One of my courses this year is a seminar on information systems. Before each class, we each write a “conversation starter” incorporating our thoughts on the papers.
A topic that has greatly interested me lately is waterfall vs. agile development. “Waterfall” projects hinge on all-or-nothing strategy: let’s spend weeks and months developing something, then release in one fell swoop (like a waterfall building at the top and then crashing down.).
In contrast, agile looks at how to release smaller bits of product to be tested by the audience and incorporated back in an iterative process.
I argue that academic publishing is still built largely on the waterfall approach (let’s spend all our time on one big paper submission and hope it gets in) rather than lean (let’s use digital media to develop a minimally viable research proposal.).
I develop this (and other questions of IS research) below:
——-
Keen in “MIS Research: Reference Disciplines and a Cumulative Tradition” offers a definition of the field from MIT’s Center for Information Systems Research: “a study of the effective design, delivery, and usage of information systems in organizations.” Given that widespread computing in organizations is new, we should not be too surprised that MIS is still finding its bearings as a research discipline. But is this novelty stifling the development of a unique discipline with its own frameworks and agendas?
Yes and no. The adoption of information systems has made good organizations better and bad organizations worse. Information systems can deliver timely, actionable data or keep bright people mired in the weeds of information. So these problems of information and organization are not new – but digitization of information has exacerbated the effects.
IS as a discipline can answer how organizations can use information as an asset – but how do we do this? This requires that IS frames itself both inside and outside academia.
My first job after graduation was in demand planning at a national specialty retailer. I was ready to be the best forecaster in the company’s history. Exponential smoothing, EOQ – you name the theory, I would be using it.
It turns out that very little of my job was built on theories of finance and economics but instead on principles of information. What data do we need? How are we going to get it? Much of my time was spent on fruitless reporting, unheeded requirements analyses, and so forth. I ended up leaving the demand planning job to go into operations finance at a public hospital, where the information problem was so much worse! It was this frustration of how companies use information that in part led me to the program.
A theme of the readings is how to place IS within the “practitioner-scholar” continuum. Can research be rigorous and not esoteric? How can IS establish itself as a discipline with a well-defined research agenda while still benefitting non-academics? One model I see here is economics. As the field has become more rigorous, it has been less understood by practitioners. I have noticed that many of the pioneers of MIS have a background in economics. Maybe they were frustrated with the way things were going in that profession? I was – when looking at graduate programs, so little of the research I saw appeared as useful social science.
So maybe we need to model IS on economics with caution. But the comparison with IS to law and medicine in Davenport and Markus can only go so far. The boundary between professional and user is not as distinct in IS. Most people with direct contact with the legal or medical professionals are themselves lawyers or doctors.
Not so information – everyone is in contact with management information systems. And most of them do not want to be a data architect. They want to get back to their jobs of being an accountant, a non-profit coordinator, a teacher, etc. Those people may also have valuable insights on information and organization, while perhaps not being so knowledgeable about the “IS artifact.” So it is more difficult to see IS as a profession than law or medicine.
This tension of how to fit IS inside management research becomes even more difficult with Big Data. There is an implicit conclusion here that “big = more.” But Ackoff had a great retort to this, even sixty years ago. Do managers really need all the information they think they do? Many managers, under the false pretense of doing “big data,” have simply added more variables to their models. And the more complex these calculations become, the more effort it takes to maintain it. Again, we see people without the necessary skill (or passion) doing “shadow IS” work under a misunderstanding of Big Data.
So, how do these tensions resolve? One meeting point between academia and trade could be publishing. While peer-reviewed journals absolutely have their place, maybe we need to look at media to help the profession outside academia. I see traditional publishing as a “waterfall” form of development where all the work is grouped into one result, which will either be acceptance or rejection. Can we adopt more “agile” based research development via digital media, consulting, practitioner-focused conferences, etc.? Maybe the breakthroughs won’t come from a blog post, but they could be a place for non-academics to collaborate. This allows practitioners to learn from and collaborate with academic research, while setting the boundaries and guidelines for IS academics.
Leave a Reply