Research Design¶

Research design is a broad term referring to the work that researchers do when planning a research activity. It is a decision-taking process (like all design activities) that aims to maximize research outcomes by adjusting the activity to the type of the phenomena under study, the opportunities available and the restrictions imposed.
A research design may take various forms depending on the domain, the objective of the research, the type of available data and the understanding one seeks to elicit.
A good point to start with is the three major research paradigms in research design:
- a) Quantitative paradigm: includes methods for measuring quantified variables (for example, temperature, income, learning performance, etc.), process relevant quantitative data and reach conclusions by testing hypotheses or implementing predictive modeling techniques.
- b) Qualitative paradigm: focuses on collecting and analyzing data about qualities (for example, people's opinions or behavior), that is, non-measurable constructs that help develop a deeper understanding and better interpretation of the subject under study. Methods like ethnography, phenomenology and phenomenography are typical examples of the qualitative approach.
- c) Mixed method paradigm: aims to apply both quantitative and qualitative analysis in a complementary fashion in order to get the best out of the two worlds. In general mixed methods research aims to elicit and synthesize the deeper understanding offered by qualitative analysis and the more rigorous testing/predicting techniques based on quantitative approaches.

Three steps in quality research design¶

(a) Applicability of the method¶

Applicability refers to whether a specific research method is appropriate to apply in a specific research situation. Not all methods are appropriate for all research settings; for example, it is meaningless to apply statistical analysis when you have a really small sample (say, 5 participants). The researcher should understand the pros/cons of the method and decide if it is apprppriate for any given situation.
There is a nice presentation of the various research design methods here but don't rush to read and memorize it. Deciding on applicability is a skill developed in time and after the researcher has tried several designs (and probably failed in some).

(b) Implementation details¶

Knowing that a method is appropriate in a situation does not mean that implementation will be without pain. Each research method requires its own "implementation know how", which typically includes: sequencing implementation actions and operations, managing any possible pecularities and "watch out for" points, and skillful use of necessary tools.
Implementation details, of course, vary depending on the type of the research method and the kind of prerequisites and standards for its reliable and efficient implementation.

(c) Reporting¶

Reporting the implementation and results of a research method (usually in a field journal and/or conference) is another kind of skill that the researcher should develop. Usually, one learns this stuff in practice by reading high quality publications, observing and understanding the rationale of presentation structure and possibly improving it.
A serious researcher does not underestimate the practicalities of reporting as it is not just a nuisance in research activity, but an essential part of scientific endeavor. Authoring high level manuscripts (scientific papers) improve the researcher's profile in the field community and -something more practical- make her life easier when dealing with reviewers' comments and criticism.

Where to start from¶

As research design is in many ways an ill-structured domain a starting researcher can not approach the field in a linear fashion, seeking to sequentially browse and understand all possible designs before attempting to implement one.
The approach shoud rather be non-linear and case-based (with a good dose of criss-crossing) as Cognitive Flexibility Theory suggests.
How does the above translate? I always suggest to my doctoral students to get familiar with and implement one well established and widely implemented research design method from each of the two major paradigms. More concretely, I suggest:
- Experimental design from the quantitative perspective, and
- Case study from the arsenal of qualitative methods.
Developing initial expertise while implementing these methods constructs the "seeds" (that is, internal representations to base further development) necessary for anchoring future experiences and understanding of other similar or even different methods.

So, what's next?¶

Having read the above, I guess you are not surprised that I go on now to discuss experimental design and, later on, hypothesis testing techniques.

Copyright¶

. Free learning material
. See full copyright and disclaimer notice