Search results

1 – 1 of 1
Article
Publication date: 24 September 2018

Carola Grebitus and Jutta Roosen

The purpose of this research is to test how varying the numbers of attributes and alternatives affects the use of heuristics and selective information processing in discrete…

Abstract

Purpose

The purpose of this research is to test how varying the numbers of attributes and alternatives affects the use of heuristics and selective information processing in discrete choice experiments (DCEs). The effects of visual attribute and alternative non-attendance (NA) on respondent choices are analyzed.

Design/methodology/approach

Two laboratory experiments that combined eye tracking and DCEs were conducted with 109 and 117 participants in the USA. The DCEs varied in task complexity by the number of product attributes and alternatives.

Findings

Results suggest that participants ignore both single attributes and entire alternatives. Increasing the number of alternatives significantly increased attribute NA. Including NA in choice modeling influenced results more in more complex DCEs.

Research limitations/implications

The current experiments did not test for choice overload. Future studies could investigate more complex designs. The choice environment affects decision-making. Future research could compare laboratory and field experiments.

Practical implications

Private and public sectors often use DCEs to determine consumer preference. Results suggest that DCEs with two alternatives are superior to DCEs with four alternatives because NA was lower in the two-alternative design.

Originality/value

This empirical research examined effects of attribute and alternative NA on choice modeling using eye tracking and DCEs with varying degrees of task complexity. Results suggest that accounting for NA reduces the risk of over- or understating the impact of attributes on choice, in that one avoids claiming significance for attributes that might not truly be preferred, and vice versa.

Details

European Journal of Marketing, vol. 52 no. 9/10
Type: Research Article
ISSN: 0309-0566

Keywords

1 – 1 of 1