- Year 2022
- Theme Financial Inclusion
- Team RV/FI Team
This project in rural Uttar Pradesh aims to enhance the adoption of digital financial services by employing behaviourally informed strategies of utilising instructional videos while simultaneously getting people to practise as they watch (Pause and Practice Approach). The findings indicate that participants exposed to videos with practice, demonstrated higher accuracy and confidence in QR code transactions, suggesting the effectiveness of these interventions.
To build our portfolio of public goods for creating teaching material in different sectors, our project team tested behaviourally informed strategies to increase the uptake of digital financial services in rural Uttar Pradesh as described above.
Indicators of interest:
Based on our pilot, we designed interventions that we hypothesised would increase successful learning of QR code transactions. Our main objectives for the interventions were:
The results indicate that while the videos are fresh in memory, respondents in treatment groups T1 and T2 are significantly more likely to get each step of the tasks right. T1 respondents were also 1.72 times more likely to complete task steps for task 1 (PhonePe) accurately two days after the intervention, while only a trend is present for T2. Accuracy for task 2 (GPay QR code transfer, novel platform) was higher for T1 and T2 as compared to the control group immediately after watching the PhonePe videos and practising. However, this task was not significantly better in either arm compared to the control two days after the intervention.
| Task 1 Immediate Outcome Odds ratio (% accuracy) | Task 1 After 2 days Odds ratio (% accuracy) | Task 2 Immediate Outcome Odds ratio (% accuracy) | Task 2 After 2 days Odds ratio (% accuracy) | |
|---|---|---|---|---|
| CONTROL | (64.44) | (64.17) | (61.81) | (63.19) |
| T1 | 3.36 (83.10) | 1.72 (77.38) | 1.71 (72.74) | 0.96 (65.12) |
| T2 | 2.78 (78.10) | 1.45 (70.48) | 1.87 (71.59) | 1.30 (69.52) |
Task 1 (PhonePe) confidence levels for completing the task independently in T1 and T2 are higher than control immediately after the intervention, before starting the task. This is to be expected as T1 and T2 participants got a chance to practise the task. More interestingly, they also state higher confidence in getting the task right at the last step, even though having reached the last step should have been an indication that they have gotten the previous steps right. Further, this persists even at the last step when testing two days after the intervention. T2 also had higher confidence before starting the task two days after the intervention.
Figure 1: Confidence graph (PhonePe)
This graph demonstrates the confidence reported by participants before performing the task and after performing the task at baseline (immediate outcome - IO) and endline (after two days - A2d). The task was performing QR code transactions on PhonePe.
Figure 2: Confidence graphs (GooglePay)
This graph demonstrates the confidence reported by participants before performing the task and after performing the task at baseline (immediate outcome - IO) and endline (after two days - A2d). The task was performing QR code transactions on Google Pay.
Mobilisation Strategy
The sample was mobilised and recruited via a convenience sample by enumerators in our Pop-Up Lab in Sewapuri for this study. They typically reached out to Anganwadi workers across different villages, requesting them to connect us to potential participants who were interested in our research. Primary owners of smartphones who have never used any form of DFS from the district of Benaras in Uttar Pradesh were eligible. The baseline survey was programmed to screen out participants who did not meet our study's eligibility criteria
We also recorded participants’ specific confidence through stated measures while they carried out the UPI payment tasks and general confidence through cognitive tasks embedded in the survey.
Experiment Flow
As part of this study, we tested variations of knowledge-based videos through a lab-in-the-field experiment conducted in villages in Sewapuri, Varanasi (Uttar Pradesh). After the mobilisation of participants in a particular village, the team would be sure to go to the village within a week’s time. The baseline survey and the intervention videos were administered on the first day. After a gap of two days (on some occasions, three days), the endline survey was conducted.
