Article Text
PDF
Statistics from Altmetric.com
Introduction
Observational studies often draw on the principles of ethnography. Ethnographic approaches to data collection were first used in anthropology and involve describing a culture and learning about the ‘native's point of view’.1 ,2 To gain an understanding of the culture being examined, the researcher is immersed in the field collecting data over a prolonged period of time, often for a year or more. Sometimes time constraints mean it is not possible to undertake a prolonged period of data collection nor is this always appropriate. However, to gain an understanding of what is happening in practice, adopting some of the principles of ethnography can be useful. (N.B. cohort, cross-sectional and case-control studies are also described as observational studies in the scientific literature. These specific methodologies are not within the scope of this paper).
Why are observational studies needed in nursing?
When using questionnaires and interviews sometimes a social desirability approach impacts on participants' responses, where they say what they think the researcher wants to hear rather than what they actually believe or do. This makes it hard to find out what is really happening in practice. Observation allows a first-hand account of behaviours, events, actions and interactions to be obtained,3 ,4 thus allowing a picture of actual practices to be captured. So if the aim of a study is to understand nursing practices, set in the context of the nursing practice environment, then an observational study should be considered.
Observational stance
Before starting data collection it is important to decide on the observational stance that best matches the study aims and study environment. Observers can adopt a detached role (non-participant observation) or can participate in the activities being observed (participant observation). In between these two stances, there are several positions that observers can occupy.1 ,5 So for example, in one study exploring parents' and children's involvement in decision-making in the hospital setting a non-participant stance was taken where the researcher sat unobtrusively in a corner and did not engage in any activities.6 In another study, the role of the ‘observer as participant’ was adopted whereby the researcher could shadow the nurse.7 The researcher did assist with basic nursing tasks such as bed-making; bed baths by, for example, handing the nurse a towel or a bar of soap; and opening packets of cotton wool and gauze when a child's surgical dressing was being changed.
Does the researcher's presence impact on practices observed?
The knowledge they are being observed often makes people self-conscious which may affect their behaviour (the Hawthorne effect). However, the results of studies carried out in several areas of nursing suggest that, as participants grow accustomed to the observer's presence, their behaviour will more closely resemble normal, everyday behaviour.8 ,9 A key strategy in minimising the effect of the researcher's presence on behaviour is taking time to build a rapport with the participants before starting data collection. So, if the study involved observing nurses, a first step might be to attend team meetings or ward handover on several occasion as a way of getting to know the potential participants and building a relationship with them. One way of testing whether the researcher's presence is impacting on the observed practices is to look for signs of habituation (reverting to normal behaviour). For example, in one study the researcher noted that once the staff on the ward on night duty began to feel comfortable with their presence, they began to leave the door of the drug room open again.10
Ethical considerations when undertaking observational research
Some of the ethical issues that should be considered when undertaking observational studies are outlined in table 1.
Ethical issues that need considering when carrying out an observational study
Data collection in observational studies
Data collected in observational studies can be qualitative, quantitative or both. Quantitative data is normally collected against a measure of best practice. If you are collecting qualitative data this will normally be recorded in field notes which will include, in a nursing context, your observations, notes relating to patient records and details of conversations you have had with participants. Using a combination of data collection methods ensures that data outside the preset codes is recorded thus providing a fuller picture of actual practices.15 ,16
Footnotes
Twitter Follow Alison Twycross at @alitwy
Competing interests None declared.
Provenance and peer review Not commissioned; internally peer reviewed.