The Role of Selective Hypothesis Testing in the Overestimation of Price-Quality Correlation

Maria L. Cronley, Wright State University
James J. Kellaris, University of Cincinnati
Frank R. Kardes, University of Cincinnati
Steven S. Posavac, University of Rochester
ABSTRACT - When consumers expect to see a relationship between price and quality, they are likely to perceive a strong relationship even when the objective relationship is weak. The present experiment shows that price-quality correlation overestimation is reduced (but not eliminated) when a large amount of price-quality data is presented in a random order, provided that consumers are sufficiently motivated to process the data thoroughly due to a low need for cognitive closure (Kruglansi and Webster 1996) or a high need to evaluate (Jarvis and Petty 1996). These interactions are consistent with the implications of a selective hypothesis testing model of data-driven information processing (Sanbonmatsu, Posavac, Kardes, and Mantel 1998).
[ to cite ]:
Maria L. Cronley, James J. Kellaris, Frank R. Kardes, and Steven S. Posavac (2001) ,"The Role of Selective Hypothesis Testing in the Overestimation of Price-Quality Correlation", in NA - Advances in Consumer Research Volume 28, eds. Mary C. Gilly and Joan Meyers-Levy, Valdosta, GA : Association for Consumer Research, Pages: 323.

Advances in Consumer Research Volume 28, 2001     Page 323

THE ROLE OF SELECTIVE HYPOTHESIS TESTING IN THE OVERESTIMATION OF PRICE-QUALITY CORRELATION

Maria L. Cronley, Wright State University

James J. Kellaris, University of Cincinnati

Frank R. Kardes, University of Cincinnati

Steven S. Posavac, University of Rochester

ABSTRACT -

When consumers expect to see a relationship between price and quality, they are likely to perceive a strong relationship even when the objective relationship is weak. The present experiment shows that price-quality correlation overestimation is reduced (but not eliminated) when a large amount of price-quality data is presented in a random order, provided that consumers are sufficiently motivated to process the data thoroughly due to a low need for cognitive closure (Kruglansi and Webster 1996) or a high need to evaluate (Jarvis and Petty 1996). These interactions are consistent with the implications of a selective hypothesis testing model of data-driven information processing (Sanbonmatsu, Posavac, Kardes, and Mantel 1998).

----------------------------------------