Teaching research methods can sometimes feel abstract for students. Terms like reliability, observer agreement and structured observation are easy enough to define, but much harder for students to genuinely understand unless they experience the problem for themselves. That is why this reliability test works so well. It takes a key methods issue and turns it into a practical, memorable classroom activity.

In non-participant observation, the researcher watches social behaviour without joining in. At first glance, students often assume this makes the method quite objective. After all, if everyone is observing the same event, surely they should record the same thing. In practice, this is rarely so simple. Different observers notice different details, interpret behaviour in different ways and may apply the same categories inconsistently. This activity helps students see that reliability is not automatic. It has to be built into the research design through clear categories, shared definitions and careful observation.

For teachers, this makes the activity particularly useful. It is easy to run, highly interactive and creates strong discussion around the difference between what students see, what they record and what they think it means. It also gives students a practical route into broader theory and methods issues such as objectivity, validity, operationalisation and the strengths and limitations of observational research.

Why this activity works so well

This task is effective because it gives the whole class a shared piece of evidence. Every student watches the same short video clip, and every student uses the same coding categories. In theory, this should produce similar findings. In reality, it usually does not. Some students will count an interruption where others see overlap in conversation. Some will identify dominance where others simply see confidence. Some will miss smaller gestures entirely. That immediate disagreement is what makes the activity so powerful.

Students begin to understand that reliability depends on more than simply watching behaviour. It depends on how clearly the researcher defines what is being counted. This helps move them beyond textbook definitions and towards a more applied understanding of research design.

It is also a useful confidence-building activity because students do not need extensive prior knowledge to take part. They only need to observe, record and compare. The theory comes alive through the task itself.

---

The activity: Reliability test using a video clip

Aim

To help students explore whether non-participant observation can produce reliable data.

Learning focus

Students should come away understanding that:

• non-participant observation can look objective, but still involves interpretation
• reliability depends on clear coding categories
• different observers may notice or classify the same behaviour differently
• structured observation can improve consistency, but does not remove all subjectivity

---

How to run the activity

Step 1: Choose a short video clip

Select a clip of social behaviour lasting around one to three minutes. It should show visible interaction between people and include behaviours that can be observed and coded. The clip should not be so busy that students cannot follow it, but it should contain enough ambiguity to create disagreement.

Step 2: Give students a coding frame

Before showing the clip, provide all students with the same categories. Keep the number fairly small so that students can focus clearly. For example:

• interruption
• agreement
• laughter
• conflict
• ignoring another person
• one person leading the interaction
• supportive behaviour
• rule-breaking

Students should use these categories to record what they observe during the clip.

Step 3: Show the clip

Play the clip once so students can get a sense of what is happening. Then play it again so they can record their observations more carefully. In some classes, a third viewing may be useful.

Step 4: Compare findings

Ask students to compare their results in pairs or small groups first, then bring the discussion to the class.

Questions to ask:

• Did everyone notice the same events?
• Which categories produced the most similar answers?
• Which categories caused disagreement?
• Did some people interpret the same action in different ways?
• Would a different researcher using the same categories get the same results?

Step 5: Improve the coding frame

Once disagreement has emerged, ask students to redraft the categories so they are clearer and easier to apply. For example, instead of “being rude,” they might write “speaks over another person before they finish their sentence.” This helps students see the importance of operational definitions.

---

What types of video work best?

The best video clips are those that contain clear social interaction, visible behaviour and some room for interpretation. You want students to be able to observe actions, but also to disagree about what exactly those actions mean. That balance is important.

1. Group discussion clips

These are often the strongest option. A short discussion between several people gives students plenty to observe, including turn-taking, interruption, leadership, silence, body language and agreement or disagreement. These clips are especially useful if you want to explore power, dominance or group dynamics.

Examples could include:

• a classroom discussion
• a team meeting
• a student group task
• a panel discussion
• a debate clip

These work well because there is enough interaction for students to code, but usually not so much movement that the task becomes chaotic.

2. Playground or break-time interaction

These clips are useful for showing informal social behaviour. Students might observe friendship groups, exclusion, conflict, cooperation or use of space. This type of clip is particularly effective if you want students to think about peer groups, status or social norms.

These also create good discussions around validity, because students can often see behaviour clearly but cannot always know the motives behind it.

3. Waiting room, queue or public-space behaviour

A quieter clip can actually be very effective. Videos showing people in queues, waiting rooms, transport settings or other public places allow students to observe norms, personal space, compliance, impatience, avoidance and reactions to others. These are often easier to code because there is less talking over one another.

4. Documentary or reality-style interaction

A short clip from a documentary or reality-based setting can be engaging for students, especially if there is tension, disagreement or role differentiation. These are useful for coding conflict, leadership, support or emotional reactions.

This option works best when students are reminded that edited footage may shape what appears important.

5. Sports or team-coaching clips

A coaching exchange, team talk or group practice scene can be excellent for observing authority, compliance, encouragement, resistance and leadership. Students can identify who takes control, who follows instructions and who is ignored or challenged.

Some suggested videos are as follows:

Socialists vs Capitalists | Middle Ground and other Jubilee Middle Ground videos are strong for turn-taking, interruption, alliance-building, facial reaction, dominance and disagreement. They usually give you several speakers and clear contrast in viewpoints. The most appropriate section for students of Cambridge OCR A Level Sociology is on the second debate point of Economic inequality is the root of most social problems from around 9:40 seconds in. Disclaimer: There is brief, but not explicit, discussion of domestic violence in this section, focusing on economic inequality as a cause of domestic violence, particularly financial violence.

Alternatively you can play a segment of this video with instructions and discussion from myself via the link below:

A second clip that can be used is Brain Games Conformity Waiting Room, Social Experiment: Cutting Queues. This focuses on conformity is a waiting room and may also be suitable for students of psychology. The main distracting feature is the voiceover.

For sports or team coaching clips, a good example to analyse is Mikel Arteta’s Emotional Dressing Room Team Talk. This is a good choice for authority, morale, leadership style and emotional influence

Additionally, you could use a selection from Men vs Women on The Island | The Island with Bear Grylls looks useful for role differentiation, conflict and group survival dynamics, though it is longer, there are several sections you might choose to use for observations.

---

What types of clips should be avoided?

Some clips are less effective for a reliability activity because they create confusion for the wrong reasons.

Try to avoid:

• scenes with too many people moving at once
• clips with poor sound or lots of overlapping speech
• heavily narrated clips where the meaning is already explained
• very emotional or dramatic scenes where students focus on the drama rather than the coding
• clips that are too long, causing students to lose focus

The aim is to test reliability in observation, not memory, and not students’ ability to cope with chaos.

---

Suggested coding categories

The best categories are visible, specific and reasonably countable. They should focus on behaviour rather than hidden motives.

Useful categories include:

• speaks first
• interrupts
• laughs
• points or gestures
• moves away
• makes eye contact
• supports another person
• ignores another person
• gives instructions
• breaks an obvious rule
• stands apart from the group
• responds to challenge

Less useful categories include vague judgements such as:

• being unfriendly
• feeling nervous
• acting strangely
• being rude

These are still valuable for discussion, because students quickly realise how subjective they are, but they are not ideal if the goal is to test reliability.

---

Differentiation ideas

This activity can be adapted quite easily for different class profiles.

For students who need more support

Use a very short clip with only two or three people. Reduce the number of categories to perhaps four or five. Focus on simple, visible behaviours such as speaking, interrupting, laughing and moving away. You could also model one category with the class before they begin.

Another useful scaffold is to provide a partially completed tally sheet with the categories already explained in simple language.

For mixed-attainment classes

Give everyone the same clip, but vary the level of support in the coding frame. Some students can use tightly defined categories, while others can be asked to apply broader ones. This allows the class to complete the same activity while working at different levels of complexity.

You could also ask some students simply to identify where disagreement happened, while more confident students explain why it happened in methodological terms.

For more confident students

Challenge students to distinguish between behaviours that are easy to code and behaviours that are harder to interpret. Ask them to consider the relationship between reliability and validity. A highly structured coding frame may improve consistency, but does it reduce the richness of the behaviour being observed?

You could also introduce the term inter-observer reliability and ask students how sociologists might test for it.

For anxious or quieter students

This is a useful activity because it allows full participation without requiring lots of public speaking. Students can code individually, compare in pairs and then contribute through written feedback or mini whiteboards. That makes it a good option for classes where some students are less comfortable with verbal discussion.

---

Extension tasks

Once the main activity is complete, there are several ways to develop it further.

One strong extension is to ask students to redesign the coding frame and then test it again on the same clip. This allows them to see whether clearer definitions improve consistency.

Another is to split the class in two and give each half a different coding frame. One group receives vague categories and the other precise ones. When findings are compared, students usually see immediately that precise categories produce greater agreement. This is a very effective way of demonstrating how reliability is built.

You could also ask students to compare non-participant observation with participant observation. Which method is more likely to be reliable? Which is more likely to be valid? This opens up wider theory and methods discussion.

---

Linking the activity to sociology

Once students have completed the task, it is worth drawing out the bigger methodological issues. This activity is not only about whether students agreed. It is about what that disagreement tells us about research itself.

This can lead to discussion such as:

• why positivists often prefer structured observation
• why interpretivists may argue that behaviour cannot be fully understood through simple categories
• whether a method can ever be fully objective
• how operational definitions shape the data researchers produce
• why observer training matters in sociological research

This makes the activity useful not just as an introduction to non-participant observation, but as a bridge into debates about objectivity, validity and the nature of sociological evidence.

---

Why teachers may want to keep using this activity

This is one of those methods activities that tends to stick with students because they experience the issue directly. Rather than being told that reliability means consistency, they discover how difficult consistency actually is. They see that two people can watch the same event and still produce different results. That realisation gives them a much stronger basis for evaluating observational methods in written work.

It also works well because it is simple to set up, adaptable across different groups and easy to revisit later in the course. It can be used as a first introduction to observation, as part of a wider methods lesson, or as revision before exams.

In short, it is a practical activity that helps students think more like sociologists. It shows that the success of non-participant observation depends not just on standing back and watching, but on making careful decisions about what is being observed, how it is defined and how consistently it is recorded.

---

Final reflection

The reliability test is a very effective way to make a key methods concept concrete. By watching the same clip and coding it individually, students quickly see that observation is not as simple or neutral as it appears. Reliability depends on observer agreement, and observer agreement depends on clear, shared categories.

That makes this activity far more than a quick classroom exercise. It becomes a way of showing students that sociological research is always shaped by the choices researchers make. Even in a method designed to appear detached and systematic, interpretation still matters.