Mark Lett (2011) 22:49–64

DOI 10.1007/s11002-010-9102-5

To seek variety or uniformity: The role of culture

in consumers’ choice in a group setting

Song Oh Yoon & Kwanho Suk & Seon Min Lee &
Eun Young Park

Published online: 27 February 2010

# Springer Science+Business Media, LLC 2011

Abstract One stream of research suggests that individuals in a collectivistic culture

tend to follow the others’ choices in a group choice setting (e.g., Markus and
Kitayama 1991). Other research, however, suggests that the choices others make
have little impact on the individuals’ choices in this cultural context (e.g., Kim and
Drolet 2003). The current research attempts to test these competing hypotheses by
examining individuals’ choices in group choice contexts (Ariely and Levav 2000).
First, comparisons of 517 meal order slips from two local restaurants in Korea with
simulated outcomes indicate a profound group-level uniformity seeking in dish
selections. Second, our field experiment reveals that members of a collectivist
culture tend to follow group members’ choices when group influence is evident.
Unlike the previous findings of studies conducted on individuals from Western
cultures, selections made in accordance with the cultural norm do not result in the
reduction of emotional well-being for the collectivist cultural individuals. Finally, we
provide more direct evidence for the hypothesized role of culture by showing that
the choices of individuals with collectivistic cultural backgrounds are associated
with higher uniformity-seeking tendency compared to those of individualistic
cultural backgrounds.

Keywords Culture . Group influence . Uniformity . Conformity . Variety seeking

S. O. Yoon : K. Suk (*) : S. M. Lee : E. Y. Park

School of Business, Korea University, Seoul 136-701, South Korea
e-mail: [email protected]
S. O. Yoon
e-mail: [email protected]
S. M. Lee
e-mail: [email protected]
E. Y. Park
e-mail: [email protected]
50 Mark Lett (2011) 22:49–64

1 Introduction

Previous research has demonstrated that consumers prefer to seek variety in their
choices (Ratner et al. 1999). Many of these findings are explained by people’s desire
to appear unique (Ariely and Levav 2000; Ratner and Kahn 2002). However, some
cross-cultural studies have challenged the general conclusion of such assessments
(Kim and Drolet 2003; Kim and Markus 1999). According to these studies, variety
seeking is less pronounced in Eastern than Western cultural members. In contrast to
the more individualistic Western culture, uniqueness and individuality take on
negative meanings in the more collectivistic Eastern culture. While past research has
provided insights into the cultural influences on variety-seeking behavior and its
underlying motivations, the implications with respect to consumer’s choice in a
group decision have yet to be clearly elucidated. This is mainly because past cross-
cultural studies have focused on individual choice situations (Kim and Markus 1999;
Kim and Sherman 2007). Thus, it is still ambiguous whether the negative meanings
associated with uniqueness in the Eastern culture will result in a mere lack of variety
seeking or in a greater conformity seeking in individual’s choice in a group setting.
Therefore, the current research focuses on examining the impact of group on
choices among consumers in collectivist cultures. Based on the findings of past
research (e.g., Kim and Drolet 2003; Markus and Kitayama 1991), we set two
competing hypotheses. One hypothesis suggests that consumers in collectivistic
cultures are likely to follow the choices of others when the choice is made in a group
setting, as being similar to others and conformity to a group is an important cultural
value in this context (e.g., Markus and Kitayama 1991). The other hypothesis
suggests that other group members’ choices would exert minimal effects on the
individual’s choice because choice is not an act of self-expression for collectivistic
cultural members (e.g., Kim and Drolet 2003). In addition, we examine the
conditions under which culture has a particularly greater influence on consumer
choice. In line with the previous research findings, culture is expected to have
greater implications on choice when individuals draw on cultural rather than
personal knowledge when they form judgment. Finally, we investigate the potential
differences between Eastern and Western cultural consumers in the emotional
consequences of aligning their choice behavior in accordance with a cultural norm.

2 Culture and variety/uniformity seeking

A sizeable body of research has demonstrated people’s desire for variety in various
domains including a choice between multiple items (Ratner and Kahn 2002),
simultaneous choices of sequentially consumed items (Simonson 1990), choices of
decision rules (Drolet 2002), and choices in group settings (Ariely and Levav 2000).
One common denominator of these findings is the people’s desire to portray a
positive self-image by appearing unique. This desire has been shown to play a role in
both individual and group decision making. For example, in individual choice
settings, Ratner and Kahn (2002) have demonstrated that people incorporate more
variety into their choices when their decisions are subject to public scrutiny.
Similarly, Drolet (2002) reported that individuals scoring high on the need for
Mark Lett (2011) 22:49–64 51

uniqueness scale tend to vary choice rules across a sequence of decisions more than
those who are scoring low. In group choice contexts, the research by Ariely and
Levav (2000) showed that individuals tend to make different choices from others to
achieve their goals of self-expression in the form of uniqueness.
However, the aforementioned phenomena were observed mostly in the USA,
which represents a highly individualistic culture. In such a cultural context,
individuals are encouraged to follow their own feelings and convictions and to
freely express their opinions. Thus, being distinctive from others generally has
positive associations, as it represents personal freedom and autonomy (Markus and
Kitayama 1991; Oyserman et al. 2002). Consequently, variety seeking is perceived
as a normative behavior in this culture. In contrast, variety seeking does not fit with
the cultural norms of many Eastern countries such as China, Japan, and Korea,
which represent highly collectivistic cultures. In such cultures where an interdepen-
dent self-view prevails, harmony and relatedness with others are more important
than being unique (Markus and Kitayama 1991; Triandis 1989). Thus, the
expression of one’s individuality is discouraged, as it is detrimental to the
maintenance of connectedness and fitting in. This suggests that variety seeking
does not have positive connotations in these cultural contexts as is the case with
individualistic cultures.
Consistent with these assumptions, Kim and Drolet (2003) showed that people
from different cultural contexts differ in the likelihood of seeking variety in the use
of choice rules. People from an individualistic culture tend to vary their choice rule
use, whereas people from a collectivistic culture do not show such a tendency.
However, in this study, there was neither a sign of variety seeking nor of consistency
seeking among the collectivist cultural members. The authors explained the results
based on different cultural assumptions about choice. Unlike people from an
individualistic culture who perceive choice as an opportunity through which the self
is expressed, those from collectivistic cultures do not perceive choice as an act of
self-expression. Therefore, collectivistic cultural members take the choice lightly and
do not attempt to convey a positive self-image through their choice. This leads to the
prediction that consumers in a collectivist culture would not necessarily follow the
choice of others in a group choice setting even if it is a socially desirable behavior
because they do not feel the pressure or need to portray their positive self-image
through their choice.
Conversely, earlier research in cross-cultural psychology suggests that even
members of a collectivist culture represent and express their cultural beliefs in their
preferences. Individuals from a collectivist culture view themselves as similar to
others because they live in a society where being like others is considered desirable
and positive (Heine and Lehman 1997; Markus and Kitayama 1991). This cultural
value, in turn, affects the formation of even the simplest preference even if there is
no pressure to behave in a culturally appropriate way. Consistent with this, Kim and
Markus (1999) showed that individualistic cultural members tend to prefer unique
and deviant options, whereas collectivistic cultural members tend to choose the
dominant option. For example, in one of their studies, European–Americans were
likely to select the colors of pens that are different from the rest of the pens
presented to them. In contrast, Chinese–Americans or Korean–Americans tended to
prefer the colors that appeared to be the choice of the majority in the group. These
52 Mark Lett (2011) 22:49–64

differences were observed despite the fact that the participants were not aware that
their choices were subject to public scrutiny. Thus, it is reasonable to predict that
consumers in a collectivistic culture will show an even more pronounced tendency
to choose similar options with others in a group choice setting when their decisions
become public.
While these two sets of research lead to different predictions regarding the impact
of the core cultural value on the choice of the collectivistic cultural consumers, these
studies are conducted in the individual choice settings, and thus, its implication for
the group decision is not clear. On one hand, East Asians’ cultural belief that choice
does not represent a medium through which an individual’s image is conveyed and
through which one is evaluated would lead to the expectation that other group
members’ choice has a minimal effect on their choices. On the other hand, positive
meanings associated with being similar to others may lead these individuals to avoid
making deviant choices from those of their group members.
We test these alternative predictions in three studies involving real restaurant
order data and two field studies which largely replicate the study of Ariely and
Levav (2000). The replication of the study settings of the previous research enables
us to make a direct comparison of our results with those based on Western
counterparts. In studies 1A and 1B, we analyze meal order slips of diners from two
local restaurants in Korea and compare the dish selection of real tables dining in a
group against what would be expected from a random sampling of population of all
individual choices across tables (i.e., simulated data representing menu selections in
the absence of any group influence). In study 2, we replicate our findings in a setting
that allows more experimental control. Finally, in study 3, we provide more direct
evidence for the hypothesized role of culture by comparing the choices of
respondents from different cultural backgrounds in one study setting.

3 Study 1A

Study 1A tests the aforementioned competing predictions in a natural setting by

analyzing real food selection data from an Italian restaurant in Korea. All procedures
and methods used in this study are similar to those of Ariely and Levav (2000, study
1) in which group-level variety/uniformity seeking was examined with the order
slips from a Chinese restaurant in the USA. As shown in many previous studies, the
USA and Korea represent strong individualistic and collectivistic cultures,
respectively (e.g., Han and Shavitt 1994; Kim and Drolet 2003; Kim and Markus
1999). Therefore, a comparison of our results with those of Ariely and Levav (2000)
would provide some insights into the cultural influence on the individual’s choice in
group contexts.

3.1 Method

Meal order slips were collected for four weekdays with the help of the restaurant
manager. The restaurant was a popular Italian restaurant near a large university in
Seoul, Korea. It served 34 main menus (mostly pasta), and each dish was priced
between 7,000 and 16,000 Korean Won (approximately between US$7.00 and US
Mark Lett (2011) 22:49–64 53

$16.00). All the dishes in this restaurant were served in individual portions and rarely
shared. All the order slips from tables of two or more diners were included in the
analysis. A total of 353 order slips (989 customers) were gathered and analyzed. The
average table size was 2.8 persons, and the table size varied from two to 13 diners.
Following Ariely and Levav (2000), the variety index (VI) was computed. This
index ranges between 0 and 1, with lower (higher) VI indicating a greater uniformity
(variety) in menu selection. For example, if all the dishes in a table are the same, the
index is equal to 0 (complete group-level uniformity). Likewise, when all the dishes
in a table are different, the index indicates 1 (complete group-level variety).
number of different options chosen within a table  1
VI ¼ :
table size  1
In order to test the uniformity/variety seeking of customers in a real group, we
compared the index of real orders with that of simulated tables, as previously done
by Ariely and Levav (2000). Simulated tables were created by reassigning each
individual’s menu selection from real groups to the tables of the same size as the
original tables. Thus, the simulated tables represent the individual diner’s order
without group influence. Such a reassignment simulation was iterated 100 times to
gain stability. The simulated menu selection of the hypothetical groups was used as
the control.
It is noteworthy that, although VI ranges from 0 to 1, the absolute VI score is not
an accurate indication of variety/uniformity seeking because VI does not adjust for
the random overlap of the choices among the group members which is affected by
such situational factors as group size, assortment size, and the distribution of choice
share among the given options. Thus, the more appropriate test of variety or
uniformity seeking would be to compare VI of the real group with that of the
simulated group since the both groups are equally affected by situational factors (i.e.,
the chance level overlap of choices are equal in both groups).

3.2 Results

First, the average VI of the real tables (M=0.82) was significantly lower than that of
the simulated outcomes (M=0.92; F(1, 349)=21.41, p<0.001). This indicates that
the group context influences the individual diner’s choice towards the direction of
group-level uniformity. This is in stark contrast with the findings of Ariely and
Levav (2000) based on American customers who displayed greater variety seeking in
real groups compared with the simulated groups. Second, no significant impact on group
size was observed in uniformity seeking. That is, the uniform choices of these customers
were prevalent regardless of the number of diners in a table (F(3, 349)=1.37, p=0.25).
The comparison between the average VI of the real and simulated tables for each table
size indicated that the VI for the real tables was significantly lower than that of the
simulated tables for all table sizes. Table 1 presents these results.1

1
The VI is lower for larger than smaller size tables for both real and simulated groups (ps <0.01).
However, the null interaction effect between the table type (real vs. simulated) and group size indicates
that this linear trend should be attributed to the nature of VI that does not adjust for the overlap of choice
by chance, rather than increasing conformity with table size.
54 Mark Lett (2011) 22:49–64

Table 1 Study 1A: the variety index for real and simulated tables

Table size N Real table VI Simulated table VI F test (df=349)

All 353 0.82 0.92 21.63***

2 diners 225 0.90 0.96 9.38***
3 diners 62 0.86 0.94 4.54**
4–5 diners 46 0.77 0.92 11.77***
6+ diners 20 0.74 0.86 3.18*

* ** ***
p<0.10; p<0.05; p<0.01

While the study provides basic evidence consistent with the uniformity-seeking
hypothesis, the results may be specific to the characteristics of this particular
restaurant. Thus, we generalize our findings in the next study by analyzing meal
order slips from a restaurant that differs in terms of the average price of dishes,
variety of menus, and type of foods they carried.

4 Study 1B

4.1 Method and results

We selected a restaurant specializing in Korean food located on the campus of a

large university in Seoul, Korea. The restaurant served only four different dishes per
day, and the types of dishes varied each day. The average prices of each dish ranged
between 5,000 and 7,000 Korean Won (approximately US$5.00 to US$7.00). Thus,
the menus in this restaurant, compared with those in study 1A, were more familiar,
were lower in price, and comprised of a more limited assortment. A total of 164
order slips (486 customers) were gathered for over two weekdays. The average table
size was 2.96 persons, and the table size varied from two to 15 diners. The methods
used to assess the uniformity/variety seeking were identical to those of study 1A,
except that the simulation was conducted separately for each day because of the
daily changes in the menu of the restaurant.
The results were largely consistent with study 1A. There was a greater group-
level uniformity for the choices made in the real groups (M=0.63) than in the
simulated tables (M=0.70; F(1, 160)=4.77, p<0.05). This pattern of uniformity
seeking occurred regardless of the size of the tables (F(3, 160)<1.0)2.

4.2 Discussion

The choice patterns observed in both studies 1A and 1B are opposite to the results
reported by Ariely and Levav (2000), which found the real groups exhibiting a

2
The overall VI in study 2B was lower than those of study 1A. As noted, the size of the menu selection in
study 1B was much smaller than in study 1A (four vs. 34), leading to greater potential for choice overlap.
Mark Lett (2011) 22:49–64 55

greater tendency to seek variety than the simulated groups among American
customers (Fig. 1). We assume that the observed difference is due to the different
core cultural values in these two countries. For Koreans who live in a highly
collectivistic culture, it is natural and almost normative to choose the option that is
similar to others, as being similar to others and not standing out from the group is a
desirable trait in this context. In contrast, Americans, whose important cultural
beliefs are uniqueness and being different from others, prefer to vary their choices
from those of their group members, as such behavior reflects positively on them in
the individualistic culture (Kim and Markus 1999).
While our results are consistent with the idea that group contexts lead Korean
customers to make more conformity choices, the restaurant order data do not provide
direct evidence for the group influence and do not allow for testing the potential
moderators. Therefore, we tested these in a more controlled setting in the next study.

5 Study 2

Study 2 was designed to achieve three important objectives. The first goal is to
replicate the previous findings in a situation that allows a more experimental control.
In this study, we manipulate the manner in which individuals choose their options,
such that in one condition they select their options under group influence (i.e.,
sequential choice condition) and in the other condition without group influence (i.e.,
independent choice condition). In addition, we measure the individual’s individu-
alism–collectivism disposition using the self-construal scale (Singelis 1994) to assess
the impact of cultural orientation.
The second goal is to identify the conditions under which cultural beliefs exert a
strong impact on individual’s choice. Past studies present that one key factor that
determines whether a culture-based effect looms larger or fades is the extent to
which a person draws on cultural versus personal knowledge when forming a
judgment (Briley and Aaker 2006; Deutsch and Gerard 1955). When cultural

1.0
0.95
0.92 0.93
0.9
0.82
0.8
VI
0.70
0.7
0.63

0.6

0.5
Study 1A Study 1B Ariely & Levav
(2000)
Real tables Simulated tables

Fig. 1 Study 1: group variety index of real and simulated groups in the current research and in study 1 of
Ariely and Levav (2000)
56 Mark Lett (2011) 22:49–64

knowledge is the only base that informs the judgment, culture exerts a strong effect
on the formation of judgments. In contrast, when people have other knowledge bases
on which their judgment can be made, such as personal experience and clear
preference, culture effect becomes weaker. Thus, it is our conjecture that the
tendency to choose the same options as the other group members will be more
pronounced when they are uncertain about their preferences.
Finally, we investigate the emotional consequence of carrying out culturally
appropriate choice behaviors among the collectivistic cultural members. Prior
research suggests that the individual’s choice in a group context reflects the
tradeoffs of different goals for the individualistic cultural members. Specifically,
these individuals sacrifice their personal satisfaction in favor of a positive self-
presentation, yielding variety seeking at a group level. Thus, people who vary their
choices from others tend to show lower satisfaction with their chosen options than
those who made choices without concern for variety (Ariely and Levav 2000;
Ratner et al. 1999; Simonson 1990). We examine whether this effect is also
applicable to the collectivistic cultural individuals who select uniform options under
group influence.

5.1 Method

A total of 122 groups (309 individuals) of undergraduate students in a large

university in Korea participated in the field experiment on a voluntary basis. The
size of group varied between two and five persons, and the average group size was
2.55 persons. Two experimenters approached students who were gathered in groups
in student lounges on campus and asked them to participate in a survey in exchange
for a small gift. The participants were informed that the survey intended to explore
college students’ preferences in various domains and that they will receive a snack
for their participation.
First, participants were given a short filler survey. After completing the filler
survey, the participants were thanked and presented with four different snacks on a
tray as a free gift. They were asked to choose one snack that they like the most. All
the snacks were imported brands contained in single-serve size packages. Less
well-known imported brands were used in order to prevent the potential effect of
personal knowledge and experience on choice. In fact, a pretest (n=13) showed
that they were not familiar with any of these snacks (M=1.93 on a nine-point scale).
In addition, the single-serve packaging ensured that the respondents did not expect to
share their choice with the other group members. By presenting on the tray the same
number of snacks for each type, we also made sure that the participants did not form
different assumptions regarding the popularity and shortage of inventory for each
snack.
The way in which group members make a snack choice differed across the
participating groups. Each group was randomly assigned to a 2 (choice task:
sequential vs. independent) × 2 (presence of product information: information vs.
no-information) between-groups conditions. The number of groups and the average
size of each group in the four experimental conditions are as follows: no-
information/independent choice (30 groups, average size = 2.53), no-information/
sequential choice (29 groups, average size = 2.69), information/independent choice
Mark Lett (2011) 22:49–64 57

(29 groups, average size = 2.34), and information/sequential choice (34 groups,
average size = 2.55). The average group size did not differ across conditions (F(3,
118)<1.0).
In the sequential choice conditions, group members chose a snack one by one in
sequence. The order of choice was randomly decided by the experimenter. The
choice order and the selected option for each member were recorded by the
experimenters. In the independent choice conditions, participants were asked to
make selection without looking at or discussing with their group members. Each
person in this condition was given an individual menu containing the names of the
four snacks and was asked to indicate on the menu the option they like the most. The
experimenters then gave the snack of their choice. In the information conditions,
respondents first tried a sample of each snack before making their choice. In the no-
information conditions, no such trial opportunity was given.
In all conditions, participants were asked to eat the snack of their choice
immediately following their selection. They were then given the second question-
naire containing the measures of satisfaction of their chosen option and individual’s
cultural orientations. Choice satisfaction was assessed by asking how satisfied they
were with the snack they selected on a nine-point scale (1 = dissatisfied, 9 =
satisfied). The participants’ cultural orientation was measured by the self-construal
scale developed by Singelis (1994). The scale measures the respondents’
collectivism (α=0.66) and individualism (α=0.72) orientation. The difference
between these two dimensions served as the participants’ cultural orientation index,
with a higher (lower) number indicating a stronger collectivism (individualism)
orientation. Furthermore, individual’s risk-seeking/avoidance disposition (Blais and
Weber 2006, α=0.75) was assessed to test the alternative account that conformity
choice is driven by collectivist members’ greater desire to avoid risk, rather than by
their collectivistic cultural orientation.

5.2 Results

5.2.1 The effects of choice order and product information on group-level uniformity/
variety seeking

We first examined the individual’s choice as a function of product information and

choice order at the group level. For each real group, we calculated the VI according
to their choice. We also simulated choices of randomly assigned groups (100
iterations) and calculated the average VI for each simulated group.
The effect of the choice task (sequential vs. independent) on VI of the real groups
differed significantly between the no-information and information conditions
(Fig. 2). When respondents made choices without sampling (i.e., no-information),
they are more likely to make similar choices as those of their group members when
the decisions were made sequentially (M=0.56) than independently (M=0.76; F(1,
118)=7.12, p<0.01). However, when participants had an opportunity to sample the
snacks before choice (i.e., information condition), whether choice was made
sequentially (M=0.67) or independently (M=0.67) did not affect their choices (F
(1, 118)<1. 0). These differences were qualified by a significant two-way interaction
(F(1, 118)=3.70, p=0.057).
58 Mark Lett (2011) 22:49–64

0.90 Real groups Simulated groups

0.80
0.76
0.72
0.71 0.70
0.69
VI 0.70 0.67 0.67

0.60
0.56

0.50
No-information/ No-information/ Information/ Information/
Sequential Independent Sequential Independent

Fig. 2 Study 2: group variety index of real and simulated groups as a function of information and choice task

Comparisons of the VI of the real group with that of the simulated group in each
experimental condition also showed that group-level uniformity seeking was
observed only in the no-information/sequential choice condition (M=0.56 vs.0.69;
F(1, 118)=2.96, p<0.10), and it was not significant in all other conditions (Fs<1.0).

5.2.2 The influence of cultural orientation on uniformity seeking

Next, we examined the relationship between the participants’ cultural orientation and
choice. In this analysis, only the participants who made choices in the second or later
order in the sequential choice were included. We conducted a logistic regression for
no-information and information conditions, separately. In each analysis, whether the
choice represents uniformity (1) or variety (0) served as the binary-dependent
variable and the individual’s cultural orientation index and choice order served as
predictor variables. If the choice was identical to any of the member’s choices in the
earlier sequence, it was coded as uniformity. If the choice did not overlap with any
of the previous choices, the choice was coded as variety. The order in the sequence
was included as a control variable, since the chance to choose the same snack as the
other members also increases as the number of preceding orders increases.
In the no-information condition, the participants’ cultural orientation was a
significant predictor of the type of choice (b=0.96, χ2 (1)=4.39, p<0.05). The
positive regression coefficient indicates that the higher the collectivistic orientation,
the greater the tendency to follow others’ choices. In addition, the order in the choice
sequence has a marginally significant effect on choice (b=0.83, χ2 (1)=3.74, p<
0.10). In the information condition, however, only the choice order had a marginally
significant impact (b=0.78, χ2 (1)=2.74, p<0.10) and the individual’s cultural
orientation did not affect choice (b=0.38, χ2 (1) =0.96, p>0.33). These findings,
when taken together, suggest that (1) cultural collectivism is positively related to the
individual’s tendency to select options identical to those of the other people in the
group and that (2) such a tendency was observed only when the participants are not
provided with product information prior to the choice.
Mark Lett (2011) 22:49–64 59

Further, individual’s risk attitude score was not significantly related to the type of
choice made by participants in both no-information and information conditions (ps>
0.20). This provides some evidence against an alternative account that the greater
conformity seeking among collectivistic cultural individuals may have been driven
by their greater motivation to reduce or avoid risk uncertainty by following others’
choice.

5.2.3 Satisfaction with choice

To investigate the potential influence of type of choice on satisfaction with the

chosen options, participants in the sequential choice group were classified as first
choosers, uniformity followers, and variety followers. First choosers are the ones
who made their choice first in the sequence. Uniformity followers are those whose
choice sequences fell second or later and whose choice was identical to any of the
earlier choices in the group. Variety followers are those whose choice sequences fell
second or later and whose choice did not overlap with any of the earlier choices in
the group. Participants in the independent choice condition (i.e., independent
choosers) were also included. The independent choosers and first choosers in the
sequential choice represent choices under no group influence; thus, they serve as the
control group.
A 2 (information) × 4 (chooser type: uniformity follower vs. variety follower vs.
first chooser vs. independent chooser) analysis of variance (ANOVA) on the
satisfaction rating showed a significant main effect of information (F(1, 301)=36.19,
p<0.001). Overall, the level of satisfaction with their chosen options was higher
when respondents made a choice with sampling (M=7.29) than without sampling
(M=6.09). While there was a significant main effect of chooser type (F(3, 301)=
2.87, p<0.05), this was qualified by the significant interaction with product
information (F(3, 301)=2.70, p<0.05). This interaction effect was further analyzed
by a series of planned contrasts (Table 2).
In the no-information conditions, the average satisfaction ratings of the variety
followers (M=5.21) were lower than those of the independent choice groups:
independent choosers (M=6.42; F(1, 301)=12.06, p<0.001) and the first choosers
(M=6.52; F(1, 301)=9.64, p<0.01). These results are consistent with an earlier
study (Ariely and Levav 2000), showing that variety seeking comes at the expense
of choice satisfaction. However, the satisfaction rating of the conformity followers
(M=6.20) did not differ significantly from either independent choosers or first

Table 2 Study 2: satisfaction as a function of information availability and type of choice

No information Information

Independent 6.42 7.12

Sequential
First choosers 6.52 7.24
Uniformity followers 6.20 7.65
Variety followers 5.21 7.15
60 Mark Lett (2011) 22:49–64

choosers (both F(1, 301)<1.0). More importantly, the satisfaction level of the
uniformity follower was substantially higher than that of the variety followers (F(1,
301)=4.48, p<0.05). In the information conditions, there were no significant
differences among the four groups (ps>0.19).

5.3 Discussion

Several conclusions can be drawn from the results of study 2. First, the higher-level
uniformity among the choices made sequentially than those made individually shows
that group influence underlies the group-level uniformity seeking. Second, group
effects have a dominating influence on choice only when individuals are not
provided with information about the options under consideration. While this result
can be interpreted as a different motivation underlying uniformity seeking, namely,
information-gathering goal, the significant relation between the individual’s
collectivism scores and the uniform choices provides evidence for our hypothesized
culture-based motivation. Finally, while the Western cultural members were shown
to seek variety for the sake of variety at the expense of their personal satisfaction, no
such signs of negative emotion emerged from the collectivist cultural individuals. In
our study, uniformity seekers were just as happy with their chosen options as those
who made individual choices. This finding is largely consistent with cross-cultural
study findings that individuals with collective cultural norm are relatively less
threatened by the restraint of one’s freedom, and it has minimal impact on personal
happiness (Iyengar and Lepper 1999). Rather, their personal happiness depends on
how much they are able to meet the expectation of others and on how they are able
to maintain a harmonious relationship with those around them.

6 Study 3

Although study 2 provided evidence that the choices representing variety/uniformity

seeking were associated with the decision makers’ cultural orientation, the evidence
was indirect because the cultural orientation was merely measured and not
experimentally manipulated. Therefore, this study provides more direct evidence
for the role of culture by comparing the choices of groups who possess different
cultural backgrounds (Korean students vs. students from individualist cultures). The
procedure of the study is similar to the no-information/sequential choice condition in
study 2. We also measure the decision maker’s knowledge on given products to offer
additional evidence regarding the role of knowledge in moderating the group effect.

6.1 Method

The study was a 2 (cultural background: collectivist vs. individualist) × 2 (product

knowledge: low vs. high) between-groups design. For groups representing
collectivistic culture, we recruited 21 groups consisting of only local Korean
students (55 participants, average group size = 2.62) from the campus of a large
university in Korea. For groups representing individualism, 19 groups (54
participants, average group size = 2.84) of exchange students who were visiting
Mark Lett (2011) 22:49–64 61

the same university were recruited from classes that were offered to these students.
The groups were formed on a voluntary basis, and the study was conducted after
class. Participants in this group identified their nationality as an individualist country
(as defined by Hofstede and Hofstede 2005) or those who had experience of living in
individualist countries for more than 5 years.3
Identical to the sequential/no-information condition in study 2, the members in
each group were asked to choose one snack from the option set containing four
different snacks in a sequential manner. In order to control for familiarity, we
selected snacks that were judged to be moderately unfamiliar to both cultural groups.
After consuming the chosen snack, participants responded to knowledge measures and
cultural orientation questions. Knowledge on the choice options was measured using
two seven-point scales assessing their familiarity with the given snacks and their choice
confidence (α=0.72). The average score of the two items served as the knowledge
index. As in the previous study, an individual’s cultural orientation index was
calculated as the difference between the individualism (α=0.86) and collectivism (α=
0.89) orientation, with the higher number indicating stronger collectivism.

6.2 Results and discussion

A preliminary analysis showed that, as expected, the average collectivism score was
significantly higher for groups with collectivistic than individualistic cultural
backgrounds (M=−0.16 vs. −0.67; F(1, 107)=6.96, p<0.01). Also, both groups
had a moderate level of knowledge on the presented snacks (Mcollectivist =4.07 vs.
Mindividualist =4.11), and this difference between the two groups in the product
knowledge was not significant (F(1, 107)<1.0). The groups were divided into low-
and high-knowledge groups according to the average knowledge index of the group
members (median = 4.25).
More germane to the current study, we tested the impact of culture and product
knowledge on the group-level VI by a 2 (cultural background) × 2 (product
knowledge) × 2 (choice: real vs. simulated) ANOVA with group size included as a
covariate.4 The analysis revealed different patterns of outcomes for the collectivist
and the individualist groups (Fig. 3). In the collectivist group, there was a significant
knowledge × choice interaction effect (F(1, 35)=4.77, p<0.05). Specifically,
individuals in the collectivist group tended to opt for uniformity options when the
product knowledge was low, as indicated by its lower VI (M=0.52) compared to that
of the simulated group (M=0.67; F(1,35)=2.83, p<0.10). However, when the level
of product knowledge was high, the VI of the collectivist group was indistinguish-
able from that of the simulated group, indicating no sign of conformity seeking

3
Initially, 29 groups (95 participants) of international students were recruited. Based on post hoc
screening, we eliminated ten groups because they included at least one member from collectivist culture
(e.g., students from China, Singapore, and Saudi Arabia). Cultural backgrounds of the individualist
participants include North America (n=40), Europe (n=10), and others (n=4).
4
Additional analysis treating knowledge as continuous variable also yielded the same results. In this
analysis, the real VI score was regressed on knowledge index for the collectivist and the individualist
groups, separately. The results showed that knowledge score was related to the VI score for collectivist
groups (b=0.27, t=1.78, p=0.08), but not for individualist groups (b=0.03, t=0.34, p=0.74).
62 Mark Lett (2011) 22:49–64

Real groups Simulated groups

0.90 0.88

0.82
0.79
0.80

0.70
0.70 0.67 0.68
0.66
VI

0.60
0.52
0.50

0.40
Low knowledge/ High knowledge/ Low knowledge/ High knowledge/
Collectivist culture Collectivist culture Individualist culture Individualist culture

Fig. 3 Study 3: group variety index of real and simulated groups as a function of cultural orientation and
knowledge

(Mreal =0.82 vs. Msimulated =0.70; F(1, 35)=1.98, p=0.17). For the individualist
groups, only the main effect of choice (actual vs. simulated) was significant (F(1,
35)=5.69, p<0.05) and no other effect emerged significant (Fs<1.0). The group-
level variety seeking occurred regardless of the level of product knowledge (Mreal =
0.83 vs. Msimulated =0.67).
Overall, the results regarding the opposite patterns of group influence in the
individualist and collectivist groups show more direct evidence that culture plays a
key role in determining the individual’s choice in a group. Also, this study replicates
the result of study 2 regarding the moderating role of knowledge in group influence
for collectivist cultural members. It is noteworthy that this moderating effect of
knowledge did not emerge in groups with individualist cultural backgrounds. That
is, individualist group members opt for choices that produced a group-level variety
even when they were relatively familiar with the given snacks, which is consistent
with Ariely and Levav (2000) who found that variety seeking among individualistic
cultural members persisted even when product information was available.

7 General discussion

We believe that our research covers an interesting and important issue in the
marketing theory that has received much attention in previous research (e.g., Ariely
and Levav 2000; Ratner and Kahn 2002). In the past, cross-cultural research has
offered two suggestions regarding the impact of the collectivistic culture on
individual’s choice in a group setting, which at first seem controversial. The current
research, building upon these two different views, shows that members of a
collective culture tend to prefer options that produce group-level uniformity seeking.
Furthermore, culture exerts the strongest impact when cultural knowledge is the only
source from which their decisions can be drawn. It is also interesting to note the
different emotional consequences of fulfilling the group-level goal among
individuals from individualist and collectivist cultures.
Mark Lett (2011) 22:49–64 63

From the managerial perspective, the current research helps to build further
understanding of consumers in collectivist cultures. As the world becomes highly
global, firms increasingly recognize the importance of cultural differences in the
consumer behavior of different cultural environments. This is reflected in the
increasing attempts among multinational firms to localize their marketing strategies
by incorporating culture-specific contents into their marketing activities. Our results
provide managers with useful insights with respect to the specific influence a group
may have on the choice of individual consumers in collective cultures and may
encourage more marketing efforts based on social influence to target these consumers.
Moreover, our research may offer some accounts for the collectivist cultural
consumers’ greater reliance on word-of-mouth and referrals (Takada and Jain 1991;
Money et al. 1998) and the relatively fast diffusion of new products and innovations in
countries with collectivistic characteristics (Van den Bulte and Stremersch 2004).
To enhance our understanding of consumers from the collectivist culture, several
issues warrant further investigation. First, while our results are in support of the idea that
individuals of the collectivist culture also take choice as a means to convey their positive
self-image to their social environment, the extent of group influence is still questionable.
For example, whether an individual will show the same level of uniformity seeking even
in an artificially formed group (i.e., without meaningful personal relations among
members) remains inconclusive. Similarly, in our studies, as well as in other studies
examining the variety-seeking behavior, the choice stimuli are exclusively hedonic
products (i.e., food), thus limiting the generalization of the findings to broader settings.
One interesting implication of our study is that meanings attached to culturally
consistent behaviors may differ by culture and, consequently, these may have
different impact on the individual’s emotional well-being. For instance, Western
individuals’ attempt to align themselves with cultural norms (e.g., variety seeking)
has detrimental effects on their personal happiness because it dampens personal
freedom and autonomy (Ariely and Levav 2000). Such adherence to cultural norms
(e.g., conformity seeking) is less likely to result in negative emotions for collectivist
cultural members, as such behavior is perceived to enhance their relatedness and
connection to others, which are the major contributors to their personal happiness
(Markus and Kitayama 1991). While our results show a mere lack of negative
emotion, they may actually enhance the personal happiness of the collectivist
cultural members in some situations (Iyengar and Lepper 1999). Future research that
will delve into this and other related issues will shed light on our understanding of
cultural influence on individual choice in a group setting.

Acknowledgements The authors gratefully acknowledge Seoung Wan Cho, Jieun Choi, Chiah Cheon,
Jinkyung Goo, Kyoungmi Kim, So Hyun Kim, Eun-Eui Lee, Jiheon Lee, Hyesung Park, and Kikyung
Park for their assistance in data collection. The authors also thank the manager of Little Province (study
1A) and the owner of Ire restaurant (study 1B) and Utpal Dholakia (Study 3) in their generous supports in
data collection. Correspondence: Song-Oh Yoon and Kwanho Suk.

References

Ariely, D., & Levav, J. (2000). Sequential choice in group settings: Taking the road less traveled and less
enjoyed. Journal of Consumer Research, 27(3), 279–290.
64 Mark Lett (2011) 22:49–64

Blais, A., & Weber, E. U. (2006). A Domain-Specific Risk-Taking (DOSPERT) Scale for adult
populations. Judgment and Decision Making, 1(1), 33–47.
Briley, D. A., & Aaker, J. L. (2006). When does culture matter? Effects of personal knowledge on the
correction of culture-based judgments. Journal of Marketing Research, 43(3), 395–408.
Deutsch, M., & Gerard, H. B. (1955). A study of normative and informational social influence upon
individual judgments. Journal of Abnormal and Social Psychology, 51(3), 626–636.
Drolet, A. (2002). Inherent rule variability in consumer choice: Changing rules for change’s sake. Journal
of Consumer Research, 29(3), 293–305.
Han, S., & Shavitt, S. (1994). Persuasion and culture: Advertising appeals in individualistic and
collectivist societies. Journal of Experimental Social Psychology, 30(4), 326–350.
Heine, S. J., & Lehman, D. R. (1997). Culture, dissonance, and self-affirmation. Personality and Social
Psychology Bulletin, 23(4), 389–400.
Hofstede, G., & Hofstede, G. J. (2005). Cultures and organizations: Software of the minds. New York:
McGraw Hill.
Iyengar, S. S., & Lepper, M. R. (1999). Rethinking the value of choice: A cultural perspective on intrinsic
motivation. Journal of Personality and Social Psychology, 76(3), 349–366.
Kim, H., & Drolet, A. (2003). Choice and self-expression: A cultural analysis of variety-seeking. Journal
of Personality and Social Psychology, 85(2), 373–382.
Kim, H., & Markus, H. R. (1999). Deviance or uniqueness, harmony or conformity? A cultural analysis.
Journal of Personality and Social Psychology, 77(4), 785–800.
Kim, H. S., & Sherman, D. K. (2007). “Express Yourself”: Culture and the effect of self-expression on
choice. Journal of Personality and Social Psychology, 92(1), 1–11.
Markus, H. R., & Kitayama, S. (1991). Culture and the self: Implications for cognitions, emotion, and
motivation. Psychological Review, 98(2), 224–253.
Money, R. B., Gilly, M. C., & Graham, J. L. (1998). Explorations of national culture and word-of-mouth
referral behavior in the purchase of industrial services in the United States and Japan. Journal of
Marketing, 62(4), 76–87.
Oyserman, D., Coon, H., & Kemmelmeier, M. (2002). Rethinking individualism and collectivism:
Evaluation of theoretical assumptions and meta-analyses. Psychological Bulletin, 128(1), 3–72.
Ratner, R. K., & Kahn, B. E. (2002). The impact of private versus public consumption on variety-seeking
behavior. Journal of Consumer Research, 29(2), 246–257.
Ratner, R. K., Kahn, B. E., & Kahneman, D. (1999). Choosing less-preferred experiences for the sake of
variety. Journal of Consumer Research, 26(1), 1–15.
Simonson, I. (1990). The effect of purchase quantity and timing on variety-seeking behavior. Journal of
Marketing Research, 27(2), 150–162.
Singelis, T. M. (1994). The measurement of independent and interdependent self-construals. Personality
and Social Psychology Bulletin, 20(5), 580–591.
Takada, H., & Jain, D. (1991). Cross-national analysis of diffusion of consumer durable goods in Pacific
Rim countries. Journal of Marketing, 55(2), 48–54.
Triandis, H. C. (1989). The self and social behavior in differing cultural contexts. Psychological Review,
96(3), 506–520.
Van den Bulte, C., & Stremersch, S. (2004). Social contagion and income heterogeneity in new product
diffusion: A meta-analytic test. Marketing Science, 23(4), 530–544.