There is a great discussion going on in LinkedIn about MROCs and the nature of the research that can be conducted using them, find it here if you are a member of LinkedIn.
I think that MROCs are essentially a qual medium and I think the difference matters. Below is the text I posted in the discussion explaining why, and setting out my thoughts.
I think the discussion about qual and quant matters, because it defines the nature of the truth we are telling to our clients. We can think about this split at two levels, IMHO
1) Traditional MR quant vs qual. Quant measures those things that can be readily measured and which it feels can be projected numerically onto some larger population. Qual identifies meanings and uses the researcher as a prism to identify which meanings are likely to be generalisable and useful.
2) The epistemological split. The post-positivists look for those relationships which can be estimated numerically and assume (as a working hypothesis) that they are genuine. The constructionists look for patterns of behaviour(assuming that these patterns either do not have a mathematical underpinning or that such underpinning is too complex to be revealed) which can lead to a truth being constructed by the observer, ideally one that can help people deal with other situations.
In social media in general, and in MROCs in particular, there are major problems with the post-positivist/quant interpretation of the truth.
1) The nature of environment is changing so fast that it is unlikely that truths are sufficiently long-lived to be measured with numbers and then projected elsewhere.
2) The power of the medium is conversation and the quantitative tools for analysing text are extremely limited and primitive.
3) The leading thinking by exponents such as Kozinets is that the researcher needs to be a participant-observer to gain real value from these communities, but that means that the truths are going to be more constructionist than post-positivist. All MROCs are, IMHO, examples of participant-observer studies, the researcher is changing both the discourse and the context.
I do not think it matters too much about how we describe the qual/quant split to end users, they simply want good insight and good advice. But I think it matters that we understand the nature of the truths we are dealing with.
I would like to come back to a point Rosie Campbell has made elsewhere, we must avoid allowing the analysis of communities to descend to simply counting the words. A Wordle is a great diagram, but it is not any form of analysis.
In the world of MROCs I would like to see polls as simply another stimulus for a qual investigation, and I think we can construct a toolkit where a survey is just another tool, to sit alongside other qualitative tools such projective techniques, and where the analysis of the quant data simply sits next to such approaches as grounded theory, semiotics, and hermeneutics - but where is does not try to claim pre-eminence.