The white-shirt effect
Learning from failed expectations
L. Prechelt Freie Universität Berlin, Berlin, Germany
Abstract
No matter how enthusiastic you are about your youngest data-scientific idea: Before you invest a lot of effort into data collection and/or analysis, consider the possibility that the data may show something other than you expect – or even nothing at all. Always hypothesize a causational model of what is going on, define fallback positions in case that model happens to be wrong, and make your investments in small steps, validating your expectations after each of them. This chapter explains those rules by a negative example in an entertaining way.
Keywords
DCDIM; Causation mechanism; Causation model; Partial checks; Face ...
Get Perspectives on Data Science for Software Engineering now with the O’Reilly learning platform.
O’Reilly members experience books, live events, courses curated by job role, and more from O’Reilly and nearly 200 top publishers.
