Observation

Definition of observation

Observation in research

In sociological research, 'observation' refers to a method in which researchers study the ongoing behaviour of their participants (or subjects). This is different from techniques such as interviews or questionnaires because observations are a study of what subjects do instead of what they say.

Observation is a primary research method. Primary research involves personally collecting the data or information being studied. This is the opposite of the secondary research method, where researchers choose to study data that has already been collected before their study begins.

Fig. 1 - Observations capture behaviour instead of words

Types of observation in sociology

There are several types of observational methods used across many social science disciplines. They're each suited to different research purposes, and have different strengths and limitations.

It's important to note that observational methods can be covert or overt.

• In covert research, the research participants don't know who the researcher is, or that there's even a researcher there at all.

• In overt research, the research participants are all aware of the researcher's presence and their role as an observer.

Participant observation

In participant observation, the researcher integrates themselves into a group to study their way of life, their culture, and how they structure their community. This technique is commonly used in ethnography.

The fact that researchers have to be integrated into the group's way of life means that they need to find a way to be let into the community.

However, many communities don't want to be studied. So, the researcher can either earn the trust of certain members and seek permission to study their way of life (overt observation), or the researcher can pretend to become a member of the group to gain access to information (covert observation).

Conducting participant observation

While conducting participant observation, the researcher should focus on capturing an accurate and authentic account of the community's way of life. This means that the researcher has to avoid influencing the behaviour of anyone in the group.

Where simply observing the crowd isn't enough, the researcher might need to ask some questions. If they're conducting covert research, they might enlist an informant. The informant will be aware of the researcher's presence and can answer questions that are not addressed by observation alone.

Taking notes is more difficult when they're acting covertly. It's common for researchers to pop into the bathroom to make a quick note of something important, or to summarise their daily observations every evening. Where the researcher's presence is known, it's relatively simple for them to take notes, because they don't need to hide the fact that they're conducting research.

Theoretical framework

Observational research falls under the paradigm of interpretivism.

Interpretivists value participant observation because the researcher has the opportunity to understand the subjective experiences and meanings of the group being studied. Instead of applying their own understandings to unfamiliar behaviours, the researcher can achieve higher levels of validity by observing actions and getting a sense of what they mean to the people who are carrying them out.

Ethical concerns

It's important to consider the moral rights and wrongs of research before we start conducting it.

Covert participant observation involves lying to the participant - it is a breach of informed consent. Also, by becoming a part of a community, the research risks their impartiality if they become attached (emotionally, financially, or otherwise) to the group. The researcher can potentially compromise their lack of bias, and thus the validity of the research as a whole. What's more, if the researcher integrates themselves into a deviant community, they could put themselves at risk of psychological or physical harm.

Non-participant observation

In non-participant observation, the researcher studies their subjects from the sidelines - they don't participate or integrate themselves into the lives of the group they are studying.

Conducting non-participant observation

Non-participant observation can be either structured or unstructured.

Structured non-participant observation involves some sort of observation schedule. Before they begin their observation, researchers make a list of behaviours that they expect to see. They then use this list to tick off what they see. Unstructured observation is the opposite of this - it involves the researcher freely noting down whatever they see.

Moreover, non-participant research can be overt. This is where the subjects are aware that they are being studied (like the headteacher sitting at the back of a class for one day each term). Or, the research can be covert, where the researcher's presence is a little more unassuming - the subjects don't know that they're being researched. For example, a researcher might be disguised as another customer in a shop, or use a one-way mirror.

As strange as it might sound, it's important for researchers to not only take note of what the subjects are doing but also what they aren't doing. For example, if a researcher was examining customer behaviour in a retail store, they might observe that people ask shopkeepers for assistance in some situations, but not others. What are those particular situations? What do customers do when they're uncomfortable asking for help?

Theoretical framework

Structured non-participant observation is generally preferred in positivism.

A coding schedule makes it possible for researchers to quantify their observational findings by marking off when and how often they see particular behaviours. For instance, a researcher studying young children's behaviour in classrooms might want to discern how often they speak without raising their hands. The researcher would mark this behaviour on their schedule every time they saw it, giving them a workable average by the end of the study.

Ethical concerns

As with covert participant observation, subjects of covert non-participant observation aren't able to give informed consent - they are essentially deceived about the occurrence or nature of the study.

Advantages and disadvantages of observational research

The different types of observational methods (participant or non-participant, covert or overt, structured or unstructured) each have their own set of advantages and disadvantages.

Advantages of observational research

• Covert participant observation is likely to have high levels of validity because:

  • Participants are being studied in their natural environment, in which their behaviour won't be swayed by the known presence of a researcher.

  • Researchers can gain the trust of their participants, and get a better idea of not only what people do, but how and why they do it. This is beneficial to making assumptions by applying their own understandings to observed behaviours.

• Non-participant research is generally cheaper and quicker to do. It doesn't require time and resources for the researcher to integrate into an unfamiliar community.

• The quantitative nature of structured observations makes it easier for researchers to make comparisons between different communities, or the same community at different times.

Disadvantages of observational research

• Michael Polanyi (1958) stated that 'all observation is theory-dependent'. What he meant is, to understand what we're observing, we already need to be equipped with a certain amount of knowledge about it.

  • For example, we might not be able to make certain inferences about a table if we didn't know what a table was supposed to look like, or function as. This is an interpretivist criticism of positivist research methods - in this case, of structured observation.

• Observations usually involve intensively studying relatively small or specific groups. Therefore, they are likely to lack:

  • representativeness,

  • reliability, and

  • generalisability.

• There's a risk of the researcher adopting the behaviours of the group that they are studying while doing overt, participant research. While this isn't inherently a risk, it could be if they are examining the behaviour of a deviant group.

• Overt observation, whether the researcher is a participant or not, risks the validity of the study due to the Hawthorne effect. This is when participants may change their behaviour because they know they are being studied.