Constrained Stochastic Extended Redundancy Analysis
Psychometrika, Volume 80, Issue 2, pp 516-534, 2015
Posted: 18 Jun 2016
Date Written: June 2015
We devise a new statistical methodology called constrained stochastic extended redundancy analysis (CSERA) to examine the comparative impact of various conceptual factors, or drivers, as well as the specific predictor variables that contribute to each driver on designated dependent variable(s). The technical details of the proposed methodology, the maximum likelihood estimation algorithm, and model selection heuristics are discussed. A sports marketing consumer psychology application is provided in a Major League Baseball (MLB) context where the effects of six conceptual drivers of game attendance and their defining predictor variables are estimated. Results compare favorably to those obtained using traditional extended redundancy analysis (ERA).
Keywords: redundancy analysis, maximum likelihood estimation, sports marketing, Major League Baseball, consumer psychology
Suggested Citation: Suggested Citation