Lessons Learned

This is arguably one of the first studies designed to juxtapose the RCM method with the

survey method, thus we can only draw tentative implications. Nonetheless, we need to

point out four potentially significant implications:

1. The four stage validation process — gathering a knowledge representation of

experts, choosing a theoretical structure to interpret those concepts, using the

evoked concepts to design large sample surveys, and finally using the survey

output to re-construct the causal maps — offers a useful approach to juxtapose

RCM and survey methods.

2. Although our results point to the empirical validity of the RCM approach in the task

of knowledge representation, they highlight the sensitivity of construct-level

representations of RCM to the underlying theoretical framework. Great care should

be exercised in the choice of theoretical frameworks to group evoked concepts.

3. Perhaps more specifically, our study suggests that the RCM approach can become

the basis of other large sample studies in exploratory contexts.

4. In discovery and evocative contexts, i.e., early stages of theory development,

where different theories may be competing for accuracy, RCMs derived from

experts may provide one method of standardizing the theoretical framework.

Conclusion

One of the significant challenges in theory development in new domains is the task of

coupling discovery with verification through normal science methods. To date, there has

been no systematic approach to accomplish this, since scholars who engage in discovery

are rarely the ones engaged in verification. As we have shown, causal mapping linked

to surveys may provide one method by which the linkage between discovery and

verification can be systematically juxtaposed. Finally, we would argue that such linkage

is consistent with the received wisdom from epistemology.