Proposing a New Logic for Data Analysis in Marketing and Consumer Behavior: Case Study Research of Large-N Survey Data for Estimating Algorithms that Accurately Profile X (Extremely High-use Consumers)

Abstract

Rather than taking a variable-oriented approach, the study here extends Ragin’s (1999) perspective on studying conjunctive paths or “causal recipes” for a limited number of cases (usually n < 30) to the study of such paths in a large number of cases (n > 300 –or, in the study here, n > 30,000). The aim here is to provide a primer in theory and practice of qualitative comparative analysis (QCA) — a method that goes beyond considering the net contributions of individual variables in influencing a dependent variable. The study aims to describe alternative conjunctive paths (estimated algorithms using Boolean algebra and set theory) that associate with a given outcome. The “outcome condition” being modeled (predicted) in the study are cases (travelers) who complete more than 24 airline trips for personal reasons annually; the label “extreme” or “X” air travelers identifies these cases. Though small in share of numbers (.2%) of Americans, X air travelers are mighty in share (7%) of air-trips for personal reasons. The study finds that completing a university/college degree is a necessary, but not sufficient, simple antecedent condition that identifies X air travelers; the conjunction of completing college and very frequent vacation travel is a sufficient, but not necessary, complex antecedent condition for identifying X air travelers. The approach can be useful in estimating conjunctive conditions leading to specific actions by consumers and executives.