4.2.3.1 - Pilot studies and the aims of piloting

What pilot studies are

Before a psychologist runs the main investigation, they may try out the procedure on a small scale. This is called a pilot study. The point is practical: it helps the researcher discover whether the study works as planned before they spend time, money and participant effort on the full version.

A pilot study might involve only a few participants, one trial of the materials, a short version of an interview, or a rehearsal of an observation schedule. It can be used before experiments, questionnaires, interviews, observations, content analyses and case-study procedures. The exact form depends on what the researcher needs to check.

The crucial exam point is that piloting is about the method, not about proving the hypothesis. A pilot may produce some data, but those data are usually too limited to be treated as evidence that the predicted effect is true. The researcher is mainly asking:

Think of a pilot study as a dress rehearsal for research. A poor rehearsal does not tell the researcher that the final performance has failed. It tells the researcher what to repair before the real performance begins.

Aims of piloting

The AQA specification names "pilot studies and the aims of piloting", so students need more than a one-line definition. In exam answers, the highest-value material is usually a specific aim linked to a specific improvement.

One aim is to check the clarity of standardised instructions. If some participants misunderstand what they are supposed to do, differences in results may reflect confusion rather than the factor the researcher intended to test. A pilot can reveal unclear wording, missing information or instructions that make participants more likely to guess the aim and change their behaviour.

A second aim is to test whether the materials and tasks are suitable. In a memory experiment, the word list might be too easy, causing a ceiling effect because most participants score near full marks. It might be too hard, causing a floor effect because most participants score near zero. Either problem makes it difficult to detect a meaningful difference between conditions.

A third aim is to check timing. The researcher may discover that a questionnaire takes 25 minutes rather than 10, that an interview runs too long for participants to stay focused, or that a repeated measures procedure causes fatigue. Timing matters because boredom, tiredness or rushed responses can threaten validity.

A fourth aim is to refine measurement. Piloting can show whether a questionnaire item is ambiguous, whether an interview question leads participants toward a particular answer, whether observation categories overlap, or whether a rating scale gives enough detail. Better measurement improves validity because the researcher is more likely to measure what they claim to measure.

A fifth aim is to check feasibility and logistics. The researcher may learn how easy it is to recruit participants, whether the recording equipment works, whether the room is quiet enough, whether consent forms are clear, and whether the debrief covers likely participant concerns. This saves resources because avoidable problems are fixed before the main data collection.

Piloting different methods

Pilot studies are often explained using experiments, but they apply across research methods. The same core idea stays the same: try out the method, find weaknesses, then improve the main study.

In an experiment, piloting can check the operationalisation of variables. Suppose a student investigates whether background noise affects reaction time. A pilot could show that the "noise" condition is not loud enough to be different from silence, or that the reaction-time app records scores inconsistently. The student could then adjust the volume, replace the app or practise the standardised procedure before collecting the real data.

In a questionnaire, piloting can check wording, layout and response options. A pilot participant might say “How often do you feel stressed?” gives no timeframe. “On how many of the past seven days did you feel stressed?” with response options “0 days”, “1–2 days”, “3–4 days” and “5–7 days” is clearer: every possible number fits exactly one category. It measures days with stress, not how severe that stress was. If severity is the aim, the researcher needs a severity measure instead.

In an interview, piloting can check whether questions produce useful detail without leading the participant. A pilot may reveal that "Do you agree that social media damages sleep?" encourages agreement. A better version might be "How, if at all, does social media affect your sleep?" This keeps the question open while still focused on the research aim.

In an observation, piloting can check behavioural categories. Categories such as “sharing resources” and “helping” may overlap. First define the unit being recorded: handing over one resource can be one act, while giving one verbal explanation is another. The schedule could define the categories as “handing an item to a peer” and “explaining a task to a peer”, with examples and rules for boundaries. Observers can then practise independently on the same material and compare their records. A child who hands over an item and then explains a task has performed two distinct acts; clear categories do not mean ignoring one of them.

Piloting can also reveal ethical issues. If a pilot participant becomes unusually embarrassed by a questionnaire, the researcher might change the wording, add a clearer warning before consent, strengthen the debrief or provide signposting to support. This links piloting to ethical design, but it does not replace formal ethical review where that is required.

The table below shows how exam answers can move from method to aim to improvement.