be no difference in the number of words recalled by participants who drink caffeine and those who drink water."

The null hypothesis is the statement that is actually tested by a statistical test. A study does not "prove" the alternative hypothesis. Instead, it decides whether to reject or retain the null hypothesis:

• If the result is statistically significant, the researcher rejects the null hypothesis and accepts the alternative hypothesis.
• If the result is not significant, the researcher retains the null hypothesis.

The null hypothesis is what the statistical test evaluates. Rejecting it means the effect is unlikely to be due to chance alone; retaining it means the evidence for an effect was not strong enough.

Understanding this now makes the later topics of significance and Type I/Type II errors far easier.

An experiment works by changing one thing and measuring another. The variables have precise names.

The independent variable (IV) is the variable the researcher manipulates or changes. The dependent variable (DV) is the variable the researcher measures to see the effect of the IV. In the caffeine study, the IV is the drink given (caffeine vs water) and the DV is the number of words recalled.

Two further types of variable can interfere with the results:

Variable	Definition	Relationship to the IV
Extraneous variable	A nuisance variable that could affect the DV if not controlled	Random; not linked to the IV
Confounding variable	A variable that provides an alternative explanation for the results	Changes systematically with the IV

An extraneous variable adds random noise — for example, small differences in how much sleep individual participants had. It affects the DV but does not favour one condition over the other, so it makes results messier rather than misleading.

A confounding variable is more dangerous because it varies with the IV. Suppose every caffeine participant was tested in the morning and every water participant in the afternoon: time of day now changes systematically with the drink, so you cannot tell whether the drink or the time caused any memory difference.

Extraneous vs confounding is the distinction examiners test most. An extraneous variable is a random nuisance unrelated to the IV. A confounding variable changes systematically with the IV and offers a rival explanation, so it is the bigger threat to validity.

Operationalisation means defining a variable clearly so that it can be objectively measured or manipulated. Vague concepts like "memory", "aggression" or "stress" cannot be measured directly, so they must be turned into something countable.

For example, "memory" can be operationalised as "the number of words recalled from a 20-word list in 2 minutes." Now two researchers running the study would measure it in the same way.

A hypothesis is only complete when both variables are operationalised. Here is a full worked example.

Scenario: A researcher wants to see whether revising in a quiet room affects test performance compared with revising with music playing. Previous research suggests background music reduces recall.

Step 1 — identify the variables.

• IV: the revision condition — quiet room vs music playing.
• DV: test performance.

Step 2 — operationalise them.

• IV (manipulated): whether the participant revises in silence or with instrumental music at 60 decibels.
• DV (measured): score out of 20 on a test taken immediately after revising.

Step 3 — write each hypothesis.

Because previous research suggests a direction (music reduces recall), a directional hypothesis is justified:

(a) Directional: "Participants who revise in a quiet room will score higher out of 20 on the test than participants who revise with instrumental music playing at 60 decibels."

If there were no clear prior direction, you would instead write:

(b) Non-directional: "There will be a difference in the test scores out of 20 between participants who revise in a quiet room and those who revise with instrumental music playing at 60 decibels."

Both name the operationalised IV (quiet vs music) and the operationalised DV (score out of 20), which is exactly what earns the marks.

1. Writing an aim when asked for a hypothesis

"To investigate whether music affects revision" is an aim, not a hypothesis — it makes no prediction. A hypothesis must predict a difference or relationship. Read the command word carefully: "write a hypothesis" needs a prediction, not a statement of purpose.

2. Failing to operationalise the variables

A hypothesis such as "caffeine improves memory" is too vague to score full marks. State how each variable is measured or manipulated: "participants who drink 200 mg of caffeine will recall more words from a 20-word list than participants who drink water."

3. Writing a directional hypothesis without justification

Only choose a directional (one-tailed) hypothesis when the scenario provides previous research pointing in a direction. If no prior direction is given, or findings conflict, a non-directional hypothesis is the appropriate choice.

4. Confusing extraneous and confounding variables

Feature	Extraneous	Confounding
Link to IV	Not linked; random	Changes systematically with the IV
Effect	Adds random noise to the DV	Offers an alternative explanation
Threat level	Lower	Higher

An extraneous variable is a random nuisance; a confounding variable varies with the IV and undermines the whole conclusion.

5. Mixing up the IV and the DV

The IV is manipulated (what the researcher changes); the DV is measured (the outcome). A quick check: the DV is almost always the score, count or time you record at the end. Getting these the wrong way round reverses the meaning of the whole hypothesis.