Yes, But is it methodological

Yes, But is it methodological? Design studies to date have been described primar- ily using a set of process descriptors (e.g., these studies are interventionist, itera- tive, process focused, collaborative, multileveled, utility oriented, and theory driven, see Shavelson, Phillips, Towne & Feuer, 2003). Unless this set of proce- dures is under-girded by a conceptual structure that forms the basis for the war- rants for their claims, design study methods do not constitute a methodology and will contribute only haphazardly to an aggregative science of learning.

For example, Russell Hulse (Nobel Laureate in physics, 1993) noted at a talk at the National Science Foundation (Hulse, 2003), that an infinite number of hypotheses may be tested. This machination of testing hypotheses (i.e., the mechanical application of the “scientific method”) does not, by itself, create useful knowledge. It is important, rather, to test powerful (i.e., con- struct-advancing) hypotheses, which often emerge from studies directed primarily at the problem of meaningfulness. 

Until we can begin to discover and model the necessary components of learning (e.g., based in cognitive neuroscience: e.g., Dehaene, 1997; OECD, 2002), design study output is prone to be dismissed as mere “narrative”4 (e.g. Shavelson et al., 2003) – that the data are merely “anecdotal.” Our descriptions cannot, over the long haul, remain at the level of contingency. When we move toward identifying what is necessary in the situation, then the methodology becomes more mature, scientifically. It is in the connection of the contingent to the necessary where the real work of theory-building occurs, in my opinion.