Information and Insurance Decisions: A Field Experiment in Singapore

Participants first complete a brief baseline survey covering demographics and basic financial information, together with several perception-based questions. After completing the baseline section, each participant is automatically and individually randomized into one of several experimental conditions within the same survey session. Some participants are shown short informational messages providing factual reminders relevant to long-term personal or financial planning, while others proceed without receiving any message and move directly to the next part of the survey. Immediately after the information exposure, or directly after baseline for those who receive no message, all participants report their intentions regarding potential life-insurance purchase or adjustments to existing coverage in the coming months, along with several secondary measures related to perceived financial preparedness.

2025-12-01

2026-03-01

The primary outcomes include both stated intentions and realized insurance behaviors. First, participants report their likelihood of purchasing life insurance or increasing their existing coverage within the next 12 months on a 1–7 scale immediately after the intervention. Second, we track actual insurance purchase behavior following the survey session, defined as whether the participant subsequently purchases new life insurance or increases existing coverage within the study’s follow-up window.

Participants complete a brief baseline survey covering demographics and basic financial information. After the baseline section, each participant is randomly assigned to one of several experimental conditions within the same survey session. Some participants receive short informational messages related to long-term personal or financial planning, while others proceed without receiving any message. All participants then answer a set of questions about their intentions regarding future life insurance decisions. The experiment is conducted at the individual level, and the intervention occurs only once within the survey flow.

Not available

Randomization is implemented automatically within the survey platform. After completing the baseline section, each participant is assigned by the computer to one of several experimental conditions using simple random assignment. No stratification or blocking is used, and the assignment is not influenced by any participant characteristics.

Individual (each participant is randomized independently).

No

0

660 individuals

Approximately 220 individuals in each group.

Power analysis based on individual-level randomization and the planned total sample size indicates that the study can detect a minimum effect size of approximately 0.30 standard deviations on the primary outcome with 80% power at alpha = 0.05. This corresponds to a difference of roughly 0.45 points on the 7-point likelihood scale. With a total planned sample of around 660 participants, the detectable effect size remains close to 0.30 SD, yielding an expected statistical power of approximately 0.77 after accounting for potential sample attrition.