Appetite 86 (2015) 61–73

Contents lists available at ScienceDirect

Appetite
j o u r n a l h o m e p a g e : w w w. e l s e v i e r. c o m / l o c a t e / a p p e t

Research report

The social facilitation of eating. A review

C. Peter Herman *
Department of Psychology, University of Toronto, Toronto, Ontario M5S 3G3, Canada

A R T I C L E I N F O A B S T R A C T

Article history: The social facilitation of eating (i.e., people eating more in groups than when alone) has been studied
Received 30 April 2014 for about three decades now. In this paper, we review the empirical research (diary studies, observa-
Received in revised form 22 September tional studies, and experimental studies) of social facilitation, attending to factors that increase or decrease
2014
socially facilitated eating. We also review the various explanations (e.g., “time extension”) that have been
Accepted 22 September 2014
Available online 26 September 2014
offered for the effect and offer our own speculations as to the underlying mechanisms. Further, we discuss
promising directions for future research on the social facilitation of eating.
© 2014 Elsevier Ltd. All rights reserved.
Keywords:
Social facilitation
Eating
Time extension
Pre-meal planning

Introduction focuses exclusively in the co-action paradigm; if food intake were

The social facilitation of eating refers to increases in food intake
when people eat together as compared to when they eat alone. Al-
though our interest is in people, it is worth noting that social
facilitation has been demonstrated in other animals, going back
several decades (e.g., Harlow, 1932, who cites even earlier studies);
and although the literature involving other animals may have some-
thing to contribute to understanding the human case, a comparison
of human and nonhuman research would take us too far afield
for present purposes, as would a discussion of the social facilita-
tion of drinking (see Stroebele & de Castro, 2004, for a review of
the drinking literature).
Zajonc (1965, p. 269) notes that social facilitation of behavior
is the “oldest paradigm of social psychology,” dating back to Triplett’s
(1897) work on the effects of the mere presence of others. Zajonc
(1965) identifies two main circumstances conducing to social fa-
cilitation – audience effects (when people behave in front of a passive
audience) and co-action effects (when people behave along with
others engaged in the same behavior). In the domain of eating, it
turns out that increased food intake occurs only in the so-called co-
action paradigm. In the audience paradigm, people usually eat
dramatically less (see the section on Noneating Observers in Herman,
Roth, & Polivy, 2003 review). Zajonc, in 1965, did not have access
to any studies involving noneating observers; indeed, in 1965 there
were as yet no audience or co-action studies of eating in humans.
Subsequent research on the social facilitation of eating in humans
* E-mail address: herman@psych.utoronto.ca
diminished in the presence of others, as happens in the audience
paradigm, modern researchers would not consider this to be a vari-
ation on social facilitation, which they regard, by definition, as an
enhancement of intake. Below we will encounter some examples
of social suppression of eating, which sometimes occurs in the co-
action paradigm, especially when one is eating with strangers.
In this paper we will examine the extent (and limits) of the social
facilitation (i.e., the enhancement) of eating as well as the mecha-
nism(s) responsible for the effect. Eating in the presence of other
people, as occurs in social facilitation situations, must be distin-
guished from eating in the presence of other people in situations
that are not classified as social facilitation situations but rather as
situations involving other types of social influence. We have already
mentioned the effect of a passive audience on eaters’ intake. Another
paradigm, modeling, involves eating in the presence of another
person who is also eating (a co-actor), but typically this co-actor
is not behaving freely but rather is an experimental confederate
acting surreptitiously on instructions from the researcher to eat a
specific amount. The naïve participant will typically eat more or less
depending on whether the model/confederate eats more or less. In
the social facilitation situation, all participants are “naïve” and are
typically free to eat more or less, allowing for the operation of re-
ciprocal influence. In some modeling studies, the two co-actors both
behave freely, but almost invariably the study is concerned with the
extent to which the co-actors eat similar amounts (rather than
whether they eat more or less than do solo eaters). In social facili-
tation studies, the comparison is typically between the amounts
eaten by solo eaters and eaters in groups (rather than whether the
people eating in groups eat similar amounts).
For purposes of this review, we supplemented our cumulative
collection of articles with a web search for articles containing “social

http://dx.doi.org/10.1016/j.appet.2014.09.016
0195-6663/© 2014 Elsevier Ltd. All rights reserved.
62 C.P. Herman/Appetite 86 (2015) 61–73
facilitation” and “eating” or “food intake.” The reference section of
all articles was further searched for additional articles.
Evidence of social facilitation of eating
The de Castro diary studies
Although not the first studies of eating in groups, John de Cas-
tro’s diary studies (see de Castro, 1997a, for a review) have a certain
historical primacy inasmuch as it was he who first used the term
“social facilitation” to describe the increase in food intake that occurs
when people eat together.1 de Castro began the diary studies in the
mid-1980s (e.g., de Castro & Kreitzman, 1985; de Castro, McCormick,
Pedersen, & Kreitzman, 1986), instructing participants to keep a
weekly diary in which various aspects of food intake, including the
social circumstances surrounding the meal, were recorded. Various
definitions of meals were used, employing different criteria for
minimum amount eaten and intervals between eating bouts, but
these different definitions usually provided similar data, and de
Castro would focus on a “representative” example, typically a
minimum of 50 kcal in a bout of eating separated from other bouts
by 45 minutes.
The specific influence of the presence and number of other people
on food intake was first explored by de Castro and de Castro (1989),
who found that people eating in groups ate substantially (on average,
44%) more than did people eating alone. “The number of people
present, then, appears to be a major factor associated with the
amount eaten in a meal and to be independent of other factors af-
fecting meal size. . . The addition of the number of people present
to the regressions . . .more than doubl[ed] the variance [in food
intake] accounted for” (p. 243). de Castro and Orozco (1990) re-
ported the same positive association between group size and per
capita food intake. de Castro, over the course of a series of studies
(and reanalyses), concluded that social facilitation was the single
most powerful influence on eating, and that “the number of people
eating with the subject . . .is the best predictor of how much food
an individual will consume” (Redd & de Castro, 1992, p. 749). Ap-
parently, when de Castro began the diary studies, he was not
especially focused on the social facilitation of eating: “neither the
subjects nor the experimenter were aware at the time of data col-
lection that the presence of other people at the meal was of particular
interest” (de Castro & Brewer, 1991, p. 124). This admission was
offered as a preemptive defense against any accusation that de Castro
was looking for evidence of social facilitation and was thus biased
in favor of finding it. Instead, it appears that de Castro became so
impressed with the power of social facilitation at least partly because
he was not expecting it and it was therefore especially surprising
to him. (We experienced a similar “epiphany” when we found that
modeling effects overrode hunger as a determinant of intake;
Goldman, Herman, & Polivy, 1991.)
de Castro, Brewer, Elmore, and Orozco (1991) examined whether
the social facilitation effect might be artifactual. For instance, meals
that tend to be larger (e.g., dinner as compared to breakfast, or meals
eaten in a restaurant as compared to at home) also tend to be the
meals eaten with others; or meals eaten with alcohol (which may
increase food intake or at least overall energy intake) tend to be
meals eaten with others; or snacks (which tend to be small) are more
likely to be eaten alone than are meals (which tend to be larger).
de Castro et al.’s (1991) data, however, indicate that the social cor-
relation (i.e., the correlation between the number of people eating
and the amount eaten [per capita]) is positive in all scenarios,
1 Schlundt and Zimering (1988) found that both obese and normal-weight people
reported difficulty controlling their food intake in social situations, but this obser-
vation was more or less parenthetical.
including breakfast, snacks, meals eaten at home, and meals eaten
without alcohol. Therefore, the social correlation is not artifactual
and indeed appears to be ubiquitous. “On the basis of these find-
ings it would seem reasonable to conclude that social facilitation
of eating is the most important and all pervasive influence on eating
yet identified” (p. 100). Feunekes, de Graaf, and van Staveren (1995a),
on the other hand, concluded from their own diary data that whereas
there was an overall correlation between group size and intake, it
was artifactual, a reflection of increasing intake from earlier to later
meals over the course of the day along with an increasing number
of other eaters from earlier to later meals over the course of the
day. The Feunekes et al. data, admittedly, are from a relatively small
sample with perhaps inadequate power to detect a within-meal
social correlation.
One possible consideration that might limit our own astonish-
ment at the size and ubiquity of the social facilitation effect is that
the diary studies were largely cumulative; that is, the studies were
not based on entirely independent samples but rather consisted of
a growing database of diary entries. By the time of de Castro (1994),
the sample size had grown to more than 500 participants. In other
words, to some extent de Castro’s social facilitation studies are really
based on one data set. Still, that cumulative data set kept proclaim-
ing the same robust message.
de Castro (1991) continued the effort to eliminate artifactual in-
terpretations of the social facilitation effect. In this analysis, he
examined whether the social correlation obtains for both weekday-
only meals and weekend-only meals, to rule out the possibility that
the social correlation arises because people eat more on the week-
ends and are also more likely to eat with others on the weekend.
Although people ate (12%) more on the weekends, and ate with (41%)
more people on the weekends, the social correlation was strong for
both weekdays (r > .3) and weekends (r = .3–.4) separately, indicat-
ing that the social facilitation of eating is not an artifact of people
eating more with more people on weekends. Also, the correla-
tions between meal size and prior and subsequent meal sizes were
negligible; thus, it is not the case that people schedule meals with
(many) others when they are most hungry.
Explorations of the social correlation
de Castro and de Castro’s (1989) social correlation – the corre-
lation between intake and the specific number of people present
at the meal –was an impressive .42. This correlation was not sig-
nificantly reduced (.30) when lone eaters were excluded from the
analysis; in other words, the effect was not simply a matter of people
in groups eating more than lone individuals; rather, the more other
people with whom one ate, the more one ate. In subsequent diary
studies in this series, the social correlation remained strong, usually
between .3 and .4.
de Castro and Brewer (1991), noting that the data underlying the
social correlation were not perfectly linear, applied a more formal
mathematical calculation and concluded that the best description
of the relation between the number of people present and intake
was a decelerating curve, essentially a power function, similar to
what Latané (1981) had described in his “social impact theory,” in
which the presence of others has a cumulative effect, but the in-
cremental size of the effect declines as the number of additional
“others” increases. “One other person present at the meal was as-
sociated with a 28% increase in meal size while 41%, 53%, 53%, 71%,
and 76% increases were associated with 2, 3, 4, 5, and 6 or more
people, respectively” (de Castro & Brewer, 1991, p. 124). In de Castro
and Brewer’s analysis, “the geometric mean of the exponents for
those [meal intake] variables that are best fit by the power func-
tion is 0.22” (p. 122). (The data also support an exponent with a
central tendency of .23, which is the value that de Castro, 1995 pro-
vides.) de Castro and Brewer reject the possibility that a ceiling effect
on intake might be responsible for the shape of the curve, arguing
 C.P. Herman/Appetite 86 (2015) 61–73
that the largest meal sizes were well below the limits of gastric
capacity.
Herman et al. (2003) concluded that whereas evidence of social
facilitation emerges regardless of whether the other people present
are friends, family or strangers – that is, people eat more in groups
than when alone – support for the social correlation (i.e., intake
varying as a function of the number of other people present) is ap-
parent only with family and friends. Thus, eating with another person
who is a stranger will increase intake, but adding more strangers
to the group will not increase intake further.
Further characterization and moderation of social facilitation
de Castro continued to fill in the portrait of social facilitation with
several additional analyses. For instance, he found that the pres-
ence of eating companions increased the duration of the meal (by
51%) without affecting the rate of intake (de Castro, 1990). That is,
people continued to eat at the same rate when in groups, but ate
for a longer period and thus consumed more food. Although in-
creased meal duration in groups was generally found, on at least
one occasion (de Castro, 1994), it did not occur: people eating with
family members (including spouses) ate larger meals but mostly
because they increased their eating rate rather than the duration
of the meal. (de Castro, 1995 claims that “the presence of other
people increases the duration of meals and not the rate of intake,”
which is not the only time that he overgeneralizes.) In the pres-
ence of friends, people ate larger meals mostly because they
increased the duration of the meal without increasing their eating
rate; in fact, the overall rate was lower when eating with friends
(de Castro, 1994). In the presence of eating companions other than
family or friends, the social correlation (i.e., more eating with more
companions) was not observed. In short, the eater’s relationship to
the eating companion makes a difference. de Castro does not wade
into the complexities of “mixed” groups (e.g., groups comprised of
some family, some friends, and some strangers).
Bellisle, Dalix, and de Castro (1999) uncovered a sex difference
in the social correlation. A diary study employing a fresh sample
of French university students found that the social correlation was
stronger for males (~.4) than for females (~.2).
In a brief summary of the diary studies, de Castro (1995) claims
that the social correlation is obtained “for low, moderate and high
restrained males and females.” It is not clear where this conten-
tion comes from – there was no mention of dietary restraint except
in de Castro (1990), in which he claims that active dieters were ex-
cluded from the analysis. It is interesting to speculate about whether
a dieter might be “socially facilitated” by a group that eats a lot and/
or whether social facilitation would occur if the group were
composed exclusively of dieters. Do groups of dieters suppress each
other’s intake or do they (sometimes) conspire to overeat? (“I’ll break
my diet if you break yours.”) Whatever limits there may be on the
social facilitation effect, de Castro (1995, p. 260) maintains that “of
all the myriad of stimuli that affect the ad libitum intake of humans,
social facilitation is the most powerful we have yet discovered.”
Other diary studies
Feunekes, de Graaf, and van Staveren (1995b) exported the de
Castro diary procedure to The Netherlands. (Interestingly, Feunekes
et al. found that meals with strangers comprised less than 1% of all
meals.) Overall, there was a .22–.27 correlation (in the two repli-
cations) between number of others present and amount eaten. (No
separate calculation was made excluding the case when no others
were present, so it may be that the presence of any others, irre-
spective of how many others there were, was enough to trigger social
facilitation, although de Castro & de Castro, 1989 already dealt with
that potential artifact.) Meal duration was positively related to meal
size. Indeed, a path analysis led to the conclusion that the in-
creased meal size in groups is mediated by the increased duration
63
of group meals (rather than by, say, the more positive atmosphere
of group meals, which does contribute to increased meal dura-
tion). Our concern here is that time (duration) is not a useful
explanation of behavior; we need to know what happens during
the (extended) meal time that causes people to eat more.
de Castro, Bellisle, Feunekes, Dalix, and de Graaf (1997) com-
pared diary records of Dutch, French, and American samples. The
Dutch ate more (and more often); and indeed, it turns out that they
were more likely to eat with others. The correlation between amount
eaten and number of eating companions, however, did not differ by
nationality (.2 < rs < .3).
Patel and Schlundt (2001) had obese women keep diaries for two
weeks, recording among other things the presence/absence of others
during eating occasions. Diarists ate much more with others than
when alone, between 24% and 33% more depending on the dia-
rists’ mood. “The impact of the social context (effect size around 50%)
is large compared to the impact of moods (effect size around 9%)”
(Patel & Schlundt, 2001, p. 116), although the impact of the social
context factor was dramatically reduced when the tendency to binge
eat was covaried out, suggesting that frequent bingers accounted
for most of the effect (even though true bulimics tend to do their
bingeing when alone rather than in company (see Herman & Polivy,
1996, for a review)). Also, “the percentages of calories from fat and
protein were greater, whereas the percentage of calories from car-
bohydrate was less in meals eaten in a social setting compared to
meals eaten alone” (p. 116). The authors speculate that “eating high
fat dessert foods like cakes may be more common with meals eaten
in a social setting” (p. 116).
Observational studies
Observational studies are studies in which people eat either alone
or in groups at their own discretion (rather than being assigned to
one condition or the other). The difference between observational
studies and diary studies is that in observational studies, the data
are collected by observers rather than diarized by the eaters them-
selves. The first study of this type was by Krantz (1979), who found
that people selected more food in a cafeteria setting when they were
accompanied by eating companions than when they ate alone. There
are three aspects of this basic finding that merit comment. First, the
social facilitation effect was a matter of initial food selection rather
than intake per se (which was not measured). Krantz assumes a cor-
respondence between selection and intake but we cannot be sure.
Secondly, the observed positive association between amount of food
selected and the presence of eating companions obtained for normal-
weight diners but not for obese diners, who selected less food when
eating with others. Finally, there being no clear prior examples in
the literature of the social facilitation of eating, its appearance among
normal-weight diners was unpredicted and came as a surprise for
Krantz.
Klesges, Bartsch, Norwood, Kautzman, and Haugrud (1984) found
that patrons at formal and fast-food restaurants ate more in groups
than when eating alone. Contrary to the findings of de Castro and
Brewer (1991), Klesges et al. (1984) found no difference in the
amount eaten in large (n > 3) versus small (n = 2 or 3) groups. In a
cafeteria setting, Young, Mizzau, Mai, Sirisegaram, and Wilson (2009),
likewise found no significant effects of groups size on the amount
of food selected, although there was a hint that for females, meal
size increased as a function of the number of female dining com-
panions and decreased as a function of the number of male dining
companions; strikingly, eating with others did not increase meal
size compared to eating alone.
Wansink and Park (2001) measured popcorn intake at a movie
theater as a function of (manipulated) package size and (mea-
sured) liking of the taste of the popcorn. Other measured variables
included the presence of others, which significantly increased
64 C.P. Herman/Appetite 86 (2015) 61–73
popcorn intake. (The authors do not comment on the fact that even
if one goes to the cinema alone, one is almost certainly in the pres-
ence of many others, raising questions about what we mean by
“eating in a group.”) The question of what constitutes a group was
further highlighted in an observational study of soldiers by Hirsch
and Kramer (1993). These researchers found that people ate more,
over a 2-week period, as a function of the proportion of meals con-
sumed socially. Interestingly, non-social eating was defined as “eating
alone or as part of an undifferentiated large group of 50 to 70 people.”
In short, eating in a crowd (as opposed to eating in a well-defined
group) does not facilitate intake. This study raises (but does not
answer) fascinating questions about what exactly constitutes the
sort of group that facilitates eating.
Experimental studies
Experimental studies of the effects of group eating preceded the
first de Castro diary studies, but we will begin our survey with the
lone de Castro diary study that involved an experimental manip-
ulation. Redd and de Castro (1992) deliberately manipulated the
social context of eating, in an independent sample of diarists, in an
explicit attempt to establish the causal impact of the presence of
other people on food intake. Participants were requested, in a within-
subjects design, to eat as often as possible (a) with others, (b) alone,
or (c) as they normally would. Food intake was significantly higher
in the “others” and “normal” groups. (Two-thirds of the “normal”
meals were eaten with others present,2 and intake was higher in
the “with others” meals than in the “alone” meals in the “normal”
condition.) This study, by experimentally manipulating the social
conditions of eating, represents an interpretive advance over the
“uncontrolled” diary studies conducted by de Castro and his col-
leagues, allowing for a clearer assessment of the causal role of the
presence of others in facilitating intake.
The first true experimental (laboratory) study of social facilita-
tion was conducted by Berry, Beatty, and Klesges (1985). Participants
were assigned to eat either alone or in groups of 3 or 4. Group intake
was much (41%) higher than was solo intake. Edelman, Engell,
Bronstein, and Hirsch (1986, Study 2) found that group diners ate
44% more than did solo diners. (The effect was stronger for over-
weight [63%] than for normal-weight diners [23%], but the group
condition by weight category interaction was not significant.) Shide
and Rolls (1991, cited in Clendenen, Herman, & Polivy, 1994) found
social facilitation of intake, but only among male friends, and prin-
cipally for dessert. Another study that failed to find conclusive (or
any) evidence of social facilitation was conducted by Bellisle and
Dalix (2001): groups of four did not eat significantly more than did
solo diners. In extenuation, the groups were composed of strang-
ers, who are less likely to display social facilitation according to de
Castro.
Clendenen et al. (1994) examined social facilitation by assign-
ing participants to one of three group size conditions in the lab: solo,
pairs, and groups of four. A large social facilitation effect emerged:
people eating with others ate 90% more than did solo diners. Con-
trary to what one would expect based on the de Castro and Brewer
power function, however, people eating in groups of four did not
eat more (703 calories) than did people eating in pairs (721 calo-
ries). Clendenen et al. (1994) conducted separate analyses on the
main course (deli foods) and dessert (cookies) and found that the
social facilitation effect observed for the entire meal was carried by
the deli foods, with no effect for cookies alone. A further analysis
examined the distinction between eating with friends versus strang-
ers (omitting solo diners). People ate considerably (44%) more with
2
It is perhaps worth noting that a recent survey (NPD Group, 2014) found that
Americans reported eating 57% of their meals alone.
friends than with strangers, irrespective of group size. This effect
was carried by cookie intake, with no friend–stranger effect for deli
food. The only hint of a “social impact” effect was among friends
eating cookies, where groups of 4 ate marginally (p < .07) more than
did pairs. Analysis of the duration of lab meals indicated that pairs
took longer to eat (1715 s) than did groups of four (1488 s), who
in turn took almost twice as long as did solo diners (767 s). Friends
did not take longer to eat than did strangers (despite the fact that
friends ate more, specifically more cookies). Restrained eaters (i.e.,
dieters) felt more comfortable eating with friends than with strang-
ers, whereas unrestrained eaters were equally comfortable in either
situation; but restrained and unrestrained eaters did not differ in
their food intake.
Kim and Kissileff (1996) had participants eat a lunch of maca-
roni and beef either alone or with a friend in a within-subjects design
including an orthogonal manipulation of glucose preload. Lunch
intake was 10% higher in the “pair” condition, and the increase in
the social condition was greater for males (16.3%) than for females
(4.6%).
Hetherington, Anderson, Norton, and Newson (2006) assigned
participants to eat alone, with two friends, with two strangers, or
alone while watching television, in a within-subjects design. Par-
ticipants in the “friends” and “TV” conditions ate the most, 18% and
14% more than when dining alone. The “strangers” condition was
intermediate, not significantly different from any of the other con-
ditions. The large increase in intake in the “friends” condition was
confined to cake (one component of the available meal), corrobo-
rating the emerging trend of social facilitation of intake among
friends being mostly a matter of dessert intake. (Hetherington et al.
note that it is rich, sweet, high-fat foods that are overeaten. That
such foods would be overeaten if they occurred earlier in the meal
rather than at dessert time is implied, but it is certainly possible
that it is when the food is eaten rather than its composition that
matters, although of course these two aspects of food are almost
always confounded.) Meal duration was longer when eating
with others, but there was no difference between “friends” and
“strangers.”
Salvy, Jarrin, Paluch, Irfan, and Pliner (2007) did not include solo
eaters in their study, but they did find that snack intake was higher
in pairs of friends than in pairs of strangers, especially among males
(see also Salvy, Howard, Read, & Mele, 2009). Similarly, Paquet et al.
(2008) found that intake was predicted by how well dining com-
panions got along with each other (communality). Koh and Pliner’s
(2009) friend pairs served themselves more food initially (and ate
more) than did stranger pairs, and the friend pairs were more likely
to take second helpings.
Some studies have examined social facilitation in young chil-
dren. Lumeng and Hillman (2007) assigned young children (2.5–
6.5 years of age) to eat their snack at school in groups of three or
nine. Groups of nine ate more, although the difference was signif-
icant only when snack duration was above the median duration. This
effect was associated with a shorter latency to start eating and a
faster eating rate. Kral, Kabay, Roe, and Rolls (2010) found no dif-
ference in intake as function of group size (two to four people) among
5–6 year olds.
Pliner, Bell, Hirsch, and Kinchla (2006) pointed out that most anal-
yses of the social facilitation effect involve two components: (a)
people in groups eat for a longer period of time and (b) people in
groups eat more. The duration of the meal and the amount eaten
are usually both left uncontrolled. Pliner et al., however, con-
trolled meal duration, assigning different-sized groups (solo diners,
pairs, groups of four) to eat for either 12 minutes or 36 minutes.
When group size did not dictate meal duration, group size did not
dictate amount eaten. That is, Pliner et al. found that meal dura-
tion had a strong effect on intake, with greater intake in the longer
meal (for the cookie component of the meal but not for pizza
 C.P. Herman/Appetite 86 (2015) 61–73
component of the meal); but group size (and intra-group familiar-
ity) did not affect intake, presumably because group size did not (and
could not) affect meal duration. Pliner et al. conclude that the reason
that groups tend to eat more than do solo diners is that groups tend
to eat longer meals and are thus exposed to palatable food cues for
a longer time. (This interpretation coincides with the conclusion that
Feunekes et al., (1995a) drew from their path analysis.) Pliner et al.
offer an explanation for Bellisle and Dalix’s (2001) failure to find a
social facilitation effect: it appears that in all conditions diners were
required to eat for at least 30 minutes, which requirement would
attenuate any effect of group size on meal duration. The social fa-
cilitation effect, then, is actually an exposure-to-food-cues effect.
When diners are allowed to “select” their own meal duration and
food intake, larger groups will “select” higher values for both, and
the greater intake reflects greater cue exposure during longer meals.
Brindal, Wilson, Mohr, and Wittert (2011), who measured meal
duration and intake in lone diners in a fast-food restaurant, found
no evidence that longer meals were larger meals. This study may
be less germane to the issue at hand than it could be, inasmuch as
the longer meals among lone diners tended to be meals in which
the diners read; it seems likely that reading directly reduces ex-
posure to whatever food cues might be in the immediate
environment (whereas talking with others in a group situation does
not interfere as much with food-cue exposure). Also, in virtually
every case, the lone diner decided in advance how much to order,
with the result that meal size was dictated by pre-meal decision-
making rather than by the circumstances of the eating episode.
When impression management trumps social facilitation
As we noted above, when a passive audience watches an indi-
vidual eating, the individual typically eats much less. Clearly, eating
in the presence of others does not necessarily facilitate intake. In
the co-action paradigm, in which the others are not merely watch-
ing but are also eating, social facilitation is more likely, but it is not
inevitable. It is assumed that in the passive audience or non-
eating observer situation, the eater becomes highly self-conscious,
and such self-consciousness may suppress behaviors, like eating,
that might be interpreted negatively (see Vartanian, Herman, &
Polivy, 2007, for an examination of how people who eat a lot are
judged negatively). It appears that self-conscious suppression of food
intake also occasionally arises in the co-action paradigm, when eaters
are concerned that their intake may be construed as excessive.
Maykovich (1978), for instance, suggests that obese people may eat
less when eating in a group including normal-weight people than
when eating alone or eating in a group of other obese people,
because the presence of (and/or observation by) normal-weight
people makes the obese more self-conscious. Maykovich found that
among regular customers at a smorgasbord restaurant, obese people
were less likely (13%) to visit with normal-weight companions than
with other heavy companions (43%) or alone (44%), presumably
because these obese diners did not want their eating to be inhib-
ited by the presence of normal-weight companions. (Of course, it
is possible that these percentages simply reflect friendship pat-
terns, with obese people more likely to associate with other obese
people; but one may fairly ask why obese people prefer the company
of other obese people, and part of the answer may be that being
with other obese people creates less aversive self-consciousness than
does being with normal-weight people.) “In order to avoid unpleas-
ant contact, the obese may huddle together. On the positive side
they may feel more at ease when they are engaged in eating”
(Maykovich, 1978, p. 459). Examination of amount eaten, however,
indicates that once obese people choose their dining companions,
they eat heartily regardless of whether the companions are over-
weight or not. Maykovich suggests that obese diners learn which
65
people are nonjudgmental and prefer their company, even if they
happen (in rare cases) to be of normal weight. Among occasional
customers at the restaurant, obese diners eating with normal-
weight companions ate much less than did obese diners eating alone
and less than did their normal-weight companions. (Whether regular
diners were more familiar with their dining companions than were
occasional diners remains unclear.)
Krantz (1979, p. 20), using similar logic, predicted “that over-
weight subjects who were accompanied while eating would be self-
conscious about eating too much and would tend to alter (suppress)
food purchases, choosing less food than when eating alone.” Indeed,
overweight individuals selected less food when eating with others
– no breakdown was provided of whether the others were over-
weight or not – than when eating alone, whereas normal-weight
individuals selected more food when eating with others than when
eating alone. Krantz attributes the “social suppression” of food se-
lection among the overweight to self-consciousness. More or less
the same pattern was found by Salvy, Coelho, Kieffer, and Epstein
(2007), who studied 6- to 10-year-old children. The overweight chil-
dren ate much (31%) more when alone than when in a group of two
overweight and two normal-weight children. Normal-weight chil-
dren ate much (64%) more when eating in a group than when alone.
In other words, normal-weight adults and children show the typical
social facilitation effect whereas overweight adults and children eat
less in (most) groups than when alone, presumably because over-
weight people are concerned about the negative inferences about
them that might arise from overeating in public. It is possible that
overweight people might not eat less (and might even eat more)
in groups if the groups were composed entirely of overweight in-
dividuals, who presumably would be seen as more lenient in how
they would interpret hearty eating among other overweight people.
Indeed, Salvy et al. (2009) found that overweight youths ate more
when paired with other overweight youths than when paired with
normal-weight youths, whereas normal-weight youths ate more
when paired with other normal-weight youths than when paired
with overweight youths.
Insofar as social facilitation effects emerge more clearly among
friends dining together than among strangers, the difference might
likewise be ascribed to impression management. As Tice, Butler,
Muraven, and Stillwell (1995) predicted and found, people are more
eager to make a positive impression on strangers than on friends;
and insofar as eating a lot interferes with making a positive im-
pression (see Vartanian et al., 2007 for a review), we should expect
people to be more likely to display socially facilitated eating among
friends than among strangers. Hetherington et al. (2006, p. 504) con-
clude flatly that “eating with strangers fails to produce a social
facilitation effect on eating,” and Salvy and Pliner (2010, p. 619) con-
clude that “it seems likely that the absence of a social facilitation
effect when individuals are eating with strangers is. . .related to im-
pression management” (see also Bellisle & Dalix, 2001; Koh & Pliner,
2009; Salvy et al., 2009; Salvy et al., 2007). Mekhmoukh, Chapelot,
and Bellisle (2012) did not observe social facilitation in a sample
of teenage boys who were “familiar” with each other but not per-
sonal friends. Péneau et al. (2009) found that eating in groups of
three actually suppressed intake (“social inhibition” rather than social
facilitation) among teenage boys and girls who were character-
ized as “individuals of the same age living in the same
neighbourhood, potentially classmates,” which would appear to make
them more like strangers than like friends. Bellisle, Dalix, Airinei,
Hercberg, and Peneau (2009) likewise found a social inhibition effect
among unacquainted women eating in groups of three. Insofar as
females are particularly concerned about impression manage-
ment in the domain of eating, perhaps we should expect to find more
evidence of social suppression among female strangers and evi-
dence of social facilitation only among females who are very
comfortable with each other.
66 C.P. Herman/Appetite 86 (2015) 61–73
Mechanisms underlying the social facilitation of eating
Prior to the de Castro studies, little effort was devoted to ex-
plaining the (few) observations of social facilitation of eating. Klesges
et al. (1984, p. 39) speculated that “perhaps eating with others pro-
vides more encouragement, support, and modeling for eating than
eating alone does. Similarly, eating alone may not be considered as
pleasant and, as a result, people may eat less.” “Eating alone is likely
to be more functional than enjoyable” (Hetherington et al., 2006,
p. 504). These proposed explanations for the effect lack specificity,
but we will pursue below the possibility that group eating is more
pleasant and therefore may lead to longer, larger meals.
In his diary studies, de Castro not only established the reality
of the social facilitation of eating, and its apparent ubiquity, but he
also attempted to evaluate possible reasons for the effect. After all,
it is not obvious why you should eat more when you eat in a group
than when you eat alone, even if it is obvious that you do so. de
Castro offered several candidate explanations for why people might
eat more when eating with others. Of course, the observational/
correlational nature of the diary data is not ideally suited for testing
explanations, but de Castro did what he could.3
Hunger/deprivation
de Castro and de Castro (1989) found that the effect of the
number of people eating together was independent of (i.e.,
uncorrelated with), among other things, pre-meal self-reported
hunger and deprivation time. In other words, it cannot be argued
that people are more likely to eat with others when they are par-
ticularly hungry, or that people gravitate toward eating in groups
as their hunger increases. Although people do eat more when they
are hungrier (according to self-reports of pre-meal hunger), “the
amount ingested with other people present is larger than can be
accounted for on the basis of prior deprivation” (p. 244). de Castro
and de Castro likewise argue against the notion that people tend
to schedule large meals with other people when they know that they
will be hungry. In other words, de Castro confirmed, as best he could,
that it is eating with others that makes you eat more rather than
your desire to eat more that leads you to eat with others.
Another possibility is that eating in the presence of others makes
you hungry (even if you were not especially hungry before the meal).
The physiological basis of this hypothesis is not easy to compre-
hend, but in any case, de Castro (1990) rejects it on the grounds that
if people were hungrier when eating in groups, they would eat more
voraciously and more quickly, which is not the case.
One possibility not considered by de Castro is that people may
want to eat a lot even though they are not especially deprived and/
or hungry, and may choose to eat in company as a way of (or as an
excuse for) eating a lot. One might imagine that dieters in partic-
ular might be looking for excuses to “overeat”; although by definition
dieters are committed to eat minimally, they are nevertheless prone
to occasional overeating and may indeed relish the opportunity to
stray from their diets (see, e.g., Herman & Polivy, 2003). As was noted
above, whether dieters were included in the diary samples is not
exactly clear. de Castro (1990) says that active dieters were ex-
cluded from the sample, but de Castro (1995) claims that the social
3
At one point, de Castro (1997b; see also de Castro, 2002) used the diary data to
establish relatively high heritabilities for the social facilitation of eating (i.e., the
number of people with whom one tends to eat is heritable, as is the slope relating
intake to the number of people present at the meal). Of course, such behavior genetic
analyses do not provide much of an explanation for the social facilitation of eating
(or any other behavior), other than to suggest that the explanation lies in unspeci-
fied genes. Moreover, heritability quotients would appear to be regularly overstated
– what is not highly heritable? – probably because of faulty assumptions underly-
ing calculation of the h2 coefficient (Winerman, 2004).
facilitation effect applies to all people regardless of their degree of
dietary restraint. (It is possible, in principle, that one could be a re-
strained eater and yet not an active dieter at the time of the study.)
In any case, de Castro and de Castro (1989), while (possibly) ruling
out the notion that people are more likely to seek eating company
when they are hungry/deprived, do not provide a positive expla-
nation of the social facilitation effect.
Arousal and emotionality
Another possible explanation for increased intake in the group
situation is arousal. Zajonc (1965), in his original analysis of social
facilitation effects, argued that the presence of others is arousing
and that arousal has the effect of increasing the probability of emis-
sion of responses that are initially dominant in the response
hierarchy. If people’s dominant (i.e., most likely) response in the pres-
ence of palatable food is to eat it, then making people (more) aroused
will have the effect of increasing even further the dominance/
likelihood of the eating response. (We have argued (Herman et al.,
2003, p. 874) that “in the presence of palatable food, and in the
absence of inhibitory forces (such as satiety), people continue to
eat indefinitely. People behave as if access to palatable food is in
and of itself a sufficient reason to eat.” Our formulation does not
address the potentiating role of arousal, but it concurs with the
premise that in the presence of palatable food people are likely to
eat it and do not necessarily need any additional incentive, such as
hunger.) de Castro (1990, p. 1129) suggests that “the presence of
other people might induce an aroused state that leads to greater
consumption.” Again, it is worth considering the special case of
dieters, for whom the dominant response in the presence of food
might not be to eat it, although arguably dieters, like everyone else,
are motivated to eat palatable food even while at the same time they
are motivated – sometimes even more motivated – to abjure pal-
atable food (see Stroebe, 2008, for a version of this view of dieters
as fundamentally conflicted).
de Castro (1990, 1994) rejects the arousal explanation of the social
facilitation effect because the presence of eating companions in-
creases intake by increasing the duration of the meal and not the
rate of ingestion. de Castro assumes that increased arousal will have
its primary impact on rate, but he does not provide direct support
for the assumption that aroused eaters will eat more enthusiasti-
cally or voraciously. Lumeng and Hillman (2007), in their study of
young children, conclude by endorsing an arousal interpretation.
Children in larger groups ate more rapidly and with a reduced latency
to begin eating, but did not take longer to eat (thus not support-
ing de Castro’s favored time-extension hypothesis). Perhaps young
children represent an intermediate state between nonhuman animals
and fully socialized adult humans. Older children (Salvy et al., 2007)
appear to have achieved the degree of socialization necessary to
display the same pattern as do adults.
Another, related possibility is that the presence of eating com-
panions induces an emotional state that facilitates intake. de Castro
(1990) includes self-reports of anxiety and elation in his analyses
but concludes that neither emotion is responsible for the social fa-
cilitation effect. “Self-rated anxiety was not found to be associated
with the number of people present” (p. 1129), and although “self-
rated elation was positively correlated with the presence of
others. . .multiple regression analyses suggested that the presence
of other people facilitates intake and increases elation indepen-
dently” and “social facilitation. . .operates independently of the
subjective state of the individual” (p. 1129). In other words, people
are generally happier when eating with others than when eating
alone, but it is not their happiness that makes them eat more in
groups. The “social” effect and the “elation” have statistically inde-
pendent effects on intake, and the effect of elation is much smaller
than the social effect.
 C.P. Herman/Appetite 86 (2015) 61–73
Disinhibition
de Castro entertains some other possible mechanisms to explain
the social facilitation of eating. Eating with others “could operate
by calling an individual’s attention to food by observing a compa-
nion’s activities. It could produce a disinhibition of eating. Observing
someone else eating may remove constraints on eating that oth-
erwise would limit the amount ingested” (de Castro, 1990, p. 1134).
The implication here is that seeing one’s eating companions eat a
lot will make one eat a lot as well, or at least allow one to eat a lot.
This “modeling” explanation of social facilitation, suggested by Berry
et al. (1985) – see Cruwys, Bevelander, and Hermans in this issue
for a general discussion of modeling effects – does not make a lot
of intuitive sense. If group member B is eating more than group
member A, then it follows that A is eating less than B; so why does
B not respond by eating less (rather than A responding by eating
more)? Is there reason to think that “overeaters” have a greater social
influence on their eating companions than do “undereaters”? (On
the contrary, a meta-analysis of modeling studies (Vartanian, Spanos,
Herman, & Polivy, in preparation) finds that “inhibiting” models are
generally more powerful than are “augmenting” models. Still, most
modeling studies do not involve friends eating together, so it remains
possible that friends may push each other to eat more even if strang-
ers push in the opposite direction.)
One unexplored possibility is that people tend to overestimate
how much their eating companions are eating. If that were the case,
then even if A and B were eating the same amount, A would think
that B is eating more (and might be motivated to match B’s “greater”
intake) and B would think that A is eating more (and be moti-
vated to match A’s “greater” intake). This scenario could conceivably
result in an “arms race” (or “intake race”) in which everyone would
end up eating more. This explanation would be hard-pressed to
account for the social facilitation of intake in nonhuman animals,
but perhaps humans display social facilitation of eating for differ-
ent reasons than other animals do. Some data bearing on this
hypothesis may be extracted from manipulation checks in model-
ing studies (in humans); participants are often asked to estimate
how much their eating companions (who are actually experimen-
tal confederates) have eaten. Is there evidence that people
overestimate their eating companions’ intake? Meanwhile, Pliner
et al. (2006) found that people eating in pairs ate more similar
amounts than did statistically “paired” solo diners. People eating
in groups of four show no more intra-group similarity than did sta-
tistically paired solo diners. Perhaps it is easier to model one other
person than three other people. How do you match your intake to
the intake of more than one other person? In this regard, it is in-
structive to note that Leone, Pliner, and Herman (2007) found that
people tended to eat a great deal when they were confronted with
evidence that other people in a similar situation ate discrepant
amounts (i.e., some people ate only a little, whereas others ate much
more); Leone et al. (2007) concluded that the absence of a clear social
norm “disinhibits” intake, with the disinhibited eaters eating more
than anyone else did. Is it possible that as the size of the group of
co-eaters increases, the intake variability (or at least the per-
ceived intake variability) of the other co-eaters increases, releasing
the individual eater from inhibitory norms? Cavazza, Graziani, and
Guidetti (2011, p. 286) refer to “the permissive guideline provided
by a shared ‘social meal’ script.”
Yet another possibility is that people are “disinhibited” by the
distracting presence of others and fail to attend to how much they
themselves have eaten, with the result that they end up eating
more than they otherwise would have. (This is the logic that Bellisle
& Dalix, 2001 and Hetherington et al., 2006 used in including a
group eating condition in their research, which they expected to
parallel a more explicit distraction condition, such as watching TV.)
Our own view (Herman et al., 2003) is that people will eat as much
67
palatable food as they can, unless or until some inhibitory factor
is applied. Our view may be reconciled with the distraction view
insofar as we interpret distraction as interfering with a customary
inhibitor of intake (e.g., a growing awareness that one has eaten
enough, or too much); in other words, distraction removes a po-
tential inhibitor and allows people to indulge in their penchant to
partake freely of palatable food. Hetherington et al. (2006) found
that eating with strangers was not as effective at producing social
facilitation as was eating with friends, despite equivalent amounts
of distraction. “Distraction is not a sufficient condition for social fa-
cilitation effects to occur” (p. 504). We would suggest that although
the presence of strangers might be distracting in some respects, it
does not overcome the self-consciousness that strangers induce.
Is there evidence that people eating in group situations are less
accurate in monitoring their own intake? And further, is there reason
to believe that people who lose track of how much they are eating
will eat more than they intended? That possibility would seem to
presuppose that people intend to eat only so much, and that their
intentions are undermined by distraction. There is no direct evi-
dence that people in general have clear intentions about how much
they are going to (allow themselves to) eat. Perhaps dieters might
be expected to have specific calorically-restrictive intentions, but
do most people? de Castro (1994) himself retains some enthusi-
asm for his version of the disinhibition explanation – with people
eating more with family and friends because they are more relaxed,
which in de Castro’s view is the same as being more disinhibited
– despite the absence of strong evidence for it.
Time extension
de Castro tends to favor one particular explanation for the social
facilitation effect – time extension. The “time extension” explana-
tion amounts to the contention that when eating in groups, people
eat for a longer period of time – de Castro (1990, 1995) consis-
tently argues that group eating results in longer meals – and
therefore eat more. Bell and Pliner (2003) observed diners in three
types of restaurants and found that meal duration varied as a direct
function of the size of the dining group (at least up to n = 5), and
that this association between group size and duration obtained re-
gardless of whether solo diners were included in the analysis or not.
The association was weaker in a fast-food restaurant than in a
worksite cafeteria or a moderately priced restaurant, but it was sig-
nificant in all three venues. (Bell and Pliner did not measure intake.)
“The presence of other people would appear to extend the du-
ration of eating and increase intake” (de Castro, 1990, p. 1134).
“People appear to eat more when other people are present because,
in social meals, people tend to eat for a longer period of time” (Redd
& de Castro, 1992, 749). In other words, more people means more
socializing, which leads to a longer meal, which leads to greater
intake. “These results suggest that the presence of other people at
a meal is associated with people lingering longer over their food
and, even though overall eating slightly slower, eating more than
when alone” (de Castro & Brewer, 1991, p. 124). Compare this for-
mulation to Edelman et al.’s (1986, p. 81) “In the social condition,
subjects talked with each other and lingered at the table.” de Castro,
however, (usually) recognizes that the correlational nature of his
data does not allow unqualified causal conclusions. “It is not clear
from the present analyses whether social factors operate to extend
the duration of intake and this in turn causes the increment in con-
sumption or whether social factors increase intake that increases
the time required to consume the meal” (de Castro, 1990, p. 1134).
(The same qualification applies to Edelman et al.’s study and to any
study where the duration of the meal and the amount eaten are both
left up to the participants. Often it is the social composition of the
group that is manipulated, with both intake and meal duration
68 C.P. Herman/Appetite 86 (2015) 61–73
uncontrolled.) For the most part, de Castro favors the “increased so-
cialization leads to longer meals and increased intake” version.
Implicit in de Castro’s formulation is that people enjoy eating
in the company of others and therefore are in no hurry to termi-
nate the meal; indeed, people will linger in pleasant company if they
can, eating more as a consequence. This generalization, however,
does not mean that longer meals are always pleasanter meals.
Hetherington et al. (2006) found that people enjoyed eating with
friends and spent more time when eating with friends; but Pliner
et al. (2006), who (unlike other researchers) deliberately manipu-
lated meal duration, found that people were much happier and liked
their eating companions more in the 12-minute-meal condition than
in the 36-minute-meal condition. In other words, being forced to
spend extended time with others – in this case, strangers – does
not produce pleasure; rather, it is enjoying the company of others
that produces longer meals.
There are at least two problems with the time-extension hy-
pothesis. First, it is somewhat surprising that de Castro promotes
the time-extension hypothesis so strongly when he himself found
(1994) that when the group consisted of family (including spouses),
greater intake was a result of a greater speed of eating rather than
a longer meal. Clendenen et al. (1994) likewise found that al-
though friends eating together ate more than did strangers eating
together, the friends’ meals were not of longer duration. In fair-
ness, Clendenen et al. did find that groups who ate the most (viz.
pairs) took longer to eat than did groups who ate less (viz. groups
of four).
Secondly, the notion that longer meals will necessarily in-
crease overall intake is not as obvious as de Castro suggests. Certainly
it is possible, in principle, that a longer meal is simply a meal that
includes longer pauses between bouts of eating that do not vary
in number or size. It may well be the case that eating with friends
is a more convivial experience,4 and that there is probably more con-
versation; but perhaps the longer duration of the meal simply reflects
a greater proportion of time spent conversing, without any in-
crease in the amount of time spent actually eating. Recall the Brindal
et al. (2011) study described above, in which solo diners who were
reading had longer meal durations but did not eat more.
Hetherington et al. (2006) found that people eating with others spent
up to 40% of their time talking (rather than eating); nevertheless,
despite this “wasted” potential eating time, these social diners did
manage to eat considerably more, thanks in part to the extended
duration of their meals. Sommer and Steele (1997) found that solo
diners were more likely to read, and that reading while eating ex-
tended the duration of the meal; but group diners’ meals were of
greater duration than solo diners’ meals, even though the solo diners
were more likely to read. (Sommer and Steele measured only the
duration of the meal, not intake.)
4
Harlow (1932) claims that “a good meal tastes better if we eat it in the company
of friends” but he does not specify whether we simply enjoy the experience more
(as de Castro, 1990, found) or whether the food literally tastes better. Bellisle and
Dalix (2001) found that people rated the meal as more palatable when they ate in
groups than when they ate alone (although the groups were composed of strang-
ers, who did not eat more than when eating alone). Boothby, Clark, and Bargh (2014)
found that a co-eater made people rate good-tasting chocolate even better (and bad-
tasting chocolate even worse); there was no measure of intake. Paquet et al. (2008)
did not find any effect of communality on the perceived taste of the food; neither
did Kim and Kissileff (1996). Bellisle et al. (1999, p. 51) found that “meals where
only one female was present were rated higher in palatability than meals con-
sumed in the presence of many women,” although (a) this finding did not apply to
males and (b) it may well have been the case that the one-woman meals were quali-
tatively different from the many-women meals. Might it be possible that eating in
company somehow disrupts the phenomenon of sensory-specific satiety and thereby
enhances the palatability of the food? Berry et al. (1985) found no difference in the
onset of sensory-specific satiety as a function of whether people ate alone or in groups.
Another, related question about the time extension model is how
exactly the longer meal produces increased intake. de Castro is not
terribly clear about what is happening during the extended meal.
Is it the case that spending a longer time at the table necessarily
involves a greater exposure to palatable food cues, and that it is such
exposure that drives food intake? (Bell and Pliner, 2003, explicitly
endorse this possibility.) Or is it the case that people simply eat at
a constant rate and therefore the more time they spend at the table
the more they will eat? “The time extension notion would predict
that the rate of intake would be the same regardless of the social
conditions but the duration of the meal would be extended when
other people were present. This should produce a higher than normal
level of stomach filling and greater satiety at the end of the meal”
(de Castro, 1990, p. 1130). It does appear to be the case that group
diners in the de Castro diary studies are more satiated after the meal.
It is not clear whether the meal ends because the food is all gone
or because the conversation has come to a natural end.
Schachter (1971), in his theory of obesity based on exposure to
external food cues, was never able to explain why fat people stopped
eating if there was still food available. de Castro likewise appears
to assume that group diners eat until all the food is gone, whereas
lone diners presumably stop well short of consuming all the food
that is available to them. We next turn to the issue of whether or
not eaters (a) leave some food uneaten, or (b) simply eat all the food
on the table and then stop, or (c) finish the food on the table and
then acquire and eat more.
Where does the additional food come from?
One of the intriguing aspects of the social facilitation literature
– an aspect that has received virtually no attention, despite our
raising it more than decade ago (Herman et al., 2003) – is the ques-
tion of where the extra food comes from. The social facilitation of
eating, by definition, entails people eating more food than they do
when eating alone. This basic finding, though, does not address the
question of the extent to which people eat all or most of what they
have been served (or of what they serve themselves) or whether
they leave some portion of their meal uneaten. The unspoken as-
sumption in discussions of social facilitation is that people leave more
(uneaten) food on their plates (or on the table) in the “alone” sit-
uation than in the “group” situation. Before we accept this
assumption, however, we must consider how much food is avail-
able to the eater in either situation. In experimental studies of food
intake, it is often – indeed, almost invariably – the case that the re-
searchers provide participants with more food than they could
possibly (or at least plausibly) eat. In such cases, we can easily
imagine that all eaters will leave some food on their plates (or in
the available serving dishes); and it then becomes a matter of
whether they leave more in the “alone” condition than in the “group”
condition. In effect, the experimenter has provided so much food
that socially facilitated eaters have no problem obtaining as much
more food as they want. In the diary studies, however, and in some
observational studies (not to mention real life), the available food
is (presumably) neither free nor unlimited, and we have to ask our-
selves how much food there is available to the eater. If “group” eaters
eat more than “alone” eaters, is it because “alone” eaters leave more
on their plates or on the table? Perhaps, but we must pursue the
matter and ask why would “alone” eaters take or buy more food
than they were going to eat? For one thing, it is wasteful. Of course,
perhaps the “alone” eaters pack up the uneaten food for a future
meal, but we have no evidence of that happening. Consider a typical
breakfast. If, as de Castro has demonstrated (de Castro et al., 1991),
“alone” eaters eat less for breakfast than do “group” eaters, is it likely
that both types of eaters start with a breakfast of the same size but
only the “group” eaters finish the available portion? Do “alone” eaters
pack up the uneaten remainder of the breakfast for a later date?
 C.P. Herman/Appetite 86 (2015) 61–73
More likely, it seems, is that the “alone” eaters serve themselves
(or order for themselves) a smaller breakfast than the “group” eaters.
Brunstrom, 2011, Brunstrom and Rogers, 2009, and Brunstrom and
Shakeshaft, 2009 have argued (and demonstrated) that people, in
serving themselves, judge how much food they are likely to eat (or
are likely to require in order to become comfortably satiated) and
provide themselves with that approximate amount, especially when
there are constraints (cost, wastage) involved. In other words, de-
cisions about the size of the meal tend to occur prior to the meal
(Fay et al., 2011; McCrickerd, Chambers, & Yeomans, 2014); Feunekes
et al. (1995b, p. 555) found that “subjects were good in estimating
beforehand the size of the meal they would consume.” Koh and
Pliner’s (2009) “participants ate nearly all the food they served them-
selves” (p. 599). Wansink and Cheney (2005) found that snackers
ate more than 90% of the snack that they served themselves. If it
is the case that people tend to eat almost all of their initial portion,
then the social facilitation effect should be evident at the begin-
ning of the meal, when the food is served – with more food being
available (per capita) for “group” eaters than for “alone” eaters. If
this speculation is correct, then the social facilitation effect is not
simply a matter of “group” eaters eating a higher proportion of the
available food but rather a matter of “group” eaters starting off with
more food. As Feunekes et al. put it (p. 555), “to some extent sub-
jects know they will consume more when more others are present.”5
Consider dinner at a restaurant. Do “alone” diners order the same
amount of food (per capita) as do “group” diners (and perhaps take
more of their dinner home with them for tomorrow)? Or is it more
likely that “group” diners order more food than do “alone” diners?
Perhaps “group” diners order some bruschetta (or a large basket of
fries) “for the table,” in addition to the normal main course. Another
possibility (cf. Clendenen et al., 1994) is that “group” diners are more
likely to order and eat dessert (but see Krantz, 1979, who found some
evidence that solo diners were more likely to select dessert). Again,
it is not the case (usually) that dessert just appears; you have to
make provision for it; and “groups” may well be more likely to order
dessert – either one large dessert “for the table” or individual des-
serts for individual diners. In either case, if the “group” eaters eat
more (dessert), it is probably because they have ordered (more)
dessert, and we are entitled to ask whether the social facilitation
effect is a matter of eating being extended as the meal is pro-
longed in groups, or whether the group, anticipating (or seeking)
a prolonged meal (because eating in groups is more convivial), is
more likely to order dessert in the first place. Could the same thing
happen at home, with dessert becoming more likely in group eating
situations?
If, as we suggest, decisions about food intake are made largely
in advance, then we may conclude that people who are about to
eat in a group, anticipating increased intake, will provide them-
selves (or the group, on a per capita basis) with more food than will
lone diners. We might call this a “feast” hypothesis, on the grounds
that group eating is more likely to resemble a feast, with an element
of excess, than is solo eating, which is more utilitarian. This is not
to deny that group eating involves longer meals (or “time exten-
sion”) but simply to argue that it is not the longer time at the table
that leads to more eating in a passive way. Rather, it may be that
the anticipation of a group meal (including extended conversa-
tion and duration) prompts the diners (individually or collectively,
or by delegation to a “meal arranger”) to provide more food in the
5
Feunekes et al., elsewhere in their paper, conclude that “with more others
present. . .there was not more food available” [p. 557], suggesting that subjects’ rec-
ognition that they would eat more in groups did not entail the provision of more
food but simply an expectation that they would eat more, and presumably leave less,
of the available food, contrary to the notion that group eaters start off with more
food and eat almost all of what they start with.
69
first place, be it all laid out initially or in sequential courses. Krantz
(1979), who found (but did not expect) social facilitation in his
normal-weight diners, suggested that “they contemplated an up-
coming ‘social event’ rather than a quick lunch” (p. 24). This
possibility squares with our personal experience of the amount of
food available at dinner parties: the more people present, the greater
the amount of food (per capita) on offer, although perhaps the math-
ematical function is asymptotic, as dictated by Latané’s (1981) social
impact theory and reflected in de Castro and Brewer’s (1991) power
function. As Cavazza et al. (2011) note, “when we invite someone
over to dinner we normally prepare more food than we think our
guests normally eat” (p. 277). Because the connotation of “feast”
may overstate the festive atmosphere of the group meal, perhaps
we may introduce a new “scientific” term to denote the anticipa-
tion of increased intake in groups – the “social precilitation” of eating.
Cavazza et al. (2011) proposed that “if the amount eaten were
really a consequence of the intention to eat, this would imply that
the social influence processes leading to the increase in food con-
sumption in social occasion [sic] must occur in the planning phase”
(p. 277). Perhaps “individuals who eat as part of a large group may
anticipate sharing food with the other members and hence order
more food” (p. 277). This hypothesis was examined in an observa-
tional study (Study 1) in an Italian restaurant (in Italy). As predicted,
the number of dishes ordered per capita (from among starters, pasta
dishes, main courses, and desserts) increased as direct function of
group size. Group size did not predict leftovers, and no one took
food home. In other words, social facilitation of intake was ob-
served, and it was evident at the time of ordering, in advance of
eating. People apparently anticipate larger meals in larger groups
and order accordingly. (Bread and wine, which were not ordered
individually, did not show a social facilitation effect; in fact, per capita
intake declined as a function of group size, perhaps in compensa-
tion for the greater intake of the individually ordered dishes.) The
researchers conclude that “a greater food intake in the presence of
other people may be a conscious, planned choice rather than an un-
conscious response to food availability” (p. 280). Put differently, “there
is a shared expectation (script/descriptive norm) that in familiar
social situations one will eat a great deal of food” (p. 287). (There
is a tautology afoot here: People expect that groups will eat more
because groups eat more. We have not quite got at the original formal
cause of increased eating in groups, even if we have identified the
proximal mechanism.)
The same positive function relating number of items ordered to
number of people in the eating group was obtained in Cavazza et al.’s
(2011) Study 2, which involved ordering a simulated meal. Cavazza
et al. argue that conformity pressures contribute to the observed
social facilitation effect; more specifically, people tend to order more
when they observe others ordering more. Such convergence of or-
dering was evident in high self-monitors (who are especially
concerned with adjusting their behavior to group norms) but not
in low self-monitors. In sum, there would appear to be an implicit
norm of greater intake in larger groups, requiring that members of
larger groups order more food (for themselves); and the dynamics
of pre-meal ordering enforce the norm. In short, “the social facili-
tation of eating has its roots in the phase of ordering food” (p. 285).
Finally, we may want to address the issue of why there is a social
facilitation effect with strangers (i.e., people eat more in groups than
solo) whereas with friends, there is not only a social facilitation effect
but a social correlation (i.e., more friends → greater intake)? Maybe
it is the case that “the more the merrier” but that this phenome-
non applies with friends and not with strangers. The more friends,
the merrier; but not the more strangers, the merrier. Groups, then,
make provision for more food than do solo diners; but only groups
of friends make provision for increasingly more food (per capita)
as the size of the group increases. Impression management (i.e., a
strong desire to avoid eating excessively) remains a factor in groups
70 C.P. Herman/Appetite 86 (2015) 61–73
of strangers, no matter how many strangers there are. Increasing
the number of strangers does not make the meal any pleasanter or
lead to greater time extension.
A proposal
It is clear that a definitive explanation for the social facilitation
of eating has not yet been achieved. Still, progress has been made
and some speculation may be in order. As we have seen, there is
reason to believe that social facilitation occurs because people who
are about to eat in groups select more food and proceed to eat it
(although in lab studies, there is always much more food available
than is needed or eaten). More specifically, people who are about
to eat with friends select and eat more food, and it is almost cer-
tainly the case that most group meals are meals with friends. (Recall
Feunekes et al.’s (1995b) observation that fewer than 1% of meals
are eaten with strangers.) As we noted earlier, in our discussion of
Cavazza et al. (2011), there appears to be an expectation that friends
eating together will eat more, and this expectation becomes self-
fulfilling when these friends, anticipating a larger (and longer) meal,
provide themselves with more food. We must now confront the ques-
tion: How did it come about that friends eating together eat more,
with the result that they come to expect to eat more on similar oc-
casions in the future? In short, what is it about eating with friends
that makes for a larger meal?
We can decompose what we know about the social facilitation
of eating into a small number of basic propositions, and propose
an explanation for each proposition in turn, based on as few ex-
planatory principles as possible:
1. We begin with the proposition that people eat more when they
eat in groups than when they eat alone. To explain this basic
effect, we invoke the expansive social meal. We propose that
meals eaten in groups of two or more people (social meals) are
treated by participants as different in kind from meals eaten alone.
(Remember that, as was noted above, meals eaten in a crowd
may not qualify as social meals in the way that meals eaten in
a group do.) Social meals are tantamount to a social occasion,
which upgrades the social meal from the spartan solo meal to
a more expansive event. There is the prospect of socializing and,
especially when the group is composed of friends, conviviality
– what Sobal (2000) and Fischler (2013) refer to as
“commensality.”6 This prospect seems to “require” more food;
although it is obviously possible to enjoy a social occasion without
food, eating with friends would appear to be a central and
perhaps essential aspect of socializing. We propose that eating
with others may provide a “justification” for eating more than
one otherwise would. People may tell themselves that it is only
a matter of common courtesy to provide friends with extra food.
(We have already quoted Cavazza et al. (2011) to the effect that
“when we invite someone over to dinner we normally prepare
more food than we think our guests normally eat” [p. 277].) We
believe (or prefer to believe) that our guests will enjoy eating
more, in line with Herman et al.’s axiom that people want to
maximize their intake of palatable food. In providing our guests
with more food than they might ordinarily eat, we are of course
providing ourselves with more food than we might ordinarily
eat; after all, it is not just other people who want to maximize
their intake of palatable food. At a restaurant, or in any situa-
tion in which we select our own meal but not other people’s
meals, the situation changes, but only slightly. For one thing, we
(or someone else) may suggest that the group order a dish or
6
One might wonder whether social facilitation of eating would occur if friends
ate at separate locations but were connected by social media.
two collectively, “for the table,” above and beyond the diners’ in-
dividual orders. Further, we may engage in a collective (albeit non-
explicit) agreement to order more food than we would order were
we eating alone, as Cavazza et al.’s research suggests. This “con-
spiracy” reflects an implicit agreement to increase intake beyond
its normal limits, an agreement that depends on mutual complic-
ity among the diners, who may not be fully aware themselves of
what is happening. To close the circle, Herman et al. (2003, p. 878),
citing Wheeler (1966), noted that “participating together in an
episode in which behavioral restraints are lifted may create a greater
fellow-feeling among group members. Thus, social facilitation of
eating may actually enhance friendship.” More work could and
should be done on the connection between eating and social soli-
darity. Perhaps the evolutionary psychologists can devise
an explanation for how eating together (and eating more food to-
gether) serves the adaptive needs of the individual and the group.
If we turn the question around and ask why lone diners do not
eat as much as possible if they have access to palatable food – “Why
is the solo meal so spartan?” – we are now in a position to suggest
that the decision to overeat requires the complicit social support
of a group of co-eaters, with hearty eating among one’s tablemates
“licensing” one’s own hearty eating, which in turn may encour-
age or allow the others to eat more. For solo diners, the meal is
not as pleasant and thus less likely to be extended; and further,
there are no co-conspirators encouraging further intake. Finally,
as Herman et al. (2003) noted, the distraction provided by so-
cializing with others may interfere with the detection of incipient
satiety signals; such interference is not likely to occur in solo diners,
who will accordingly be more responsive to satiety signals. Non-
social distractions such as listening to a detective story (Bellisle &
Dalix, 2001) have also been shown to increase intake, possibly by
disrupting the perception of satiety signals. In sum, eating in a
group may not only potentiate intake but may even interfere with
the detection of satiety signals that would ordinarily limit intake
even in hearty eaters. Put differently, social eating may remove
or interfere with inhibitions (e.g., the perception of satiety) that
otherwise limit food intake. As we shall see, eating with others
imposes certain inhibitions while removing others; more gener-
ally, the effect of others may be analyzed in terms of how the effect
of the presence and behavior of others inhibits or disinhibits eating.
Beyond the basic social facilitation effect – people eat more in groups
than when alone – there are other considerations that a cogent
theory of socially facilitated eating must explain. As noted above,
we are inclined to explain the observed variations in the extent
of social facilitation in terms of inhibitory social influence effects
(see Herman et al., 2003, for a prior attempt to explain social in-
fluence effects in terms of inhibitory processes). We postulate that
the presence and behavior of other people while one is eating can
limit one’s intake, for at least two reasons. Being watched and/or
evaluated by others tends to induce self-consciousness, which in
turns inhibits eating.7 Likewise, being aware that other people have
eaten or are eating only a certain amount establishes an upper
limit for one’s own intake beyond which one risks being re-
garded as eating excessively, with negative consequences for how
one is evaluated. Both reasons, then, converge on the notion that
one must not risk being evaluated negatively; and eating a lot
(or more than the others are eating) risks such an evaluation.
Note that in the typical co-eating situation, the other people
7
Heatherton and Baumeister (1991) argued that binge eating is a consequence
of aversive self-awareness, a variant on self-consciousness. Accordingly, we confine
our claims about the inhibitory effect of self-consciousness on food intake to non-
pathological eaters. It is also worth mentioning that even binge eaters tend to curtail
their binges when they are in the presence of others.
 C.P. Herman/Appetite 86 (2015) 61–73
present serve both as an evaluative threat and as guides to the
limits of non-excessive eating. With these inhibitory influ-
ences in mind, we can now turn to the other basic features of
socially facilitated eating.
2. Our second basic proposition is that eating with friends and family
leads to greater intake than does eating with strangers. To explain
this effect, we focus on the reduction of intake in the presence
of strangers and propose that the effect of strangers is particu-
larly inhibitory: we are more concerned with making a good
impression on strangers than on our own family and friends, and
so we are correspondingly more concerned to avoid conveying
the negative impression associated with eating excessively.
Strangers inhibit our intake by making us more self-conscious
about eating “properly.” Note also that in a situation that in-
cludes people who are strangers to us, we are likely to serve as
strangers to them. The likely consequence is that not only do these
strangers threaten to evaluate us, but they themselves will prob-
ably eat only a limited amount of food for fear of negative
evaluation by us. Thus, we and our stranger co-eaters are both
likely to set a lower bar for what constitutes excessive eating.
(See the paper by Kaisari and Higgs in this issue for a nice
example of co-eating strangers eating less than co-eating friends.)
It is worth mentioning that the distinction between friend and
stranger is not necessarily hard-and-fast. There are degrees of
friendship and degrees of “strangership,” but these nuances are
generally ignored in the literature on social facilitation.8
3. The third basic proposition is the social correlation: the more
people in the group, the greater the per capita intake. This aspect
of social facilitation is more difficult to explain, but we may offer
some suggestions. For one thing, it may be that the amount food
(per capita) that is provided increases as a function of group size.
As the group size increases, the appropriate amount of food to
order or arrange for may become more difficult to estimate, and
those who are ordering may err on the side of over-ordering. This
relative abandon with larger groups may been seen as a type of
disinhibition; as the group size increases, the meal arrangers
become careless, caring less about the consequences of over-
ordering. Looking at the phenomenon from a slightly different
angle, we may suggest that as the group size increases, any given
individual may experience greater deindividuation (Diener, Lusk,
DeFour, & Flax, 1980), with less (perceived) focus by others on
how much that individual is eating. (Whether the extent of such
“deindividuation” differs depending on the group consisting of
friends or strangers is worth pondering. Some evidence sug-
gests that the social correlation is more prominent among friends
than among strangers, suggesting perhaps that adding strang-
ers to the eating situation may not reduce one’s self-
consciousness.) Another consideration is that as group size
increases, so does the likelihood of different people eating dif-
ferent amounts; and as we have seen, if the “intake norms” in
the group are highly variable, people may exceed those norms
with apparent impunity, as if “anything goes” (Leone et al., 2007).
Again, this effect may be construed in terms of (dis)inhibition,
with clear norms serving to constrain intake to a certain nor-
mative level, whereas ambiguous norms appear to remove the
inhibitory effect of intake norms altogether.
The preceding three propositions comprise the most widely ac-
cepted empirical generalizations about social facilitation of eating.
Other generalizations (e.g., males displaying greater social facili-
tation of eating than do females), as they garner empirical support
and acceptance, will likewise demand explanation. We are (mildly)
8 We thank an anonymous reviewer for making this point.
71
confident that the basic principles employed above to explain our
initial three propositions may be adapted to explain other phe-
nomena as they achieve propositional status.
Future directions
Significant progress has been made in the description and
analysis of the social facilitation of eating, but many questions
remain. As we have seen, de Castro’s time-extension hypothesis,
which is the prevailing “explanation” of social facilitation, does not
really explain the phenomenon. Time, after all, is never an ade-
quate explanation of behavior. Pliner et al. (2006) suggest that with
extended time, the diner undergoes extended exposure to food cues,
which drives eating. This analysis takes us only so far, however, and
does not adequately engage with the evident fact that social facili-
tation depends critically on pre-meal decisions about how much
food will be available to eat. (This pre-meal decision-making obtains
in real life, but perhaps not in the lab, where the available food is
virtually unlimited.) These competing views of what is happening
during socially facilitated eating deserve research attention, with
critical experiments designed to help us to distinguish between these
(and other) alternative explanations. Researchers’ ingenuity will be
challenged.
Other questions abound. For instance, people may eat more in
a group, but what exactly constitutes a group? Some research (e.g.,
Hirsch & Kramer, 1993) demands that we distinguish between a
group and a crowd (with the diner “lost in a crowd” behaving like
a solo diner). Is the distinction between a group and a crowd merely
a matter of numbers or is there some specifiable bond that sepa-
rates a group from a crowd? Another set of questions concerns the
types of relationships that conduce to social facilitation (e.g., friends)
or inhibit intake (e.g., strangers). The friend–stranger distinction
evokes some intermediate categories (acquaintances, family) that
complicate things. We may need to identify what operative char-
acteristics apply to friends versus strangers (comfort? lack of self-
consciousness?) and explore how these attributes apply to the
intermediate relationships and to groups containing both friends
and strangers. The same sort of analysis could be applied to the
obese–normal distinction that was explored in early research (Krantz,
1979; Maykovich, 1978). What about dieters, who bring with them
to the eating occasion an additional motive (weight maintenance
or loss), which ought to dampen social facilitation, but who also bring
with them a propensity for disinhibited eating when circum-
stances (including social circumstances) allow (see Herman et al.,
2005, for a review)? Does socially facilitated eating depend on the
palatability of the available food? If the food were mediocre, would
we observe a differential decline in intake among groups of friends,
or might the social dynamics of the situation lead diners to ignore
the quality of the food?
There have been very few studies of social facilitation in chil-
dren. Is there a particular age or stage of development at which
socially facilitated eating first emerges? Perhaps the developmen-
tal emergence of social facilitation could tell us something about
the underlying mechanism. For instance, if socially facilitated eating
emerges at the same time as responsiveness to social norms, that
might provide support for the social-normative view of the phe-
nomenon. Of course, in this paper we have ignored socially facilitated
eating in non-human animals. A great integrative challenge awaits
anyone who attempts to integrate the human and nonhuman lit-
eratures. Zajonc (1965) tried it, but that was before there were any
studies of social facilitation of eating in humans. Another ques-
tion concerns the possible contribution of socially facilitated eating
to the obesity epidemic. If eating with others increases intake, we
may well proceed to ask whether people have become more likely
to eat with others during recent years in which the obesity epi-
demic has been observed. Standing back, we can see that there
72 C.P. Herman/Appetite 86 (2015) 61–73
remains a great deal of work to do in understanding the social fa-
cilitation of eating; and insofar as socially facilitated eating may be
contributing to the obesity epidemic, it could be claimed that such
work is of more than merely theoretical urgency.
References
Bell, R., & Pliner P. L. (2003). Time to eat: the relationship between the number of
people eating and meal duration in three lunch settings. Appetite, 41, 215–218.
Bellisle, F., & Dalix, A. M. (2001). Cognitive restraint can be offset by distraction, leading
to increased meal intake in women. The American Journal of Clinical Nutrition,
74, 197–200.
Bellisle, F., Dalix, A. M., Airinei, G., Hercberg, S., & Peneau, S. (2009). Influence of dietary
restraint and environmental factors on meal size in normal-weight women. A
laboratory study. Appetite, 53, 309–313.
Bellisle, F., Dalix, A. M., & de Castro, J. M. (1999). Eating patterns in French subjects
studied by the ‘weekly food diary’ method. Appetite, 32, 46–52.
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.
Boothby, E. J., Clark, M. S., & Bargh, J. A. (2014). Shared experiences are amplified.
Psychological Science, doi:10.1177/0956797614551162. In press.
Brindal, E., Wilson, C., Mohr, P., & Wittert, G. (2011). Does meal duration predict
amount consumed in lone diners? An evaluation of the time-extension
hypothesis. Appetite, 57, 77–79.
Brunstrom, J. M. (2011). The control of meal size in human subjects. A role for
expected satiety, expected satiation and premeal planning. Proceedings of the
Nutrition Society, 70, 155–161.
Brunstrom, J. M., & Rogers, P. J. (2009). How many calories are on our plate? Expected
fullness, not liking, determines meal-size selection. Obesity, 17, 1884–1890.
Brunstrom, J. M., & Shakeshaft, N. G. (2009). Measuring affective (liking) and
nonaffective (expected satiety) determinants of portion size and food reward.
Appetite, 52, 108–114.
Cavazza, N., Graziani, A. R., & Guidetti, M. (2011). Looking for the right amount to
eat at the restaurant. Social influence effects when ordering. Social Influence, 6,
274–290.
Clendenen, V. I., Herman, C. P., & Polivy, J. (1994). Social facilitation of eating among
friends and strangers. Appetite, 23, 1–13.
de Castro, J. M. (1990). Social facilitation of duration and size but not rate of the
spontaneous meal intake of humans. Physiology & Behavior, 47, 1129–1135.
de Castro, J. M. (1991). Social facilitation of the spontaneous meal size humans occurs
on both weekdays and weekends. Physiology & Behavior, 49, 1289–1291.
de Castro, J. M. (1994). Family and friends produce greater social facilitation of food
intake than other companions. Physiology & Behavior, 56, 445–455.
de Castro, J. M. (1995). Social facilitation of food intake in humans. Appetite, 24, 260.
de Castro, J. M. (1997a). Socio-cultural determinants of meal size and frequency. British
Journal of Nutrition, 77, S39–S55.
de Castro, J. M. (1997b). Inheritance of social influences on eating and drinking in
humans. Nutrition Research, 17, 631–648.
de Castro, J. M. (2002). Independence of heritable influences on the food intake of
free-living humans. Nutrition (Burbank, Los Angeles County, Calif.), 18, 11–16.
de Castro, J. M., Bellisle, F., Feunekes, G. L. J., Dalix, A.-M., & de Graaf, C. (1997). Culture
and meal patterns. A comparison of the food intake of free-living Americans,
Dutch, and French students. Nutrition Research, 17, 807–829.
de Castro, J. M., & Brewer, E. M. (1991). The amount eaten in meals by humans is a
power function of the number of people present. Physiology & Behavior, 51,
121–125.
de Castro, J. M., Brewer, E. M., Elmore, D. K., & Orozco, S. (1991). Social facilitation
of the spontaneous meal size of humans is independent of time, place, alcohol,
or snacks. Appetite, 15, 89–101.
de Castro, J. M., & de Castro, E. S. (1989). Spontaneous meal patterns in humans:
Influence of the presence of other people. American Journal of Clinical Nutrition,
50, 237–247.
de Castro, J. M., & Kreitzman, S. N. (1985). A microregulatory analysis of spontaneous
human feeding patterns. Physiology & Behavior, 35, 329–335.
de Castro, J. M., McCormick, J., Pedersen, M., & Kreitzman, S. N. (1986). Spontaneous
human meal patterns are related to preprandial factors regardless of natural
environmental constraints. Physiology & Behavior, 38, 25–29.
de Castro, J. M., & Orozco, S. (1990). Moderate alcohol intake and spontaneous eating
patterns of humans. Evidence of unregulated supplementation. American Journal
of Clinical Nutrition, 52, 246–253.
Diener, E., Lusk, R., DeFour, D., & Flax, R. (1980). Deindividuation. Effects of group
size, density, number of observers, and group member similarity on self-
consciousness and disinhibited behavior. Journal of Personality and Social
Psychology, 39, 449–459.
Edelman, B., Engell, D., Bronstein, P., & Hirsch, E. (1986). Environmental effects on
the intake of overweight and normal-weight men. Appetite, 7, 71–83.
Fay, S. H., Ferriday, D., Hinton, E. C., Shakeshaft, N. G., Rogers, P. J., & Brunstrom,
J. M. (2011). What determines real-world meal size? Evidence for pre-meal
planning. Appetite, 56, 284–289.
Feunekes, G. I., de Graaf, C., & van Staveren, W. A. (1995a). No clear evidence of social
facilitation of spontaneous meal size in Dutch students. Appetite, 24, 188.
Feunekes, G. I., de Graaf, C., & van Staveren, W. A. (1995b). Social facilitation of food
intake is mediated by meal duration. Physiology & Behavior, 58, 551–558.
Fischler, C. (2013). Commensality, society and culture. Social Science Information, 50,
528–548.
Goldman, S. J., Herman, C. P., & Polivy, J. (1991). Is the effect of a social model
attenuated by hunger? Appetite, 17, 129–140.
Harlow, H. F. (1932). Social facilitation of feeding in the albino rat. Journal of Genetic
Psychology, 43, 211–221.
Heatherton, T. F., & Baumeister, R. F. (1991). Binge eating as escape from self-
awareness. Psychological Bulletin, 110, 86–108.
Herman, C. P., & Polivy, J. (1996). What does abnormal eating tell us about normal
eating? In H. Meiselman & H. MacFie (Eds.), Food choice, acceptance, and
consumption (pp. 207–238). London: Blackie Academic & Professional.
Herman, C. P., & Polivy, J. (2003). Dieting as an exercise in behavioral economics. In
G. Loewenstein, D. Read, & R. F. Baumeister (Eds.), Time and decision. Economic
and psychological perspectives on intertemporal choice (pp. 459–490). New York:
Russell Sage Foundation.
Herman, C. P., Polivy, J., & Leone, T. (2005). The psychology of overeating. In D. Mela
(Ed.), Food, diet and obesity. Cambridge, UK: Woodhead Publishing, pp. 115–136.
Herman, C. P., Roth, D. A., & Polivy, J. (2003). Effects of the presence of others on food
intake. A normative interpretation. Psychological Bulletin, 129, 873–886.
Hetherington, M. M., Anderson, A. S., Norton, G. N., & Newson, L. (2006). Situational
effects on meal intake. A comparison of eating alone and eating with others.
Physiology & Behavior, 88, 498–505.
Hirsch, E. S., & Kramer, F. M. (1993). Situational influences on food intake. In B. M.
Marriott (Ed.), Nutritional needs in hot environments (pp. 215–243). Washington,
DC: National Academy Press.
Kim, J. Y., & Kissileff, H. R. (1996). The effect of social setting on response
to a preloading manipulation in non-obese women and men. Appetite, 27, 25–
40.
Klesges, R. C., Bartsch, D., Norwood, J. D., Kautzman, D., & Haugrud, S. (1984). The
effects of selected social and environmental variables on the eating behavior of
adults in the natural environment. International Journal of Eating Disorders, 3,
35–41.
Koh, J., & Pliner, P. (2009). The effects of degree of acquaintance, plate size, and sharing
on food intake. Appetite, 52, 595–602.
Kral, T. V., Kabay, A. C., Roe, L. S., & Rolls, B. J. (2010). Effects of doubling the portion
size of fruit and vegetable side dishes on children’s intake at a meal. Obesity (Silver
Spring), 18, 521–527.
Krantz, D. S. (1979). Naturalistic study of social influences on meal size
among moderately obese and non-obese subjects. Psychosomatic Medicine, 41,
19–27.
Latané, B. (1981). The psychology of social impact. American Psychologist, 36, 343–355.
Leone, T., Pliner, P., & Herman, C. P. (2007). Influence of clear versus ambiguous
normative information on food intake. Appetite, 49, 58–65.
Lumeng, J. C., & Hillman, K. H. (2007). Eating in larger groups increases food
consumption. Archives of Disease in Childhood, 92, 384–387.
Maykovich, M. K. (1978). Social constraints in eating patterns among the obese and
overweight. Social Problems, 25, 453–460.
McCrickerd, K., Chambers, L., & Yeomans, M. R. (2014). Does modifying the thick
texture and creamy flavour of a drink change portion size selection and intake?
Appetite, 73, 114–120.
Mekhmoukh, A., Chapelot, D., & Bellisle, F. (2012). Influence of environmental factors
on meal intake in overweight and normal-weight male adolescents. A laboratory
study. Appetite, 59, 90–95.
NPD Group (2014). Consumers are alone over half of eating occasions as a result of
changing lifestyles and more single-person households. https://www.npd.com/
wps/portal/npd/us/news/press-releases/consumers-are-alone-over-half-of
-eating-occasions-as-a-result-of-changing-lifestyles-and-more-single-person
-households-reports-npd/.
Paquet, C., St.-Arnaud-McKenzie, D., Ma, Z. F., Kergoat, M. J., Ferland, G., & Dubé, L.
(2008). More than just not being alone. The number, nature, and complementarity
of meal-time social interactions influence food intake in hospitalized elderly
patients. The Gerontologist, 48, 603–611.
Patel, K. A., & Schlundt, D. G. (2001). Impact of moods and social context on eating
behavior. Appetite, 36, 111–118.
Péneau, S., Mekhmoukh, A., Chapelot, D., Dalix, A.-M., Airinei, G., Hercberg, S., et al.
(2009). Influence of environmental factors on food intake and choice of beverage
during meals in teenagers. A laboratory study. British Journal of Nutrition, 102,
1854–1859.
Pliner, P., Bell, R., Hirsch, E. S., & Kinchla, M. (2006). Meal duration mediates the effect
of “social facilitation” on eating in humans. Appetite, 46, 189–198.
Redd, E. M., & de Castro, J. M. (1992). Social facilitation of eating. Effects of instructions
to eat alone or with others. Physiology & Behavior, 52, 749–754.
Salvy, S.-J., Coelho, J. S., Kieffer, E., & Epstein, L. H. (2007). Effects of social contexts
on overweight and normal-weight children’s food intake. Physiology & Behavior,
92, 840–846.
Salvy, S.-J., Howard, M., Read, M., & Mele, E. (2009). The presence of friends increases
food intake in youth. American Journal of Clinical Nutrition, 90, 282–287.
Salvy, S.-J., Jarrin, D., Paluch, R., Irfan, N., & Pliner, P. (2007). Effects of social influence
on eating in couples, friends and strangers. Appetite, 49, 92–99.
Salvy, S.-J., & Pliner, P. (2010). Social influences on eating in children and adults. In
L. Dubé, A. Bechara, A. Dagher, A. Drewnowski, J. Lebel, P. James, et al. (Eds.),
Obesity prevention. The role of brain and society on individual behavior
(pp. 617–627). Elsevier.
 C.P. Herman/Appetite 86 (2015) 61–73
Schachter, S. (1971). Some extraordinary facts about obese humans and rats. American
Psychologist, 26, 129–144.
Schlundt, D. G., & Zimering, R. T. (1988). The dieter’s inventory of eating temptations.
A measure of weight control competence. Addictive Behaviors, 13, 151–164.
Shide, D. J., & Rolls, B. J. (1991). Social facilitation of caloric intake in humans by friends
but not by strangers. International Journal of Obesity, 15, 8 [Abstract].
Sobal, J. (2000). Sociability and the meal. Facilitation, commensality, and interaction.
In H. Meiselman (Ed.), Dimensions of the meal. The science, culture, business, and
art of eating (pp. 119–133). Gaithersburg, MD: Aspen.
Sommer, R., & Steele, J. (1997). Social effects on duration in restaurants. Appetite, 29,
25–30.
Stroebe, W. (2008). Dieting, overweight, and obesity. Self-regulation in a food-rich
environment. Washington, DC: American Psychological Association Press.
Stroebele, N., & de Castro, J. M. (2004). Effect of ambience on food intake and food
choice. Nutrition (Burbank, Los Angeles County, Calif.), 20, 821–838.
Tice, D. M., Butler, J. L., Muraven, M. B., & Stillwell, A. M. (1995). When modesty
prevails. Differential favorability of self-presentation to friends and strangers.
Journal of Personality and Social Psychology, 69, 1120–1138.
73
Triplett, N. (1897). The dynamogenic factors in pacemaking and competition. American
Journal of Psychology, 9, 507–533.
Vartanian, L., Herman, C. P., & Polivy, J. (2007). Consumption stereotypes and
impression management. How you are what you eat. Appetite, 48, 265–277.
Vartanian, L., Spanos, S., Herman, C. P., & Polivy, J. (in preparation). Modeling of food
intake: A meta-analytic review. University of New South Wales.
Wansink, B., & Cheney, M. M. (2005). Super bowls: Serving bowl size and food
consumption. Journal of the American Medical Association, 293, 1727–1728.
Wansink, B., & Park, S. B. (2001). At the movies. How external cues and perceived
taste impact consumption volume. Food Quality and Preference, 12, 69–74.
Wheeler, L. (1966). Toward a theory of behavioral contagion. Psychological Review,
73, 179–192.
Winerman, L. (2004). A second look at twin studies. APA Monitor, 35, 46.
Young, M. E., Mizzau, M., Mai, N. T., Sirisegaram, A., & Wilson, M. (2009). Food for
thought. What you eat depends on your sex and eating companions. Appetite,
53, 268–271.
Zajonc, R. B. (1965). Social facilitation. Science, 149, 269–274.