Position Bias in Best‐worst Scaling Surveys: A Case Study on Trust in Institutions

Danny Campbell,Seda Erdem

doi:10.1093/ajae/aau112

Danny Campbell, Seda Erdem

Open Access

https://doi.org/10.1093/ajae/aau112

Copy DOI

Abstract

AbstractThis paper investigates the effect of items’ physical position in the best‐worst scaling technique. Although the best‐worst scaling technique has been widely used in many fields, the literature has largely overlooked the phenomenon of consumers’ adoption of processing strategies while making their best‐worst choices. We examine this issue in the context of consumers’ trust in institutions to provide information about a new food technology, nanotechnology, and its use in food processing. Our results show that approximately half of the consumers used position as a schematic cue when making choices. We find the position bias was particularly strong when consumers chose their most trustworthy institution compared to their least trustworthy institution. In light of our findings, we recommend that researchers in the field be aware of the possibility of position bias when designing best‐worst scaling surveys. We also encourage researchers who have already collected best‐worst data to investigate whether their data shows such heuristics.

Highlights

This paper investigates the effect of items’ physical position in the best-worst scaling technique
Our findings show that: (1) the choices made by around half of our sample were not sensitive to any position bias; (2) the probability of an institution being chosen depends on the institution itself, and on its position in the bestworst scaling (BWS) choice task; (3) the institution positioned at the top of the choice task stands a significantly higher chance of been identified as the most trustworthy; and (4) not accommodating for position bias has significant implications on choice predictions, priority-setting, and the model fit
We separately report the estimated trust coefficients for the institutions used in the BWS survey and position-specific constants—where, for normalization, the final item and position s = 5 are arbitrarily set to the base level and, are omitted from the table

Summary

Introduction

This paper investigates the effect of items’ physical position in the best-worst scaling technique. The best-worst scaling technique has been widely used in many fields, the literature has largely overlooked the phenomenon of consumers’ adoption of processing strategies while making their best-worst choices. We examine this issue in the context of consumers’ trust in institutions to provide information about a new food technology, nanotechnology, and its use in food processing. A number of these processing strategies have been studied in the stated preference literature, mainly in discrete choice experiments. We examine whether respondents use item position as a schematic cue when making best-worst choices.

Methods

Results

Conclusion