Mixed and Mixology

I cannot get out of my head the recent round of interviews my department conducted for two new openings–one for the mixed method position, the other for quant. In a way, the event has made an impression on me…It brought the sense of urgency and with it the potential of generative anxiety, the good kind that motivates. Dr. Dedrick kept offering little comments on the process–receiving nearly a hundred of applications to finally narrowing the search down to a handful of candidates, arranging for the final interviews, flying the finalists in, showing them around, meeting the faculty, administrators, and students… I wonder if someone enjoys this… I panic just thinking about how one day I may be in this position… mixed methods would be amazing…

And then I think about how I fell in love with research methods when I took my first RM class as psych undergrad, then the second… they were quant classes, too. At USF, I discovered qualitative and philosophy. How I have grown. This semester I am taking advance measurement class, single case experiments class, and a post qual class, and all I can think about how different and yet, the same they seem. At least qual does not pretend that the researcher is not a crucial part of the study. Subjectivity is a feature, not a condition that needs to be kept in check… Last night Dr. Ferron explained the essence of the statistical approach to analysis in such easy, palatable terms (both Dedrick and Ferron do that–they make quantitative methods friendly somehow, not so sterile) that I must question why do we have so many purists in research methods? Of course, it is a rhetorical question–there is a great deal of culture in academic training among other things.

In stats, we compare our observed scores to theoretical curves to prove or to disprove our null hypotheses. We design our studies using the logic tied to assumptions of normality, homogeneity of variance, independence… then there is a neverending tug of war between type I and type II errors that must be balanced well enough to convince the others who (also thanks to their training) only loosely (empirically or theoretically, or somewhere in between) agree on thresholds between statistically significant and not significant  results. And what of the general convention of 0.05 alpha? Arbitrary, but widely accepted. The data conceived, then collected, then analyzed as a model–observed scores, true scores… some of the concepts are “squishy” (in Dr. Dedrick’s own words).

It is ALL so damn squishy!

Barad’s application of quantum physics acknowledges and celebrates the squishiness. I like it. So as I think about “mixed methods” designs, they are just like a trail mix–the chunks that never quite blend together. Within most designs, quant and qual carefully observe their methodological boundaries. The methodologies are preserved and limited by conventions and traditions that decide what a researcher can and cannot do. Researchers are obliged to stick with conventions or else be scrutinized in terms of rigor, validity, integrity. They must be bold enough to answer methodological examination from BOTH qualitatively and quantitatively oriented peers. The risks are obvious. But what of risks? RISK is a construct, it does not have to be a variable or a sure limitation of research. Does it? Just think: risk, too, is a highly subjective term. It is socially constructed, much like validity, objectivity, and other measures that weigh a study and pronounce its value. So what if it is dealt with reflexively? I have nothing yet, just an abstract vision (which is yet, somehow, seems solidified) of my future employment as a methodologist.

I should totally try and conceive MIXED methodology designs as a cocktail, liquid and blended, not mixed like the trail mix.

So what if I do a single case experiment but analyze it qualitatively AND quantitatively?