Pilot studies improve research before it begins, while experimental designs decide how participants are placed in conditions. Both choices affect the quality, efficiency, and credibility of psychological investigations.

Pilot studies

Before collecting the main set of data, psychologists often carry out a pilot study to test whether their method works in practice rather than only in theory.

A pilot study uses the planned procedure on a much smaller scale. It is not the main study itself, and its purpose is improvement rather than final data collection.

The main aims of piloting are to:

• check whether the instructions are clear

• find out whether participants understand the task

• identify problems with materials, questions, or equipment

• estimate how long the procedure is likely to take

• test whether the planned method for recording responses is workable

• reveal practical problems before time and money are spent on the full investigation

Piloting is especially useful because a procedure that seems straightforward to the researcher may confuse participants or produce unusable data. For example, instructions may be too vague, questions may be interpreted in different ways, or the task may be too difficult or too easy.

After a pilot, researchers may change the wording of instructions, alter the order of tasks, replace unclear materials, or adjust the timing of the study. If the pilot shows major weaknesses, the procedure can be redesigned before the full investigation begins. This makes the main study more efficient and reduces the chance of avoidable errors.

A pilot study can also show that a method is unrealistic.

If participants lose interest quickly, fail to complete the task, or misunderstand what they are supposed to do, the researcher has early warning that the study needs revision.

Experimental designs

Once a study has been planned, the researcher must decide how participants will be used across the conditions of the experiment.

In AQA Psychology, the three key experimental designs are repeated measures, independent groups, and matched pairs. Each design changes the balance between control, fairness, practicality, and the number of participants needed.

Repeated measures design

In a repeated measures design, the same participants take part in every condition of the experiment.

Because the same people are used throughout, differences between individuals are less of a problem. Intelligence, personality, motivation, or past experience are kept more constant because each person is effectively being compared with themselves.

Strengths of repeated measures include:

• fewer participants are needed

• participant differences are reduced

• comparisons between conditions may be clearer

However, repeated measures can create order effects, where performance in one condition changes performance in another.

A participant may improve through practice, become tired, or become bored. This means the order in which conditions are experienced can influence the results.

Independent groups design

In an independent groups design, different participants take part in each condition.

This design avoids order effects because each person completes only one condition. Their performance cannot be changed by earlier experience in another condition.

Strengths of independent groups include:

• no order effects

• reduced risk that one condition directly influences another

• suitable when the task cannot be repeated without changing behavior

Its main weakness is that the participants in one condition may differ from those in the other condition in important ways. These participant variables may affect the results, making it harder to know whether the difference between conditions was caused by the manipulation or by differences between the people in each group. Independent groups also usually require more participants than repeated measures.

Matched pairs design

A matched pairs design tries to reduce participant differences by pairing people who are similar on relevant characteristics, then placing one person from each pair in each condition.

The aim is to make the groups more alike than they would be in a standard independent groups design. If matching is done well, differences between conditions are less likely to be caused by pre-existing differences between participants.

Strengths of matched pairs include:

• no order effects, because each participant does only one condition

• better control of participant differences than ordinary independent groups

• useful when a researcher knows which participant characteristics are likely to affect results

There are also clear limitations:

• matching takes time and effort

• it may be difficult to find close matches

• participants cannot be matched perfectly on every relevant variable

• if one person in a pair withdraws, the matching arrangement may be disrupted

Choosing between designs

No design is always best. The choice depends on the practical demands of the study and the kind of problem the researcher is trying to avoid.

• Use repeated measures when participant differences are likely to be a major issue and the task can be completed more than once without serious carryover effects.

• Use independent groups when taking part in one condition would clearly affect performance in another condition.

• Use matched pairs when participant differences are important, but using the same participants in every condition would create serious problems.

A strong researcher chooses the design that best fits the procedure, the participants, and the kind of bias most likely to threaten the findings.