Banking for Baby Boomers – A Survey Experiment on Internet Banking Training Take-Up
How to cite
Sarrazin, Erik; Hartinger, Katharina; Streich, David. 2026. "Banking for Baby Boomers – A Survey Experiment on Internet Banking Training Take-Up." AEA RCT Registry. AEARCTR-0017891. https://doi.org/10.1257/rct.17891-1.2
Download: BibTeX · RIS — for reference managers without a browser translator. Zotero and most others pick up this page's citation metadata directly.
* Primary investigator
Abstract
Digitalization in banking is leaving elderly clients at risk of losing access to financial services, yet little is known about how to effectively encourage technology adoption at an advanced age. In a companion field experiment (AEARCTR-0013985), we find that while internet banking training is highly effective conditional on participation, take-up of training among the group of elderly non-adopters is very low (2% initial response rate). This survey experiment investigates how the design of training invitation materials can increase older adults’ willingness to participate in internet banking training. We conduct an online survey experiment with around 1,000 German respondents aged 50-75 who do not regularly use internet banking. Respondents are presented with a hypothetical invitation to attend a free internet banking training course. We experimentally vary three dimensions of the invitation: (1) whether the invitation emphasizes the convenience benefits of internet banking usage or addresses common security concerns and barriers (benefit vs. cost framing), (2) whether the invitation includes a testimonial from a previous (hypothetical) training participant, and (3) whether the testimonial features a person of the same or different gender as the respondent (gender matching). Our primary outcome is stated willingness to participate in the advertised training. Secondary outcomes capture perceived benefits, perceived costs and risks of internet banking, trust in the offering institution, the intention to recommend the training to peers, and information demand (whether respondents click on a link to real-world internet banking training resources). Heterogeneity analyses examine whether treatment effects differ by gender, age, education, income, banking activity, responsibility for household finances, and stated barriers to internet banking adoption.
Intervention
The intervention consists of experimentally varied invitation flyers for a free internet banking training. All respondents view a flyer inviting them to participate in a hypothetical internet banking course. The flyer is designed to closely resemble the actual invitation materials used in the companion field experiment (AEARCTR-0013985). We vary three treatment dimensions.
Treatment Dimension 1: Content/Framing
Respondents are randomly assigned to one of three content framing conditions:
Control group: The invitation contains a neutral description of the training program without additional framing. It provides basic information about the course (training content, target audience, format, registration deadline) without specifically highlighting benefits or addressing barriers.
Benefits (convenience) framing: In addition to the baseline information, the invitation highlights the practical advantages of internet banking under the header “Why internet banking?”. These include: 24/7 availability, full transparency over account balances and transactions, fast processing of transfers and standing orders, and self-management capabilities (e.g., updating personal data, blocking cards, booking advisory appointments).
Costs (security) framing: In addition to the baseline information, the invitation addresses common barriers and fears under the header “Why internet banking?”. These include: no prior knowledge required, one-time setup effort, TÜV-certified security and fraud protection training, and the continued availability of branch services.
Treatment Dimension 2: Testimonial
A subset of respondents in the benefit and cost framing conditions additionally receives a testimonial from a fictitious previous training participant. The testimonial is displayed alongside an AI-generated photograph of the endorser, depicting an older adult. Testimonials are framing-consistent:
The benefits testimonial highlights convenience gains from training (checking balances from home).
The costs testimonial addresses the reduction of fears and barriers through training (gaining confidence).
Both testimonials have been selected from a set of six testimonials (three for the benefits and three for the costs treatments) in a pre-test based on comprehensibility.
Treatment Dimension 3: Gender Matching
Among respondents who receive a testimonial, we additionally vary whether the depicted endorser is of the same or different gender as the respondent. The photograph pairs (one male, one female) are selected through a pre-test on Prolific to ensure high and comparable levels of perceived trustworthiness, sympathy, and representativeness.
Experimental design
We employ a between-subjects design. Respondents are randomly assigned to one of seven treatment arms that result from the combination of three treatment dimensions - leading to the following target group sizes:
1) No Framing // No Testimonial // N=300
2) Benefits Framing // No Testimonial // N=150
3) Costs Framing // No Testimonial // N=150
4) Benefits Framing // Same gender Testimonial // N=100
5) Benefits Framing // Different gender Testimonial // N=100
6) Costs Framing // Same gender Testimonial // N=100
7) Costs Framing // Different gender Testimonial // N=100
The control group is oversampled (N=300) relative to the individual testimonial arms (N=100 each) to provide sufficient statistical power for pairwise comparisons between each treatment arm and the control, as well as for pooled comparisons of all benefit arms versus all cost arms versus control.
We specify three main analyses, corresponding to the three treatment dimensions:
1) Effect of Content Framing
We test whether benefit or cost framing increases willingness to participate in internet banking training relative to the control group. Specifically, we compare: (a) pooled benefit arms (Arms 2, 4, 5; N=350) vs. control (Arm 1; N=300), (b) pooled cost arms (Arms 3, 6, 7; N=350) vs. control, and (c) non-testimonial benefit arm (Arm 2; N=150) vs. non-testimonial cost arm (Arm 3; N=150) vs. control (Arm 1; N=300). We estimate OLS, logit, and probit regression models of the primary and secondary outcomes on treatment indicators.
2) Effect of Testimonials
We test whether the addition of a testimonial increases the effect of benefit and cost framings. We compare: (a) testimonial benefit arms (Arms 4+5; N=200) vs. non-testimonial benefit arm (Arm 2; N=150), and (b) testimonial cost arms (Arms 6+7; N=200) vs. non-testimonial cost arm (Arm 3; N=150).
3) Effect of Gender Matching
We test whether same-gender testimonials are more effective than different-gender testimonials. We compare: (a) all same-gender arms (Arms 4+6; N=200) vs. all different-gender arms (Arms 5+7; N=200), (b) same-gender benefit arm (Arm 4; N=100) vs. different-gender benefit arm (Arm 5; N=100), and (c) same-gender cost arm (Arm 6; N=100) vs. different-gender cost arm (Arm 7; N=100).
The survey follows a structured sequence of sections:
1. Screening: Age (numeric free text), frequency of internet banking use (5-point scale: never to always), German language proficiency (6-point scale), and gender. Respondents who use internet banking “always” or “frequently”, respondents with insufficient German language skills, and respondents that do not state their gender are screened out, as the intervention targets non- or infrequent users. Respondents reporting diverse/non-binary gender are assigned to the control group since they cannot be randomized into the gender-specific treatment arms.
2. Scenario presentation: Respondents view the randomized invitation flyer from their bank, adapted to the assigned treatment condition.
3. Outcome measurement: Primary and secondary outcome items are collected (see above).
4. Demographics: Highest educational attainment (6 categories), employment status (6 categories), and monthly net household income (6 brackets).
5. Internet banking barriers and attitudes: Reasons for not using internet banking more frequently (multiple choice, max. 3, randomized order), household financial responsibility, technology openness, and general risk attitude.
6. Banking activity: Number of monthly transfers, primary banking channel, distance to nearest branch, frequency of branch visits, and experience with branch closures.
7. Training preferences: Preferred training format (online video, at home with guidance from a trusted person, or in-branch with other customers) and an open-ended question on conditions for regular internet banking use.
8. Information provision: Respondents are presented with a clickable link to a government-sponsored website containing information on internet banking. Whether the respondent clicks on this link is recorded as a secondary outcome.
Design details
Cited by
No citation data available for this registration yet.