All framing effects are not created equal: Low convergent validity between two classic measurements of framing

Shanshan Zhen, Rongjun Yu*

*Corresponding author for this work

Research output: Contribution to journalArticlepeer-review

7 Citations (Scopus)

Abstract

Human risk-taking attitudes can be influenced by two logically equivalent but descriptively different frames, termed the framing effect. The classic hypothetical vignette-based task (Asian disease problem) and a recently developed reward-based gambling task have been widely used to assess individual differences in the framing effect. Previous studies treat framing bias as a stable trait that has genetic basis. However, these two paradigms differ in terms of task domain (loss vs. gain) and task context (vignette-based vs. reward-based) and the convergent validity of these measurements remains unknown. Here, we developed a vignette-based task and a gambling task in both gain and loss domains and tested correlations of the framing effect among these tasks in 159 young adults. Our results revealed no significant correlation between the vignette-based task in the loss domain and the gambling task in the gain domain, indicating low convergent validity. The current findings raise the question of how to measure the framing effect precisely, especially in individual difference studies using large samples and expensive neuroscience methods. Our results suggest that the framing effect is influenced by both task domain and task context and future research should be cautious about the operationalization of the framing effect.

Original languageEnglish
Article number30071
Number of pages9
JournalScientific Reports
Volume6
DOIs
Publication statusPublished - 20 Jul 2016

Scopus Subject Areas

  • General

Fingerprint

Dive into the research topics of 'All framing effects are not created equal: Low convergent validity between two classic measurements of framing'. Together they form a unique fingerprint.

Cite this