A distinction that misses the point

Devi Jankowicz writes.

In the May issue (‘News’) Professor Daryl O’Connor and Professor Andy Tolmie are reported expressing concern on behalf of the BPS about the way in which certain journals are banning research papers based on qualitative methodology – a concern that I share – but why on earth should one have to choose between qualitative and quantitative approaches in the first place?

Using a highly quantitative technique such as factor analysis, we still have to make a qualitative judgement in order to choose names for the factors so as to best express the way in which variance has been partitioned across the variables. Conversely, at the end of a definitively qualitative procedure such as content analysis, we engage in quantitative procedures when we ask ‘which categories are more important/comprise the largest frequency of categorised items’. And of course we check the reliability of both the category set, and the coding of meanings to categories, by applying a quantitative procedure (percentage agreement, Cohen’s kappa, the Perreault-Leigh Index, whatever) to the qualitative judgements that have been made, independently of one another, by the researcher and a collaborator. The qualitative and the quantitative should go hand in hand if either is to be credible.

Some methodology journals (e.g. Journal of Mixed Methods Research; Qualitative Health Research) agonise, in my view unnecessarily, over the qualitative–quantitative divide, and related terminologies of ‘objectivity versus subjectivity’, with little attention being paid to a genuine, and very important, distinction: between constructivist and positivist epistemologies, and their associated phenomenological versus realist ontological stance.

Subject-matter journals express positivist or constructivist preferences since they deal with particular kinds of research question best addressed from one or other of these epistemological positions – fair enough.

But a ‘qualitative versus quantitative’ distinction is not synonymous with the constructivist/positivist distinction. There is an association, to be sure, which we convey to our undergraduates when defining the latter, but the two distinctions aren’t isomorphic; and to base a journal policy on the rejection of qualitative (or quantitative!) data collection and analysis techniques is a move into obscurantism which misses the point.

Let’s applaud O’Connor’s and Tolmie’s expressions of concern.

Devi Jankowicz
Visiting Professor, Heriot-Watt University

BPS Members can discuss this article

Already a member? Or Create an account

Not a member? Find out about becoming a member or subscriber