About

By

We dissect and discuss data-centric decision making in publicized business cases from two different perspectives: academic and practice. One of the authors, Gorkem Turgut (G.T.) Ozer, is a decision sciences professor currently at the Paul College of Business and Economics, where his research and teaching focus on causal modeling using statistics, machine learning, and deep learning in big data. The other author, Duygu Dagli, is a data science and pricing executive who currently serves as Director of Data Science - Pricing & Promotions Strategy. Each of the authors also publishes solo articles that focus on providing insights into data-centric modeling and decision making.

Ideally, we'd like to invite comments and feedback here, but we have so many social media tools these days and it's hard to keep up with a conversation on all of them. As a solution, we post each article link on LinkedIn (see G.T. Ozer's profile here and Duygu Dagli's profile here). We look forward to your comments and feedback there.


Disclaimer: All opinions, views, and commentary expressed on this blog are solely our own and do not represent or reflect the official position, policies, or views of our current or former employers, clients, or any other organization or entity with which we are or have been affiliated. Any experiences, anecdotes, or examples shared are based on our personal perspectives and recollections of past events and should not be construed as current or factual representations of any organization or entity. This blog is not affiliated with, sponsored, endorsed, or authorized by any employer or organization.

Other popular articles

How to (and not to) log transform zero

By
Image
Image courtesy of the authors: Survey results of the papers with log zero in the American Economic Review Academic's take Log transformation is widely used in linear models for several reasons: Making data "behave" or conform to parametric assumptions, calculating elasticity, etc. The figure above shows that nearly 40% of the empirical papers in a selected journal used a log specification and 36% had the problem of the log of zero. When an outcome variable naturally has zeros, however, log transformation is tricky. In most cases, the solution is to instinctively add a positive constant to each value in the outcome variable. One popular idea is to add 1 so that raw zeros remain as log-transformed zeros. Another is to add a very small constant, especially if the scale of the variable is small. Well, the bad news is that these are all arbitrary choices that bias the resulting estimates. If a model is correlational, a small bias due to the transformation may not be a big con...
Read more

What if parallel trends are not so parallel?

By
Image
Image courtesy of github.com/asheshrambachan/honestdid Academic's take In difference-in-differences models, parallel trends are often treated as a make-or-break assumption, and the assumption is not even testable. Why not? One misconception is that comparing the treated and control units before treatment is a test of the assumption. In reality, the assumption is not limited to the pretreatment period but for the entire counterfactual. Comparing trends in the pretreatment period is only a plausibility test. That is, "parallel pretreatment trends" and "parallel counterfactual trends" are not the same. We can observe the former, but we need the latter. So, what we have is not what we want, as is usually the case. Even testing for parallel pretreatment trends alone is tricky because of potential power issues (evidence of absence is not the absence of evidence!) and other reasons, including a potential sensitivity to the choice of functional form / transformations ...
Read more

Explaining the unexplainable Part II: SHAP and SAGE

By
Image
Image courtesy of iancovert.com Academic's take This post is an exception to close the loop we opened with  our post on LIME . Starting in 2025, we're changing the scope of Data Duets to focus on business cases (as opposed to methods). Having discovered an excellent write-up  that explains both SHAP (SHapley Additive exPlanations) and SAGE (Shapley Additive Global importancE), I will focus on the why questions, possible links to counterfactuals and causality, and the implications for data centricity. Why do we need SHAP and SAGE? The need for explainability methods stems clearly from the fact that most ensemble methods are black boxes in nature: they minimize prediction error but they obscure how the predictions are made. This is problematic because some predictions are only as good as the underlying conditions are favorable. For example, Zillow's iBuying was a notorious failure , likely due to a lack of clarity and transparency in how the predictions ...
Read more