Topic XII: Research Methods

Research is fundamentally a developed capability (see Behavioral Manifold:). For all its familiarity as various methods (e.g., experiment, survey, census, history, content analysis), its functionality lies in the production of observations (, ) that can be used in making and taking steps – by one’s self or others, now or later, so that Og(On) replaces Og as the guide for Mg. (An observation can be quite extensive. One of communication’s virtues is that it serves as the “carbon” for constructing such products — that is, if languages have been developed, as procedural tools, for this purpose. Which, in some profusion, they probably were!).

Seeing research as a developed capability implies (helpfully) that it might yet be further developed, so as to make our store of observations more complete and more accurate – assuming that as we do so we would also further develop languages to make these observations more useful: more accessible and more applicable. We do not just want more and more observations, lest we have even more pollution (: Ps) and impediments than we have now.

Observations and research when applicable speak to the matter of effectiveness, to our concern with problem solving. Research dealing with behavior (as the observed) always involves problem solving, not just by the observer but by the observed: Recall that the behavioral problem (of needed capacity and capability) exists in any and every step made and/or taken (I: Pbeh). Unfortunately, far too much behavioral research focuses primarily on the particulars of situational problem (Psit) and behavioral solution (Sbeh), employing order-of-things methods to the neglect of the Nature of Things – in violation of the balance requisite.

Compositional change, the path to control via problem solving, is incompletely and inaccurately served by simplistic notions of consequentiality. Yet the components of “all that it takes,” especially the extensive core of agency capabilities needed, is frequently subverted – or relegated to the status of circumstances (e.g., “independent variables” and “dependent variables” – and “mediating variables” to cover contingent relationships). Timing, so crucial to construction’s “recipes” – invented or followed — is neglected). Applicability is not well served.

Research often is understood to emphasize questioning at the expense, even the exclusion, of problem solving. This as though investments in problem solving could and should be made outside and after the reign of research (as implied in the distinction, and division of labor, between science and engineering). This does not work well for research on human behavior, of course, because an important segment of that research needs to be about initiated development: Development and research not just research and development.

In light of behavioral aspects we have covered up to now we need to take a fresh look at our research methods. Consider two familiar (at least by name) kinds of research … and then we shall look at a third kind more appropriate to problem solving.

“Basic research” – inquiry into the order of things. On matters behavioral, these methods are best typified by the test of statistical significance (to a 5% criterion): that observed relationships (in physical time-space or in attributional N-dimensional space) are not due to chance. Because of the working assumption of an underlying order to Everything, it is thus assumed that research results will add up. (All non-chance contributions accepted, so to speak.)

Much is said to be basic research on behaviors that does not involve problem solving at all – except in the solving of all the observational problems entailed in trying to answer the research question. N-dimensional space is often employed; it consists of the various attributes assigned a body (a “person who”). Steps, when considered at all, are gratuitously assigned as body properties. They, like other attributes thus generated, are then sorted into “independent” and “dependent” variables – a modest bow to consequentiality. “Causal” thinking runs rampant.

“Applied research” – inquiry into the efficacy of behavioral solutions (:Sbeh). Here a solution (perhaps competing solutions) to a situational problem (Psit) has been developed. Two objectives usually to be found are: 1/ Will the solution beat chance: Is it better than not doing anything? 2/ Is one solution better than another by more than a chance margin: Is there a competitive advantage to be had?

Beating chance and/or the competition has its merits. But what if one wants not to test the (limited number of) available solutions? What if one is faced with producing a new solution? What if the criterion is not 5% statistical prediction but 100% productive success – that is: all that it takes? We have all sorts of orderings that we have come up with that are better than chance (e.g., legal statutes as de facto laws of behavior), with or without formal research. But how much better could they – should they – be?

“Applicable research” –observation of the developing solution (Sbeh) to the behavioral problem (Pbeh) as it pertains to a particular situational problem (Psit). For this we need to restore the dynamic balance between problem solving and inquiry in our research paradigm. (See Diagrams.)

(c) 2010 R. F. Carter


Supporting Audio

Part 1
Part 2
Part 3

Supporting Diagrams

Click to Download Diagram 1
(x) Downloads
Click to Download Diagram 2
(x) Downloads