Digital Commons Network^™

Two Essays On Using Data Manipulation To Justify Biased Reports, Cody Lu

Doctoral Dissertations

An important aspect of data-generated metrics is the degree to which they can be manipulated by users. Agents providing reports often (a) have incentive conflicts with principals and (b) have access to and an ability to manipulate data used in those reports. This study investigates how data manipulation affects advisors’ tendency to provide biased recommendations when incentives between advisors and advisees are misaligned and aligned. Drawing on theories in deception and persuasion, I posit that, when incentives are misaligned, advisors will provide a more biased recommendation when the evidence used to support their recommendation is more manipulable. I also predict …