Last Updated on October 9, 2026 by Karl Thompson
Non-participant observation is where a researcher watches a group without taking part in its activities: a ‘fly on the wall’ approach. The example most students already know is an Ofsted inspector sitting at the back of a lesson, watching and taking notes.
This post covers:
- What non-participant observation is, and its four main types
- Whether it produces qualitative or quantitative data
- Observation schedules, including Flanders’ classic system for recording classroom talk
- Examples of non-participant observation research
- The theoretical, practical and ethical advantages and disadvantages
- How it compares with participant observation
- How it is used to research education, and how it comes up in the exam
Non-participant observation is part of the research methods topic in AQA A-level sociology. It is examined mainly in the theory and methods section of Paper 3, and on Paper 1, where it can come up in either theory and methods or the Methods in Context question.
What is non-participant observation?
In non-participant observation the researcher observes individuals or groups but does not get involved in their lives. They do not join in, take on a role in the group or build relationships with the people being studied. They watch, and they record what they see.
This is the key difference from participant observation, where the researcher joins the group and takes part in its day-to-day activities. The participant observer aims to see the world through the eyes of the people being studied. The non-participant observer stays outside it.
Non-participant observation can be done in person, with the researcher physically present, or by recording, for example by filming a classroom and analysing the footage afterwards.
Types of non-participant observation
Every non-participant observation involves two choices. The first is whether the observation is structured or unstructured. The second is whether it is overt (the people being observed know about it) or covert (they don’t).
A structured observation uses an observation schedule: a list of behaviours decided in advance, which the researcher looks for and records, often as tallies. An unstructured observation has no schedule. The researcher simply observes and notes down anything that stands out.
| Overt | Covert | |
|---|---|---|
| Structured | An Ofsted-style lesson observation, with the observer recording behaviour against set categories | Counting particular behaviours from hidden-camera footage, without the people filmed knowing |
| Unstructured | A researcher openly sitting in on staff meetings and noting whatever seems significant | Observing how people behave in a public place, such as a high street, without telling them |
To be honest, it’s probably not possible to do an observation with no structure at all, at least if you’re going to write up your results at some point, because writing them up gives some kind of structure to what you’ve observed. ‘Unstructured’ observation is really less structured and more informal observation.
Covert non-participant observation is harder to do than it sounds. Sitting at the back of a classroom or a workplace without anyone asking who you are is difficult in practice. More realistically, covert non-participant observation means hidden cameras or observing people in public places.
Is non-participant observation qualitative or quantitative?
It can be either, and it depends on the structure. A structured observation produces quantitative data: counts of how often each behaviour occurred, which can be compared across classrooms, schools or researchers. An unstructured observation produces qualitative data, in the form of written field notes.
This is why positivists are more comfortable with structured non-participant observation than with most other observational methods. It produces numbers, it can be standardised, and it can be repeated. Interpretivists are generally less keen, because the observer never finds out what the behaviour means to the people doing it. If they want observation, they usually prefer participant observation.
Overt or covert? A common exam mistake
Students often assume that non-participant observation must be covert, because the researcher is not ‘one of the group’. It isn’t. Most non-participant observation is overt: the teacher knows the inspector is there.
This matters in the exam. In AQA’s report on the November 2020 paper, examiners noted that many answers to the non-participant observation question drifted into a general comparison of overt and covert observation, and some claimed non-participant observation is always covert. Neither earned much credit. Stay focused on the non-participant part of the method.
Observation schedules: Flanders’ Interaction Analysis Categories
The best-known example of a structured observation schedule in education research is Ned Flanders’ Interaction Analysis Categories, published in Analyzing Teaching Behavior (Flanders, 1970). It was designed to measure interaction between teachers and pupils quantitatively.
The observer records what kind of talk is happening every three seconds, placing each observation in one of ten categories:
Teacher talk
- Teacher accepts pupils’ feelings
- Teacher praises or encourages pupils
- Teacher accepts or uses pupils’ ideas
- Teacher asks questions
- Teacher lectures
- Teacher gives directions
- Teacher criticises pupils or justifies authority
Pupil talk
- Pupils talk in response to the teacher
- Pupils initiate talk
Silence
- Silence or confusion
Research using the system produced what became known as the ‘rule of two-thirds’: about two-thirds of classroom time is spent talking, the teacher does about two-thirds of that talking, and about two-thirds of the teacher’s talk is ‘direct’ (lecturing, giving directions and controlling pupils).
The strengths and limitations of the method are both visible here. The schedule is standardised, so another researcher could use it in another classroom and compare the results. HOWEVER, it only records verbal behaviour, and the categories were decided before the observation began. Anything that doesn’t fit one of the ten boxes simply isn’t recorded.
Examples of non-participant observation
Ofsted lesson observations
Ofsted inspectors observe lessons as part of school inspections. This is a form of overt, mostly structured, non-participant observation. Inspectors watch rather than teach, and they look for particular things.
Note one change, since older textbooks describe this differently: Ofsted’s guidance to schools states that inspectors do not grade individual lessons (Ofsted, 2018). Lesson observations feed into the overall judgement of the school rather than producing a score for each lesson.
A researcher who uses existing Ofsted reports is using secondary data, not doing non-participant observation themselves. HOWEVER, Ofsted-style observation is relatively easy to adapt for a researcher’s own primary research in schools.
The ORACLE study of primary classrooms
ORACLE (Observational Research and Classroom Learning Evaluation) was a large British study of primary classrooms that began in 1975 and ran for five years, reported in Inside the Primary Classroom (Galton, Simon and Croll, 1980). It used systematic, structured observation of teachers and pupils.
What makes ORACLE useful for evaluating the method is what happened next. Two decades later, the study was repeated in over two-thirds of the original schools, using the same observation instruments (Galton et al., 1999). That is reliability in action. Because the schedule was standardised, the researchers could go back and measure what had changed.
Advantages of non-participant observation
Theoretical advantages
Reliability. Structured non-participant observation should have better reliability than participant observation because the researcher is less involved. With a standard schedule, another researcher can repeat the observation looking for exactly the same things, as the ORACLE replication shows.
Comparison. If observations are structured, it is relatively easy to compare classrooms, schools or groups, and to look for patterns and correlations. This is what positivists want from a method.
Less disturbance than participant observation. It is easier for the researcher to blend into the background than in participant observation, which should mean people act more naturally over time. The researcher is also less at risk of ‘going native’, because they never become part of the group.
Seeing what people do. As with all observation, the researcher records what people actually do rather than what they say they do in an interview or questionnaire.
Practical advantages
Quicker and cheaper. Non-participant observations are generally cheaper and quicker to do than participatory methods, because the researcher does not have to get to know the respondents or spend months building trust.
Less training. Structured observations don’t require much training on the part of the researcher. Once the schedule is designed, the observer’s job is to record against it.
Funding. Because it is quicker and produces quantifiable results, structured observation is more likely to attract funding than long, open-ended ethnographic research.
Ethical advantages
Overt non-participant observation involves no deception. People can be told what is being observed and why, and can give informed consent. The researcher is also not drawn into the group’s activities, so does not face the covert participant observer’s dilemma of having to join in with illegal or harmful behaviour to keep their cover.
Disadvantages of non-participant observation
Theoretical disadvantages
Validity: you can’t ask why. This is the most important limitation. Non-participant observation lacks validity because you cannot ask people why they are acting the way they are. You see the behaviour but not the meaning behind it, which is exactly what interpretivists want to understand.
The Hawthorne effect. In overt observation, people may act differently because they know they are being watched. Teachers and pupils behave differently when an inspector is in the room, as anyone who has been in a school during an Ofsted inspection will tell you.
Observer bias and inter-observer reliability. Different observers may not classify the same behaviour in the same way. One researcher’s ‘off-task chatter’ is another’s ‘collaborative discussion’. A schedule reduces this problem but does not remove it, because someone still has to decide what counts as each category.
Unstructured observation brings its own problems. If done carefully and thoughtfully, unstructured observation can reveal hidden norms of behaviour. HOWEVER, the researcher may be overwhelmed with data, and what gets written down depends on what the researcher happens to find significant.
Representativeness. Samples can be biased. In schools, for example, you are likely to end up with a self-selecting sample, because better schools and teachers are more likely to agree to be observed than weaker ones.
Practical disadvantages
Access. Although gaining access is easier than with participant observation, it can still be difficult. Schools, workplaces and other organisations have gatekeepers who may refuse.
You can only see what you can see. A single observer cannot record everything happening in a busy setting, such as a class of thirty pupils. Filming helps, because footage can be watched again and checked by a second researcher, but it raises its own ethical questions.
Ethical disadvantages
Disempowering. Being observed can be disempowering for the people involved. The observer is detached and acts as an expert judging others, which is how many teachers experience Ofsted inspections.
Consent can be partial. An organisation may give permission for the research without the individuals being observed agreeing to it. A school might allow an observer in without asking the pupils.
Covert observation raises privacy issues. Hidden cameras and covert observation in public places mean people are being studied without their knowledge or consent.
Non-participant observation compared with participant observation
| Non-participant observation | Participant observation | |
|---|---|---|
| Researcher’s role | Watches from outside the group | Joins the group and takes part |
| Type of data | Quantitative if structured, qualitative if unstructured | Mainly qualitative |
| Validity | Limited: sees behaviour but not its meaning | High: aims for empathetic understanding (verstehen) |
| Reliability | Good if structured | Low: depends on one researcher’s skills and relationships |
| Hawthorne effect | A problem if overt | Reduced over time, or avoided if covert |
| Time and cost | Relatively quick and cheap | Very time-consuming |
| Ethics | Fewer problems if overt | Serious problems if covert |
| Usually preferred by | Positivists (structured version) | Interpretivists |
For the undercover version of participant observation and its ethics, see covert participant observation.
Non-participant observation in education
The classroom is the natural home of non-participant observation. Lesson observation is already part of school life, through Ofsted inspections and schools’ own internal observations, so a researcher sitting at the back of a lesson is less unusual than almost anywhere else.
Applied to education, the main issues are these. Practically, gaining access still means persuading head teachers and individual teachers, and the sample is likely to be self-selecting. Theoretically, a structured schedule gives reliable, comparable data on things like how much time is teacher talk, or how often pupils are praised, but it cannot tell you what pupils think about their lessons or why they behave as they do. The Hawthorne effect is a particular problem, because teachers who know they are being observed may teach their ‘inspection lesson’. Ethically, the main issues are consent, especially from pupils, and the disempowering experience of being judged.
For a worked exam answer applying all of this, see Evaluate the strengths and limitations of using non-participant observation to investigate pupil behaviour in schools (20).
Applying this in the exam
Non-participant observation can come up in three ways.
Paper 3, theory and methods. As a method in its own right, or as part of a question on observation, quantitative methods or the positivist approach to research. The comparison with participant observation is the most useful thing to have ready.
Paper 1, theory and methods. In the 10-mark ‘Outline and explain two…’ question at the end of the paper, which can be on any methods topic. For example, you might be asked for two advantages or two problems of using non-participant observation. Remember there are no marks for evaluation on these questions.
Paper 1, Methods in Context. Non-participant observation was the Methods in Context question in November 2020, applied to investigating pupil behaviour in schools. Examiners reported that the weaker answers wrote about observation in general and didn’t apply the method to pupil behaviour. Stronger answers applied the specific features of non-participant observation to the topic. For example, a detached outsider is less likely to understand the meanings pupils attach to behaviour within pro- or anti-school subcultures.
Students often struggle to find much to say about non-participant observation because it seems simple. The easy marks come from the structured/unstructured and overt/covert distinctions. The good marks come from applying the ‘you can’t ask why’ problem to whatever the question is about.
Theory and Methods Revision Resources
If you like this sort of thing, then you might like my Theory and Methods revision bundle.
[IMAGE: Theory and Methods revision bundle cover, linked to https://revisesociology.sellfy.store/p/2qyB/]
Written by Karl Thompson, who taught A-level sociology for eighteen years and was an AQA examiner across all three papers from 2010 to 2018.
Signposting and Related Posts
This post has been written primarily for students studying the research methods topic as part of the AQA’s A-level sociology course.
Non-participant observation is usually taught alongside Overt Participant Observation: Advantages and Disadvantages, which covers the method where the researcher joins in, and Covert Participant Observation: Examples and Ethics, which covers the undercover version.
Field experiments are closely related to non-participant observation, since the researcher often observes the effects of a change without taking part.
The key concepts used throughout this post are explained in more depth in the posts on validity, reliability and positivism and interpretivism.
For the rest of the topic, see the research methods hub.
Please click here to return to the homepage – ReviseSociology.com
Sources
Bryman, A. (2016) Social Research Methods, Oxford University Press
AQA (2020) Report on the Examination: A-level Sociology 7192/1, November 2020
Flanders, N. (1970) Analyzing Teaching Behavior
Galton, M., Simon, B. and Croll, P. (1980) Inside the Primary Classroom, Routledge and Kegan Paul
Galton, M., Hargreaves, L., Comber, C., Wall, D. and Pell, A. (1999) Inside the Primary Classroom: 20 Years On, Routledge
Ofsted (2018) Ofsted inspections: clarification for schools