Societies 2013, 3, 128–146; doi10.3390soc3010128 soci.docx

Societies 2013, 3, 128–146; doi:10.3390/soc3010128
societies
ISSN 2075-4698
www.mdpi.com/journal/societies
Article
Culture Matters: Individualism vs. Collectivism in Conflict
Decision-Making
Rebecca LeFebvre and Volker Franke *
International Conflict Management, Kennesaw State University,
1000 Chastain Road, Kennesaw,
Georgia 30144, USA
* Author to whom correspondence should be addressed; E-Mail:
[email protected];
Tel.: +1-678-797-2931; Fax: +1-770-423-6312.
Received: 25 January 2013; in revised form: 1 March 2013 /
Accepted: 4 March 2013 /
Published: 12 March 2013
Abstract: Does culture matter in decision-making? Existing

literature largely assumes that
the cognitive processes that inform decision-making are
universally applicable, while only
very few studies indicate that cultural norms and values shape
cognitive processes. Using
survey based quasi-experimental design, this research shows
that subjects with higher
levels of individualism tend to be more rational in their decision
processing, while those
with higher levels of collectivism tend to be more dependent
and less likely to betray the
interests of members of more central ingroups in favor of less
central ingroups. Furthermore,
the results indicate that in conflict settings that seem familiar,
individuals are more likely to
compromise in order to achieve peace.
Keywords: individualism; collectivism; social identity; culture;
decision-making
1. Introduction
In August of 2012, Erick Barrondo became Guatemala’s first
ever Olympic medal winner. After
finishing second in the men’s twenty kilometer walk race,
Barrondo said, “It’s well known that

Guatemala has problems with guns and knives. I hope that this
medal inspires the kids at home to put
down guns and knives and pick up a pair of trainers instead”
[1]. Even today, Guatemala feels the
effects of its 36-year civil war that ended in 1996, as
individuals throughout the country make
decisions regularly on whether to cooperate or fight, and
whether to pursue peaceful activities or use
violence. The country’s history of armed conflict and polarized
politics created a complex social
OPEN ACCESS
Societies 2013, 3 129
environment prompting the presence of a United Nations (UN)
peacekeeping operation, MINUGUA,
for ten years ending in 2004. In its final report on the mission,
the UN noted that “the most difficult
challenge for MINUGUA was to operate in the multi-ethnic,
multi-cultural and multilingual
environment of Guatemala” [2]. This raises the question of how
significant the impact of culture is on
the decisions made not only by deployed peacekeeping troops,

but also by the individuals living in the
conflict society [3]. Decision-making as an activity is
considered to be a rational universal practice in
human behavior, as people of all cultures have problems and
opportunities that require making choices
from among alternatives. Most research studies presume that the
cognitive processing behind decision
making is also universal, although a small set of literature
suggests that cultural variation does make a
difference in cognitive processes [4]. If the argument for
cultural variation can be more fully
supported, particularly in the context of conflict, then
culturally-relevant constraints in decision-making
environments become more applicable to models of decision
processing and conflict prevention.
The main research question of this study is: how do cultural
norms and values affect decision-making
in conflict situations? Using a comparative case study, we
examined cross-country differences in
decision-making processes between sample populations in the
U.S. and Ghana. We selected these two
countries based on significant cultural differences, with the
U.S. being highly individualist and Ghana
being highly collectivist [5]. The design of our study centered

on a survey that examined key
individual attributes such as individualism-collectivism and
styles of decision making in response to a
quasi-experimental decision scenario presented to a sample of
university students from each country.
The article is divided into four main sections. The first section
provides a review of the current
literature on conflict theory, decision theory, culture, and group
identity. The second section presents a
theoretical framework and develops a set of testable hypotheses.
The third section provides a
discussion of the methodology and data collection techniques
and the fourth section presents our
findings and some preliminary suggestions for future research.
2. Decision-Making in Conflict
In this section, we review relevant literature from conflict
resolution theory, decision theory,
cultural dimensions, and group identity in an attempt to
systematically explore the impact of cultural
dimensions on decision-making processes, especially in the
context of conflict resolution
and transformation.
2.1. Conflict Resolution Theory

Conflict resolution emerged as a field of study in the 1950s and
1960s when a group of scholars
began to develop specific approaches and techniques for
analyzing the cultural generalizability of
conflict resolution [6]. Early theorists in the field of conflict
resolution did not find cultural variation to
be particularly relevant. For example, Burton and Sandole
characterized conflict resolution approaches
as generic, based on universal behaviors thought to transcend
cultural differences, because the root of
conflict was seen not through the impact of social institutions
and cultural values, but rather through
the drive to satisfy basic human needs [7]. Anthropologists such
as Avruch and Black brought
increased awareness to the culture question, challenging some
of the assumptions made about the
universality of human nature. Avruch and Black introduced the
notion of ethnoconflict theory which
precedes discussion of interests, needs, and values in the
analysis of conflict by focusing on the

implicit knowledge contained within a culture that is often
taken for granted and may rarely be
verbalized [8]. They refer to this knowledge as “local common
sense”. The importance of cultural
relevance to conflict resolution theory emerged partly as a
result of specific case studies that
demonstrated the ineffectiveness of taking North American-
based conflict resolution techniques and
trying to transfer them to other parts of the world without
accounting for local common sense [9,10].
Conflict resolution techniques often involve various aspects of
negotiation, and in the study of
negotiation processes, we find that cultural variation becomes
apparent while building relationships,
communicating, setting goals, and reaching agreements [11].
Cohen considered negotiations to be a
special case of communications, with all the inherent traps of
misunderstandings and
misinterpretations [12]. In international negotiations, the
potential for miscommunication is
characteristic of intercultural interaction in general, for in order
to have true communication, the
parties must have a common base of semantic assumptions.
People who have no shared experience or

common history have no guarantee that the intended meaning by
the sender is decoded correctly by the
receiver. In this sense, culture becomes a key variable when
people from different cultures interact.
Although there are differences of opinion on the relevance of
culture to conflict resolution, the field
has evolved to a point where culture is considered to be an
important factor in settings of interactive
conflict resolution. However, just how does culture factor in,
and how does it affect the individual
level, shaping how people make personal decisions about
whether to participate in conflict escalation
or resolution activities?
2.2. Decision Theory
How do people go about making a decision? In particular, we
are interested in how people make
decisions when faced with a situation of conflict. We briefly
review four types of decision theory.
2.2.1. Theories of Rationality
The rational choice theory was the dominant paradigm in
decision making from World War II until
the late twentieth century. Rational choice theory makes three
broad assumptions: actors know what

they want, are able to order their wants, and will choose the
best means to reach their desired ends [13].
For instance, in expected utility theory, different people
attribute different value to risk, so each seeks
to maximize their own “expected utility” rather than
maximizing an overall “expected value” that is
mathematically calculated [14]. In other words, people make a
rational choice by analyzing the costs,
benefits, and risks and seeking to optimize their outcome based
on their own preference for risk.
One flaw in any rational approach is the assumption that people
have perfect information and the
perfect ability to calculate all the costs and benefits before
making a decision [15]. In order to
accommodate the more imperfect real world, Simon introduced
his famous satisficing theory to
include shortcuts in the optimization process [16], allowing
individuals to set a threshold and accept
the first choice that crosses their threshold, leaving the rest of
the alternatives without analysis.
Some scholars believe that rationality provides the only
scientific approach to social theory [13].
Theories of rationality do simplify or generalize the problem at
hand, and generalization is seen to be

of great value because it allows for prediction. However, critics
of the rational choice approach argue
that it is inherently inconsistent with observed human behavior
[17–20]. For example, more lottery
tickets are sold as the size of the prize increases, indicating that
the size of the jackpot is more
important to players than the probability of actually winning
[21].
2.2.2. Theories of Emotion
Although much of the criticism against rational choice stems
from cognitive theory (described in
the next section), there are a few emotion-based decision factors
worth mentioning. The concept of
affect in decision-making is one that arises when emotions such
as fear are particularly strong. Affect
enables a person to make decisions quickly in the face of
danger, but can also cloud judgment [15].
Gordon and Arian studied the relationship between feelings of
threat and decision processing from
data on the Arab/Israeli conflict as well as local U.S. conflict
situations such as neighborhood crime [22].

They found that when people felt threatened, their decision
process was dominated by emotion, not
logic or rational considerations. Their findings showed that the
stronger the level of threat, the more
belligerent the policy choice; the lower the level of threat, the
more pacific the policy choice. They
argued that even though logic has as role in decision-making,
much of the process is driven by emotion.
2.2.3. Theories of Cognition
Observation has shown that people often make choices that are
not rational, and many social
scientists have pursued an explanation for this unpredictable
behavior via a cognitive approach.
Cognitive approaches focus on human processing of
information, including how individuals gather
information and then use it to evaluate situations [23]. This
approach has led to a model of human
behavior broadly referred to as “behavioral decision theory”,
which shows that people use cognitive
shortcuts and preferences when processing information,
deviating predictably from rational choice
theories [18].
Kahneman and Tversky researched why people seem to exhibit

inconsistent behavior when making
decisions in risky situations and published their ground-
breaking research on prospect theory in 1979 [19].
In prospect theory, the objects of choice are prospects with
value assigned in terms of gains and losses,
rather than final outcomes. People react differently depending
on how the situation is presented in
terms of those gains and losses. Underlying the work on
prospect theory are the concepts of heuristics
and biases. Heuristics are essentially cognitive shortcuts, which
reduce complex tasks to simpler
operations of judgment. Judgmental heuristics, however, can
introduce error and lead to suboptimal
outcomes [18,23–25]. These errors are usually referred to as
biases, which can vary depending on the
heuristic being used. Three commonly discussed heuristics are:
availability, representativeness, and
adjustment and anchoring. The availability heuristic is based on
information that is most accessible to
a person or occurrences that can be most easily brought to mind.
In the representativeness heuristic,
others are grouped into types with assumed similarity of
characteristics, as in stereotypes. Using the
adjustment and anchoring heuristic means starting from an

initial value that is known and making
adjustments from there based on the current situation [25].
There are many other heuristics that have
been identified in political science and psychological literature,
with more than fifty distinct heuristics
named [26]. The danger of this seemingly endless proliferation
of heuristics is that keeping track of
them all and discovering how individuals coordinate these
multiple judgment strategies becomes
overwhelmingly complex. A more productive approach would
be to focus on the cognitive processes
behind the decisions rather than on the individual heuristics
themselves.
There are varied examples in the literature of how social
scientists have applied aspects of cognitive
theory to specific situations. The application of the cognitive
approach to domestic politics and voting
practices is the most prevalent [26,27]. Other applications
include understanding how heuristics are
used by political elites [23] and security policy makers [24].
The application of heuristics to decision

making in situations of conflict has been studied only
infrequently.
2.2.4. Blended Theories
One can imagine various ways to combine theories of
rationality, emotion, and cognition, but there
is one particular blended approach that has received the most
attention in the literature: poliheuristic
theory. Poliheuristic theory offers an alternative to the
traditional rational actor model by integrating
aspects of both the cognitive and rational approaches to
decision-making. Poliheuristic theory is
conceptualized as a two-stage decision process. The first stage
is a cognitive process, which screens
the possible alternatives and narrows the choices by eliminating
options based on one or more
heuristics. In the second stage, the remaining alternatives are
then evaluated in a rational way in order
to minimize risk and maximize benefit [28–30]. Poliheuristic
theory has made important contributions
to understanding the decision making process in the realm of
foreign policy [31], but its application to
other types of decisions made by individuals and groups has
been limited [29].

2.3. Cultural Dimensions
The 1948 Universal Declaration of Human Rights states that
“everyone has the right to a
nationality,” but what determines nationality? Is it culture?
From a legal point of view, nationality is
about reflecting an individual’s genuine link with a country,
based on objective factors such as place of
birth, descent, and residency [32]. On the other hand, culture,
even though it is often associated with a
country, is more difficult to define and measure [33]. One
common definition of culture is based on the
anthropological conception of a learned system of meanings
rooted in symbols and language that allow
people to adapt to their environment and interpret their
experiences [34]. Hofstede’s 1980 study
defined culture as “the collective programming of the mind
which distinguishes the members of one
human group from another” [35].
In attempts to understand and categorize national cultures, a
number of scholars have narrowed
their focus to the study of values. Values are principles that
give order and guidance to people in their
thoughts and actions as they face common human problems and
issues [36]. By observing the

principles that different social groups use in their thoughts and
actions, researchers have been able to
infer the values shared by various social groups. Societies can
then be categorized and compared based
on commonly shared values. In 2008, Thomas reviewed the five
major frameworks that have emerged
out of these value studies, each allowing for the categorization
and comparison of national cultures: the
Kluckhohn and Strodtbeck framework, Hofstede’s model, the
Schwartz Value Survey, Trompennars’s
value dimensions, and the Global Leadership and Organizational
Behavior Effectiveness (GLOBE)
study [37]. Each of these five studies categorized culture in
terms of value measurements, deriving a
set of four to nine dimensions depending on the framework. The
only dimension to appear in all five
values studies is the individualism-collectivism dimension. The
individualism-collectivism dimension
has been used extensively in the study of social behaviors,
especially in efforts to predict behavioral

patterns [37–40]. For example, conducting a meta-analysis of
253 studies on individualism-collectivism,
Osyerman et al. concluded that the individualism-collectivism
construct does impact basic psychological
processing, and that cultural differences in the dimension
“provide a powerful explanatory tool for
understanding the variability in the behavior of individuals in
different parts of the world” [39].
The term individualism is used to define the degree to which
members of society define their self-image
as an individual or as part of a larger group. On the other hand,
those who define themselves from the
social and collective aspects of the self-concept are described
with the term collectivism. Individualism
and collectivism are sometimes seen as opposite ends of a
single continuum, but it is more accurate to
describe them as worldviews that make different aspects of the
self-concept salient. The core elements
of individualism are independence and uniqueness, whereas the
core elements of collectivism are duty
to in-group and maintaining harmony. Triandis writes about the
prototypical social relationships that
describe each of these two constructs [41]. For collectivism, the
prototypical relationship is the family.

The family cares for its members and cooperates together, often
acting as a single unit with common
goals. Each member of the family has a well-defined role and
status determined by position within the
group. Family members have strong emotional ties to one
another and are linked typically for life. For
individualism, the prototypical relationship is the marketplace
where an individual makes a payment
and receives a good or service in return. The relationships are
emotionally distant and although
members of the market interact frequently, each member
maintains his or her own distinct identity.
The marketplace encourages competition, and status is usually
determined by individual achievement
and success and not by membership in a particular group.
The individualism-collectivism construct is useful as a
mechanism for systematically describing
ways in which cultures differ. The construct is helpful for
understanding how culture influences
not only what people think but also how they think. How could
the measured difference in
individualism-collectivism play out in terms of decision
making? Chen and Li explored cultural
differences of decision-making between individualist and

collectivist societies by looking at Chinese
(collectivist) versus Australian (individualistic) cultures [42].
They found that the Chinese were less
cooperative with foreigners than with Chinese, whereas
Australians were equally cooperative with
members of either group. However, such research studies into
the cultural impacts on decision making
are rare and there are many unresolved and unanswered
questions in this area. Do individualist and
collectivist cultures differ in the values they apply to decisions
made in the face of conflict? Do such
differences enhance or aggravate peacekeeping efforts that are
initiated in conflict prone zones? Can
peacekeeping personnel be trained to manage cultural
differences and achieve peace more efficiently?
There is a need for further research to gain an improved
understanding of how cultural differences
manifest themselves in the decision making process,
particularly in conflict contexts. The well-
established differences in the individualism-collectivism
dimension between the two countries make
the U.S. and Ghana an excellent test case for a cross-country
comparative study on culture and its
impact on decision-making.

2.4. Social/Group Identity
Most generally, identity represents “the process by which the
person seeks to integrate his (sic)
various statuses and roles, as well as his diverse experiences,
into a coherent image of self” [43].
Individuals draw on multiple, sometimes even competing
subidentities (e.g., religious, political, social,
ethnic, or occupational) to derive their self-conceptions. These
subidentities become consequential for
behavior in situations when their salience is invoked [44]. A
person’s various subidentities form
specific links between the self and his or her membership in
social groups. Hofman specified salience
as the probability by which a subidentity is remembered and
activated in a given context [45].
Prolonged salience upgrades the subidentity in the “prominence
hierarchy” thereby enhancing its
“centrality” and the degree to which it connects with other
subidentities. The more central a

subidentity is to an individual’s self-conception and the more
interconnected it is with other
subidentities, the more committed the individual will be to
preserving and enhancing that identity and
to display attitudes, values, and social behaviors consistent with
it.
In the present context, the concept of “social identity” refers to
“that part of individuals’ self
concept which derives from knowledge of their membership in a
social group (or groups) together with
the value and emotional significance attached to that
membership” [46]. Theories of social identity are
typically based on three premises:
(1) people are motivated to create and maintain a positive self-
concept;
(2) the self-concept derives largely from group identifications;
and
(3) people establish positive social identities through normative
comparisons between favorable in-groups
and unfavorable out-groups [47].
Social identity research has demonstrated that individuals tend
to invoke their group identifications
in many decision contexts, since the norms, values, stereotypes
and behavior patterns associated with a

particular identity provide a sense of certainty and may inform
their choice among decision
alternatives [48–54].
3. Theory and Hypotheses
The main research question of this study asks with regard to
cultural variation: (1) how are
decisions made (the decision-making process), and (2) what are
the actual decisions (the decision-making
outcome). We operationalized the decision-making process
using the scale developed by Spicer &
Adler-Smith which identifies a self-reported general decision
making style (GDMS) falling into one of
five categories: rational, intuitive, dependent, avoidant, or
spontaneous [55]. A rational style is a
logical and structured approach to decision making. An intuitive
style relies upon hunches, feelings
and impressions. A dependent style relies upon the direction
and support of others. An avoidant style
tends to postpone or avoid making decisions. In addition, a
spontaneous style is impulsive and prone to
making spur of the moment decisions.
The independent variable, “cultural norms and values”, was
measured analyzing: (1) the cultural

traits of the individual decision maker, (2) the society that the
individual lives in, and (3) the cultural
context of the conflict setting. Recognizing that Hofstede’s
cultural dimensions of decision-making
operate at the country level and variation occurs across people
within each country, we measured the
cultural traits of the decision maker by employing the Auckland
Individualism Collectivism Scale
(AICS) [56]. Individualists are those that exhibit traits of
uniqueness, responsibility, and
competitiveness, while collectivists seek advice and harmony.
To measure societal aspects, we collected data using survey-
based quasi-experimental design at an
American university and compared results to data collected at a
university in Ghana, The cultural
dimensions of these two countries vary across Hofstede’s
measures, as shown in Figure 1, with the
most marked difference seen in the individualism-collectivism
score [5]. Even if individuals within a
country do not align with Hofstede’s dimensions for that

country, we anticipated that living in a
society with particular cultural dimensions would affect how the
individual behaved.
Figure 1. Hofstede’s Cultural Dimensions for the U.S. and
Ghana.
Source: The Hofstede Center 2012 [5].
For the final cultural aspect, the setting of a particular conflict
scenario was varied randomly
between two descriptions that differed distinctly in the
individualist/collectivist nature of the groups
involved. The conflict scenario was written as a short fictional
vignette depicting a mining operation
under protest. Participants were asked to take on a specific role
within the vignette and make decisions
that would potentially affect themselves and the various groups
to which they belonged (see below for
a more detailed description of the scenario). The vignette itself
was designed to make salient different
layers of social identity, as shown in Figure 2. Both versions of
the vignette followed this layered
model. The key difference between the versions was in the
description of the overall community—
individualist versus collectivist—and reference to friends and

family.
We hypothesized that each of the three aspects of culture would
impact decision making. First, the
individualist-collectivist nature of a person should affect their
decision-making style. The desire for
uniqueness and the competitive drive are likely to reflect a more
rational decision making style, while
the desire for harmony and the drive to seek advice from others
are likely to result in a more dependent
decision making style for collectivists. We tested the following
hypotheses:
H1: Individuals with more individualist traits are more rational
in their decision-making.
H2: Individuals with more collectivist traits are more dependent
in their decision-making.
With regard to the second cultural aspect, society, we
hypothesized that people living in a more
collectivist society (Ghana over the U.S.) would be more likely
to prioritize the needs and interests of
their more central ingroups, namely those identity groups that
are at more salient at the inner layers.

H3: Individuals from a more collectivist society are more likely
to prioritize the interests of
members of social groups that are closer to the core of their
salient identities.
We also anticipated an effect of the cultural setting of the
scenario, expecting that when decision
makers were presented with a conflict scenario in a cultural
setting that did not align with their own
cultural experiences, their style of decision-making would
adjust and become more avoidant.
Additionally, we expected that when decision makers were
presented with a scenario in a cultural
setting that generally felt similar to their own social
surroundings, their decisions would more often
lean toward cooperation and the pursuit of peace, rather than
continued protests. This type of effect can
be considered a “representativeness” heuristic where the
opposition is judged to be more trustworthy
because they represent something more familiar.
H4a: The more individualistic a person is, the more avoidant
his/her decision-making style will
become in collectivist contexts.
H4b: The more collectivist a person is, the more avoidant

his/her decision-making style will
become in individualist contexts.
H5: When the setting of the story assimilates one’s own cultural
context, decision makers more
often choose to cooperate and achieve peace, than resist and
continue protesting.
Figure 2. Layers of Group Identity in the Vignette.
4. Methods and Data
Data was collected from 469 undergraduate students across two
universities: Kennesaw State
University (KSU) located in Kennesaw, Georgia where 265
respondents participated through an
on-line survey system implemented through introductory
psychology classes, and the University of
Cape Coast (UCC) in Ghana, where 204 participants completed
a paper survey. UCC participants
came from three classes: 84 participants from an introductory
level history class; 67 participants from
a junior level sociology class; and 53 students from a senior

level business class. The resulting samples
from KSU and UCC were compared on key demographic
characteristics, such as gender, age, religion,
ethnicity, major, and country of origin, indicating that both
samples were similar in terms of age,
percent native born, religion, and distribution of ethnicities with
one dominant ethnicity and one
secondary ethnicity. The largest demographic difference was in
terms of gender, where KSU had an
unusually high participation of females, much higher than the
university average.
Selecting our sample from student populations presents a well-
recognized limitation to the
generalizability of our findings, because college students are
not representative of society or the public
at large. However, the primary goal of this experiment is not to
generalize to society, but to test for the
impact of culture on decision-making in a laboratory-like
setting, controlling as much as possible for
other, non-cultural, decision factors [57]. Surveying student
participants is justified by the fact that
they represent likely future leaders [58], a population of interest
as exemplar carriers of social identity
and values and, as such, a valid target group for research into

the cultural dynamics of local conflict
and its resolution and/or transformation.
Employing the quasi-experiment previously described,
participants were randomly assigned one of
two versions of the story: one set in an individualistic culture
and the other set in a collectivist culture.
Prior to being presented with the vignette, participants were
asked to complete the AICS instrument to
determine their individualism-collectivism traits, and the GDMS
instrument to determine the general
decision making style. After reading the vignette, participants
were presented with two decisions to
make. The first decision was whether to reveal fellow protest
members who were dangerously
sabotaging the mine, and the second decision was whether to
take an offer from the company CEO that
would provide additional benefits to the protest group, but not
all the miners. Each of these decisions
was presented as a binary choice. Finally, the GDMS instrument
was administered again in modified
form, with references to the particular decision just made in
order to assess the extent to which their
decision was affected by type of setting/cultural context. The
flow of the experiment is represented in

Figure 3.
Figure 3. Experimental Design.
The AICS instrument consisted of 26 items measuring traits
divided between individualism and
collectivism that had been previously validated and shown to be
reliable with α > 0.70 [59]. Responses
were given on a 5-point Likert-type scale ranging from strongly
disagree to strongly agree. Based on
participant feedback from a pretest conducted earlier, the most
repetitive questions in the survey were
removed, resulting in 14-item instrument with a slightly reduced
reliability of α > 0.67. See Appendix
A for the 14-item question set. The GDMS instrument consisted
of 25 items with five items identified for
each of the five decision-making styles and a previous
validation shown to be reliable with α > 0.67 [55].
Again based on participant feedback from the earlier pretest, the
most repetitive questions in the

survey were removed, resulting in 17-item instrument with a
slightly reduced reliability of α > 0.64.
See Appendix B for this 17-item question set.
5. Findings
We tested the first two hypotheses using regression analysis. H1
anticipates a positive relationship
between individualism and rational decision making, and this
was reflected in the results. The
correlation between the rational decision-making style and
individualism was statistically significant
(p < 0.05), remaining so when controlled by age, gender,
country, and religion1. Another factor that
emerged as significant was the country location, with
participants from the U.S. being less rational,
whereas those from Ghana being more rational. H2 anticipates a
correlation between dependent
decision making and collectivism, as reflected by the negative
sign on the individualism coefficient
(see Table 1). The correlation between a dependent decision-
making style and collectivism was
significant (p < 0.01), remaining so when controlling for age,
gender, country, and religion. The
coefficient for individualism-collectivism was small and again
country location emerged as a

significant factor with a larger coefficient value indicating that
those from the U.S. were less
dependent and those from Ghana tended to be more dependent.
No issues of heteroskedasticity were
present in the models. Adjusted R-squared was 4% and 5% for
the rational and dependent models
respectively, with statistical significance for the F-stat of both,
indicating robust models. Overall, our
statistical analysis showed support for both H1 and H2,
revealing significant relationships between
individualism and rational decision-making (H1), and
collectivism and dependent decision-making
(H2). The small coefficients and low R-squared values suggest,
however, that individualism-collectivism
traits are only a small part of the decision-making equation.
Tables 1 and 2 show a summary of the results.
H3 anticipates that in a more collectivist society (Ghana),
decision makers will prioritize interest of
members of more central ingroups over those of less central
ingroups. This hypothesis was tested via
the first decision made after reading the vignette. In this case,
the participants must decide whether to
reveal the names of the people sabotaging the mine (the inner
group circle of protestors) to the head of

the mining company (the most outer group circle, see Figure 2).
Revealing their names could get the
people into trouble with the law or cause them to lose their
jobs. However, not revealing their names
puts all the miners in physical danger (the group circle between
the other two layers), and puts the
company at risk (the most outer group circle). There were many
factors for participants to consider,
and the dilemma had no straightforward or obvious answer, but
a general indicator of prioritizing inner
1 Due to the dominance of the Christian religion in both
samples (93% in Ghana and 77% in the U.S.), we controlled for
religion as a dummy variable of Christian or not-Christian.
group circle interests should show in the decision not to reveal
the names. We therefore hypothesized
that participants from the U.S. were more likely to reveal the
names, and participants from Ghana were
less likely to reveal the names. The results of our logistic
regression did support this hypothesis when
controlling for age, gender, and religion (p < 0.05). Gender also

emerged as a significant factor (p < 0.05),
with men more often deciding to reveal the names. See Table 3
for the results.
Table 1. OLS Regression Results for Rational Decision-Making
Style.
DV=Rational I II III IV V
Individualism
(variable of interest)
0.108
(0.028) **
0.122
(0.012) **
0.121
(0.013) **
0.103
(0.034) **
0.110
(0.026) **
Age
−0.325
(0.128)
−0.325
(0.129)

−0.380
(0.075) *
−0.391
(0.065) *
GenderMale
−0.463
(0.853)
−2.475
(0.338)
−1.858
(0.472)
USParticipant
−7.087
(0.004) ***
−5.750
(0.023) **
ReligionChristian
7.218
(0.025) **

Adjusted R-Squared 0.008 0.013 0.011 0.028 0.037
F-Stat
4.867
(0.028) **
3.872
(0.022) **
2.561
(0.054) *
4.035
(0.003) ***
4.268
(0.001) ***
N 469 425 424 424 424
P-value in parentheses; *** = significant at 0.01; ** =
significant at 0.05; * = significant at 0.1.
Table 2. OLS Regression Results for Dependent Decision-
Making Style.
DV=Dependent I II III IV V
Individualism
−0.152
(0.003) ***
−0.139
(0.013) **

−0.141
(0.011) **
−0.172
(0.002) ***
−0.169
(0.002) ***
Age
−0.064
(0.793)
−0.081
(0.741)
−0.172
(0.477)
−0.178
(0.462)
GenderMale
3.121
(0.279)
−0.197
(0.946)

0.135
(0.964)
USParticipant
−11.692
(0.001) ***
−10.972
(0.001) ***
ReligionChristian
3.888
(0.291)
Adjusted R-Squared
0.016 0.011 0.011 0.049 0.049
F-Stat
8.747
(0.003) ***
3.284
(0.038) **
2.607
(0.051) *

6.407
(0.001) ***
5.351
(0.001) ***
N 469 425 424 424 424
Table 3. Logit Regression of Country Location and Choice to
Reveal Saboteurs.
DV=Reveal I II III IV
USParticipant
0.310
(0.115)
(0.387)
(0.057) *
0.494
(0.020) **
0.520
(0.018) **

Age
0.021
(0.227)
0.021
(0.252)
0.021
(0.269)
GenderMale
0.455
(0.044) **
0.465
(0.041) **
Religion: Christian
0.134
(0.633)
% of cases correctly predicted 60% 59% 58% 58%
N 446 417 416 416
significant at .05; * = significant at 0.1.

Statistical analysis did not support H4a and H4b, indicating that
there was no effect of individualist
versus collectivist cultural context on the tendency to avoid
decision-making. However, an individual’s
country of residence emerged as a factor that affected the
avoidance tendency. Individuals from Ghana
were more likely to want to avoid making the vignette-based
decisions altogether.
Finally, H5 was tested to determine if the familiarity of the
vignette setting had an effect on the
decision to compromise with the company management and take
an offer that met part but not all of
the protesting group’s stated objectives. Participants were asked
if the location of the vignette was
similar to where they lived. Analysis showed that the
coefficient for the familiarity of the cultural
context did have the hypothesized effect and was significantly
correlated to the decision to cooperate,
remaining so when controlled by age, gender, country, and
religion (p < 0.05). See Table 4 for a
summary of the results.
Table 4. Logit Regression Results for Similarity-of-Setting.
DV=Cooperate I II III IV V

LocationFeltSimilar
0.562
(0.037) **
0.634
(0.021) **
0.613
(0.029) **
0.742
(0.012) **
0.743
(0.013) **
Age
−0.044
(0.049) **
−0.044
(0.048) **
−0.038
(0.092) *
−0.039
(0.059) *
GenderMale

0.060
(0.797)
0.128
(0.596)
0.130
(0.588)
USParticipant 0.357 (0.133) 0.364 (0.133)
Religion: Christian 0.032 (0.916)
% of cases correctly predicted 67% 67% 68% 67% 66%
N 432 410 409 409 409
6. Discussion and Conclusions
Analysis of the data collected in the U.S. and Ghana indicate
that culture does indeed have an
impact on decision making, showing indications of cultural
impact across individual traits, societal
traits, and conflict setting. First, the two hypotheses regarding
personal traits were well-supported by

the data, indicating that cultural traits such as individualism and
collectivism can have an effect on the
general decision making style of individuals. Those who are
more individualist tend to be more
rational, and those who are more collectivist tend to be more
dependent. This is an important finding
for those who work in mediation and conflict resolution. When
attempting to bring parties together to
form an agreement, one ought to understand the cultural traits
of the parties involved and tailor one’s
approach accordingly. Our findings lead us to conclude that we
ought not to expect collectivists to
make long lasting decisions without the involvement of others
in their group or individualists to toss
aside the thoroughly researched and logical choice to go with a
snap decision.
Second, the nature of the society one lives in informs the
decision making process. This study
shows support for the third hypothesis that those from a
collectivist society (Ghana) are more likely to
prioritize interests of members of more central social identity
groups by, in this case, not revealing the
sabotage of associates. The country location emerged several
times throughout the study as a relevant

factor in determining decision making styles including rational,
dependent, and in some cases avoidant
thinking. This finding has particular relevance to peacekeeping
operations. Preparation and training for
peacekeeping missions should be built on an awareness of the
cultural traits of the target society, and
particularly on the central group identities informing the parties
to the conflict. Our research suggests a
tendency for collectivist societies to want to solve group
problems within the most central and salient
ingroup identities and a general reluctance to go to outer circles
to find a solution. This potentially
makes collectivist societies less prone to accepting external
intervention, and is an area worthy of
further research.
Third and finally, our analysis indicates that the cultural setting
of a situation can alter decision-making
processes, but not in all cases. The nature of the conflict setting
itself did not appear to have an impact,
showing no support for the fourth set of hypotheses. However,
the fifth hypothesis was supported,
showing an effect of the feeling of familiarity to a conflict
setting on the tendency toward peace. This
may be an indication that it is advisable to deploy peacekeepers

from collectivist societies to conflict
settings in collectivist societies because their familiarity will
aid their effectiveness. Similarly,
peacekeepers from individualist societies should be sent to
conflict settings in individualist societies.
The importance of aligning cultural traits of peacekeepers with
cultural traits of conflict societies needs
to be explored more thoroughly in future research.
While the survey method we used is effective for collecting
large amounts of standardized data
suitable for quantitative analysis, there is a minor limitation to
the generalizability of our data because
of the use of college students as previously discussed. However,
this limitation did not prevent us from
revealing the impact of culture on decision-making in our
particular experimental setting. A further
limitation of our approach is that the artificial nature of the
questionnaire poses potential threats to
validity. Having respondents self-report their perceptions
measured in terms of responses to
predetermined Likert-type statements limits the accuracy with
which people can respond. However, by
presenting all participants in a group with the same stimulus in
a controlled setting, experiments and

surveys allow for a level of standardization that is effective in
eliminating most concerns for reliability.
The field of international conflict theory has recognized the
value of considering cultural variation
in applications of conflict management such as cross-cultural
negotiations and interactive conflict
resolution, yet little research has been done to date to
understand the implications of cultural variation
on individual decision making processes. Decision theory has
marginalized the impact of cultural
dimensions to the cognitive process, assuming that findings are
universal to human behavior.
Conversely, theories of ethnoconflict emphasize “local common
sense”, or the implied cultural
constraints in a society, which can be part of the unconscious
cognitive processing. Understanding how
local common sense plays into the decision-making process
displayed during conflict situations can
play a key role in effective conflict transformation efforts.
The results of this research lead us to conclude that there are

cultural constraints that have an effect
on the decision making process. Consequently, the broader
assumption often made in the literature that
decision-making has universal characteristics must be tempered.
The cultural traits of individuals as
well as the general cultural characteristics of the society in
which they live factor into the decision
making process. This study has focused on quantitative
measures attempting to reveal often
subconscious factors such as one’s own traits and tendencies,
the salience of one’s group identities,
and the effects of where one lives. However, there is also likely
to be important information in the
conscious part of the decision-making process that could be
discovered simply by asking people “why”
they make certain decisions. Such a study would be able to
reveal trends in justification factors and
rationale that become salient when making decisions. This is a
prime area for future research that
could be used to begin building a conceptual model of the
decision-making process in conflict
situations from the point of view of cultural constraints.
Further research is also warranted in other locations of the
world. This study was limited to just the

U.S. and Ghana in an initial cross-country comparison.
Incorporating other countries that have
similarities and differences in key cultural dimensions would
allow for further comparisons and
refinement of the conceptual model. In addition, conducting
follow-on research on non-student
populations within the same country would allow for an
informative within-country comparison to
highlight cultural differences between populations aside from
the country location factor. For example,
it would be useful to study the decision-making of a population
experienced in the realm of
peacekeeping and conflict resolution to compare with the
typically inexperienced university student
population. How does the conflict decision-making approach
vary between these groups? What is the
effect of professional experiences, for instance in the military
or police, in a person’s decision-making
process and outcome? Future research should investigate
decision theory specifically in the context of
conflict resolution practices, peacekeeping strategy
development, and pre-deployment troop training to
assess the extent to which individuals with different
demographic, professional or identity

backgrounds make decisions on whether to fight or to cooperate.
Central to any such analysis is the
need to gain a thorough understanding of the deeper effects of
culture on cognitive processing and
decision-making. Only through such studies will we be able to
help in heeding Erick Barronodo’s call
to the youth of Guatemala, and understand why some youth
choose to pick up guns and knives and
others do not.
References
1. Maidment, N. Athletics: Choose Trainers Not Guns, Says
Barrondo. Reuters, 4 August 2012.
Available online: http://www.reuters.com/article/2012/08/04/us-
oly-athl-atm20w-barrondo-
idUSBRE8730M520120804 (accessed on 1 March 2012).
2. United Nations Verification Mission in Guatemala: Report of
the Secretary-General. United
Nations General Assembly, 2005, Document A/59/746.
Available online: http://daccess-dds-

ny.un.org/doc/UNDOC/GEN/N05/277/50/PDF/N0527750.pdf
(accessed on 1 March 2012).
3. Rubenstein, R.A.; Keller, D.M.; Scherger, M.E. Culture and
interoperability in integrated
missions. Int. Peacekeep. 2008, 15, 540–555.
4. Mann, L.; Redford, M.; Burnett, P.; Ford, S.; Bond, M.;
Leung, K.; Nakamura, H.; Vaughan, G.;
Yang, K. Cross-cultural differences in self-reported decision-
making style and confidence. Int. J.
Psychol. 1998, 33, 325–335.
5. Hofstede, G. Geert Hofstede Cultural Dimensions, 2012.
Available online: http://geert-
hofstede.com (accessed on 1 March 2012).
6. Fisher, R.J. Interactive Conflict Resolution; Syracuse
University Press: Syracuse, New York, NY,
USA, 1997.
7. Burton, J.W.; Sandole, D.J.D. Generic theory: The basis of
conflict resolution. Negot. J. 1986, 2,
333–344.
8. Avruch, K.; Black, P. The Culture Question and Conflict
Resolution. Peace Change 1991, 16,
22–45.

9. Lederach, J.P.; Wehr, P. Mediating conflict in central
America. J. Peace Res. 1991, 28, 85–98.
10. Rubenstein, R.E. Dispute resolution on the eastern frontier:
Some questions for modern
missionaries. Negot. J. 1992, 8, 331–346.
11. Moore, C.W.; Woodrow, P.J. Handbook of Global and
Multicultural Negotiation; Jossey-Bass:
San Francisco, CA, USA, 2010.
12. Cohen, R. Negotiating Across Cultures: Communication
Obstacles in International Diplomacy;
US Institute of Peace Press: Washington, DC, USA, 1991.
13. Riker, W.H. The political psychology of rational choice
theory. Polit. Psychol. 1995, 16, 23–44.
14. Schoemaker, P.J.H. The expected utility model: Its Variants,
purposes, evidence and limitations.
J. Econ. Lit. 1982, 20, 529–563.
15. Franke, V. Decision-Making Under Uncertainty: Using Case
Studies for Teaching Strategy in
Complex Environments. J. Mil. Strateg. Stud. 2011, 13.
Available online:
http://www.jmss.org/jmss/index.php/jmss/article/view/385
(accessed on 6 March 2013).
16. Simon, H.A. A Behavioral model of rational choice. Q. J.

Econ. 1955, 69, 99–118.
17. Berejikian, J.D. A Cognitive Theory of Deterrence. J. Peace
Res. 2002, 39, 165–183.
18. Gowda, M.V.R. Heuristics, biases, and the regulation of
risk. Policy Sci. 1999, 32, 59–78.
19. Kahneman, D.; Tversky, A. Prospect theory: An analysis of
decision under risk. Econometrica
1979, 47, 263–292.
20. Kurtz, C.F.; Snowden, D.J.T. The new dynamics of strategy:
Sense-making in a complex and
complicated world. IBM Syst. J. 2003, 42, 462.
21. Cook, P.J.; Clotfelter, C.T. The peculiar scale economies of
Lotto. Am. Econ. Rev. 1993, 83,
634–643.
22. Gordon, C.; Arian, A. Threat and decision making. J.
Conflict Resolut. 2001, 45, 196–215.
23. Miler, K.C. The limitations of heuristics for political elites.
Polit. Psychol. 2009, 30, 863–894.
24. Kanwisher, N. Cognitive heuristics and American security
policy. J. Conflict Resolut. 1989, 33,
652–675.

25. Tversky, A.; Kahneman, D. Judgment under uncertainty:
Heuristics and biases. Science 1974,
185, 1124–1131.
26. McGraw, K.M. Contributions of the cognitive approach to
political psychology. Polit. Psychol.
2000, 21, 805–832.
27. Lau, R.R.; Redlawsk, D.P. Advantages and Disadvantages of
Cognitive Heuristics in Political
Decision Making. Am. J. Polit. Sci. 2001, 45, 951–971.
28. Goertz, G. Constraints, compromises, and decision making.
J. Conflict Resolut. 2004, 48, 14–37.
29. Mintz, A. How do leaders make decisions? A poliheuristic
perspective. J. Conflict Resolut. 2004,
48, 3–13.
30. Stern, E. Contextualizing and critiquing the poliheuristic
theory. J. Conflict Resolut. 2004, 48,
105–126.
31. Keller, J.W.; Yang, Y.E. Leadership style, decision context,
and the poliheuristic theory of
decision making: An experimental analysis. J. Conflict Resolut.
2008, 52, 687–712.

32. Batchelor, C.A. Statelessness and the problem of resolving
nationality status. Int. J. Refugee Law
1998, 10, 156–183.
33. Armstrong, R.W. The relationship between culture and
perception of ethical problems in
international marketing. J. Bus. Ethics 1996, 15, 1199–1208.
34. Rubenstein, R. Cross-cultural considerations in complex
peace operations. Negot. J. 2003, 19,
29–49.
35. Hofstede, G. Culture’s Consequences: International
Differences in Work Related Values. Sage
Publications, Inc: Beverly Hills, CA, USA, 1980.
36. Kluckhohn, F.R.; Strodtbeck, F.L. Variations in Value
Orientations; Row, Peterson, and
Company: Evanston, IL, USA, 1961.
37. Thomas, D.C. Comparing Cultures: Systematically
Describing Cultural Differences. In Cross-
cultural Management: Essential Concepts; SAGE Publications,
Inc: Thousand Oaks, CA, USA,
2008; pp. 47–69.
38. Kirkman, B.L. A quarter century of culture’s consequences:
A review of empirical research

incorporating Hofstede’s cultural values framework. J. Int. Bus.
Stud. 2006, 37, 285–320.
39. Oyserman, D.; Coon, H.M.; Kemmelmeier, M. Rethinking
individualism and collectivism:
Evaluation of theoretical assumptions and meta-analyses.
Psychol. Bull. 2002, 128, 3–72.
40. Taras, V.; Kirkman, B.L.; Steel, P. Examining the impact of
culture’s consequences: A three-
decade, multilevel, meta-analytic review of Hofstede’s cultural
value dimensions. J. Appl.
Psychol. 2010, 95, 405–439.
41. Triandis, H.C. Individualism and collectivism as cultural
syndromes. Cross-Cult. Res. 1993, 27,
155–180.
42. Chen, X.; Li, S. Cross-national differences in cooperative
decision-making in mixed-motive
business contexts: The mediating effect of vertical and
horizontal individualism. J. Int. Bus. Stud.
2005, 36, 622–636.
43. Epstein, A. Ethos and Identity; Tavistock: London, UK,
1978.
44. Stryker, S. Identity Salience and role performance: The
relevance of symbolic interaction theory

for family research. J. Marriage Fam. 1968, 30, 558–564.
45. Hofman, J. E. Social Identity and Intergroup Conflict: An
Israeli View. In The Social Psychology
of Intergroup Conflict: Theory, Research and Applications;
Stroebe, W., Kruglanski, A.W.,
Bar-Tal, D., Hewstone, M., Eds.; Springer Verlag: New York,
NY, USA, 1988; pp. 89–102.
46. Tajfel, H. Differentiation between Social Groups: Studies in
the Social Psychology of Intergroup
Relations; Academic Press: London, UK, 1978.
47. Franke, V. Preparing for Peace: Military Identity, Value
Orientation, and Professional Military
Education; Praeger Publishers: Westport, CT, USA, 1999.
48. Abrams, D.; Hogg, M.A. Social Identity and Social
Cognition; Blackwell: Oxford, UK, 1999.
49. Hogg, M.A. Intergroup Processes, Groups Structure and
Social Identity. In Social Groups and
Identities: Developing the Legacy of Henri Tajfel; Robinson,
W.P., Ed.; Butterworth-Heinemann:

Oxford, UK, 1996.
50. Hogg, M.A.; Abrams, D. Social Identifications: A Social
Psychology of Intergroup Relations and
Group Processes; Routledge: New York, NY, USA, 1988.
51. Sherif, M.; Harvey, O.J.; White, B.J.; Hood, W.R.; Sheif,
C.W. The Robbers Cave Experiment:
Intergroup Conflict and Cooperation; Wesleyan University
Press: Middletown, CT, USA, 1988.
52. Sidanius, J. The psychology of group conflict and the
dynamics of oppression: A social
dominance perspective. In Explorations in Political Psychology;
Iyengar, S., McGuire, W.J., Eds.;
Duke University Press: London, UK, 1993.
53. Sidanius, J.; Haley, H. Person-organization congruence and
the maintenance of group-based
social hierarchy: A social dominance perspective. Group
Process. Interg. 2005, 8, 187–203.
54. Sniderman, P.M.; Brody, R.A.; Tetlock, P.E. Reasoning and
Choice: Explorations in Political
Psychology; Cambridge University Press: New York, NY, USA,
1991.
55. Spicer, D.P.; Sadler-Smith, E. An examination of the
general decision making style questionnaire

in two UK samples. J. Manage. Psychol. 2005, 20, 137–149.
56. Shulruf, B.; Hattie, J.; Dixon, R. Development of a new
measurement tool for individualism and
collectivism. J. Psychoeduc. Assess. 2007, 25, 385–401.
57. Mook, D.G. In defense of external invalidity. Am. Psychol.
1983, 38, 379–387.
58. Besley, T.; Reynal-Querol, M. Do Democracies Select More
Educated Leaders? Am. Polit. Sci.
Rev. 2011, 105, 552–566.
59. Shulruf, B.; Alesi, M.; Ciochina, L.; Faria, L.; Hattie, J.;
Hong, F.; Pepi, A.M.; Watkins, D.
Measuring collectivism and individualism in the third
millennium. Soc. Behav. Personal. 2011,
39, 173–188.
Appendix A: Reduced Auckland Individualism Collectivism
Scale (AICS) Questionnaire
The following questionnaire is used to measure individualist
and collectivist traits. Items are
indicated with I or C to indicate they are part of the
individualist or collectivist index respectively.
Each question is implemented on a scale of strongly agree,
agree, disagree, strongly disagree, or
don’t know.

1. I define myself as a competitive person. (I)
2. Before I make a major decision I seek advice from people
close to me. (C)
3. I believe that competition is part of human nature. (I)
4. I consider my friends’ opinions before taking important
actions. (C)
5. I like to be accurate when I communicate. (I)
6. It is important to consult close friends and get their ideas
before making a decision. (C)
7. I ask the advice of my friends before making career related
decisions. (C)
8. I sacrifice my self-interest for the benefit of my group. (C)
9. I prefer using indirect language rather than upset my friends.
(C)
10. I take responsibility for my own actions. (I)
11. My personal identity independent of others is very
important to me. (I)
12. Winning is very important to me. (I)

13. I see myself as “my own person.” (I)
14. I consult my family before making an important decision.
(C)
Appendix B: Reduced General Decision Making Style (GDMS)
Questionnaire
The following questionnaire is used to measure general decision
making styles. Items are indicated
with an R, I, D, A, or S to indicate they are part of the rational,
intuitive, dependent, avoidant, or
spontaneous index respectively. Each question is implemented
on a scale of strongly agree, agree,
disagree, strongly disagree, or don’t know.
1. When I make decisions, I tend to rely on my intuition. (I)
2. I rarely make important decisions without consulting other
people. (D)
3. When I make a decision, it is more important for me to feel
the decision is right than to have a
rational explanation for it. (I)
4. I double-check my information sources to be sure I have the
right facts before making decisions. (R)
5. I make decisions in a logical and systematic way. (R)
6. When making decisions I do what feels natural at the
moment. (S)

7. I like to have someone steer me in the right direction when I
am faced with important decisions. (D)
8. My decision making requires careful thought. (R)
9. When making a decision, I trust my inner feelings and
reactions. (I)
10. When making a decision, I consider various options in terms
of a specified goal. (R)
11. I avoid making important decisions until the pressure is on.
(A)
12. I often make impulsive decisions. (S)
13. I often need the assistance of other people when making
important decisions. (D)
14. I often put off making important decisions. (A)
15. If I have the support of others, it is easier for me to make
important decisions. (D)
16. I generally make important decisions at the last minute. (A)
17. I make quick decisions. (S)
© 2013 by the authors; licensee MDPI, Basel, Switzerland. This
article is an open access article
distributed under the terms and conditions of the Creative
Commons Attribution license
(http://creativecommons.org/licenses/by/3.0/).

Pol-03 discussion (two pages add 10)
Sidney Milkis, in "The President and Political Parties," states
that "the relationship between the presidency and the American
party system has always been difficult."
Apply this concept to a president and analyze why this was the
case.
Pol- Research paper 8 pages add (45)
Please submit your final 3000-3500 word research paper.
CITATION AND REFERENCE STYLE
Assignments completed in a narrative essay or composition
format must follow the Turabian Style Manual. The following
link may
help: http://www.press.uchicago.edu/books/turabian/turabian_ci
tationguide.html
Possible topics for the paper include:
-War Powers Act
-executive privilege
-executive orders
-divided government
-White House Staff
-the Cabinet
-Congressional supremacy
-Congressional committee structure

-the Speakership
-Congressional leadership
-legislative veto
-legislative controls on the bureaucracy
-Congressional budget process
-government corporations
-regulatory
Originality of attachments will be verified by Turnitin.com.
Both you and your instructor will receive the results.
Supporting Materials
· ChicagoManualofStyleSheet2010.doc (60 KB
Group Decis Negot (2012) 21:79–98
DOI 10.1007/s10726-010-9190-x
Cultural Differentiation of Negotiating Agents
Gert Jan Hofstede · Catholijn M. Jonker ·
Tim Verwaart
Published online: 26 February 2010
© Springer Science+Business Media B.V. 2010
Abstract Negotiations proceed differently across cultures. For

realistic modeling
of agents in multicultural negotiations, the agents must display
culturally differenti-
ated behavior. This paper presents an agent-based simulation
model that tackles these
challenges, based on Hofstede’s model of national cultures. The
context is a trade
network for goods with a hidden quality attribute. The
negotiation model is based on
the ABMP negotiation architecture and applies a utility function
that includes market
value, quality preference and risk attitude. The five dimensions
of Hofstede’s model
are the basis for the modification of ABMP parameters and
weight factors in the util-
ity function. The agents can observe each other’s group
membership and status. This
information is used, along with the indices of Hofstede’s
dimensions, to differentiate
behavior in different cultural settings. The paper presents
results of test runs that ver-
ify the implementation of the model. The model helps to explain
behaviors of actors
in international trade networks. It proves that Hofstede’s
dimensions can be used to
generate culturally differentiated agents. Further validations of
the model with case
studies from literature and experiments have yet to be
conducted. Extensions can make
this model a useful tool for training traders who engage in
cross-cultural negotiation
and for implementation in negotiation support systems.
G. J. Hofstede
Wageningen University, Hollandseweg 1, 6706 KN
Wageningen, The Netherlands

C. M. Jonker
Delft Technical University, Mekelweg 4, 2628 CD Delft, The
Netherlands
T. Verwaart (B)
LEI Wageningen UR, Postbus 29703, 2502 LS Den Haag, The
Netherlands
e-mail: [email protected]
123
80 G. J. Hofstede et al.
Keywords Multi-agent simulation · Culture · Bargaining ·
Negotiation ·
Trade network
1 Introduction
Anybody with experience in international trade knows that
bargaining practices differ
across the world. Multinational companies sometimes work with
different price lists
for different countries. To give just one example: whereas
German buyers want to
know exactly how much the products cost, Arabs need to have
room for bargaining.
In order to sell at the same price, the selling company needs to
adapt its offer to the
varying bargaining practices. This means that a single piece of
advice about how to
bargain, or a single model to describe bargaining, are obviously
not valid across the
world unless culture is taken into account.

‘Culture’ is a notion with many meanings, some of which are
contested in some
disciplines. However, the leading paradigm today is widely
accepted and used in both
practice and academia. According to it, culture refers to the
unwritten rules of society.
Culture is that which makes a group cohesive by smoothing
communication. It is a
phenomenon that is specific to a group, not to an individual. Its
essence is unconscious
shared systems of values, and it is transmitted in early youth
through example and edu-
cation. As a result it is stable across centuries in spite of huge
changes in environment
and technology. Cultural differences show no signs of
diminishing in the Information
Age (Hofstede and Hofstede 2005).
Within the literature various basic dimensions can be found
according to which
societies differ from one another. Of these, the most widely
used is Hofstede’s model
(Hofstede 2001), (Hofstede and Hofstede 2005). His work is
accessible, sparse, and
based on a very large, very well stratified sample that continues
to give it great explan-
atory value. No other model matches society-level variables so
well to date (Smith
2004).
This paper describes an agent-based model for bargaining in the
context of trade.
The agents follow common sense strategies such as maximizing
gain, seeking good
quality, and minimizing risk. But they also have models of how

to behave in an appro-
priate manner. These models are based on Hofstede’s five
dimensions of culture. The
challenge that we take up is the one posed by De Rosis et al.
(2004), who suggested to
investigate the feasibility of Hofstede’s model for building
culturally consistent agent
characters. An agent-based model of bargaining in which the
agents are cultured offers
several promises. It can help understand the dynamics of
international negotiations in
trade. It could also serve as a training tool for aspiring
international traders.
The paper first briefly introduces Hofstede’s model of five
dimensions of culture.
Next, the ABMP (Agent-Based Market Place, Jonker and Treur
2001) negotiation
model that we adopt is presented. We show how this model can
be used in agent-
based simulations. We also discuss the limited subset of
negotiation situations that
are considered in this article. In the third section we link culture
and negotiation by
describing the influence of each of Hofstede’s dimensions of
culture on negotiators’
practices and preferences. This section sets the scene for the
presentation of the rules
for our cultured agents in the fourth section. Section five shows
example runs with the
123
Cultural Differentiation of Negotiating Agents 81

model and discusses them. Finally we discuss the model and
how to proceed, since
this model forms the basis of future research and tools.
2 Hofstede’s Five Dimensions of Culture
Each human society has found a different pattern of response to
the problems of social
life. In some societies, groups are permanent and close-knit
while in others, group
membership is volatile and voluntary. In some, leadership style
is usually autocratic
and in others, participative. Research has shown and repeatedly
confirmed that basic
tendencies to deal with a few central issues of social life are
stable across the genera-
tions in societies (Hofstede and Hofstede 2005). They are,
because they are instilled
into a society’s members from birth. As a baby and as a toddler,
a child is primed as
a social being. Once a child sets foot into the wider society as a
teenager, its basic
cultural orientation is firmly in place.
This research stream has led to dimension models of culture.
The most widely
used of these is the five-dimension model by Hofstede. The five
dimensions are about
five issues that relate to our basic drives. They will be
introduced briefly in order to
use them further on in the text. Note that these are not
personality traits, but societal
patterns! Also note that the picture drawn here is necessarily
simplified. It presents
the two caricatured extremes of each dimension. In reality,

almost all cultures have
intermediate positions on almost all dimensions. The
dimensions are introduced in the
following subsections.
2.1 Collectivism Versus Individualism
This dimension is about affiliation. To a collectivist (e.g., East
Asian, most non-Western
countries) mindset, fixed membership of a single group in which
all members are inter-
dependent is the natural state of being human. No member of
the natural group can be
cast aside. This means that maintaining harmony is crucial.
To an individualist (e.g., North-American, North-west
European) mindset, self-
sufficiency is the natural state of being. Everybody should be
judged in the same way,
whether or not the person is a group member. Honest people
speak their minds, even
if that means open disagreement.
2.2 Hierarchy: Large Versus Small Power Distance
This dimension is about dominance as an ascribed quality. It has
to do with author-
ity as seen from below. Are parents, teachers, priests and bosses
held in awe, and
is autocratic leadership expected? Then we have a society of
large power distance
(e.g., China, India, Russia.).
Or is leadership a role that could change from one person to
another with ease,
and are all people born equal? In that case, the society is one of

small power distance
(e.g., Anglo and Germanic countries).
123
2.3 Aggression and Gender: Masculinity Versus Femininity
This dimension is about assertive dominance, about muscle
power, and about the emo-
tional roles of the two sexes. In what is called a masculine
society (e.g., Japan, Anglo
countries), men in particular are supposed to be fighters.
Women are supposed to be
cheerleaders to the men’s fight—but they have to be tough too.
Men try to look and act
like real men and women try to look and act like real women.
These are tough societies,
with strong-handed police and military and with heavy
punishment for offenders.
In what is called feminine societies (e.g., Scandinavian
countries), both men and
women are supposed to be peace-loving and consensus seeking
and their social behav-
iors are not strongly different. Criminals should be helped, not
punished.
2.4 Otherness and Truth: Uncertainty Avoidance
This dimension is about how to cope with the unknowable.
Some societies are termed
uncertainty avoiding (e.g., Arab, Latin and Slavic countries).

They tend to have strict
rules and rituals about things that are strange or different, such
as religious rules and
food taboos, or strange sexual practices. In these societies, the
distinction between
clean and dirty is important. In fact they feel that any
distinction should be a sharp
one. They are concerned about right and wrong, about theory,
about arguing for its own
sake. They like to show their emotions, particularly anxiety,
verbally and non-verbally.
Other societies are termed uncertainty tolerant (e.g., Anglo
countries, China, Scan-
dinavia, Vietnam). They are relaxed and curious about strange
things and people, and
not worried about establishing strict classification schemes for
everything. They value
exploratory behaviors and novel experiences, and a relaxed
communication style.
2.5 Short Versus Long-Term Gratification of Needs
This dimension is about all the basic human drives. Which drive
should get precedence,
one that presses now or one that might become pressing in ten
years? Some societies
live for today, and these are termed short-term oriented.
Behaving in an appropriate
manner and respecting conventions is important in these
societies, as well as ‘keeping
up with the Joneses’ as the Americans have it. There are strong
opinions about good
and bad, and these are believed to be immutable.
Other societies live for the future; these are termed long-term

oriented (e.g. China,
Japan). Reasoning is pragmatic, and principles are adapted to
context. Good and bad
are seen as complementary and changeable. Planning, foresight
and perseverance are
valued. On the downside, this could lead to stinginess and
calculation.
2.6 Five Dimensions, One World
So far in this text, the dimensions of culture have been isolated
from one another in an
artificial way. In reality, cultures have a recognizable feel to
them, a Gestalt that can
be described, albeit only roughly, by its combination of
dimension scores. The five
123
dimensions are no more than abstractions that capture main
behavioral trends. Cultures
have ‘Gestalts’ of behavior. Experienced negotiators know the
range of behaviors that
they can expect from negotiators from other parts of the world.
They also know how
gender, age, status and personality can affect the negotiation
style of people from these
parts of the world.
In Hofstede et al. (2006, 2008a,b,c, 2009) the influence of each
of the dimensions
on trade processes was modeled separately; a slightly artificial,

but also necessary
intermediate step to model agents differentiated along the
Hofstede dimensions. Rec-
onciling these dimensional models into one believable model
that shows the ‘whole
negotiator’, although still abstracting from personality, is the
aim of this article.
3 Negotiation
In bilateral negotiation, two parties aim at reaching a joint
agreement. They do so by
exchanging various offers or bids using e.g. an alternating
offers protocol (Osborne
and Rubinstein 1994) called the “negotiation dance” in Raiffa et
al. (2002). Negoti-
ation is a complex emotional decision-making process aiming to
reach an agreement
to exchange goods or services (Thompson 2005).
3.1 Agent Models for Negotiation
The literature on automated negotiation contains a number of
agent models for nego-
tiation. The focus of that literature is on reaching deals that are
Pareto-efficient (i.e.,
neither can improve without making the situation worse for the
other). Furthermore,
some aim at reaching fair outcomes, i.e., in which the deal is
equally good for both
parties. The strategies differ in whether or not they take
knowledge about the domain,
and/or opponent into account. Examples of strategies that do not
use any domain
or opponent knowledge can be found in Faratin et al. (1998) and
Jonker and Treur

(2001). Other strategies try to learn the opponent’s preferences,
see e.g., (Coehoorn
and Jennings 2004) and (Hindriks and Tykhonov 2008). The
work presented in this
paper aims to develop models of actual human behavior. It does
not aim to develop an
optimal bargaining strategy that can outperform human
negotiators or other agents.
3.2 Focus on Interpersonal Bargaining
This work focuses on a specific type of negotiations: two
persons bargaining about
business transactions. Gaming simulations form the context of
the bargaining ses-
sions. The gaming simulations are designed as tools in supply
chains and networks
research (Meijer et al. 2006). In this setting participants
negotiate a transaction of a
commodity with either base quality or a superior quality. The
real quality is known to
the supplier and invisible to the customer. The customer can
either trust the supplier’s
quality statement, or request third-party testing at the cost of a
fee. A customer may
negotiate that the supplier provide certified quality by third-
party testing and have the
commodity delivered with the test report as a certificate. So, the
relevant attributes for
comparing bids are price, quality, and certification.
123

If the quality is above base level and the transaction is not
certified, the customer
is exposed to the risk of supplier’s opportunism. The valuations
of quality and risk
have a rational component that can be calculated from market
value and probabil-
ity of deceit. Furthermore, they have a subjective component
that is influenced by a
trader’s personality and culture. The rational component of the
valuation of quality is
the difference in market price with the price of base quality.
The rational component
of the valuation of risk is the product of probability of deceit
and value of the quality
attribute. The subjective valuation comes in addition to the
rational value. For quality,
it is the trader’s quality preference, for instance because of the
societal status that
results from trading high quality products. For risk, it is an
agent’s risk aversion. To a
risk-averse agent, the absence of risk has a value by itself.
In a culturally homogeneous society, not all agents have equal
quality preference
and risk aversion. However, significant differences between
cultures exist in the aver-
age values of these parameters.
3.3 Agent-Based Market Place (ABMP) and its Application in
the Agents
For the agents’ negotiation strategy we chose ABMP (Jonker
and Treur 2001), because
its similarity to human negotiations has been validated (Bosse et
al. 2004). The ABMP

process is an exchange of bids, starting with a bid by one of the
partners. The other
partner evaluates the bid using a utility function that maps a
weighted linear combi-
nation of bid attributes to the interval [0, 1]. The weight factors
in the utility function
represent an agent’s preferences. The utility function used in
this research is elaborated
in a following subsection.
ABMP is a concession strategy. An agent prepares a bid that is
a concession to
its previous bid. Concession factor γ and negotiation speed β are
the parameters that
govern the concession making.
Concession factor γ is the fraction of the opening bid’s utility
that the agent is
willing to give in during the negotiation. It determines the
minimum utility that is
acceptable to an agent, also called the reservation value.
Negotiation speed β is the fraction of difference between the
agent’s previous bid
and the minimum utility that an agent uses to determine the
target utility of its next bid.
After calculation of the utility of a partner’s bid and the target
utility of its own next
bid, the agent decides whether to accept partners bid or not,
governed by the utility
gap parameter ω.
Acceptable utility gap ω is the maximal difference between own
target utility and
last partner’s bid’s utility for which an agent will accept

partner’s bid.
If the target utility minus the partner’s last bid’s utility is
greater than the acceptable
utility gap, the agent does not accept and has to decide about its
next action. It can
terminate the negotiation for several reasons. First, partner’s
bid may be interpreted
as unrealistic if its utility is too far below the minimum utility.
Second, an agent may
be unsatisfied by the progress in partner’s bids. Third, there
may be no more room for
123
a substantial change of attributes to make a bid with the target
utility. In the latter case
the agent terminates the negotiation. In the first two cases the
probability that an agent
terminates the negotiation depends on the impatience parameter
ι.
Impatience ι is the probability that an agent will terminate the
negotiation if (a) the
utility of partner’s bid is less than the cut-off value or (b)
progress in the last three
rounds is less than the minimal progress required. In the present
model the cut-off
value φ is computed from minimum utility μ:
φ = (1 − ι)μ.

Minimal progress ϕ over three rounds of negotiation is
computed as
ϕ = (1 − μ)ι.
So, the decisions whether to accept a bid or not and whether to
continue or not depend
on partner’s bid, own last bid, partner’s progress, and the values
of parameters γ , β, ω,
and ι. Evaluation of bids involves the utility function discussed
in the next subsection.
This utility function is also used in planning a new bid that has
the target utility, taking
the agent’s quality preference and risk aversion into account.
3.4 The utility function
The agent model applies a utility function as proposed by
Tykhonov et al. (2008):
U (b) = wv V (b) + wq Q(b) + wr R(b),
with 0 ≤ wi ≤ 1, i = v, q, r, and
∑
wi = 1.
V (b) represents the business value of a bid. A customer agent
calculates it as
Vc =
1 − v + q − rc
2
,
with v representing the price of the bid, mapped to the interval

[0, 1] (0 represents
minimal market price for base quality of the commodity; 1
represents maximal market
price for top quality); q in [0, 1] (0 represents base quality; 1
top quality); customers
cost of risk is calculated as
rc = (1 − c)(1 − t )q,
with c = 1 representing presence and c = 0 absence of a quality
certificate; t rep-
resents the customers trust in the supplier, defined as the
customer’s estimate of the
probability that the supplier will cooperate and deliver
according to contract, even if
the supplier has the motive and the opportunity to defect.
123
For a supplier the business value of a bid is calculated as
Vs =
1 + v − q − rs
2
,
with suppliers cost of risk
rs = c f,
where f stands for the certification fee scaled to the same ratio

as v.
Both customers and suppliers may have a preference in excess
of the market value
for dealing top quality rather than base quality products. A
trader’s preference for
dealing top quality, even if profits from base quality trade are
superior, is represented
by wq ; in the present model Q(b) is computed as:
Q = q − 0.5.
Some traders may be risk-averse, in which case wr is positive.
In the present simulation
suppliers are informed about the actual quality level, so
Rs = 0.
For risk-avoiding customers the absence of risk may have a
value in itself, which is
represented as follows in the present model:
Rc = 0.5 − rc.
4 Culture and Bargaining
Hofstede et al. (2006, 2008a,b,c, 2009) modeled the influence
of culture on trade
processes for each of the five dimensions separately. From these
papers, the narra-
tive descriptions of the influences of the dimensions on trade
negotiations—i.e. the
bargaining about transactions—are cited below.
4.1 Power Distance (Hofstede et al. 2009)

Traders from egalitarian cultures may have different ways to
negotiate, but they will
always negotiate. Traders from large power distance cultures on
the other hand are
not used to negotiate seriously. The powerful dictate the
conditions. The less powerful
have to accept. In feminine or collectivist cultures the powerful
may exercise restraint,
or the lower ranked may successfully plead for compassion, but
this is not a joint deci-
sion making process like a negotiation is. The higher ranked
partner decides. When
people from hierarchical cultures are forced to negotiate,
because they are in a position
of equal status or trade with foreigners, the negotiations often
end in a game of power.
The higher ranked in hierarchical societies prefer top quality
commodities to stress
their position. They accept risk, because they do not expect the
lower ranked to deceive
123
them. The lower ranked on the other hand, avoid risk and
protect themselves by settling
for base quality commodities.
A trader from a culture with large power distance expects a
lower ranked busi-
ness partner to accept his conditions rapidly. If the lower ranked
partner has the same

cultural background, there is no problem and the rights of the
higher ranked will be
recognized and respected: the lower ranked will be modest and
give in easily. How-
ever, a trader from an egalitarian culture will not give in to the
pressure if his status
is lower, but will either react furiously (e.g., break off
negotiations) or simply ignore
the pressure (make a counterproposal), in which case the
opponent will be furious.
If a trader from a culture with large power distance negotiates
with a foreigner and
assumes the foreigner to have a higher status, he may give in
more easily than the
foreigner expected. In that case the foreigner may be happy, but
his opponent will
not have his fair share. If both are from hierarchical cultures but
do not perceive one
another’s hierarchical position they may make misattributions
resulting in one of them
being dominated or stopping the negotiations.
4.2 Uncertainty Avoidance (Hofstede et al. 2008b)
Uncertainty avoiding traders have an emotional style of
negotiation, making sure that
the opponents understand their feelings. They will not adapt
their behavior to their
opponent’s. They are quality-minded and avoid risk in business
transactions, espe-
cially when dealing with strangers. For uncertainty avoiding
traders, time is money.
They want to go directly to their target, and are impatient. After
a few unsuccessful
iterations, the uncertainty avoiding trader will break off the

negotiation.
Uncertainty tolerant traders on the other hand have a relaxed
style of negotiation.
They try to adapt their behavior to their counterparts, although
they are not prepared
to come to an agreement at all cost. They do not show their
emotions and may be
disconcerted if their opponents do. They are careful not to be
more yielding than their
counterparts are, not especially modest, and are ready to break
off negotiations in case
of insufficient progress.
4.3 Individualism and Collectivism (Hofstede et al. 2008a)
For a collectivistic trader negotiation has to be preceded by the
formation of a relation-
ship. If that goes wrong there will be no negotiation. During the
negotiation, collectivist
traders discriminate between in-group and out-group partners.
They feel obliged to be
more conceding to an in-group partner, are more hesitant to
break off negotiations with
in-group partners, and will try to maintain harmony as long as
the opponent follows
the in-group rules. Breaking the rules asks for a reaction. The
style of that reaction
may be furious, or they might never explicitly say anything, but
just avoid the other
from now on. The reply to a proposal from an in-group partner
will be modest, but
there is no need to be modest to an out-group partner. If an out-
group partner replies
with no or small concession, negotiation is likely to be broken
off, where an in-group

partner or an acquainted relation would get a second chance.
123
Responsibility for in-group welfare and compliance with in-
group rules always
play a prominent role in a collectivistic culture. A collectivist
will accept benefits for
his in-group rather than his personal advantage as a convincing
argument.
Individualists have one thing in mind during negotiations: their
own personal inter-
est. Depending on their personality and incentives, this might be
the material advantage
of the deal in question, or the development of new long-term
trusting relations with
perspectives of future deals, or just the pleasant conversation
during the negotiations,
or the satisfaction of winning the game, but one thing stands for
sure: individualists
pursue private interests. So individualist traders are not very
modest in their negoti-
ations, nor will they give in for the purpose of maintaining
harmony. If they are not
aware of the cultural differences when trading with
collectivists, they may be upset
by the lack of explicit communication, or they may upset their
opponents by being
too explicit, or by talking business before the relationship has
been established and
acknowledged. They are not particularly patient or impatient

negotiators, but behave
patiently as long as it serves their interest.
4.4 Masculinity and Femininity (Hofstede et al. 2006)
The dimension of masculinity versus femininity can be
interpreted as a preference for
performance versus cooperation. A performance oriented trader
(masculine culture)
is interested in fast trades, with as many top quality goods as
possible in one trade.
This trader is rather impatient, and if bids are too far off from
his profile, he will walk
away quickly. The performance oriented sticks to the contract
of the deal, deceive the
trade partner to the limits of the contract without any
compunction, and expects the
partner to do so too. Each subsequent negotiation will be dealt
with without taking
past trustworthiness into account. Each new contract will be set
up from scratch. The
trader learns from mistakes to make sure that the contract will
not lead to new and
uncomfortable surprises on his side.
A cooperation oriented trader (feminine culture) is interested in
the relationship with
the trade partner; building trust is important. The amount of
goods or quality level is
not of the most interest, because the relationship built during
negotiation might pay off
in future negotiations. Given the interest in the relationship
with the trade partner, a
first negotiation with a trade partner will take time that is
willingly spent by the trader.
During such negotiations, the trader appreciates a negotiation

process in which both
partners show a willingness to accommodate the other over
time. Past negotiations do
play an important role in subsequent negotiations. The trader is
perfectly willing to
see the current negotiation as a kind of continuation of the
previous one. If the trade is
about the same kind of commodity, the trader will start the
negotiation from the deal
of the last one. If the other accepts, then the deal can be made
in one round and in
seconds, whereas the first deal might have taken a lot of rounds
and lots of time.
4.5 Long- Versus Short-Term Orientation (Hofstede et al.
2008c)
Long term oriented negotiators are pragmatic and take the
bigger picture. They tend
to see one bargaining instance as a small step in a long process,
and their decisions
123
will be led by their estimation of the profitability or other
success chances of that
longer process. Long-term oriented traders show patience. They
do not rapidly break
off negotiations. They do not overcharge, but they do not
rapidly give in.
Short term oriented negotiators, on the other hand, think in

terms of moral princi-
ples and apply them to the situation that is before them here and
now. They are very
reliable when it comes to following standards of
appropriateness of behavior, but this
can make them disregard the ulterior consequences of their
actions. They are conced-
ing and patient with high-status partners and do not show them
distrust. Otherwise
they follow an opportunistic quality strategy.
5 Modeling Culture in ABMP
The model of the effects of culture on ABMP parameters and
utility weight factors is
based on the narrative descriptions in the preceding section. The
descriptions indicate
if a parameter is to be increased or decreased along each of
Hofstede’s dimensions.
Table 1 summarizes the direction of the effects (increasing
versus decreasing).
Some cultural dimensions have a direct effect on the parameter
values, but in other
cases the influence depends on the relationship with the partner:
– The societal status of an agent and that of its partner affect
behavior in societies
where power distance matters.
– Also in short-term oriented societies, partner’s status is
relevant.
Table 1 Influence of culture on the utility weight factors and
ABMP parameters

Culture type Conditions wq wr γ β ω ι
Large power d. Self status high +
Self status low −
Higher partner + + +! −
Lower partner −
Small power d.
Uncertainty av. Similar partner + + + +
Different p. + ++ + +
Uncer. tolerant
Individualistic
Collectivistic Ingroup partner + −
Outgroup p. + −
Masculine + + + +
Feminine − − −
Long-t. oriented − −
Short-t. oriented General +
High partners + − + −
wq Quality preference; wr Risk aversion; γ Concession factor; β
Negotiation speed; ω Acceptable utility
gap; ι Impatience; + Increased parameter value; − Decreased; +!
Increased every negotiation round
123

– Members of uncertainty avoiding societies distrust strangers
more than people they
are familiar with.
– Common group membership and group distance are important
in collectivistic
societies.
In the model the agents are labeled with tags that indicate status
and group member-
ship. The tags are visible to other agents so that they can
estimate status difference
and group distance. The model combines effects of culture with
effects of status and
group membership.
The effects on negotiation parameters and weight factors are
modeled as follows.
(1) The Hofstede indices PDI, UAI, IDV, MAS, and LTO
position national cultures
on the five dimensions. They are known for many countries
(see, e.g., Hofstede
2001). Let H, A, I, M, and L, respectively, represent the indices
as real values
scaled to [0,1], so that H = 1 represent maximal power distance,
(1 − H ) = 1
represent maximal egalitarianism, etc. Thus, there are 10
cultural stereotypes.
(2) Status and group distance are represented as real values in
[0,1]. Where sta-
tus, status difference and group distance are relevant, the effect

is conditional
upon the value of a cultural index. For instance, the product (1
− I)g represent the
effect of group distance g in conjunction with collectivism (1 −
I); in a maximally
individualistic society, 1 − I = 0, so group distance g has no
effect.
(3) As indicated in Table 1, some of the cultural stereotypes
may have a positive,
monotonously increasing, effect on a particular parameter
value; other stereo-
types may have a negative, monotonously decreasing, effect.
(4) For each parameter there may be a set of positive effects and
a set of negative
effects. The joint effect of a set of effects working with equal
sign is weakly dis-
junctive, i.e. the joint effect equals the maximum of the effects
of the individual
dimensions1 (e.g., if we model statement “people from
uncertainty avoiding or
masculine societies prefer rapid negotiations”, then the joint
effect is the maxi-
mum of the effects of uncertainty avoiding and masculinity).
(5) The resulting joint negative effects are assumed to
compensate for joint positive
effects, vice versa: the effect on parameter x is the difference of
the joint positive
and joint negative effects.
ex = ex + − ex −.
One can, for instance, represent the effect “in hierarchical
societies parameter

x is increased in case of status difference, unless the society is
collectivistic or
feminine” as follows:
ex = ex ,H (si −s j )
[
H (si − s j )
] − max {ex ,(1−I )g [(1 − I )g] , ex ,1−M (1 − M)
}
,
1 Weak disjunction is taken as the formalism to combine effects
of cultural dimensions working in the same
direction, because it takes only the stronger of the dimensions
into account. This is to be preferred to, for
instance, linear combinations, or (weighted) geometrical
averaging, because in those cases a strong effect
of a high value of one dimension would be conditional upon
high values of the other dimensions. Weak
disjunction is also preferred to stronger forms of disjunction,
because combination of moderate values of
several dimensions must not have a strong joint effect.
123
Where ex ,D represent a function that computes the effect of D
on x. No actual
evidence for the form or the range of the functions is available.
In the current
model we assume linear relations with range [0,1] for ex ,D , so
the above example

would reduce to
ex = H (si − s j ) − max [(1 − I )g, 1 − M ] .
(6) Joint positive and joint negative effects deduced from Table
1 are presented in
Table 2.
(7) The actual value of x is assumed to be a function that maps
xT (the typical value
of x) and the effect ex to a range [x L , x H ], x L ≤ xT ≤ x H .
In the current model
we take a simple approach and use linear interpolation:
x = xT +
ex + |ex |
2
(x H − xT ) +
ex − |ex |
2
(xT − x L ).
(8) The utility weights wi , i = v, q, r , are to be normalized for
proper functioning
of ABMP. Culturally adjusted values of w′q , and w′r (see Table
2) are relative to
w′v = 1, so:
wi =
w′i
1 + w′q + w′r
.

(9) The culturally adjusted parameter values and weight factors
resulting from rules
(1)–(8) are used in the ABMP evaluation of bids, in the
decisions (to accept or
not; to continue or not; see Hofstede et al. 2006), and in the
planning of a new
bid. In each round of the negotiation, the parameters are
recomputed.
The next section presents results obtained from this model.
6 Test Runs
The models discussed in the previous sections, including the
ABMP architecture, are
implemented in a multi-agent simulation, where agents can
select trade partners, nego-
tiate business transactions with price, quality, and certification
as attributes, deliver
truthfully or opportunistically, have deliveries tested for
quality, and update beliefs
about partners according to experiences in negotiations and
testing. The simulation
environment is implemented in Cormas.2
The model was tested for correct implementation. The observed
variables were:
– The number of successful negotiations (i.e. terminated with a
contract) in runs of
200 time steps with a population of 8 supplier agents and 8
customers;
– The percentage of negotiations that failed, i.e. that were
terminated by one of the
agents before agreement was reached;

2 http://cormas.cirad.fr/indexeng.htm
123
http://cormas.cirad.fr/indexeng.htm
Table 2 Formulas for the effects of culture on negotiation
parameters; H, A, I, M, and L represent the
scaled Hofstede indices; si represent the agent’s societal status
in [0, 1]; sj partner’s status; group distance
g between the agent and partner is computed from agent labels,
with 0 representing minimal distance, 1
maximal distance; ρ represent the round number in the current
negotiation
x Increasing effect ex + of culture on x Decreasing effect ex −
of culture
w′q max (H si , A, M, 1 − L) max
[
H (1 − si ) , 1 − M, L
]
w′r max
⎡
⎣
H
(

s j − si
)
, A,
√
Ag,
(1 − I ) g, M
⎤
⎦ max
[
H
(
si − s j
)
, (1 − L) s j
]
γ max
⎡
⎣
H
(
s j − si
)
, (1 − I ) (1 − g) ,

(1 − L) s j
⎤
⎦
β max [ A, M ] max [(1 − I ) g, 1 − M ]
ω H
(
s j − si
)
ρ
ι max ( A, M) max
⎡
⎢
⎢
⎢
H
(
s j − si
)
, (1 − I ) (1 − g) ,
(1 − M) , L , (1 − L) s j
⎤
⎥

⎥
⎥
– The percentage of successful negotiations leading to top
quality transactions, as a
measure of willingness to accept risk.
In all test runs, the agents were set to be neutral with respect to
trust, i.e. they had no
information whether their partners were trustworthy or not (t =
0.5). The option of
certification was switched of, so that agents were forced to
accept risk if they bought
top quality products.
The following hypotheses about agents behavior in this
environment are formu-
lated on the basis of the narrative descriptions in this article’s
section on culture and
bargaining.
H1 In hierarchical societies (large power distance), high-ranked
agents buy top qual-
ity products; low-raked agents buy basic quality products to
protect themselves
from risk.
H2 In hierarchical societies, high-ranked agents are more
successful trading with
low-ranked than with equal-status partners, because lower-
ranked yield.
H3 In uncertainty avoiding societies, high quality products are
preferred, but agents
are risk-avoiding and impatient. Therefore, trade proceeds less
smooth in uncer-

Societies 2013, 3, 128–146; doi10.3390soc3010128 soci.docx

Societies 2013, 3, 128–146; doi10.3390soc3010128 soci.docx

Recommended

Recommended

More Related Content

Similar to Societies 2013, 3, 128–146; doi10.3390soc3010128 soci.docx

Similar to Societies 2013, 3, 128–146; doi10.3390soc3010128 soci.docx (20)

More from whitneyleman54422

More from whitneyleman54422 (20)

Recently uploaded

Recently uploaded (20)

Societies 2013, 3, 128–146; doi10.3390soc3010128 soci.docx