J Saw 2006 528
J Saw 2006 528
J Saw 2006 528
Mindless Eating:
Rachael Goldstein
Lehigh University
Mindless Eating 2
Abstract
Do portion sizes affect how much food you eat? Recent studies have shown that when
people are given larger portions of food, they tend to eat more than when they are served a
smaller portion. Past research has overlooked the influence of gender in individuals’ sensitivity
to portion sizes. Are females less influenced than males by changes in portion sizing as a result
of increased sensitivity to health and diet? We tested the hypothesis that the phenomenon in
which people eat more food when presented with larger portions, demonstrated in previous
research, will be true of males to a greater extent than of females. To test this idea, we presented
college students with either small or large bags of popcorn that they were free to eat during
cartoon clips that lasted for 15 minutes. We measured the initial and final weight (in grams) of
the bags. The final weight of each person’s bag was subtracted from the initial weight to
determine how much popcorn each person consumed. We predicted a main effect of portion size:
Participant s will eat more when given larger portion sizes. Consistent with our prediction,
participants consumed more popcorn from the large portion. However, the difference between
the means was not significant. We also predicted an interaction between portion size and gender,
with the effect of portion size on the amount of popcorn consumed depending upon gender. For
males, we expected to see a large difference in the amount of popcorn eaten when given a larger
portion size. In contrast, for females, we expected that the amount of popcorn consumed from a
larger portion would be only slightly greater than the amount eaten from a smaller portion of
popcorn. Consistent with our hypothesis, men ate more from the large bag of popcorn than
women. However, the difference between the means was not significant.
Mindless Eating 3
Mindless Eating: How Differences in Portion Size Influence Popcorn Consumption For Males
and Females
The prevalence of overweight and obesity has increased dramatically among adults and
children in the U.S. in recent years (Young & Nestle, 2002). With obesity becoming an
increasing medical concern in this country, researchers have investigated its causes. It has been
reported that the rise in body weight is caused by increased energy intake (Young &Nestle,
2002). If Americans are consuming more energy than in the past, it becomes important to
determine the cause of this increase. Do portion sizes affect how much food people consume?
Young and Nestle (2002) indicated that larger portions both contain more energy and
encourage people to eat more. Fischer, Rolls, and Birch (2003) further the finding that people
eat more when given larger portion sizes through their research with preschool aged children.
These researchers wanted to determine how multiple instances of exposure to large portion sizes
of an entrée effects a child’s ability to select portion sizes, food intake, and the awareness of
portion sizes. They evaluated 30 children at two sessions of lunches that were either given an
age-appropriate portion or a large portion of an entrée. Results showed that when children were
given large portions of an entrée, there was an increase in entrée and total energy intake.
Specifically, the children did not eat less of the other foods served at the meal, even though they
were given larger portions. The more responsive children were to portion size, the more likely
they were to overeat. These results clearly indicated that large portions sizes produced excessive
intake at meals.
Past research, such as the study completed by Fischer et al. (2003), indicated that
environmental cues (packaging and container size) increased the consumption of appealing
foods. Do environmental cues increase the consumption of less favorable foods? Wansink and
Mindless Eating 4
Participants were either given medium or large bags of popcorn that were either fresh or stale.
Wansink and Kim (2005) found that participants ate more from the large bag of popcorn
regardless of the quality of the popcorn. These results indicate that portion sizes can increase
The extensive research on portion sizes has developed a strong link between portion
sizes, food consumption, and the rise in obesity. Although there have been many studies that
support the relationship between these components, past research has overlooked the influence of
gender on individuals’ sensitivity to portion sizes. Are females less influenced than males by
First, it is essential to distinguish the biological differences that create a difference in the
amount of food males and females consume. Gender differences in food intake first appear in
adolescence. Boys, with their greater average body weight and higher resting metabolic rate,
mainly require a higher energy intake tha n do girls (Rolls et al., 1991). The difference in energy
intake is the highest during adolescence (Rolls et al., 1991). For instance, the Food and Nutrition
Board of the National Academy of Sciences has recommended energy of 3,000 kcal/day for 15 to
18 year old males involved in light-to- moderate activity, whereas the recommended energy
intake for females in the same range and activity level is 2, 200 kcal/day (Rolls et al., 1991). By
adulthood, the energy needs of men are still greater than women; however, this difference is
Beyond the biological distinctions that differentiate the overall amount of food that
females and males consume, sociocultural factors influence the gender difference in food
consumption. For instance, studies revised by Rolls et al. (1991) have shown that there are
Mindless Eating 5
socialization differences between men and women concerning mealtime etiquette. Specifically,
women may have been taught to take small bites, which would decrease the time between bites,
decrease eating rate, and increase the total number of bites. Rolls, Pirraglia, Stoner, & Laster
(1990; as cited in Rolls, Fedoroff, & Guthrie, 1991) completed a study that supports the gender
difference in eating styles. These researchers hypothesized that the way food is presented will
have different effects on food consumption in males and females. The experimenters presented
participants with sandwiches that were either whole units or in cocktail size pieces. The results
showed that men ate more overall and more of the whole sandwiches, whereas women did not
show a difference between conditions. The findings suggest that a feminine eating style involves
eating less than men and at a slower rate than men, despite the mode of presentation.
Further, additional studies have demonstrated the pressure for women to eat less in
accordance with a gender appropriate role. Chaiken and Pliner (1987; as cited in Rolls, Fedoroff,
&Guthrie, 1991) tested the hypothesis that “eating lightly” is a sex-role-appropriate behavior for
women. As a result, women who ate smaller amounts of food would be considered more
feminine than women who ate large meals. Male and female participants read a food diary
designated to either a male or female target that was portrayed as eating a small meal or a large
meal. The ratings of the male target were unaffected by the difference in portion size, whereas
the female target who ate the smaller meal was considered to be more feminine, more concerned
about her appearance, more attractive, and more likely to display stereotypically feminine
personality traits. These results show the strong cultural pressure women face to eat according to
the societal food consumption expectations for females. Ultimately, the pressure women face to
eat less then men can lead women to become more cautious of their eating habits.
Mindless Eating 6
In addition to the societal pressure for women to eat less, there are gender differences in
satisfaction with body weight and shape. Throughout history, there have always been certain
ideals that have influenced the way women view and treat their bodies. Women’s status has been
more dependent on physical characteristics than that of men, ultimately causing more pressure
for women to conform to the current physical ideal (Rolls et al., 1991). Data from the Miss
America pageant and women’s magazines indicated a shift toward a thinner standard in body
weight for women in recent years (Rolls et al., 1991). The increased cultural emphasis upon the
ideal feminine figure has encouraged many women to diet in order to achieve this goal. This
idealized image has led women to become less satisfied with their body shape than men and has
led women to perceive themselves as being overweight (Rolls et al.,1991). However, men are
more likely to see themselves as underweight and want to gain weight (Rolls et al., 1991). Both
the biological and cultural influences upon women to “eat lightly” and to fit an idealized image
have made women more sens itive to food consumption. With the increased sensitivity women
experience towards food consumption, changes in portion sizes might have a very small impact
upon the actual amount of food women consume. Yet, with men’s ove rall greater satisfaction
with body weight and shape, men might be more affected and consume more when given larger
portions.
The hypothesis we will test, then, is that the phenomenon in which people eat more food
when presented with larger portions, demonstrated in previous research, will be true of males to
a greater extent than of females. To test this idea, we presented college students with bags of
popcorn that they were free to eat. Participants were either given a small bag of popcorn or a
large bag of popcorn. After participants were given bags of popcorn, they were shown two
cartoon clips that lasted for 15 minutes. We measured the initial and final weight (in grams) of
Mindless Eating 7
the bags. The final weight of each person’s bag was subtracted from the initial weight to
determine how much popcorn each person consumed. If our hypothesis is correct, then there
should be a main effect of portion size: Participants will eat more when given larger portion
sizes. There should also be an interaction between portion size and gender, with the effect of
portion size on the amount of popcorn consumed depending upon gender. For males, we
expected to see a big difference in the amount of popcorn eaten when given a larger portion size.
In contrast, for females, we expected that the amount of popcorn consumed from a larger portion
would be only slightly greater than the amount eaten from a smaller portion of popcorn.
Although not of primary interest, we also predicted a main effect of gender, such that males will
Method
Participants
pool participated for course credit. Nineteen participants were male and 21 were female.
Design
The first independent variable was portion size, with two levels: small bag and large bag.
The second independent variable was gender, also with two levels: male and female. The
dependent variable was the amount of popcorn consumed, which was measured in grams.
The experiment used a between-subjects design, with subjects randomly assigned to the
levels of the first independent variable, portion size. Gender was recorded. Nine subjects were in
the Small Portion, Male condition, 11 subjects were in the Small Portion, Female condition, 10
subjects were in the Large Portion, Male condition, and 10 subjects were in the Large Portion,
Female condition.
Mindless Eating 8
Materials
We used economy size bags of packaged prepared popcorn that we separated into
individual, weighed portions. Participants received either a large bag of popcorn, which weighed
160 grams, or a small bag of popcorn, which weighed 80 grams. Participants watched two
randomly assigned Tom & Jerry cartoon clips (a children’s cartoon series) that lasted for 15
minutes while eating popcorn. A scale was used to measure the initial and final weight (in
grams) of all the bags. Also, participants were given an informed consent form, which presented
Procedure
Before participants arrived, the experimenters filled the bags with either a small portion
of popcorn (80 g) or a large portion of popcorn (160 g). The bottom of each bag was marked
with the participant number. The initial weight of the bags was measured using a scale (in
grams). We randomly assigned participants to conditions and kept a tally of the number of
participants per condition in order to balance the number of participants in each condition.
Participants tested in the same group were all given the same amount of popcorn.
The experimenters verbally instructed participants that they would be watching a short 15
minute cartoon clip. They were told that they could enjoy some popcorn during the movie if they
wished. Lastly, the experimenters told participants to wait patiently when the cartoon clip ended
for further instructions. The experimenters distributed the bags of popcorn. Then, two randomly
assigned cartoon clips of Tom & Jerry totaling 15 minutes were shown. Next, the experimenters
instructed participants to remain seated while the bags of popcorn were being collected. The
experimenters collected the bags from each person, marking the gender of the participant on the
Mindless Eating 9
bottom of the bag. Then, the participants were orally debriefed and asked to take a written
debriefing form on their way out. Participants were thanked for their participation in the study.
After participants left the room the study was conducted in, we measured the final weights of
the bags. The weight of each person’s remaining popcorn was subtracted from their initial weight
Results
The mean weights of the popcorn were calculated in grams for each of the four conditions.
If our hypothesis is correct, then there should be a main effect of portion size: Participants
should eat more when given larger portion sizes. The pattern of the means is consistent with this
prediction. As Table 1 and Figure 1 show, participants ate more from the large bag of popcorn
than from the large bag of popcorn. Howeve r, the difference between the means was not
significant, F(1,36) =.14, p > .50. The results provide weak support for our hypothesis that
people eat more when given a larger portion size. The interaction reported next contributes to the
We also predicted an interaction between portion size and gender, with the effect of
portion size on the amount of popcorn consumed depending upon gender. For males, we
expected to see a big difference in the amount of popcorn eaten when given a larger portion size.
In contrast, for females, we expected that the amount of popcorn consumed from a larger portion
would be only slightly greater than the amount eaten from a smaller portion of popcorn. The
pattern of the means is partially consistent with our predictions. As Table 1 and Figure 1 show,
for males, there was a big difference in the amount of popcorn consumed when they were given a
large bag of popcorn. However, females actually consumed slightly less popcorn when given a
Mindless Eating 10
large bag. The interaction between portion size and gender was not significant, F (1,36) = .846,
p > .1. The results for the interaction effect provide weak support for our hypothesis.
Finally, we predicted a main effect of gender, with males consuming more overall than
females. The pattern of means is consistent with this prediction, as shown in Table 1 and Figure
1, because males consumed more popcorn than females. The difference between the means was
significant, F(1, 36) = 8.474, p < .01. This outcome supports past research that men have a
greater body mass and require a higher energy intake than women. Ultimately, this causes men to
Discussion
We found a trend in the pattern of the means that supports the idea that people will eat more
when given larger portion sizes. The trend supports past research completed by Young and
Nestle (2002), who found that when children were given large portion sizes there was an
excessive intake at meals. Also, the trend supports the research conducted by Wansink and Kim
(2005) who found that portion sizes can increase consumption for foods that are less appealing.
Although our results provide weak support for our hypothesis that people eat more when given
larger portion sizes, our sample size was small and we most likely would have obtained more
The trend in the pattern of the means implicates the need for serious health policy regarding
portion sizes in the U.S.; these implications could be further strengthened if more people
participated in our study. Young & Nestle (2002) found that the sizes of current marketplace
foods almost always exceeded the sizes of those offered in the past. National chain restaurants
promoted large-size items on their menus (Young & Nestle, 2002). These increasing portion
sizes have exceeded the federal standards for dietary guidance and food labels (Young & Nestle,
Mindless Eating 11
2002). The increase in portion sizes in the marketplace suggests the need for greater attention to
food portion size as a factor in weight management. It is essential that the government makes
public health efforts to explain the relationship of portion size to caloric intake and weight gain
(Young & Nestle, 2002). In addition, it is important for federal agencies to make serving size
definitions more consistent and comprehensible (Young & Nestle, 2002). Thus, educational and
other public health programs are crucial to address the effects of the trend towards larger portion
sizes.
Additionally, we found a trend in the pattern of the means that supports the idea that the
difference in the amount of popcorn consumed between the small and large bags is more
pronounced for males. This trend supports past research that women are more sensitive to health,
nutrition, and body image. Chaiken and Pliner (1987; as cited in Rolls et al., 1991) tested the
hypothesis that “eating lightly” is sex-role-appropriate behavior for women. The results of the
study indicated that women were expected to consume less food in concordance with their
gender role. Additionally, Rolls et al. (1991) found that women’s status has been more dependent
on physical characteristics than that of men, ultimately causing more pressure for women to
conform to the current physical ideal. This increased sensitivity for women towards food
consumption is supported by the pattern of means which show that women are less affected by a
change to a larger portion size. Like the predicted main effect of portion size, the results of our
interaction between portion size and gender would likely have been more significant if more
It is interesting to note that the pattern of the means did not fully follow our specific
prediction for the pattern of interaction between portion size and gender. The means showed a
large increase in consumption for men when they were given a large portion. However, for
Mindless Eating 12
women, there was actually a decrease in the amount of popcorn consumed from the small to the
large portion rather than a slight increase. It is possible that women eat less from the large
portion because our sample was composed of college students. College students face more
pressure to fit an idealized body image than other groups. If the decrease for our females is
specific to the college population, our findings might be difficult to generalize to the public. If a
more diverse sample of participants was used, such as women not in college and women older
than college age students, in addition to a more ethnically diverse group, the results might be
more representative of the public. If it turned out to be generally true that women eat less from
the larger portion size, it is difficult to explain why there is a constant rise in the obesity rate
It is important to examine the trend in the means which supports the idea that females are
more sensitive to eating, nutrition, and health. Female sensitivity was seen throughout the
experiment because many women would not consume any of the popcorn given to them. The
trend in the means in addition to more people participating in our study, have important health
policy implications for women. Rolls et al. (1991) have suggested that the increased pressure on
women in Western cultures to achieve and maintain a low body weight may contribute to the
increase in the prevalence of eating disorders. It is important for future health policy to focus on
ways to promote a healthy body image among women. Also, future health campaigns should
focus on helping women to take pride in their bodies and also provide ways that women can take
Lastly, our results showed a main effect of gender with males eating more overall than
females. This supports past research about the biological differences between the males and
females regarding food consumption. Rolls et al. (1991) found that males with their greater
Mindless Eating 13
average body weight and higher resting metabolic rate, mainly require a higher energy intake
than girls.
It is essential to look at the constructs of our experiment to see how future researchers
could structure the experiment better in order to obtain more significant results. First, as
previously mentioned, our experiment requires a larger sample size than the size utilized in our
experiment. In addition, the time the experiment was conducted presented another problem.
Cartoon clips were shown in the morning and also around meal times. Participants might not
have been hungry in the morning and also might have consumed a main meal directly before the
experiment. This might have caused participants to eat very small amounts or simply nothing. In
order to determine the effect of the time of day upon participants’ popcorn consumption, it would
be helpful to distribute questionnaires to participants after the experiment, asking how hungry
they were at the time of the experiment. Also, the length of the cartoon clips could have affected
the amount of popcorn participants consumed. Cartoon clips totaling 15 minutes long might not
have been long enough fo r participants to consume a significant amount of popcorn. It would be
advantageous to show a movie that is at least 30 minutes, which would give participants time to
consume more popcorn. For instance, in the experiment completed by Wansink & Kim (2005),
the researchers measured popcorn consumption after a 119 minute movie session. The length of
the movie could have helped the researchers to receive significant results. Lastly, the number of
participants in each session of the experiment could have affected our results. Each group had no
more than five participants and some sessions only had one participant. Eating in a group could
In addition to the design of our experiment, a limitation our study is the generalizability
of our results to other groups. College students might be more sensitive to issues regarding
Mindless Eating 14
eating, body image, and nutrition than people that are not in college. Hart and Ollendick (1985;
as cited in Rolls et al., 1991) found five times the frequency of bulimia in women enrolled in
universities compared to working women. This indicates that college students might face more
social pressure concerning body image, which ultimately affects eating patterns. Rolls et al.
(1991) indicated that the environment is a risk factor for developing an eating disorder. It has
been suggested that the incidence of bulimia is greater in boarding schools and college dorms
(Rolls et al., 1991). Thus, a sample of college students might consume less food than the other
Overall, there were trends in the patterns of the means that supported our hypothesis. The
trends suggest that people eat more when given a larger portion size. Also, the trends suggest that
the difference in the amount of popcorn consumed between the small and large bags is more
pronounced for males. In addition, our results showed a main effect of gender. Specifically,
males consumed more popcorn overall than women due to biological differences. The trend
regarding the interaction between portion size and gender suggests that women are more
sensitive to health, nutrition, and body image. In order to strengthen our experiment, it is
important to look at the vario us constructs of our experiment, such as the length of our cartoon
clip, time of day the experiment was given, and social pressures. Also, one limitation of our
study is the generalizability of our results to other groups. In sum, the trends we found provide
changes in portion sizes. Using a larger sample size would strengthen our trends and provide
References
Chaiken, S., & Pliner, P. (1987). Women, but not men, are what they eat; The effect of meal size
and gender on perceived femininity and masculinity. Journal of Personality and Social
Fischer, J.O., Rolls, B. & Birch, L.L. (2003). Children’s bite size and intake of an entrée are
greater with large portions than with age-appropriate of self-selected portions. American
Hart, K.J. & Ollendick, T. H. (1985). Prevalence of bulimia in working and university women.
Rolls, B.J., Fedoroff, I.C., & Guthrie, J.F. (1991). Gender differences in eating behavior and
Rolls, B.J., Pirraglia, P.A., Stoner, S.A., & Laster, L. (1990). Effect of sandwich size on food
Wansink, B. & Kim, J. (2005). Bad popcorn in big buckets: portion size can influence intake as
Young, L.R. & Nestle, M. (2002). The contribution of expanding portion sizes to the US obesity
Table 1
Mean Amount of Popcorn Consumed (in Grams) as a Function of Portion Size and Gender
_________________________________________________________
Portion Size
___________________________________________________________
____________________________________________________________
X 11.85 14.55
Mindless Eating 17
Figure 1. Mean Amount of Popcorn Consumed (in Grams) as a Function of Portion Size and
Gender
Mindless Eating 18
30
Popcorn consumed (in grams)
25
20
male
15
female
10
0
small large
Portion size