Content uploaded by Brian Wansink
Author content
All content in this area was uploaded by Brian Wansink on Jan 21, 2016
Content may be subject to copyright.
Environmental Persuaders and the 200 Daily Food Decisions
Brian Wansink, Ph.D.
Cornell University
Jeffery Sobal, Ph.D.
Cornell University
May 3, 2006
Corresponding Author: Brian Wansink, 110 Warren Hall, Cornell University, Ithaca,
NY 14853-7801. Tel: 217-244-0208 & 607-255-5024. E-mail: Wansink@Cornell.edu
Key words: Mindless eating, Food-related decisions, estimation, Obesity, Meal Cessation
Word count (excluding tables):
The published version of this working paper is:
“Wansink, Brian and Jeffery Sobal (2007), “Mindless Eating: The 200 Daily Food
Decisions We Overlook,” Environment and Behavior, 39:1 (January), 106-23
2
Environmental Persuaders and the 200 Daily Food Decisions
Abstract
How aware are people of the number of food-related decisions they make in a day
and how the environment influences these decisions? Study 1 surveyed 139 people
showed they grossly underestimated the number of food-related decisions they made – by
an averaged of over 220 decisions – particularly in initiation and cessation of eating.
Study 2 content analyzed 749 debriefing comments of controlled field studies. Although
the people in these studies overserved and overate 31% more food as a result of having
been given an exaggerated environmental cue (large bowl, large spoon, etc.), 52% denied
having eaten more, and 45% attributed it to other reasons (such as hunger). These studies
underscore two key points: First, we are aware of only a fraction of the food decisions
we make. Second, we are either unaware of how our environment influences these
decisions or we are unwilling to acknowledge it.
3
Environmental Persuaders and Today’s 200 Food Decisions
Many food-related decisions occur in distracting environments and may lead to
relatively “mindless eating.” This would explain why people often cannot really explain
why they ate six chocolates from the office candy dish, ate two bites of chicken for every
one bite of cole-slaw at lunch, or why they consumed three helpings of potatoes for
dinner (Wansink 2004).
Food choice decisions often focus on what is eaten, while food consumption
decisions are a subset of food choice which focuses more specifically on volume
decisions. The former determine what we eat (soup or salad); the latter determine how
much we eat (half of the bowl or all of it). Yet environmental factors (such as package
size, plate shape, lighting, variety, or the presence of others – Stroebele and de Castro
2004) can increase food consumption volume far more than many people realize.
Here we investigate one of the ironies of food consumption research. Whereas people
will acknowledge that environmental factors influence others, they often wrongly believe
they themselves are unaffected. This suggests environmental influences occur at a basic
level at which people are not aware or do not monitor. Understanding these influences on
consumption volume has immediate implications for research, nutrition education, and
consumer welfare. This article examines two of the reasons environmental factors may
influence consumption intake and why they do so.
Environmental Influences of Overserving and Overeating
4
Environmental drivers of food consumption can be categorized as relating to the eating
environment and the food environment (see Figure 1). The eating environment refers to
the ambient factors that are independent of food, such as atmosphere, the effort of
obtaining food, the time of day, the social interactions that occur, and the distractions that
may be taking place, and others. In contrast, the food environment refers to the food itself
and to factors that directly relate to the way food is provided or presented, such as its
salience, structure, package or portion size, whether it is stockpiled, and how it is served.
An academic distinction has often been made between overserving and overeating.
While the size of a serving bowl might offer a visual trick that influences how much a
person serves oneself, a food-regulation perspective would argue that it has no influence
on the actual amount consumed. That is, if one overserves himself, he will stop eating
when full. In practice, there is a strong link between how much one serves and how
much one eats. One study showed there is a 92% corrleation between the two behaviors
(Wansink and Cheney 2005).
[Insert Figure 1]
Both the eating and food categories of environments contribute directly to
consumption volume. Additionally, they can also contribute indirectly because they
suggest consumption norms and inhibit consumption monitoring. For instance, dining
with a friend can have a direct impact on consumption because of the longer duration of
the meal (Strobele and de Castro 2004; French, Story, Jeffrey 2001; de Castro 1994;
2000; de Castro and Brewer 1992). Communal eating can also have an indirect impact
on consumption volume because of the intake norms set by the friend---who cleans his
5
plate and orders a dessert---and because the enjoyment of his or her company distracts
one away from accurately monitoring consumption.
Although research has effectively identified many of the environmental factors that
influence consumption (e.g., Stroeble and de Castro 2004; Wansink 2004), it has less
effectively explained why they do so. Two promising starting points involve consumption
norms and consumption monitoring. While consumption norms and consumption
monitoring have been generally posited as mediating consumption intake (Wansink
2004), they have not been examined in detail to identify the extent to which they operate.
Are We Aware of How Many Food-related Decisions We Make?
The ability to monitor consumption can help reduce discrepancies between perceived
and actual consumption levels (see Figure 1). The influence of environmental factors on
consumption is magnified because they can bias or confuse estimates of how much
someone has eaten, or even the number of times someone thinks they are actively making
decisions about starting or stoping an eating episode.
Not surprisingly, a major determinant of how much one eats is often whether the
person deliberately paid attention to (or attempted to monitor) how much he or she ate
(Arkes 1991; Polivy et al 1986; Polivy and Herman 2002). In lieu of monitoring how
much one is eating, people can use cues or rules-of-thumb (such as eating until a bowl is
empty) to gauge how much they will eat (Wansink, Painter, and North 2005).
Unfortunately, using such cues and rules-of-thumb can yield inaccurate estimates and
surprises. In one study, unknowing diners were served tomato soup in bowls that were
refilled through concealed tubing that ran through the table and into the bottom of the
6
bowls. People eating from these “bottomless” bowls consumed 76% more soup than
those eating from normal bowls, but estimated that they ate only 4.8 calories more
(Wansink, Painter, and North 2005).
Are We Aware of the Consumption Norms that Have Lead Us to Overeat?
People can be very impressionable when it comes to how much they will eat. There is
a flexible range in how much food an individual can eat (Herman and Polivy 1984), and
someone can often “make room for more” (Berry, Beatty, and Klesges 1985) and be
influenced by consumption norms around them (see Figure 1).
For many individuals, determining how much to eat or drink is a mundane and
relatively low-involvement behavior that is a nuisance to monitor continually and
accurately, so they instead rely on consumption norms to help them determine how much
they should consume. Consumption can be further influenced by other norms or cues that
are present in the environment. Many seemingly isolated influences of consumption---
such as package size, variety, plate size, or the presence of others---may involve or
suggest a perceptual consumption norm that influences how much individuals will eat or
drink (Wansink 2004). The use of consumption norms, as with normative benchmarks in
other situations, may be relatively automatic and may often occur outside of conscious
awareness (Schwarz 1996; 1998).
The trouble with the impact of consumption norms is that they occur at such a low-
level of consciousness that people may be unaware of how much influence they have.
For this reason, we are likely to be less vigalent when consumption norms are being
communicated. Even when consumption norms do influence us, there is anecdotale
7
evidence that people are generally either unware of their influence or they are unwilling
to acknowledge it (Wansink, Painter, and North 2005).
Past evidence of the presence or the absence of this awareness has sometimes
been suggested in the context of lab experiments (Vartanian and Herman 2005; Wansink
and Cheney 2005). The problem with trying to generalize from such artificial contexts is
that people are generally aware that some manipulation has occurred, and their
willingness to accurately acknowledge their duplicity may lead them to deny any
influence, simply out of reactance (Meiselman 1992). This phenomenon can best be
observed in the context of controlled field studies conducted in natural environments.
Building upon Figure 1, two studies investigated the two mediating factors of
consumption monitoring and consumption norms. In addressing our ability to effectively
monitor our consumption, Study 1 provides preliminary evidence about whether we are
aware of how many food-related decisions we make. To address our awareness of the
influence of consumption norms, Study 2 content analyzes debriefing data from seven
studies of environmental cues.
Study 1.
Are We Aware of How Many Food-related Decisions We Make?
The purpose of this study was to provide an initial examination of how many
food-related decisions a person makes in contrast to how many they believe they make.
Method. One hundred and fifty adults who had been involved in earlier studies were
contacted through email and were asked a series of questions related to food-related
8
decisions. They were initially asked to estimate how many total decisions about foods
and beverages they make in one day. They were then asked six questions about snacks,
six questions about meals, and six questions about beverages. The numbers from these 18
different questions were aggregated, and self-reported questions about height and weight
were asked and used to calculate a relative weight as Body Mass Index (BMI) for each
participant. Following the guidelines of the Center for Disease Control and the World
Health Organization (WHO 1998), participants were classified as normal weight if their
BMI was below 25 kg/m,² as overweight if their BMI was higher than 25 kg/m², and
obese if it exceeded 30 kg/m².
Results. Of the 150 participants recruited, 139 (93%) completed the study. The
average participant initially estimated they made 14.76 food and beverage-related
decisions in the day (see Table 1). Upon aggregating the total number of decisions they
made upon greater reflection, it was found that they instead made 219.0 decisions, which
is significantly higher (t=178, p<.001) than their initial global estimate.
While the typical person estimated they made around 15 food and beverage
decisions in a day, the average that was calculated from subsequent questioning was 219,
approximately 200 more. Part of these inconsistencies are due to a tendency for people to
consider only food choice decisions as actual food decisions. For example, a snack
deliberation in front of a vending machine would not be counted as a food-related
decision by many people unless it resulted in an actual purchase. In general, most of the
food decisions people neglect to consider as decisions are those involving initiation and
cessation.
[Insert Table 1]
9
Interestingly, these calculation estimates vary between BMI categories. There was
a significant a J-shaped relationship between weight and food-related thoughts. A spline
regression indicated that both the legs of this J-shaped relationship were significant at the
p<.05 level.
[Insert Figure 2]
Discussion. Given that people so dramatically underestimate the number of meal-
related decisions they make in a day, perhaps it is not unfair to say we often engage in
mindless eating. Each of these small decision points is a point where a person can be
unknowingly influenced by environmental cues. Given the interest in better controlling
our food intake, people need to be more aware of the number of decisions that influence
what they eat as well as when they start and when they stop eating.
In addition, the possibility of a J-shaped relationship between weight and food-
related thoughts merits more investigation. Although not significant, it still suggests that
obese people (BMI>30) may be qualitatively different than those who are simply
overweight (BMI 25-30). When grouped together for analysis, which is often the case,
aggregation of overweight and obese people could obscure important differences. For
instance if the estimates of the two groups were collapsed, they would look almost
identical to that of the normal weight people.
At the core of mindless eating is the idea that we make many more food-related
decisions than we are aware of having made. While some decisions focus on the choice
of particular foods, many more decisions involve the initiation and cessation of eating
10
(Rozin et al 2003). If people were more conscious of the number of food-related
decisions they make in a day, they could be more vigilant of how their environment is
influencing them (French, Story, Jeffrey 2001).
Study 2.
Are We Aware of the Consumption Norms that Have Lead Us to Overeat?
Study 1 suggeted that we make a much larger number of food-related decisions than
most of us realize. Each of these decisions that we are not consciously aware of provides
an opportunity for being unknowingly influenced by environmental cues. In Study 2 we
investigate whether people 1) are aware of overconsuming, or 2) aware of being impacted
by these cues after the cues and their general impact is made salient.
Method. Study 2 involved a content-analysis of seven controlled field studies
which investigated how environmental factors such as bowl size, spoon size, and glass
shapes influenced how much people consumed in natural environments when randomly
assigned to an exaggerated treatment condition. To assess the awareness of these factors,
the qualitative data collected during the post-experiment debriefings was coded using
content analysis procedures (Webber 1989; Neuendorf 2002). Across all of these studies,
the same two questions were asked of those in the exaggerated (big bowl, big spoon, etc.)
treatment conditions:
1. “How much did you eat compared to what is typical for you?”
11
2. “In this study, you were in a group that was given [a larger bowl]. Those
people in your group ate an average of 20-50% more than the others. Why do you think
you might have eaten more?”
The answers to the first question about amount eaten were coded as either “less
than,” “about the same,” or “more than.” The second question about explanations for
overeating was coded as to 1) .they denied eating more, 2) they attributed it to hunger, 3)
they attributed it to the intervention, or 4) an other explanation (being in an exciting
situation, etc.).
Results. In total, 749 people were involved in these field studies with roughly
half of them being in the exaggerated environmental cue conditions. Among this
treatment group, although the average increase in consumption over the control was 31%.
However, an average of 73% of the participants believed they ate as much as they
normally ate. Of those remaining, an average of twice as many believed they had eaten
less compared to those thinking they might have eaten more (19% vs. 8%). For the 8%
of people to have eaten enough to fully account for this 31% increase (excluding those
who claimed to have eaten less), each would have had to eat 387% more than the average
member in the control group.
[Insert Table 2]
When told of their treatment groups’ bias, and when asked why they might have
eaten more, an average of 52% claimed they did not eat more, and 31% said that if they
did eat more, it was because they were hungry. An average of only 2% of the participants
believed they had eaten more because of the environmental cue that had been specifically
named. Fifteen percent claimed they ate more for miscellaneous reasons, such as because
12
it was a special occasion (the Super Bowl, or a celebratory ice cream social) or because it
was “free.”
Discussion. Although the typical study in this sample lead the average participant
to consume 31% more when presented with an exaggerated environmental cue, only 2%
believed they had been personally influenced by this cue. Furthermore, only 9% of the
participants believed they had eaten more than they normally would have eaten in that
situation. Even when confronted with how much their group had overeaten, over half of
the participants denied that they had been influenced. Of those remaining 48% of
participants who did believe it conceivable that they had possibly overeaten, and 96% of
these believed it must have been because they were hungry or for another reason
unrelated to the actual environmental cue itself.
Lab studies have often found that people either do not believe they were
influenced by external cues or to not want to admit this was the case (Vartanian and
Herman 2005). While such studies have not been systematically evaluated, their
anecdotal evidence has often been discounted because of their demand effects. Using
field studies, we show here that people claim to be unaware of these factors increasing
their consumption. Even when confronted with empirical findings, most participants in
environmental manipulations continue to disavow the findings or to look for alternative
explanations. Although these results do not fully disentangle unawareness from denial,
the consistency of the findings across studies point to a strong systematic influence that
goes beyond what people either know or will confess to.
13
General Discussion
The environment influences food-related decisions consistently through the day.
There are two problems with this. First, we are not aware of how many decisions we
make that are being influenced. Second, we are not aware or unwilling to acknowledge
that the environment has any impact on us at all.
These findings show that people tended to not acknowledge their own
susceptibility to manipulations of the food environment. This is consistent with other
psychological work that shows that people tend to have flawed self-assessments, be
overconfident, and overestimate their own capabilities (Dunning 2005). Thus broader
tendencies to be self-confident and competent reveal themselves in food intake decisions,
which may lead to overconsumption and overweight.
These data suggest that many people engage in mindless eating where they are not
consciously aware of the effects of the environment on how much food or beverage they
consume. Other food research has found that people make “automatic” food choices
where they unconsciously eat without considering what or how much food they select
and consume (Furst et al 1996). Automaticity is an important part of everyday behavior
(Uleman & Bargh 1989), and appears to be involved in food consumption as well.
Increasing mindfulness (Langer 1990) may facilitate healthier food choices.
Useful future research could characterize social and psychological characteristics
that predict peoples’ perceptions of eating decisions and that acknowledge environmental
influences. This type of investigation could help identify audiences and mechanisms that
could be used to make eating more salient and make people more mindful of influences
of the built environment on how much food and drink they actually consume.
14
References
Arkes, H.R. (1991). Costs and benefits of judgment errors---implications for debiasing.
Psychological Bulletin 110:486—98.
Berry, S.L., Beatty, W.W., & Klesges, R.C. (1985). Sensory and social influences on ice-
cream consumption by males and females in a laboratory setting. Appetite 6:41—
45.
De Castro, J.M. (2000). Eating Behavior: Lessons From the Real World of Humans.
Ingestive Behavior and Obesity 16: 800-13.
De Castro, J.M., & Brewer, E. (1992). The amount eaten in meals by humans is a power
function of the number of people present. Physiology & Behavior 51: 121-5.
De Castro, J.M.. (1994). Family and Friends Produce Greater Social Facilitation of Food-
Intake Than Other Companions. Physiology & Behavior 56: 445-55.
Dunning, D. (2005). Self-insight: Roadblocks and detours on the path to knowing thyself.
New York: Psychology Press.
French, S.A., Story, M., & Jeffery, R.W. (2001). Environmental influences on eating and
physical activity. Annual Review of Public Health 22:309-325.
Furst, T., Connors, M., Bisogni, C.A., Sobal, J., & Falk, L.W. (1996). Food choice: A
conceptual model of the process. Appetite 26:247-266.
Herman, C.P., & Polivy, J. (1984). A boundary model for the regulation of eating. In
Eating and Its Disorders, ed. A.J. Stunkard, & E. Stellar, pp. 141--56. New York:
Raven.
Langer, E. (1990). Mindfulness. New York: Addison Wesley.
Meiselman, H.L. (1992). Obstacles to studying real people eating real meals in real
situations. Appetite 19(1):84-86.
Neuendorf, K.A. (2002). The content analysis guidebook. Thousand Oaks, CA: Sage.
Polivy, J., Herman, C.P., Hackett, R., & Kuleshnyk, I. (1986). The effects of self-
attention and public attention on eating in restrained and unrestrained subjects.
Journal of Personality and Social Psychology. 50:1203—24.
Polivy, J., & Herman, C.P. (2002). Causes of eating disorders. Annual Review of Psychology
53:187—213.
15
Rozin, P., Kabnick, K., Pete, E., Fischler, C., & Shields, C. (2003). The ecology of
eating: Smaller portion sizes in France than in the United States help explain the
French paradox. Psychological Science 14:450-454.
Schwarz, N. (1996). Cognition and Communication: Judgmental Biases, Research
Methods and the Logic of Conversation. Mahwah, New Jersey: Lawrence
Erlbaum Associates, Publishers.
Schwarz, N. (1998). Warmer and More Social: Recent Developments in Cognitive Social
Psychology Annual Review of Sociology 24: 239-64
Stroebele, N., & de Castro, J.M. (2004). Effect of ambience on food intake and food
choice. Nutrition 20:821-838.
Uleman, J.S., & Bargh, J.A. (eds). (1989). Unintended thought. New York: Guilford
Press.
Vartanian, L. & Herman, C.P. (2005). Unawareness and Denial in Intake Interventions,
unpublished working paper.
Wansink, B, (2004), Environmental factors that increase the food intake and consumption
volume of unknowing consumers, Annual Review of Nutrition 24: 455-479.
Wansink, B., & Cheney M.M. (2005), “Super Bowls: Serving Bowl Size and Food
Consumption,” JAMA – Journal of the American Medical Association, 293:14
(April 13), 1727-1728.
Wansink, B., & Kim, J. (2005). Bad Popcorn in Big Buckets: Portion Size Can Influence
Intake as Much as Taste. Journal of Nutrition Education and Behavior, 37:5 (Sept-
Oct), 242-5.
Wansink, B., & Park, S.B. (2001). At the Movies: How External Cues and Perceived Taste
Impact Consumption Volume. Food Quality and Preference, 12:1 (January), 69-74.
Wansink, B., & van Ittersum, K. (2003). Bottoms Up! The Influence of Elongation and
Pouring on Consumption Volume. Journal of Consumer Research, 30:3
(December), 455-463.
Wansink, B., Painter, J.E., & North, J. (2005), Bottomless Bowls: Why Visual Cues of
Portion Size May Influence Intake. Obesity Research, 13:1 (January), 93-100.
16
Wansink, B., & van Ittersum, K. (2005). Shape of Glass and Amount of Alcohol Poured:
Comparative Study of Effect of Practice and Concentration. BMJ – British
Medical Journal, 331:7531 (December 24) 1512-1514.
Webber, R.P. (1989). Basic content analysis. Newbury Park, CA: Sage.
WHO. (1998). Obesity: Preventing and managing a global epidemic. Geneva: World
Health Organization.
17
Figure 1.
Environmental Influences on Overserving and Overeating
Environmental Influences on
Food Intake
The Food Environment
-- Salience of food
-- Structure and variety of food
assortments
-- Size of food packages and
portions
-- Stockpiling of food
-- Serving containers
The Eating Environment
-- Eating atmosphere
-- Eating effort
-- Eating with others
-- Eating distractions
Overserving or
Overeating
Consumption
Monitoring:
Unawareness of How
Many Food-related
Decisions We Make
Consumption
Norms:
Unwillingness to
Believe Consumption
Norms Influence Us
18
Figure 2.
The Calculated Number of Daily Food- and Beverage-related Decisions
Number of
Decisions
19
Table 1.
All People Underestimate How Many Food-related Decisions They Make
Body Mass Index (BMI)
F-Value
(p-value)
Normal
Weight
(BMI<
25)
(n = 71)
Over-
weight
(BMI
25-30)
(n = 38)
Obese
(BMI>
30)
(n=30)
Average
“How many total food- and
beverage-related decisions do
you make in one day?”
14.81
12.21
17.90
14.76
0.13 (.877)
Actual (calculated) number of
snack-related decisions
52.5
28.4
53.5
46.1
0.59 (.554)
Actual (calculated) number of
meal-related decisions
75.4
63.3
120.1
82.5
2.65 (.075)
Actual (calculated) number of
beverage-related decisions
94.0
93.8
129.1
102.7
0.99 (.372)
Total (calculated) number of
food- and beverage-related
decisions
201.6
187.3
289.1
219.0
2.16 (.120)
20
Table 2.
Participants in Seven Field Studies Deny the Influence Interventions Have on their Intake Behavior1
Sample and
Context of Study
Intervention and
Findings
“How much did you eat
compared to what is typical for
you?”
“In this study, you were in a group that was given [a larger
container]. Those people in your group ate an average of 20-50%
more than the others. Why do you think you might have eaten
more?”2
Less
About the
Same
More
“I didn’t eat
more”
“I was
hungry”
“The
(intervention)
influenced me”
Other
40 MBA students
at a Super Bowl
party in a bar
(Wansink &
Cheney 2005)
Those serving
themselves Chex
Mix from 4-liter
bowls served
53% more than
those serving
from 2-liter
bowls
23%
57%
20%
63%
31%
3%
3%
83 nutrition
experts at an ice
cream social to
celebrate
promotion of a
colleague
(Wansink, van
Ittersum, and
Painter 2006)
Those given 3 oz
ice cream spoons
served 14.5%
more ice cream
than those given
2 oz spoons.
25%
71%
4%
45%
34%
3%
18%
161 teenagers at
a summer
nutrition camp
(Wansink & van
Ittersum 2003)
Those given wide
glasses poured
77% more juice
than those given
tall glasses
holding the same
32%
65%
3%
78%
19%
0%
3%
21
volume
86 Philadelphia
bartenders
working at their
bar
(Wansink &van
Ittersum 2005)
Those given 10
oz tumbers
poured 25%
more alcohol
than those give
10 oz highball
glasses
0%
97%
3%
88%
--
0%
12%
143 evening
moviegoers in
Feasterville, PA
(Wansink &Kim
2005)
Those given
large popcorn
buckets ate 45%
more than those
given medium
buckets
14%
78%
8%
12%
79%
2%
7%
161 afternoon
moviegoers in
Mt. Prospect, IL
(Wansink &Park
2001)
Those given
large buckets ate
48% more than
those given
medium buckets
19%
75%
6%
15%
77%
5%
3%
86 people in a
Laundromat
(Wansink 1996)
Those given
large bottles of
detergent used
32% more than
those given
medium bottles
19%
65%
11%
24%
--
11%
65%
Average across all studies
19%
73%
8%
52%
31%
2%
15%
1 Answers are from those in the treatment group who received the intervention that resulted in greater consumption
2 The specific intervention in the study was noted at this point. Here, the example of larger bowls was used.
