AQA Syllabus focus:
'Observational design: behavioural categories, event sampling and time sampling; questionnaire construction and interview design.'
This subtopic focuses on how psychologists plan observations, questionnaires, and interviews so data collection is clear, focused, and usable. Good design improves consistency and reduces unclear or biased responses.
Observational design
Observational design is the planning stage of an observation. The researcher decides exactly what behavior will be recorded and how it will be sampled. Good design turns broad ideas like aggression or attention into specific, observable acts.
Behavioral categories
When psychologists create an observation schedule, they usually break behavior into behavioral categories.
Behavioral categories: Clearly defined, observable, and recordable units of behavior used in an observation.
Good categories should be:
Objective: based on visible actions, not guesses about thoughts or feelings
Operationalized: written so each category has a precise meaning
Mutually exclusive: one behavior should fit one category at a time
Exhaustive: together, the categories should cover the likely behaviors being studied
A category such as being disruptive is too vague because observers may interpret it differently. More precise categories, such as calling out without permission or leaving seat, are easier to record consistently. Categories are usually placed on a recording sheet so the observer can tick, tally, or code behavior quickly while the observation is happening.
Event sampling and time sampling
After deciding the categories, the researcher chooses a sampling method. This determines when behavior will be recorded.
Event sampling: Recording every time a specified behavior or event occurs during the observation period.
Event sampling is useful when the target behavior is important and may happen irregularly. It allows the researcher to capture the full frequency of behaviors such as shouting, helping, or smiling. However, if many events happen quickly, the observer may miss some of them.
A different approach is to record behavior at selected time intervals.
Time sampling: Recording behavior at predetermined time intervals, such as every 30 seconds or every minute.
Time sampling is useful when an observation lasts a long time or when continuous recording would be impractical. The observer records what is happening at each chosen moment or during each brief interval. This makes recording easier to manage and can show patterns across time. However, behavior that happens between sampling points may be missed, so the length of the interval must be chosen carefully.
The recording sheet must match the sampling decision. Event sampling needs space for repeated tallies, while time sampling needs the intervals clearly marked beside the behavioral categories.
Questionnaire construction
A questionnaire should be designed so participants understand what is being asked and can respond in a way that produces useful data. Each question should link directly to the research aim.
Questionnaires often combine:
Closed questions, which give fixed response options and are easy to compare
Open questions, which allow fuller answers and can reveal unexpected information
The wording of questions is crucial. Good questions are:
Clear and simple
Focused on one idea at a time
Free from leading wording
Free from double negatives
Suitable for the age and background of the participants
A question that asks two things at once can be difficult to answer accurately. Emotionally loaded wording can also push respondents toward a particular response. Response options need care too.
A sample Likert-scale format showing a statement followed by ordered response categories (e.g., from strong disagreement to strong agreement). It illustrates how closed questions can standardise responses, making comparisons and quantitative analysis easier while still capturing degree of attitude. Source
Rating scales should cover the likely range of answers and be arranged logically.
Question order also matters. Early questions should usually be easy and straightforward so participants understand the task and continue responding. More sensitive or detailed items are often better placed later. Clear instructions and a neat layout help reduce confusion and missing data.
The overall length of a questionnaire also matters. If it is too long, participants may rush, skip items, or repeat the same answer pattern. A shorter questionnaire with focused items usually produces more complete and thoughtful responses.
Interview design
An interview needs structure, even when the researcher wants detailed answers. The researcher should prepare an interview schedule, which is the planned set of questions, prompts, and order used in the interview.
Interview schedule: A prepared plan showing the questions to be asked, their order, and any prompts the interviewer may use.
A well-designed interview schedule keeps the interview focused on the research aim. Questions should move in a logical sequence, often from broad and simple topics to more specific ones. This can help participants feel comfortable and reduce confusion.
Good interview questions are similar to good questionnaire items: they should be clear, neutral, and easy to understand. The interviewer should avoid:
Leading questions, which suggest the desired answer
Complex wording, which may confuse participants
Multiple questions at once, which make answers hard to interpret
Prompts should also be planned carefully. Useful prompts encourage the participant to say more without changing the meaning of the answer. If the interview is meant to be highly standardized, the same wording and order should be used for every participant. If some flexibility is allowed, the interviewer still needs to stay focused on the same topics so the data remain relevant.
Interview design also includes an opening statement explaining the topic and how the interview will proceed. This helps participants understand what kind of answer is expected and reduces awkward pauses or off-topic responses.
Recording decisions are also part of interview design. The researcher must decide whether responses will be written down on the schedule, audio-recorded, or both. This affects how much detail can be captured and how easy the data will be to review later.
Practice Questions
Outline what is meant by behavioral categories in observational design. (2 marks)
1 mark for stating that they are specific, observable behaviors or actions.
1 mark for stating that they are defined in advance so they can be recorded consistently or objectively.
A psychologist wants to observe distracted behavior during a lesson and then ask students about their study habits. Explain how the psychologist could design the observation using time sampling and behavioral categories, and describe two features of a well-designed questionnaire. (6 marks)
1 mark for explaining that time sampling records behavior at predetermined intervals.
1 mark for application, such as recording student behavior every 30 seconds or every minute.
1 mark for explaining that behavioral categories should be specific and observable.
1 mark for application, such as categories like talking to others, looking away from work, or leaving seat.
1 mark for one questionnaire design feature, such as clear wording, one idea per question, or non-leading questions.
1 mark for a second questionnaire design feature, such as logical question order, suitable response options, or clear instructions.
FAQ
There is no fixed number. Use enough categories to capture the target behavior but not so many that the observer cannot keep up in real time.
Start with the core behaviors linked to the aim.
Remove overlapping categories.
If events are being missed, the schedule is probably too complicated.
Momentary time sampling records what is happening at the exact sampling point, like a snapshot.
Interval recording asks whether the behavior happened at any time during the whole interval. Momentary methods are usually quicker, while interval recording can capture more behavior but may make brief actions seem more common than they really are.
A midpoint can be useful when a genuinely neutral answer is meaningful, such as “neither agree nor disagree.”
An even number of options removes the midpoint and forces a choice. The best option depends on the topic, but the scale should always be balanced and clearly labeled.
This can reduce the chance that participants simply tick the same response all the way down the page without reading each item carefully.
However, negatively worded items should be used carefully. If they are awkward or contain double negatives, they can confuse respondents and reduce the quality of the data.
Researchers can prepare a short list of neutral prompts in advance, such as “Could you say a little more about that?” or “What happened next?”
This keeps the interviewer supportive but consistent. The key is that prompts should invite elaboration without suggesting what the participant ought to say.
