Thursday, November 23, 2006

Quantitative and Qualitative Research Compared

A slide from a presentation to a graduate course at the University of Toronto's Centre for Industrial Relations. A course supervisor considered qualitative research unscientific, and so in response I created this graphic illustration of the remarkable overlap between quantitative and qualitative work.

It seems to me that this gentle caricature is a particularly effective way of comparing quantitative and qualitative research approaches because it exposes their methodological and epistemological underpinnings.

In the scatterplot shown on the left, numerous cases are plotted and the suggestion of a downward trend is observed -- but only if the data are 'fit' to a line on an assumption that trends exist in the sample. Ultimately outliers are quarantined and removed from analysis.

The poem shown on the right illustrates a single word ("measure") that diffuses like smoke in the direction of the wind. In this case the outliers are what give shape and meaning to the data. In qualitative work the researcher is much more suspicious of efforts to 'fit' data to a curve.

The intent here is not to argue that one research method is preferable to the other, but rather to underscore the reality that quantitative and qualitative approaches are equally suited to different kinds of research. Quantitative methods are best used with large samples and many cases where the objective is to prescribe or predict the behaviour of a population. Qualitative research, conversely, involves inquiry from the inside and an interest in studying entities in their natural settings. Qualitative research typically involves fewer cases but many variables, and is characterized by complexity, interrelation, and a holistic approach.

A picture worth a thousand words? I think so. Please do not use without permission and attribution.


Anonymous said...

So what you are saying is that they both may be used with the same group of people or places but they simply measure things as either direct data ie numerical way while qualitative deals more with the abstract.

Amy Lavender Harris said...

I'd say that's a reasonable characterization.

I'd add, thought, that quantitative measurements focus on large populations and are most useful when there is a need for prediction or prescription. Gaging how large airplane seats should be, or how effective (or dangerous) a new drug is, or how effective a teaching tool is -- these situations benefit from quantitative analysis.

But the experience of flying in an airplane (as part of research focusing on safety), or the benefits of a palliative drug to a dying patient, or a student's response to a particular learning style -- these situations fit well with qualitative approaches.

My summary is that quantitative measurement is most interested in central tendency (the phenomenon where large sample sizes tend to congregate around an average), while qualitative approaches are as likely to be interested in outliers as they are with the middle. (It might surprise you, though, to learn that qualitative researchers also use numbers and statistical tools, albeit in different ways).

One approach isn't more useful than another; they just do slightly different things.

And in my opinion, research is best served when both approaches are applied in a study.