Author: Charlotte Kennedy / Editors: Govind Oliver, Charlie Reynard / Codes: / Published: 03/10/2019
Research methods: A brief overview
There are so many different ways you could research a topic that you’re interested in, how on earth do you choose between them all? Below, we give you a brief overview of the different research methods available. In the coming series ‘Explaining Research Methods in Emergency Medicine’ (ERM-EM) we will delve a little deeper into specific approaches, using recently published Emergency Medicine studies as examples.
This blog series aims to help guide those starting out in research but it is equally applicable to those of you reading and interpreting research for your clinical practice. Understanding research methods can help answer questions such as: are the studies findings applicable to me? Was the author’s choice of methods appropriate to answer the question? Could it have been investigated another way? Should this change my practice?
Where to begin: quantitative vs. qualitative methods
The first step to deciding which research method to use is to think about what the research question actually is. What is it that you want to find out?
In broad terms, there are two different over-arching methodological approaches: these are qualitative and quantitative. Generally speaking, these equate to understanding vs. magnitude.
If you are interested in understanding a phenomenon, a qualitative approach is likely to be most appropriate. Qualitative research endeavors to investigate an experience or situation, exploring the nature of what is occurring and attempting to gain a deeper insight into it. Research questions of this type are usually framed around asking “why”, with researchers aiming to increase their depth of understanding of the topic.
If you want to know about the magnitude of a phenomenon, quantitative methodologies will provide the best answers. Quantitative research investigates topics in ways that can be directly measured or quantified. This might be examining an objective outcome, for example how often something occurs (e.g. the prevalence of a disease), or the relationship between two or more things (e.g. the association between a risk factor and a disease outcome). Different statistical approaches are then used to help understand and describe what is being measured.
| A general overview of the difference between qualitative and quantitative research | ||
| Qualitative | Quantitative | |
| Perspective | Subjective | Objective |
| Research process | Exploratory | Confirmatory |
In reality, there are often a number of different approaches for any one question and sometimes a mixture of both qualitative and quantitative methods, otherwise known as a mixed methods approach, is most suitable. When you’ve decided whether you want to use qualitative or quantitative methods (or both!) you can look at the different methods in each category and decide which will best answer your research question.
Once you’ve decided that qualitative research is the way to go, you need to pin down what you’re trying to learn about or understand and think about how you can realistically achieve this. There are a number of different types of qualitative research methods but the three most commonly used are:
- Phenomenology
This type of research aims to describe people’s experiences of something and to look at how they have made sense of that experience. The aim isn’t to produce a theory or model of the phenomenon by breaking it down into separate issues but to understand all of the experience as a whole, in all its complexity. An example of this might be research that aims to develop a deeper understanding of what patients with life-limiting disorders experience when they visit the Emergency Department (ED).
- Ethnography
This type of research aims to understand cultures and describe the structure and function of the particular group of people being studied. ‘Culture’ in this context has a wider meaning and might include communities, organisations or social groups. So this type of research might aim to understand the culture of an ED by looking at different staff factors such as interactions between individuals, friendship groups or the effect of seniority.
- Grounded theory.
This type of research uses the information gathered during the study of the phenomenon to identify themes and develop theories that are constantly updated as more information is gathered. In this way, the research constantly informs the theories generated rather than one set of theories being ‘tested’. An example of this might be where a researcher wants to understand factors affecting burn-out in Emergency Medicine doctors; they might start off with no preconceptions of what affects people but identify key themes as the research is conducted that they can then explore in more detail as the research progresses.
There are lots of other methods that can be used in qualitative research (too many to list here!) and we will touch on these in the coming series but the above 3 are the most used in healthcare research and hopefully give you a taster as to how useful qualitative research can be.
Once you’ve decided you want to use quantitative methods, you need to think about whether your research will be interventional or observational. The easiest way to tell the difference between these is to ask yourself whether you’re going to change anything about the patient journey. If so, your research is interventional, irrespective of what this change is.
Interventional studies are usually conducted as randomised control trials (RCTs). In a randomised control trial, patients are randomly allocated to different interventions. Interventions can be medication-related but might also be things like smoking cessation advice or different types of follow-up. The new intervention is usually compared to either standard management or to no intervention at all. RCTs are useful because the randomisation process (hopefully) ensures that if you do find differences between groups, these aren’t due to the way patients have been allocated but are due to your intervention. Over time, if enough people are recruited and the allocation is truly random, confounding factors (even those that you didn’t know about!) should end up balanced between each group, meaning that your intervention is the only thing that is different between them.
If you’re not changing anything about your patient journey and you’re just observing what would happen naturally, your research is observational. Observational research can be done in a few different ways. The key study types are:
- Cohort studies.
Cohort studies identify specific groups of patients at the beginning of their patient journey and follow them through it. Patients are usually identified by their exposure to something (a risk factor or an intervention). Groups of patients (exposed vs. unexposed) are then followed forwards to see if they develop the disease / condition of interest. The groups should be as equally matched as possible in every other way. From this you can calculate relative risk.
Cohort studies can also be used to look at the prognosis of a disease. People with the disease are followed forward and divided into those that develop the outcome of interest and those that don’t. This gives you the incidence of your outcome of interest.
- Case control studies.
If you already have a group of patients with the disease of interest and you want to study them, this is the method for you. Case control studies identify patients that already have the disease and look backwards for potential causes. You need a control group without the problem but matched in as many other ways as possible. You then look for differences in exposure between the groups that might suggest potential causes of the condition of interest. From this you can calculate the odds ratio.
- Cross-sectional surveys.
Cross-sectional methods create a ‘snap-shot’ across a population at a single time point, looking at both exposure and outcome at that same point in time. They’re usually used for identifying potential associations that can then be studied more rigorously using other methods. They can also be used to calculate prevalence i.e. the number of cases of a disease or condition present in the population at any one time.
- Case reports or a case series.
Are you trying to study a rare disease or intervention? Then look no further! Case reports are used to describe findings in either one patient or a series of patients. Due to the small number of patients involved, it is often difficult to generalise findings to a wider population but they can be useful when collecting observations on very rare phenomenon.
Final thoughts
We hope this has been a useful introduction and overview of the main aspects of qualitative and quantitative research. Our lists of research methods are definitely not exhaustive. We’ll cover some of the rarer methods along with the more common ones in our ERM-EM series. We will look in turn at different quantitative and qualitative research methods and use published EM studies to illustrate key points. We will also try to explain some of the idiosyncrasies around interpreting and applying the results with each one.
Watch this space…