Nudging Towards Healthier Eating: Insights from an Experiment on Unit-Bias & Convenience
Do you decide what you eat? In a field experiment we tested the effect of integrating unit-bias and the convenience principle in the classic setting of the coffee break.
Many of us are struggling to stay on the narrow path of healthy choices. At work, at meetings, and in the car on the way home we are constantly tempted to snack a little bit—to kill boredom, to have a good time, or to get some quick energy to stay awake and aware. Each single occurrence does obviously not make a difference, but over time one snack becomes more, and slowly but surely our waistline expands.
For more than a decade we have run a long series of experiments to see if simple nudges can change how much we eat. In this experiment we teamed up with the restaurant Søren K and got close to the participants at a big conference for Danish Communicators held at the Royal Library of Copenhagen. We wanted to test the effect of integrating unit-bias and the convenience principle in the classic setting of the coffee break—and as a bonus we caught the Editor in chief at Weekendavisen, Martin Krasnik, and the Danish politician, Holger K. Nielsen, with their fingers in the proverbial cookie jar.
Fact Box: Nudge
A nudge is a function of every attempt to affect people’s assessment, choice, or behaviour in a predictive direction under the assumption (1) that cognitive limitations and bias as well as habits affect our individual and social behaviour, and (2) that it works by making use of those as an integrated part of such experiments. [1]
The Experiment: Unit Bias & Convenience
Long-lasting research streams in the applied behavioural sciences has documented that small, but seemingly insignificant features of our foodscapes can significantly influence how much we end up eating. Research has looked into everything from plate size to the width and height of glasses, and even if it is better to use spoons instead of tongs in the salad bar [2]. This time we experimented with contextual factors that are present in almost every foodscape: the size of the units in which food is served and the convenience eating it.
Our experiment investigated unit bias and convenience in the setting of the universal ‘coffee break buffet’ served at a two-day conference held by Danish Communicators. Just like almost any other conference coffee break buffet, this one offered a combination of fruit and cake – and the way that this serving was offered was the classic configuration viz. quadratic brownies the size of a ‘full hand’ and whole fresh apples.
Fact Box: Unit Bias
Unit bias is a psychological mechanism that indicates that the portion size we are presented with is suitable to eat [3]. Unit bias is our tendency to accept that just one unit or portion is a suitable amount for us to eat [3]. One of the classic examples of unit bias is when we make the plate size smaller in a self-service buffet. We tend to accept that one full plate is a suitable amount of food, so when the plate size is smaller, we eat less [4].
Our question was whether this classic configuration is optimal if we wish to boost the intake of apples and limit the calorie intake resulting from eating cake? To find out, the restaurant Søren K kindly agreed to serve their ‘business as usual’ classical configuration at the first day of the conference, while we observed the choice of each person visiting the buffet using well-placed hidden cameras. The next day the same buffet was served – for the same guests (within-subjects-design) – at the same time… but with the slight difference that the units of brownies were cut in half and the apples were cut into convenient wedges.
Figure 1. Illustration showing the standard setup of brownies and apples on day 1, and the change on day 2 with units of brownies cut in half and the apples cut into convenient wedges.
Take your Organisation
11 – 13 November, 2024 | COPENHAGEN
JOIN OUR NEXT MASTERCLASS
Results and findings: Smaller Pieces Work
The results from this ‘withing subjects’ experiment showed that the small change in size and convenience lead to a significant change in the conference participants’ intake of apples and cake, respectively. The average intake of apples per person increased from approx. 12.7 grams to 20.4 grams – this corresponds to people eating 60.8 % more apple when it was served in smaller pieces than when served as whole apples. The opposite pattern appeared for the intake of cake; that fell from approx. 93.9 grams to 61.2 grams in average per person. Thus, 34.8 % less cake was eaten when the pieces were cut in half.
Figure 2. Average intake (grams) per person. More apple was eaten on average person, and less cake was eaten on average per person.
Besides this result, which replicate one of our earlier experiments [5], our method of measuring (within subjects design) gave us the possibility to measure something that has not previously been accounted for; namely how the intake is distributed among each individual. And this is where it gets really interesting: with the whole apples, only 32.9 % of the participants ate some apple; whereas when the apples were cut in smaller pieces, 85.3 % of the conference participants ate apple. We also saw a decrease in the share of conference participants who ate cake; namely 83.5 % on Day 1 and 74.7 % on Day 2, when the cake was cut in smaller pieces.
Watch the video from our experiment
Some complexities of interaction
Our experiment teaches us a number of things – but it does not show us how the Behavioural Insights of Unit Bias and Convenience may interact. For unit bias to work, it is known that the food portions should be within a reasonable size range. In other words, that means when we are presented with a whole baking pan of cake, the portion size is too big for us to eat it all. On the other hand, a cake crumb would be too small for us, and we would most probably eat more crumbs to be satisfied. Somewhere in between those two extremes there will be a number of different cake sizes that we accept as a suitable portion.
Figure 3. An illustration showing unit bias at work, emphasizing that food portions should fall within a reasonable size range.
This also means that if a food portion, as with the breadcrumbs, is not big enough to make the unit bias come into force, people will just take more units. This is actually what might be the case when we cut the apples in smaller pieces. The apple pieces were small enough that the conference participants chose to take more pieces on their plate.
Fact Box: The Principle of Convenience
The principle of convenience describes our tendency to make the choices that are most convenient for us. Sometimes one could even be so cheeky to call us lazy. That means, the more convenient a particular type of food is to eat, the more we eat of it. The opposite also applies – when was the last time you ate a whole coconut, for example?
However, our experimental design does not make it possible to distinguish this effect from the effect of the principle of convenience. In our setting there are some aspects of a whole apple that make it less convenient to eat (and more convenient in other settings). First, it can be difficult to get rid of the core when you have finished the apple. Second, it is not always easy to take suitable sizes of a whole apple – and if you get too much in your mouth it is not always so appetizing to look at, if you stand among other people. By cutting the apple in pieces, it becomes easier to eat and you don’t have to worry about how to get rid of the core afterwards.
Figure 4. An illustration showing the convenience principle at work, emphasizing that food options should be easy to eat.
Thus, for our experiment we cannot identify the potential interaction effect – and the same goes for their potential derivatives: eating more apple, might lead you to eat less cake, and being surrounded by people eating more apple and less cake, might make you adopt that behaviour as well. What is positive, though, is that for the setting studied all these effects point in the same direction. So, at the applied level, this need not concern us much, when considering how to rearrange the classic configuration of the coffee break buffet.
Conclusion
Ultimately, the experiment confirms the general point that similar experiments working with Unit Bias and the Principle of Convenience have indicated:
Serve food in suitable and convenient portion sizes relative to what you consider a healthy intake.
In that way we can, to a large extent, facilitate far more appropriate food intake than we ourselves are able to handle, when left alone to take into account all aspects of the complex contexts on the narrow path of healthy choices.
Read the References
[3] Geier, A. B., Rozin, P., & Doros, G. (2006). Unit bias: A new heuristic that helps explain the effect of portion size on food intake. Psychological Science, 17(6), 521-525.
[4] Wansink, Brian (2007). “Portion Size Me: Downsizing our consumption norms”. Journal of the American dietetic association. 20 (10): 1–4.
[5] Hansen, P.G.; Skov, KL; Skov, LR; Jespersen, AM; Schmidt, K (2015) Apples vs. Brownies : A Field Experiment in Rearranging Conference Snacking Buffets to Reduce Short-Term Energy. Journal of Foodservice Business Research, Vol. 19, Nr. 1, 2015.
11 – 13 November, 2024 | COPENHAGEN
JOIN OUR NEXT MASTERCLASS