Reynolds continues his exploration of theory building in the last half of his book. Building theories is essential for taking the state of IS further (and getting published), and the fundamentals delineated here are important foundational information.
Breaking out the three types of theories (laws, axioms, and process) was instructive, but I reached an "aha" when Reynolds pointed out that on causal process or axiomatic theory put into a causal process form could "provide a sense of understanding" (page 107). Sets of law are not temporal in nature - they are timeless. However, it is necessary to observe cause-and-effect in order to really understand a phenomenon, and this must happen in a sequence of steps (process).
In our search for generalized truth, the section on "How To Get a New Idea" (page 151) is quite instructive. It made me feel that all was not lost when Reynolds declared that those responsible for developing the new idea were "not superbrains" and were typically "above average" (I grew up in Lake Wobegon). While a superbrain is not required, the ability to discern a "good" idea from a "bad" one is key. As in so many other aspects of life, the ability to succeed is partially based on talent, but there are many other "emotional intelligence"-type factors involved.
Finally, Reynolds' explanation of research methods is a nice introduction to research activity. From the small amount of IS-specific research that I have read, it appears that surveys are the dominant methodology. While surveys are in many cases the only way for an IS researcher to gather data, my practical experience leads me to believe that observation is the most trustworthy of data-gathering activities. The problem is that many companies do not want an outsider to observe too many of their operations, don't have the time to spend with the observers, or want to "tailor" their data via a survey. IS research can suffer from spurious data, and, sadly, sometimes does in published journal articles.
Comments
Eric,
I too found some sense of happiness in knowing that it doesn't take a genius to come up with a good idea - whew! :) Joking aside I would have to agree. From the research we have read it appears that people look at previous research and ask the next logical question (at least that is the case if research is really cumulative which I believe it is overall). Those people need passion and above average intelligence but that's it -- well -- a little determination doesn't hurt either!
As much as such research would benefit companies, I can see your point about firms being resistant to outsiders studying their operations.
I also liked the discussion on new idea generation. Of the few IS papers I've read I'm disappointed at the lack of new ideas. Certainly there are many reasons but you make a good point about the talent and emotional intelligence factors.
I agree with you that surveys are the dominant methodology in IS research. However, observation can be done in many ways in our field that makes it easier than others. Observation can be automated though the use of computer application. I remember in a class that I took during my undergrad about computer network management that an observer machine (computer) can be set within the network that can observe all the activities on the network. So, at the end you can have a look at the data form that machine.