Money vs. Pearls of Wisdom: personalising incentives for behavioral change

Participants are nudge through motivational messages and economic incentives using a mobile application, aiming to increase their level of physical activity. This intervention is deployed over a 6 month period.

2017-06-01

2018-01-11

Determining the effects of economic incentives and motivational messages on the participant's level of physical activity.

The study follows a between-subject design with 4 conditions: Fix incentive(FIX), Motivational Message(POW), Lottery(LOT) and Control(CON).
The study will be longitudinal with a 6 months observation period.
Participants will be provided with a mobile application that will retrieve the steps information from their phone.

Hypotheses
H1. Participants in POW, LOT, and FIX will walk more than participants in CON condition.
Rationale: The informative content of POW messages and the money offered in the other two conditions should effectively motivate people to walk more.

H2. Participants high IMI will walk more than participants with low IMI.
Rationale: The IMI survey measures effectively the inner disposition towards walking/running. Therefore participants who are already well disposed towards this activities will perform them more than those who are not.

H3a. Participants in POW with high IMI will walk more than participants in POW with low IMI.
Rationale: POW incentives are designed to resonate with the inner values of people who are already well disposed towards walking. Therefore they should have a maximum impact with users who have scored high on the IMI survey.

H3b. Participants in LOT and FIX with high IMI will walk less than participants in LOT and FIX with low IMI.
Rationale: LOT and FIX incentives are designed to help participants who have not internalized the value of having an active lifestyle. Therefore they should have a maximum impact with users who have scored low on the IMI survey.

H3c. Participants in POW with high IMI will walk more than participants in LOT and FIX with low IMI.
Rationale: In absolute terms incentives designed to support participants who have already internalized the value of having an active lifestyle will produce greater effects than incentives designed to persuade participants who have not yet internalized these values.

H4. Participants in LOT will walk more than participants in FIX.
Rationale: While the incentive offered to people in FIX condition is purely extrinsic, the incentive offered in LOT mixes the extrinsic value of the payoff of the lottery and the entertainment benefit of participating in a game. Together these two factors should have a greater effect than the money alone offered in the FIX condition. (need reference)

H5. Participants in FIX_high should walk more than participants in FIX_low.
Rationale: The right price point for rewarding the activity of walking more than 10K steps a day is above or equal to 0.60 CHF. The current market value is set to 0.50 CHF.


H6. Participants in POW with high IMI and participants in LOT and FIX with low IMI will increase their steps during the course of the experiment.
Rationale: Participants who are exposed to the right incentive for their status of internalization of the values of walking/running at the beginning of the experiment will be effectively nudged towards an increased physical activity.

H7. Participants in POW with low IMI and participants in LOT and FIX with high IMI will not increase, or even decrease, their steps during the course of the experiment.
Rationale: Exposing participants to the wrong incentive mechanism for their status of internalization of the values of walking/running at the beginning of the experiment will not be nudged towards an increased physical activity. Offering an extrinsic incentive to those who already value running/walking might even have a negative effect over time.

H8. Participants in POW with high IMI will present higher levels of physical activity than participants in LOT and FIX with low IMI after the notifications have stopped.
Rationale: Participants in POW with high IMI have been effectively nudged towards finding inherent satisfaction from the activity of running/walking. This will be a deeper state of motivation than that that could reached by participants who started from a state that required extrinsic motivation and therefore the healthy behavior will have higher chance of continuing in time for those participants who have internalized its value the most.

H9. Participants in LOT and FIX with low IMI will present higher levels of physical activity than participants in CON with low IMI after the notifications have stopped.
Rationale: Participants that required extrinsic motivation will be effectively nudged towards a deeper state of motivation than those who started from the same level of motivation but were not exposed to the nudged. As a consequence the healthy behavior will have higher chances of continuing in time.

H10. Participants will walk more in the weeks after they installed the app when compared to their personal baseline measured from 1st of June. This effect will fade out for the CON condition during the course of the experiment.
Rationale: The simple fact of receiving more notifications could make people more aware of walking and have a positive effect (novelty effect). This will fade out during the course of the experiment for participants in the control condition as they will habituate to the notifications.

Dependent Variable:
Number of steps/ Physical activity

Independent Variables:
Fix incentive Condition
Motivational Message Condition
Lottery Condition
Control Condition

Participants are assigned to each group condition based on their Intrinsic Motivation Inventory Score.

1

No

~250 participants

~250 participants

~50 participants control condition, ~50 participants POW condition, ~50 participants fix incentive condition, ~50 participants Lottery condition