Background

An increasing part of the population is overweight and obese in Australia. Two main factors are the cause of this:

less physical exercise
more high-energy foods are consumed

I would like to concentrate in my nudge challenge of the food consumption aspect. My nudge addresses overweight/obese individuals who know they have a self control problem and want to do something about it. The nudge will be applied by a health fund during an education course at a public(health fund) kitchen with individuals invited.

Desired Behaviour Change

The nudge once applied should expect that the user group:

is able to resist high energy food
understands how to plan his food intake
understands how to plan his meals, recipes and knows what ingredients he needs
understands how much money he saves by planned cooking his own food.

Decision Making Process

One of the major reasons for individuals, is the easy and convenient access to fast food. In my opinion, hyperbolic discounting adds an additional level of difficulty for individuals to follow through with cooking their own food: the smaller sooner reward (the unhealthy food) which can be easily consumed is preferred, instead of preparing healthy food by cooking for them self. It is seen as:

less convenient to obtain (need to be bought from the local supermarket)
even seen pricier, since the ingredients are seen as is, instead of being shared among many dishes and recipes.

Furthermore, peer pressure sets in after the consumption since individuals are being judged on their appearence.

My decision map (upstream to downstream from left to right):

Know what recipes the individual prefers -> find a corresponding recipe -> find supermarket/special shops where to by ingredients -> buy additional cookware if needed -> buy ingredients -> cook food -> clean dishes

The iterative approach involves around:

Know what recipes the individual prefers -> find the recipe -> buy ingredients -> cook -> clean dishes

Nudge

My proposed nudge is uses a de-biasing strategy: individuals are subscribed to a free education course by their health fund on how to plan, cook, healthy meals. If they subscribe, they are allowed to bring a friend of their choice and to spread the word through social networks. That adds additional peer pressure nudge (friend, social network) to follow through with the course.

At the beginning of the course, a plan is created by the individual on what they usually consume during the day. This is kept to the individual. In the middle or end of the course, the individual is motivated by cumulating the costs of his consumption and compares it with the cooking he has done. It should outline how much money they would have spend(lost) if they would have not followed the course.

The nudge has the following goals:

become aware of the unhealthy diet
understand how to plan your meals ahead (fix one bottleneck as identified in the decision map)
use combinations of recipes to effectively use the bought ingredients
understand that cooking fresh food can be as convenient (fixes another bottleneck identified in the decision map).
save more money with cooking fresh food than spending it on fast food.

Experiment

Type: Field experiment Design:

Factor: Amount of weight loss + following through with the course.
Type: 2-level condition design: control, treatment - between participants
control group: no invitation to the course, no weight monitoring
- Expected outcome: no weight loss
treatment: individuals receive invitation for the course and the invitation for the app for weight monitoring
- Expected outcome: measurable weight loss, no drop outs.

Data analysis:

Using ANOVA, I'm expecting a main effect between the three groups, that is the highest result (weight loss) in the last group.

Concluding Comments

The course can be adjusting depending of the effectiveness of the nudge. Perhaps there is no need to add additional peer pressure, or perhaps an additional webapp would help individuals to compare their course outcome with fellow peers and makes the course more effective. This can making scaling the course easier, if people would like to follow up with the course online instead of going to a public kitchen of the health fund.

For the experiment, the groups have to be significantly large to avoid background variables adding errors into the effect (eg. a TV show motivates the control group to loose weight as well) and it needs to be repeated after a couple of months with roughly the same effects.