# What are the methods of data collection in psychology?

> Written with AI · [Psychology: research methods and statistics](https://lessoons.com/en/psychology-research-methods) · Grade 10 · 2026-11-14 · https://lessoons.com/en/psychology-research-methods/what-are-the-methods-of-data-collection-in-psychology

The main methods of data collection in psychology, besides experiments, are observation, interviews and questionnaires. Observation records what people actually do; interviews and questionnaires are self-report methods that ask people about themselves. Each can be more or less structured, and each has its own bias, such as people behaving differently when watched or answering to look good.

## The topic explained

### Watching or asking

Some questions cannot be put in a laboratory: how children play in a kindergarten yard, or why teenagers choose a subject. So, besides experiments, psychology uses three methods of data collection. Observation records what people do. Interviews and questionnaires ask people about their thoughts, feelings and behavior, so they are called self-report methods. Each can also be used inside an experiment, to measure the dependent variable.

The choice depends on the question. Observation shows behavior directly, but not the reasons behind it. Self-report reaches thoughts and feelings that no one can see, but depends on people being able and willing to report them accurately. A common mistake is to treat self-report as a direct window into the mind; it is what people say, which is not always what they do.

### Observation: kinds and checks

Observations differ on several choices. In a naturalistic observation behavior is watched where it normally happens; in a controlled observation the setting is arranged by the researcher. In a structured observation the observer counts set behavior categories, such as sharing, pushing and talking, often in fixed time intervals; an unstructured observation writes down whatever seems important. A participant observer joins the group being studied, and an observation can be overt, with people knowing, or covert, without their knowledge, which raises ethical questions about consent.

Two problems need checks. People who know they are watched may change their behavior, called reactivity or the Hawthorne effect, after studies at the Hawthorne factory in the 1920s and 1930s whose original reports some scholars now doubt. And observers may see what they expect. Using clear categories and two observers whose tallies are compared, inter-rater reliability, reduces the second problem.

### Interviews: from structured to unstructured

In a structured interview every person gets exactly the same questions in the same order, so answers can be compared and added up. A semi-structured interview has set questions but lets the interviewer follow up on interesting answers. An unstructured interview is closer to a guided conversation around a topic, giving rich, personal answers that are hard to compare. Interviews collect detail a form cannot, but the interviewer's tone, appearance or reactions can shape what people say, an interviewer effect, and analyzing long answers takes time.

Imagine a study of how students feel after moving to a new school. A structured interview would ask every student the same ten questions, making it easy to say how many felt lonely in the first month. A semi-structured interview would start from the same questions but let a student who mentions a sports team explain how it helped. The trade-off runs through every method: the more structure, the easier it is to compare and repeat; the less structure, the closer the answers come to the person's own experience.

### Questionnaires and the wish to look good

A questionnaire reaches many people quickly and cheaply. Closed questions offer fixed answers and give quantitative data that are easy to count, such as a Likert scale, developed by Rensis Likert in his 1932 dissertation, where people rate their agreement with a statement from 1 to 5. Open questions let people answer in their own words and give qualitative data that are richer but harder to analyze.

Every self-report faces social desirability bias: answering in a way that looks good, reporting more good behavior and less bad behavior than is true. Researchers reduce it with anonymous questionnaires, neutral wording and avoiding leading questions, and some add a scale that detects it, such as the one Douglas Crowne and David Marlowe built in 1960.

## Practice: methods of data collection

For 10th grade psychology students. Answer each question from the page, then check your work against the answers. All the data are imaginary.

1. **Name the kind of observation in each description, using the terms on the page.** — Counts acts from a checklist: a structured observation (naturalistic, in a yard). Joins a group in secret: a covert participant observation. Watches via a one-way mirror: a covert, non-participant observation, usually controlled if the room is set up by the researcher.
2. **Write one closed and one open questionnaire item about how students prepare for exams.** — A good answer gives a closed item with fixed answers, for example "How many days before an exam do you start studying? (0–1, 2–3, 4–7, more than 7)", and an open item, for example "Describe how you usually prepare for an exam." The closed item gives quantitative data; the open one gives qualitative data.
3. **Two observers recorded the same 30 minutes. Is their observation reliable? Which category needs work, and what could the researcher do?** — Mostly: they agree on sharing (9 and 9) and talking (15 and 14), but not on pushing (3 and 10), so that category is unreliable. The researcher should define pushing more exactly, with examples of what counts, train the observers together and test agreement again.
4. **What kind of question and what kind of data does this Likert item give? Name one advantage and one limitation.** — A closed question giving quantitative data. Advantage: answers from many people are quick to collect and easy to compare and average. Limitation: it does not show why people agree or disagree, and people may avoid the extreme options or answer to look good.
5. **A questionnaire asks: "How many hours do you study each day?" Explain why the answers may be too high, and give two ways to reduce the problem.** — Social desirability bias: studying is seen as good, so people may over-report it to look good. Ways to reduce it: make the questionnaire anonymous; word the question neutrally, for example offering ranges that include low amounts; or add a social desirability scale to detect such answers.
6. **Rewrite the leading question in a neutral form, and explain what makes the original leading.** — For example: "How do you find the new schedule?" or "How tiring or not tiring do you find the new schedule?" The original suggests the expected answer, that the schedule is tiring, and pushes people to agree.
7. **Choose a method for each question and justify it: (a) How often do children share toys in kindergarten? (b) Why do students choose the psychology track?** — A good answer chooses (a) a structured naturalistic observation with clear behavior categories and two observers, since young children cannot report this accurately and their behavior can be seen; and (b) a semi-structured interview or a questionnaire with open questions, since reasons cannot be observed and need the students' own words, noting the risk of social desirability.

## Common questions

### What is the difference between observation and self-report?

Observation records what people do, as seen by a researcher. Self-report methods, interviews and questionnaires, ask people to describe their own thoughts, feelings and behavior. Observation shows behavior but not its reasons; self-report reaches inner experience but depends on accurate and honest answers.

### What is the difference between structured and unstructured observation?

In a structured observation the observer records set behavior categories, often by counting them in fixed time intervals, which makes results comparable and reliable. In an unstructured observation the observer writes down whatever seems important, which gives rich detail but is harder to compare and more open to bias.

### What is social desirability bias?

It is the tendency to answer self-report questions in a way that will be viewed favorably, reporting more good behavior and less bad behavior than is true. Anonymous questionnaires, neutral wording and special scales help to reduce or detect it.

### What is a Likert scale?

It is a common closed questionnaire format in which people rate their agreement with a statement, usually on five or seven points from strongly disagree to strongly agree. Rensis Likert developed it in his 1932 doctoral dissertation on measuring attitudes.

### Which is better, an interview or a questionnaire?

Neither is always better. A questionnaire reaches many people quickly and gives comparable answers; an interview gives depth and lets the researcher follow up, but takes longer and can be shaped by the interviewer. The research question decides.
