We saw that observation might be used for data collection in both qualitative and quantitative research and that the approach to using it in each is different for reasons related to the types of data it is used to collect and its quality. We also considered that observation forms a useful data collection tool in mixed methods research — although how it is used to collect qualitative and quantitative data does vary.
We identified that qualitative observation is important in identifying how people behave and what they actually do in reality as opposed to what they say they do. We identified that observation allows researchers to study people in a natural setting, but that it is fraught with issues of bias and misinterpretation as observers may sometimes lack objectivity.
Why use qualitative observation?
There are a number of good reasons that qualitative observation is a useful data collection method. These include:
- The area of research is not well defined and needs to be explored first.
- The topic of research is too complex to study using quantitative approaches
- The topic of research needs to be approached using multiple methods and triangulation because it involves complex human activity.
Qualitative observation is widely used where the research expects the data generated to be both rich (detailed and in-depth) and plentiful. In this sense, the researcher must act as a novice and assume that everything they see is important because by making detailed observations they will allow the reality of what they are observing to emerge (Smit and Onwuegbuzie, 2018).
Naturalistic observation
As the name suggests, naturalistic observation is observation that takes place in the natural environment of the observed subject. This does not mean that the observation takes place out in nature, rather it takes place where the people being observed are naturally found. For example, Hunter et al (2023) used naturalistic observation, as well as other methods, to explore decision-making processes among intensive care nurses in the intensive care unit — their natural setting.
Natural behaviours in work or elsewhere are frequently studied by social scientists and psychologists, as well as nurse researchers, because they give a realistic insight into the spontaneous and natural behaviours of individuals. Such observations may be very informative on their own, or might be used to prime future research, e.g. a quantitative enquiry (Creswell and Poth, 2017).
Unlike in the quantitative observation methods, which may include an intervention, qualitative naturalistic observations are undertaken without any preconceptions and the observations tend to be unstructured (Parahoo, 2014). That is to say, qualitative observations are very inductive in their approach to data collection. In qualitative observational, research observation can take place just to see what information might emerge (Ellis, 2022).
Naturalistic observations are often constrained by the fact that they are hard to reproduce due to inter-observer variability. The same observer may notice different things at different times, even when observing the same situation. Notably, naturalistic observation might involve the observers in literally observing what is going on; however, their very presence can alter the ways in which people behave (Pope and Mays, 2020). Hence much qualitative observational research involves the researcher becoming part of the group being observed and participating in their activities so that people normalise to their presence — although there are issues with this, as we will see.
Participant observation
Participant observation is a variation on the theme of naturalistic observation where the observer participates in the activity that they are observing. There are two main approaches to participant observation: overt and covert.
In overt participant observation, the researcher makes it clear what they are doing, usually gains some form of consent, and may observe a group over a period of time or as a one off or for a limited time. Observing over a period of time, even overtly, does mean that the people being observed become used to seeing the researcher and revert to their usual ways of behaving (Ellis, 2022). Covert observation is much less common in health and social care research because of issues with obtaining consent.
Participant observation is commonly used in nursing research, for example, in a study looking at meaning in life for older people, Hupkens et al (2019) observed 197 community nursing interactions with older people over a 33-month period. While the observers limited their participation to joining in conversations and supporting with nursing tasks, these were undertaken to make the observation appear more natural and prevent the conversations becoming uncomfortable.
An issue with long-term participant observation is that the researcher may start to take on the values of the group under study as they become an insider. This process occurs as the researcher immerses themselves in the world of the people they are researching. So while immersion enables the researcher to see more of the nuance and complexities of the social world that they are studying (Jones and Smith, 2017), it can also mean that they lose some objectivity and fail to observe what is going on around them.
With long-term observation, those being observed can start to normalise to the researcher’s presence. When this happens, participants may forget that they are participating in research. While this may be good for the data collection as people act as they would naturally, it does mean that there are concerns over the validity of their ongoing consent (Pope and Mays, 2020).
Recording and analysing observations
A key challenge of observational research is recording what is seen. On the one hand, the researcher may collect a lot of data during the observation of an interaction between a nurse and a patient, for example, but on the other hand may miss a lot of what is going on by trying to write all of what they see in real-time. Waiting until after the event to record their observations may mean that the researcher forgets some of what they have observed or that they start to apply interpretations, which means they record things which did not happen (Ciesielska et al, 2018).
Observation as a data source can be quite voluminous as the observer tries to capture and describe their observations. It is often best that any data collected using qualitative observations are analysed using a structured approach, such as that which might be enabled using a framework. Frameworks enable the researcher to present a lot of data in a structured and logical fashion such that patterns can start to emerge from the data in a way that both the researcher and readers can understand (Jones and Smith, 2017).
Conclusions
Qualitative observational research is about capturing how people behave in their natural environments, which in the case of health and social care, is usually their working environment. We have seen that it is used to research topics too complex to be pigeon-holed into other forms of research approaches or where multiple research approaches need to be applied to gain a meaningful understanding of a topic.Qualitative observation can be overt or covert and that it may be long or short-term and often involves the researcher immersing themselves in the life of a group of people, such as those working in a particular care setting. It can be difficult to capture all of the data that a researcher observes and furthermore, the data collected is best analysed using a framework to allow meanings to be elicited.
