By Jens Marvin Bruun

These days, Lars Qvortrup, Dean of DPU, speaks out for the theory that "educational research must not be either or, but both and. Both qualitative and quantitative, both based on science with tests and evaluations and on experience".

The wish for a connection between the qualitative and quantitative (hereafter Q+Q) relates to the present evidence discussion where Qvortrup has pointed out the duplicity of the English evidence and the French évidence in several connections. As Qvortrup elegantly expresses it when talking about the difference: "In one place - as 'evidence' - it connotes numbers, natural science and causality. When something is evident, it is true in relation to highly formalised truth criteria. "In the other place - as 'évidence' - it connotes experiences and practice.

When something is evident, it is a matter of course".  To Qvortrup, the Q+Q problem also relates to a specific type of product which the qualitative research is ostensibly unable to deliver.  As the saying goes: "It is very characteristic that qualitative research has predominantly been practised at the Danish School of Education - research where the ambition was to be descriptive.

However, methodically speaking, it could not have been generalising".  Combined with the wish for increased generalisation, Qvortrup demonstrates that "one of the problems we need to approach is that evidence and évidence become/are presented as enemies".

If we have a look at the outside world outside the Danish School of Education, it is significant that Q+Q - under the name 'Mixed Methods' - are about to establish worldwide as new pragmatic paradigm. So let us sum up what they are trying to fraternise in today's pragmatics and evidence demands. 

Quite often, the relations between Q+Q are mentioned as a discussion of method. The intelligent comment heron is typically that a more fundamental epistemological and ontological discussion hides under the method discussion. This discussion points out that the Q+Q discussion cannot only be considered a discussion of method turning the qualitative and the quantitative into a combination problem. It also involves research which brings up an integration problem. The relation between Q and Q is typically stylised by a wide range of opposite combinations:

It is a well-known fact that these opposites tend to divide the waters as it will be difficult to be at both sides at the same time, though it will not always be impossible. And the discussions and ambitions in the relations between the quantitative and the qualitative can be even more complex as the opposites may show up at many levels in the research process. For instance in relation to:

In other words, Q+Q is very far from being unambiguous, and they may occur at different levels and in different ways where the combination or integration possibilities vary. An attempt to integrate Q+Q at a given level will thus often involve a situation where the complete group of opposites is included.  It may therefore be difficult to isolate the Q+Q research problem into delimited shares.

Anyhow, the basic question, when considering the Q+Q research, is whether it involves some kind of synergy effect, i.e. whether the research result of a Q+Q survey is more than the sum of the respective or qualitative surveys.

A mapping of the area concludes that, opposite the ambition and the outlook, the Q+Q research is often just a mechanical combination of the two types, but that an actual integration is largely non-existent. This is due to the fact that Q+Q projects ostensibly never reflect on the relation between epistemology and ontology, but only combine methods. In practice, Q+Q research is unchartered territory. 

This is probably a consequence of the fact that the traditional line of separation - in the traditional way of thinking - does not permit exceeding.

So the way ahead seems difficult but perhaps a Q+Q directory could be found in the old stronghold of the quantitative camp. In the article "What is evidence in education and in occupation? ("Hvad er evidens i uddannelser og i professioner?") , Professor Peter Allerup starts by commenting Georg Rasch for stating that all surveys are fundamentally qualitative. It may be a provoking statement, however, it is interesting that Rasch, of all, took this view as his statistical theory has a great impact on the quantitative methods in the present international surveys such as PISA, which takes the form of the Rasch scales.

Paradoxically, according to Rasch, the outcome is after all only greater in terms of quality when we work more in terms of quantity. I will not challenge Rasch's expertise; I will not pretend to know everything about Rasch, however, one could point out that the quantitative research always depends on a quantification of something that is qualitative.

In other words, a numerical scale must be designed if one should be in a position to even consider assessing anything. In this connection, one could just point out that the best functioning quantified background variables, such as 'number of books at home' or 'number of years of experience', function so well statistically because the quantitative unit is a quantified size which is basically qualitative. In the French way, it is for instance evident that it is not only about the number of books, however, also about some extent of reading.

In other words, it is obvious to call attention, in all its banality, to the fact that statistics may be considered a construction. And in principle also consider themselves a construction. From this point of view, the problem is probably 'only' that statistics are often used as if they were causally explanatory, i.e. as an objective method with a realistic legal ontology. It is not necessarily the statistics' fault, but in principle the inductive - qualitative - statistics are interpreted according to the commonplace misunderstanding saying that when one is to assess something using an objective method, then something valid is assessed.

Thereby, a wrong conclusion from method to ontology, from induction to deduction, occurs. One could probably go far in the Q+Q integration if the abusers would admit that the reliable scales' latent variable - in a statistical sense - correspond to the analytical object of the qualitative research, and therefore the validity is in principle always up for discussion. To the abuser, it is only a difficult admission of disease as the extensional logic is reduced. And it is a problem when the extensional logic is the target. To those who understand evidence in the French way, the plea of the evident - on the other side of the Channel - is in itself ideological, which is known from the general popular learning; that statistics can be abused for anything.

Incidentally, it should be noted that the classic quantitative virtues of objectivity, repeatability, reliability and validity may very well remain as aspects of a research ethos; also for non-essentialistic research. Even if the virtues represent impossible ideals, they may never the less still function as ideals.

When it makes sense to refer to Rash, it is because he - by admitting that everything is "fundamentally" qualitative - may calculate that the probability of the stochastic research variable in the form of the incident Q+Q research will appear as nil multiplied by an infinite number of occurrences. In other words: P (X = K+K) = 0 x 8.

However, one should remember that the probability of Q+Q's prevalence depends on the normal distribution of the user groups' need and of DPU's need for accommodating its surrounding world in a fast and flexible way. And if the users wish for a combination of Q+Q, well then Q+Q will of course occur abundantly - if not in reality then at least as a true discursive construction.

As mentioned before, the idea of the Q+Q research often relates to an idea that a synergetic situation could arise as more quantity and less quality would occur at the same time. However, the point of the above considerations is that you could, with advantage, construct a meeting place for Q+Q by reassessing the extensional logic.

A solution could be that Q+Q may take place in practice as both less quantitative and less quantitative. This would require that one should abstain, in terms of extensional logical, from 'the total objectivity' which arises by linking essentialistic epistemology and essentialistic ontology.

Moreover, one should refrain from 'the total subjectivity' which arises by linking constructivistic epistemology and constructivistic ontology. The field of possibility should then be a research practice where the extensional logic is in the dissent which may arise on the one side perceiving the individual as being more than individual, viz. as the usual or typical.

And on the other side by perceiving the quantitative generalisation as less than universal, viz. as a normal distribution depending on the population - or as a dissent between the social and the cultural factors.

The decisive point by having such a research should thus be that it actually becomes more trustworthy as it reflects on less extensional logic and less predication effect.

In principle, it is not impossible that Q+Q could meet, however, such a meeting place is based on a simultaneous weakening of two of the dominating ideologies of the time: The individualisation on the one side and the evidence-based generalisation on the other side. In other words, the meeting will be in the nature of an ambiguous admission of the qualitative's quantity and the quantitative's quality. This is not a very interesting perspective for the research policy or for the political research so the meeting place will probably not be crowded.

1) www.information.dk/153716

2) See note 1

3) See note 2

4) Alan Bryman: Barriers to Integrating Quantitative and Qualitative Research. Journal of Mixed Methods Research, Vol 1, Sage Publications 2007

5) http://cvustork.dk/hvaderevidensiu1055.asp

About

Jens Marvin Bruun

Assistant Professor at the Department of Philosophy of Education, the Danish School of Education. During recent years, Jens Marvin Bruun has worked with international comparative surveys. He has, among others, worked as project manager of the Danish participation in the international research project Civic Education Study.