There is a body of literature about research designs which focus on the methods to be used, nuances, traditions, authority, experts, paradigms, or schools of thought related to each method. This all makes it sound so complex? As a new researcher, the question in our minds is – what design should I choose for successful completion of my dissertation? The short and clear answer to this question is -the research design should follow from and be able to answer the research question. Through this article, I aim to present a brief decision making guide for students to make it easy for them to choose a research design which will help them answer their research question and to finish their dissertation on time.
So what exactly do we mean by research design and why is it important? In plain terms, research design is the plan and procedure for research that gives direction and systematic approach to any research project. It includes assumptions made during study planning, methods for data collection, data analysis, presentation of results and drawing of conclusions. So it is the soul of any research project.
Design should stem from the question being asked by researchers
In the first article of the Physiotimes research series, I explored means of formulating a research question, which is then further refined through a literature review. So once the question has been decided, it is time to explore the various means through which it can be answered.
If the research question is ‘What is the relationship between work stress and risk of cardio vascular mortality?, the design should allow us to measure the level of work stress a set of employees are experiencing at their work place and their baseline cardiovascular status and then follow them for a period of time (till end of study) to assess whether higher work stress had any association with cardiovascular mortality through use of appropriate statistical correlational techniques. Such a design is non-experimental or observational design, where you are not influencing the outcome but rather just observing a set of participants over a period of time.
If the research question is ‘What are the factors which Asian patients perceive to be important in uptake of physiotherapy services?’, the design should allow us to help identify and characterise these factors through in-depth discussion with Asian patients, therefore use of qualitative methods like indepth semi structured interviews or focus groups might work best in this scenario.
If the research question is ‘In patients with low back pain, is the use of manual mobilization superior to therapeutic ultrasound to alleviate pain and increase functional capability?’, the design should allow us to compare the two treatments – manual mobilization vs therapeutic ultrasound’ in a group of low back pain patients, therefore use of experimental, randomised design would provide the best answer to our question. The name randomised control trial should not be the major focus, but rather the choice of this design stems from the fact that a non-biased comparison of two treatment approaches is required which can follow up the patients for a period of time to determine the effect of these treatments on the chosen outcomes of pain alleviation and increased functional capability.
I will utilize one final example of a research question which explores -‘What is the effectiveness and acceptance of a novel mobilization technique in management of lateral epicondylitis? In this question, two aspects of the new technique need to be explored – effectiveness as well as acceptance. Now as discussed in the previous question, use of randomized experimental designs is optimal for evaluating effectiveness of interventions but it will not answer the question about acceptance of the technique by the patients and therapists. We can utilise semi structured interviews with patients and therapists to obtain their perception of the technique. The combined results will therefore answer the research question completely. Such designs which utilise various different approaches are called as mixed methods research designs.
This highlights the need for a careful consideration of your research question as the research design selection should also be based on the same. While I have suggested that these questions be answered through such methods as described above, other designs may also be used to answer these questions. For e.g. the question about role of motivation as a barrier to adherence in physiotherapy can be explored through indepth interviews with patients, through a focus group with patients, through an observational study of a group of patients, where they are surveyed about their motivation and subsequent adherence is observed.
Other considerations in choosing research design
As discussed above, while the most important element in choosing a research design is the research question, there are other aspects which need consideration as well. The research design approach chosen should fulfil these three criteria as well – reliability, replication and validity.
Reliability is an indication of the ability of a system to perform and maintain its functions consistently in routine circumstances as well as hostile or unexpected circumstances. Reliability is particularly important in quantitative research and may refer to the statistical reliability of a set of data as well as the experimental reliability of an experiment (Bryman 2004).
Replication refers to the ability of an experiment or study to be repeatable by other researchers. A study must be replicable in order that the reliability of a measure or a concept can be determined. This means that the design and methods that you choose should be clear enough for someone else to be able to repeat the study in similar or different circumstances.
Validity is described as the degree to which a research study measures what it intends to measure. Face validity refers to whether a technique looks as if it should measure the variable it intends to measure. For example, a method where a participant is required to click a button as soon as a stimulus appears and this time is measured appears to have face validity for measuring reaction time. Construct validity relates to whether the measurements of a variable in a study behave in exactly the same way as the variable itself for e.g. does an IQ test really measure variations in intelligence? Internal validity refers to- if we suggest that x causes y, can we be sure that it is x that is responsible for the variation in y and not something else?, while external validity relates to whether the results of the study can be generalised to the whole population or are those restricted to the specific experimental/ research participants.
There are other aspects of research design which will be decided based on time and budget. While the sample size calculation for an RCT may say that you need 1100 participants, it is not feasible to recruit or manage these numbers in a BPT or MPT dissertation, so you have to realistically think about the numbers that are available within your own chosen timeframe and then go from there. Budget is also an important consideration as it leads to several constraints in planning and conducting the study. For example, while your research question may demand that you undertake indepth interviews with expert physiotherapists in a subject area, it may not be practically feasible for you to visit them to complete the interviews and instead you would have to undertake telephonic interview or with the advent of online technology, use Skype or other such interface. These will have own set of ethical issues which will need to be addressed while planning your research.
Summary of research designs and when to use them
The main purposes of a research are exploration, description or explanation of a phenomenon. Exploratory research tried to understand aspects of a phenomenon which are not well understood or previously explored. Descriptive research is aimed at providing an account of reality while explanatory research attempts to find causal relationships among the variables that are measured in the study.
In general, there are three main research design paradigms – quantitative, qualitative and mixed methods. Wikipedia defines quantitative research as the systematic empirical investigation of social phenomena via statistical, mathematical or computational techniques, while qualitative research as a design which aims to gain an indepth understanding of human behaviour. Mixed methods research designs utilize a combination of quantitative and qualitative methodologies to answer the research question.
The quantitative designs are classified as descriptive (key focus is to describe – ‘What is e.g the purpose of this research is to describe the current attendance rates in outpatient physiotherapy clinics), correlational (key focus is to explore relationships amongst study variables for e.g. Is old age related to increased nonattendance at outpatient physiotherapy clinics), quasi experimental (key focus is to test casuality with sub-optimal control e.g. test the effectiveness of a motivational cognitive bahavioural program in older age patients in enhancing adherence at outpatient physiotherapy clinics) and experimental (key focus is to test casuality with optimal controls e.g. effectiveness of telephone reminders in improving attendance rates in outpatient physiotherapy clinics). I have only included a broad overview of types of experimental design to aid decision making. Each of these quantitative designs has further sub-classifications e.g factorial, nested or cross over experimental designs.
Qualitative research is by definition exploratory, and it is used when we don’t know what to expect, to define the problem or develop an approach to the problem. It’s also used to go deeper into issues of interest and explore nuances related to the problem at hand. Common data collection methods used in qualitative research are focus groups, triads, dyads, in-depth interviews, uninterrupted observation, bulletin boards, and ethnographic participation/observations.
Mixed methods research involves use of concurrent or sequential use of various qualitative and quantitative data collection and analysis methods to answer a research question. For.e.g my PhD involves use of mixed methods approach to evaluate factors related to attendance and adherence in outpatient musculoskeletal physiotherapy. So I used a mixed methods research synthesis, followed by a prospective cohort study and finally interviews and focus groups with patients and physiotherapists to collect data which will be analyzed to provide a detailed perspective of the determinants of adherence.
In summary, qualitative research can be used to develop an initial understanding of an issue or problem, look for range of ideas and feelings about something, understand different perspectives between groups, understand factors that influence decision making, help in designing or explaining the results of a quantitative study. Quantitative research is generally used to find whether there is consensus on an issue, project results to a larger population, evaluate cause-effect relationships and test hypotheses.