Observation Methods

Watching behaviour in its setting

Observation is a data-collection method that records behaviour directly rather than relying on self-report. It focuses on what people actually do, not what they say they do. It can be participant or non-participant, structured or unstructured, and overt or covert. By entering the field, the researcher captures layers of reality inaccessible to surveys or interviews, but important limitations such as observer effects and observer bias must be carefully managed.

What Is Observation?

Observation is a data-collection method in which the researcher systematically watches and records individuals, groups, or events in controlled or natural settings. Its key distinction from surveys and interviews is that the researcher obtains behavioural data from a primary source without requiring participants to interpret their own actions. This largely eliminates self-report biases such as social desirability and recall error. Observation is widely used across disciplines including education, health sciences, organizational behaviour, anthropology, and sociology.

Main Types and How It Works

Observation is classified along four main dimensions. (1) In participant observation the researcher joins the group and collects data from within; in non-participant observation the researcher remains outside. (2) Structured observation counts specific behaviours using pre-designed coding schemes or checklists; unstructured observation notes emerging patterns freely in the field. (3) In overt observation participants know they are being watched; in covert observation they do not. Each dimension entails distinct methodological and ethical choices. A systematic observation process includes defining the focus, obtaining ethical approval, piloting the instrument, establishing a data-recording protocol, and testing inter-rater reliability.

A Concrete Example: Classroom Observation

Consider a researcher studying interaction patterns in primary school mathematics teaching. Using a structured approach, the researcher enters the classroom and records teacher-student interactions (questioning, feedback, individual support) into a pre-designed coding form every five minutes. A non-participant position allows the natural teaching environment to be maintained while the teacher remains the observed subject. Choosing overt rather than covert observation fulfils ethical responsibilities, but the researcher accounts for behaviour changes linked to observer presence (the Hawthorne effect) by planning multiple sessions and tracking change over time.

Common Pitfalls and Good Practice

The most common challenges in observation are as follows. Observer effect occurs when participants alter their behaviour because they know they are being watched; scheduling acclimatisation periods reduces this. Observer bias happens when the researcher's own expectations or values colour the observation; structured coding schemes and multiple observers lower this risk. Covert observations raise serious ethical concerns around privacy and informed consent. Observations reflect a limited context, so caution is needed when generalising. Reflexivity — the researcher's regular examination of how their own presence and perspective shape the data — is a fundamental practice that enhances the quality of observational research.

Key terms

Participant Observation: Observation in which the researcher joins the studied group to collect data from within.
Structured Observation: Observation that uses a pre-designed coding scheme to systematically count specific behaviours.
Observer Effect: Participants changing their behaviour because they know they are being observed.
Observer Bias: Systematic distortion of observation data by the researcher's own expectations or values.
Inter-rater Reliability: Degree to which multiple observers independently code the same behaviours consistently.