Experimental evidence on Educational Tips and Warnings within a Digital Payment App

How to cite

Jose, Anu; Us-Salam, Danish; Kelly, Jane; King, Michael; Gulesci, Selim. 2024. "Experimental evidence on Educational Tips and Warnings within a Digital Payment App." AEA RCT Registry. AEARCTR-0015071. https://doi.org/10.1257/rct.15071-1.1

Download: BibTeX · RIS — for reference managers without a browser translator. Zotero and most others pick up this page's citation metadata directly.

Anu Jose* , Danish Us-Salam , Jane Kelly , Michael King , Selim Gulesci

* Primary investigator

Registry ID
AEARCTR-0015071
DOI
https://doi.org/10.1257/rct.15071-1.1
First registered
Last updated
Status
In development
Trial dates
2024-12-21 – 2025-05-31
Countries
Private
JEL codes
D18, G4, D14
Registry record
http://www.socialscienceregistry.org/trials/15071

Abstract

Financial scams impose huge costs on consumers and the financial industry. Consumer education is viewed as one part of a wider toolkit to address fraud, particularly given the growth in authorised push payment fraud (APP), where consumers unwittingly authorise digital payments to fraudsters (BPFI, UK Finance). Through an online trial, we intend to test the causal effects of providing generic or scam specific fraud education tips, or a visual fraud warning, just before the payment confirmation step within a simulated banking app. We will disentangle which, if any, of these interventions assist consumers to detect fraudulent scenarios, and whether any of the methods are more or less effective. Additionally, we will test which methods are associated with greater consumer comprehension and overconfidence, as well as whether effectiveness varies with participants online risky payment behaviour. Our results will aid policymakers in their efforts to protect consumers from payment fraud.

Experimental design

Control Group (C) - Standard payment journey - Participants in this group will complete a standard mobile payment journey consisting of no more than 2–3 payment steps for each scenario. This group serves as the baseline comparison, allowing for the measurement of differences between the control and treatment groups.
Treatment 1 (T1) - Generic fraud tips within the mobile payment journey - Participants will complete a standard mobile payment journey similar to the control group but with the addition of generic fraud tips embedded within the journey. These tips will be displayed just before the payment confirmation step. The generic fraud tips have been designed based on fraud prevention material available on the Central Bank of Ireland’s website and provide broad guidance on identifying fraudulent activities.
Treatment 2 (T2) - Fraud specific tips within the mobile payment journey - Participants in this group will receive fraud tips tailored to the specific type of fraud presented in each scenario. These scenario specific tips are embedded within the payment journey and are shown just before the payment confirmation step. The content of these tips has been developed by reviewing the tell-tale signs of common fraud types, ensuring relevance and specificity for each scenario.
Treatment 3 (T3) – Visual Fraud Warning within the mobile payment journey - In this group, participants will complete a standard mobile payment journey similar to the control group but with the addition of a visual fraud warning. This warning embeds behavioural principles (Wang et al., 2022) and is designed to alert participants that a specific payment may be fraudulent. There will be a small element of text, warning participants that they could lose their money but without any specific capacity building tips. Unlike the generic or scenario-specific fraud education tips, the visual warning employs a visual approach to emphasise the risk associated with certain payments.

Design details

Randomization method
Participants are randomised to either a control group or one of the three treatment groups based on randomisation done by the Gorilla experiment builder tool.
Randomization unit
The unit of randomisation is at individual level
Observations
3000 nationally representative individuals
Arms
750 in each of the experimental arms
Primary outcomes
1. Proportion of participants who confirm a payment for legitimate and illegitimate scenarios separately. 2. Proportion of participants who confirm a payment for each scenario.
Secondary outcomes
1. Comprehension in identifying scams – a fraud comprehension module which measures overall fraud detection literacy of participants. 2. Confidence in Payment Decisions - A self-reported measure of how confident participants are in their payment decisions 3. Measure of the proportion of participants who successfully cancelled all payments for illegitimate scenarios and confirmed all payments for legitimate scenarios.

Cited by

No citation data available for this registration yet.

Keywords