Curseu, Schruijer - The Effects of Framing on Inter-group Negotiation.pdf - Psychologia - koni86

Group Decis Negot (2008) 17:347–362

DOI 10.1007/s10726-007-9098-2

The Effects of Framing on Inter-group Negotiation

Petru Lucian Curseu

Sandra Schruijer

Published online: 19 October 2007

Abstract The present paper explores the way in which groups cognitively represent infor-

mation framed as danger and the way in which such collective cognitive representations

inﬂuence group performance during inter-group negotiations. One hundred and two partici-

pants were distributed over 34 three-person groups and were involved in a negotiation game

developed by Lewicki et al. (1999, Negotiation: readings, exercises and cases. McGraw-Hill,

Boston). The groups were organized in 17 pairs and each pair played the negotiation game

in two rounds. The game rules and the available resources were the same for both groups,

but one of the groups in each pair received the game information framed as “danger”, while

the other group in the pair received a neutral framing. The groups with a “danger” frame

developed a more defensive strategy during negotiations, adopted more often a collabora-

tive approach and had a significantly lower performance as compared to the groups in the

non-framing condition.

Keywords Inter-group negotiation

Framing

Cognitive representations

The framing of information as danger has a strong and pervasive effect on human behavior

because perceived threats induce fear and negative emotions (Jarymowicz and Bar-Tal 2006) ,

and because the human cognitive system is very sensitive to negatively framed information

(Fiske and Taylor 1991; Ito et al. 1998; Ito and Cacioppo 2005) , especially to information

signaling danger (Curseu 2003; Miclea and Curseu 2003) . Such a sensitivity serves an evo-

lutionary function (Baumeister et al. 2001; Miclea and Curseu 2003) . At the individual level

of analysis, the effects of this sensitivity have been extensively studied in a variety of tasks

from decision-making (Ito et al. 1998) to social judgments (Yzerbyt and Leyens 1991; Lupfer

et al. 2000; Ito and Cacioppo 2005) . However, the effects of negatively framed information

)

Department of Organisation Studies, Tilburg University, Room P1.161, Warandelaan 2, P.O. Box 90153,

Tilburg, NL 5000LE, The Netherlands

e-mail: P.L.Curseu@uvt.nl

S. Schruijer

Utrecht School of Governance, University of Utrecht, Utrecht, The Netherlands

123

P. L . C u r ¸seu (

348

P. L. Curseu, S. Schruijer

(e.g. information framed as danger) on information processing in groups received little or no

attention in the literature. In this paper we will empirically explore the effects of framing in

terms of danger on group decision making in inter-group negotiation situations.

The impact of framing on individual decisions has been extensively studied in the litera-

ture. Some scholars (Levin et al. 1998) rightfully argue that “the framing effect” is one of the

most proliﬁc areas in individual decision-making research. The framing effect stands for the

phenomenon that small phrasing changes of decisional alternatives, with similar expected out-

comes, induce a speciﬁc preference for one of the alternatives (Levin et al. 1998; Kühberger

1998; Tversky and Kahneman 1981) . One key construct in the explanation of the framing

effect is that of cognitive or mental representation.

Gingerenzer et al. (1991) consider the framing effect to result from the construction of a

mental model of the decisional situation. This model is built around knowledge describing

the decisional situation and the cognitive representations that are activated from the long-

term memory (LTM). According to this model, a decision results from a comparison between

the data describing the current situation and preexisting cognitive representations. A similar

model suggests that the framing effect occurs because the alternatives are embedded in a cog-

nitive causal schema (Jou et al. 1996; Olekalns and Smith 2005) . In this approach, the manner

in which the alternatives are presented activates congruent schemata from the LTM, schemata

that lead to a selective processing of the available information and ﬁnally to a decision-mak-

ing bias. Therefore, the causal schema provides the decision-maker with a referent about

alternatives and about the consequences of the alternatives, which will ultimately inﬂuence

the outcome of the decision-making process (Rohrbaugh and Shanteau 1999) or the outcome

of the negotiation (Olekalns 1994, 1997; Olekalns and Smith 2005) . Both explanatory mod-

els are similar in that they consider the role of cognitive representations in determining the

decisional output.

A particular situation in which the framing effect is important is the situation in which

the information is framed in negative terms inducing a so-called negativity bias (Fiske and

Taylor 1991; Ito et al. 1998; Ito and Cacioppo 2005) . The negativity bias refers to the higher

sensitivity of the human cognitive system to negatively framed information. In the process of

impression formation, expectancy disconﬁrming negative information overrules the effect of

disconﬁrming positive information (Lupfer et al. 2000) . For example, information that sig-

nals danger or threats outweighs the information that signals hope (Jarymowicz and Bar-Tal

2006) . Miclea and Curseu (2003) showed that individuals are more willing to allocate own as

well as external resources to address a situation which is framed as danger and is perceived as

a threat as compared to a situation framed as an opportunity or a problem solving situation.

Therefore, the human cognitive system seems more prepared to mobilize resources in order

to process negatively framed information.

Group decision-making is also affected by information framed as danger. Groups are

more willing to allocate external resources, and to get personally involved in solving an issue

framed as danger as compared to dealing with an issue framed as a problem or an opportunity

(Curseu 2003) . The classical framing effect has been replicated in group studies too (Paese

et al. 1993) , the explanation being that groups tend to overuse the shared information; group

discussions consist primarily of information that is held by all group members (Stasser and

Titus 1985, 1987) and shared individual preferences are accentuated in group settings (Lamm

1988) . The information sampling model (Stasser and Titus1985, 1987) and the group polar-

ization model (Lamm 1988) suggest that an individual tendency of processing information

in a particular way will most probably be accentuated in group settings. In other words, the

use of a general decision making causal schema held by the individual group members will

be accentuated in group settings. It is therefore expected that the negativity bias will hold at

123

Framing on Inter-group Negotiation

349

the group level too. The group as a socio-cognitive system is expected to be more sensitive to

information framed as danger and to develop defensive strategies to deal with the perceived

threat.

The concept of collective cognitive representation (Curseu 2003, 2006; Hinsz et al. 1997)

is essential in explaining the framing effect and the negativity bias at the group level. Within

a group, each member behaves according to the representations (schemata) he/she has about

the other group members, about the group’s task and the context of the group. The behav-

ior of each group member is a social stimulus for the other group members. This stimulus

affects the individual representations. When representations change, individual behavior also

changes. Thus the process of structuring the collective representations is a dynamic process.

We consider the moment that group members reach consensus to be the point when the rep-

resentation becomes stable. At that moment, the group will most probably make a decision or

identify a solution to a problem in line with this representation (Curseu 2006) . Groups store

these representations and use them to make decisions in similar situations. Using previously

developed representations about a situation reduces the time to reach consensus and make a

decision, but at the same time imposes constraints on the extent to which groups analyze the

available information.

When two groups negotiate, they exchange information that will be further on used to make

decisions within groups and develop strategies for further negotiations (Eden and Ackerman

2001; Olekalns 2002) . When one of the groups provides relevant information, the members

of the other party will use existing schemata to make sense of this information (Gingerenzer

et al. 1991; Olekalns and Smith 2005) . During group discussions, the group as a whole will

develop a collective cognitive representation about the situation. Cognitive schemata that are

congruent with this situation are activated, leading to a selective processing of the available

information. Schema-congruent information is analyzed in detail, while schema-incongruent

information is ignored and sometimes, when schema-congruent information does not exists

it is produced (Fiske and Taylor 1991) . Due to this selective information processing the

performance of groups may be impaired.

When information is framed in negative terms, information processing is preferred that

is consistent with this frame. It is therefore expected that during inter-group negotiations,

groups that receive the negotiation information framed as danger will develop more defensive

strategies as compared to groups not having received a danger frame. This is because the

former group members focus more on processing the information consistent with the danger

frame and the perceived threat. Focusing exclusively on the information that signals dangers

will most probably reduce the amount of relevant information processed and the number of

alternative strategies used in the negotiation. Therefore the hypothesis of this study is:

In inter-group negotiations groups that receive information framed as danger will adopt

a more defensive strategy and will have a lower performance in the negotiation than

the groups that receive the information framed in neutral terms.

We designed an experimental task in which two groups receive the same body of knowl-

edge yet framed differently. The groups were asked to interact in a negotiation exercise,

after carefully considering the available information. Our prediction was that the collective

representation developed by the two groups would differ and therefore the outcomes of the

negotiation game would be different. The hypothesis reﬂects the effects of framing on the

negotiation outcomes for the group that receives a danger framed information. Because in

each negotiation there are two groups involved (negotiation as a social interaction process

Olekalns 2002 ), we study the behavior of the other party involved in the negotiation in an

123

350

P. L. Curseu, S. Schruijer

exploratory way. Research on individual negotiations showed that the behavior of one party

inﬂuences the behavior of the other party involved in the negotiation (see for details the

meta-analysis of Druckman 1994 ). Therefore, one of the questions to be answered by the

present study is: does the defensive behavior of the framed group generate a more offensive

strategy and behavior of the other group?

1Method

1.1 Participants

The participants, 102 undergraduates (86 women) studying psychology at a Romanian

University (“Babes-Bolyai” University Cluj-Napoca), were distributed over 34 groups. They

were told they would participate in a negotiation exercise. We videotaped the meetings. Two

groups in two negotiation sessions did not agree to be ﬁlmed. In this case we made extensive

notes of their strategy and their negotiation behavior.

1.2 Procedure

The 34 groups were organized in 17 pairs. A negotiation pair consisted of two groups of three

members. At the beginning of the experimental session, some general information about the

negotiation game was given to both groups from each pair. They were told that they would

participate in a negotiation game and each of them received the rules of the game on an A4

format paper. The respondents were instructed to read the rules carefully and to develop a

group strategy how to approach the situation. There was no time constraint for developing

their strategy. The length of the strategy development process was between 1 and 2½ h. The

next step was the real negotiation.

The game is an adapted version of the “disarmament excercise” developed by Lewicki

et al. (1999) . One of the groups in a pair received the game information framed as danger,

while the other group received the information with a neutral frame. Basically, the decision

space was similar to both groups. The difference consisted only in the fact that half of the

groups received the information that they are a small country that only recently gained its

independence and that they are put into the situation to confront another group that represents

a big and powerful country. Apart from this, all teams received the same information. It was

explicitly stated that all groups have the same number of weapons, have the same rules and

the same rights during negotiations. It is important to mention here that although the framing

was done in terms of power (the term power was used in the game text), this is not a power

manipulation because in terms of absolute power (Wolfe and McGinn 2005) , both groups

had the same resources and the same alternatives. It was furthermore clearly stated that the

small country was independent.

The negotiation game was organized in two rounds. Each round consisted of seven steps

of 3 min length, during which the groups had to make decisions about the activation of

their “weapons”. There was a 30-s break between the steps, during which the groups could

call for negotiation or could declare an attack. Throughout the game, the groups recorded

their decisions at the end of each step. The experimenter played a mediating role during the

negotiations.

Each group could gain a certain number of points based on its own decision and the other

group’s decision. For each group, the maximum gain could be obtained when the group had

a higher number of activated weapons compared to the other group and declared attack. The

123

Framing on Inter-group Negotiation

351

most certain gain could be obtained if no group declared attack and if at the end of the seven

steps they had no weapons activated. Therefore, a normative analysis of the game proves that

disarmament and collaboration (deactivating as many weapons as possible) is the optimal

strategy for both parties since in this case, every party keeps its own initial number of points

and adds to that a certain number of points gained from a third party, in this case the World

Bank.

After the seven steps, the groups received feedback from the experimenter regarding their

gained points. Then they had the opportunity to revise their strategy for the next sequence of

the game, and they spent 30–60 min to reconsider their position. After that, the negotiation

game repeated itself under the same speciﬁed rules.

The mediator was informed about the strategy the groups were going to use. The group

decision-making was realized under time constraints. Each group had 3 min to decide about

the activation of their weapons during each step of the game. The detailed description of the

decision situation is presented in Appendix 1.

In sum, the independent variable of this study was the type of information received by

the groups (danger framing versus no-framing) and the dependent variable were: the num-

ber of points gained by each team, the number of attacks declared on the other group, and

the numbers of requested negotiations. In operational terms and related to the framing of

the hypothesis, a smaller number of attacks declared and a larger number of negotiations

requested are indicative for a defensive attitude, while the number of points gained is indic-

ative for group performance in the negotiation game. In order to check the content of the

group representations, the themes that emerged during the negotiations are also used in the

analyses. A quantitative coding strategy described by Weingart et al. (2004) was used for

the themes that emerged during negotiation. The results of these analyses were mostly used

for the exploratory part of our study. The ﬁrst author made a preliminary analysis of the

themes that emerged in negotiations, these themes were listed, and then both authors dis-

cussed and clustered them based on their content similarity in 15 general themes (for a list

see Table 2 ). The frequencies of these ﬁnal themes were then recorded for each game round.

2Results

The summary of the results is presented in Table 1 . Groups that received no framing gained a

significantly higher number of points than the other groups (Mann–Whitney test, Z

2.605,

009). The groups in the framing condition declared only three attacks in each of

the game rounds while the groups which received no framing declared 11 and 12 attacks,

respectively.

Groups that received the danger-framed information developed a more defensive strategy

compared to the no-framing condition. This conclusion is also supported by the analysis of

Ta b l e 1 The results for the two rounds of the negotiation game

Round

Group

Average points

Average activated

No. of

No. of negotiations

gained

weapons

attacks

requested

Round 1

Danger- framing

5.92

12.76

No-framing

10.54

14.58

Round 2

Danger- framing

5.68

13.41

No-framing

11.42

12.88

123

Curseu, Schruijer - The Effects of Framing on Inter-group Negotiation.pdf

Plik z chomika:

Inne pliki z tego folderu:

Inne foldery tego chomika: