We regret to inform you that we will no longer be publishing Information Management. It has been an honor to provide you with the insights and connections to move your career forward. We wish you continued success on your professional journey and welcome you to explore our other titles at www.arizent.com/brands.

Quantitative Social Science and BI

My company, OpenBI, is just concluding its college hiring for the first half of 2009. A business/actuary grad started in January, while a economics/math major begins June 1. Perhaps the tight market helped, but we feel quite fortunate to have added two outstanding newbies to our staff. At this point, I head up OpenBI's college recruiting. It's probably the part of my job I enjoy most, especially the fall campus interview visits. I suppose it makes me feel young. I think I'm somewhat biased though, looking for candidates that remind me of myself at age 22. 
I was a quantitative social science (QSS) major as an undergrad. On the social science side, I studied lots of economics, political science, psychology and sociology. Math, statistics, operations research, and computer/systems science were my main quantitative interests. Back then, QSS was a role-your-own major; now, many top schools have well-developed programs. And BI should take notice.
Though I didn't realize it at the time, a QSS background is ideal for business intelligence (BI). Business is certainly an economic, social, and pychological endeavor, so the social sciences seem great prep for further work in the world of commerce. Business schools would agree. A noticable percentage of B-school faculty are social scientists, and social science/ humanities undegrad majors constitute the largest proportion of acceptances to prestigious MBA progams like Harvard and Stanford. Of course, it goes without saying that a strong math/stat, systems, and programming background is sine qua non for a productive career in BI.
I find the intersection of the social and quantitative sciences of particular interest – and pertinence for BI. Social scientists generally promote theories about human behavior. Theories take the form of "the more (or less) of condition A, the more (or less) of behavior B", or, alternatively, "A caused B". Utility theory from economics, which says that individuals align their behavior with  expected economic benefit or utility, is an example. Scientists use methods, both qualitative and quantitative, to put their theories to test. Included in said methods are specific designs such as randomized experiments or interview panels, which  help distinguish among competing explanations – and therefore "prove" the correcteness of the proposed theory.
Business strategies can be likened to social science theories. As with theory, strategy posits the cause and effect of business behavior in the form of: the more (or less) of condition A, the more (or less) of desired business outcome B. An example  would be the prospective benefit of a new employee CRM training program, which would hopefully result in higher quality interactions with customers, Those improved customer interactions would, in turn, promote loyalty, ultinately leading to higher profits.
A major role of business intelligence is to measure the performance of an organization's strategies – to determine if and to what extent such initiatives work. Many of the methods and designs used by social scientists can also be productively deployed in business. And with the current open access to much of the academic world, it's not that difficult to stay on top of the latest developments.
I follow two QSS blogs to inspire my BI methodology thinking. The first is the Social Science Statistics Blog, from the Institute for Quantitative Social Sciences (IQSS) at Harvard University. Several years ago, I had the good fortune to meet Gary King,  Professor of Government and Director of the IQSS at Harvard. Gary graciously agreed to a few interview columns with me for Information Management, and I've attempted to keep up with his seemingly endless productivity ever since. Among their many research accomplishments, Gary and the IQSS are significant contributors to the open source R Project for Statistical Computing.
I discovered the Statistical Modeling, Causal Inference, and Social Science blog of Andrew Gelman of Columbia University after purchasing his excellent statistical analysis text: Data Analysis Using Regression and Multilevel/Hierarchical Models. Like Gary King, Gelman is both a political scientist and statistician, obsessed with design, methods and analyses of political data. And also like King, with whom he's collaborated on research, Gelman is a big proponent of R.
I'd encourage IM readers to  investigate these and other sources of the latest social science methodology developments and consider applying the findings to their work. Subsequent columns will highlight several of the more notable contributions of these blogs to BI.

For reprint and licensing requests for this article, click here.