mind the medium: a qualitative analysis of email negotiation

23
Group Decis Negot DOI 10.1007/s10726-014-9393-7 Mind the Medium: A Qualitative Analysis of Email Negotiation Jennifer D. Parlamis · Ingmar Geiger © Springer Science+Business Media Dordrecht 2014 Abstract Using qualitative analysis of email transcripts, this research investigated the behavioral differences in more or less successful email negotiations. We hypothesized that proactive and reactive medium management, relationship building, positive and negative emotion transmission along with integrative and distributive behaviors would influence joint gain and subjective value in email negotiation dyads. The hypotheses were tested on simulated buyer-seller email negotiations (n = 52 dyads) from a US and a German university. Ordinary least squares regression revealed that value creating behaviors and the total amount of communication increased joint gain while reactive medium management decreased joint gain. Controlling for individual gain and individ- ual target profit, negotiators’ global subjective value of the negotiation was negatively impacted by distributive negotiation behaviors and reactive medium management, as revealed by hierarchical linear modeling. Practical implications and future research are discussed. Keywords E-negotiation · Joint gain · Subjective value · Communication mode · Content analysis J. D. Parlamis (B ) School of Management, University of San Francisco, 2130 Fulton Street, San Francisco, CA 94117, USA e-mail: [email protected] I. Geiger Marketing Department, School of Business and Economics, Freie Universität Berlin, Otto-von-Simson-Str. 13, 14195 Berlin, Germany e-mail: [email protected] 123

Upload: ingmar

Post on 25-Jan-2017

214 views

Category:

Documents


1 download

TRANSCRIPT

Group Decis NegotDOI 10.1007/s10726-014-9393-7

Mind the Medium: A Qualitative Analysis of EmailNegotiation

Jennifer D. Parlamis · Ingmar Geiger

© Springer Science+Business Media Dordrecht 2014

Abstract Using qualitative analysis of email transcripts, this research investigated thebehavioral differences in more or less successful email negotiations. We hypothesizedthat proactive and reactive medium management, relationship building, positive andnegative emotion transmission along with integrative and distributive behaviors wouldinfluence joint gain and subjective value in email negotiation dyads. The hypotheseswere tested on simulated buyer-seller email negotiations (n = 52 dyads) from a US anda German university. Ordinary least squares regression revealed that value creatingbehaviors and the total amount of communication increased joint gain while reactivemedium management decreased joint gain. Controlling for individual gain and individ-ual target profit, negotiators’ global subjective value of the negotiation was negativelyimpacted by distributive negotiation behaviors and reactive medium management, asrevealed by hierarchical linear modeling. Practical implications and future researchare discussed.

Keywords E-negotiation · Joint gain · Subjective value · Communication mode ·Content analysis

J. D. Parlamis (B)School of Management, University of San Francisco, 2130 Fulton Street, San Francisco,CA 94117, USAe-mail: [email protected]

I. GeigerMarketing Department, School of Business and Economics, Freie Universität Berlin,Otto-von-Simson-Str. 13, 14195 Berlin, Germanye-mail: [email protected]

123

J. D. Parlamis, I. Geiger

1 Introduction

Email has long been considered a medium best suited for simple, concrete tasks ratherthan complex, nuanced communication (e.g., Daft and Lengel 1984, 1986). However,with a growing emphasis on a global marketplace, the expectation of constant con-nectedness, and the general ease of access, email has become a common mediumfor negotiations, a complex process of give and take requiring coordination and per-sonal interaction to arrive at a mutually acceptable agreement (Lewicki et al. 2010;Nadler and Shestowsky 2006). Indeed, research has shown that using email to con-duct negotiations can pose significant challenges and may lead to less than optimalsolutions and behaviors (e.g., Stuhlmacher and Citera 2005; Moore et al. 1999; Mor-ris et al. 2002). For example, individuals are more likely to deceive a counterpartwhen using email (Naquin et al. 2010), have more difficulty building trust and rapport(Galin et al. 2007; Morris et al. 2002) and experience greater hostility (Stuhlmacherand Citera 2005). Furthermore, outcomes such as joint gain, settlement success, andsatisfaction have been found to suffer when negotiations take place over email as com-pared to face-to-face (Galin et al. 2007; Naquin and Paulsen 2003). That said, otherevidence suggests that computer-mediated communication may be better for nego-tiations than once thought. Some research shows higher individual and joint gains(Stuhlmacher and Citera 2005; Croson 1999) and higher satisfaction in electroni-cally mediated negotiations as compared to face-to-face negotiations (Geiger 2012).Moreover, email, because it is asynchronous, reviewable, and revisable, may mitigatetime, linguistic and cultural differences leading to benefits for international businessnegotiations (Ambrose et al. 2008; Pesendorfer and Koeszegi 2006; Utz 2007).

While comparisons across communication modes (e.g., email, text, face-to-face,Skype) are essential to understanding the impact of technology on negotiationprocesses and outcomes, and help elucidate theoretical foundations that explain thecomplexities of technology and negotiations, research to date has been inconclusiveand equivocal. As such, a shift in research focus seems appropriate. Examining theemail communication mode explicitly may clarify previous findings and augment the-ory development. Importantly, while generally considered more problematic, not allemail negotiations end in impasse or have low levels of satisfaction. Determiningwhy this occurs or identifying behaviors that predict success in email negotiationscould lead to significant practical implications for e-negotiators. Indeed, researchershave begun to investigate successful and unsuccessful email negotiations (e.g.,Kersten and Zhang 2003; Hine et al. 2009) showing emotion and language use differ-ences across more and less successful e-negotiations (Hine et al. 2009) and offeringpractical implications.

Determining what makes email negotiations successful has been given limited atten-tion in the literature and is a primary aim of this research.

2 Theoretical Background

Most theories regarding the sub-optimal outcomes in email negotiation focus on char-acteristics of the communication environment that are lacking or costly. For exam-

123

Mind the Medium

ple email lacks co-presence, sequentiality, co-temporality, simultaneity, visibility,and audibility and may be costly in terms receiving, understanding, starting-up, dis-playing, or repairing communication, which all take more time and effort in email(Clark and Brennan 1991). Media richness theory and the task-media fit hypothesis(Daft and Lengel 1986; Trevino et al. 1987; Hollingshead et al. 1993; Straus andMcGrath 1994) are seminal theories often applied to help explain why email nego-tiations tend to yield less positive results than face-to-face negotiations. In partic-ular, these theories propose that text-based media, such as email, restrict the levelof non-verbal communication and limit the depth of information exchanged, lead-ing to greater possibility of misunderstanding and inhibition of the requisite personalconnections that help foster success in negotiations. Other theorists suggest that thedispute-exacerbating properties of email (such as excess attention, reduced socialcues, diminished feedback, and email length) contribute to conflict intensification by,among other things, weakening interpersonal ties and causing more aggressive tacticuse (Friedman and Currall 2003).

The email medium may be lean and lack characteristics of a face-to-face environ-ment, but recently researchers suggest (Geiger and Parlamis 2014) that negotiatorsmay encounter problems not because of differences in the inherent characteristics ofthe medium but rather due to the lack of suitability of email for all aspects of thecommunication process. Media Synchronicity Theory, as put forward by Dennis et al.(2008), suggests that most tasks where two parties interact to accomplish some goalrequire the use of conveyance processes (i.e., expression and transmission of infor-mation) along with convergence processes (i.e., arriving at a common understanding).In those authors’ view, different media are better suited for different processes (e.g.,email is better for conveyance and face-to-face communication is better for conver-gence) but both are important for interactions such as negotiation. Therefore, theyargue that complex tasks requiring both conveyance and convergence processes willbe most successful when individuals working together use multiple media.

If multi-media options are not available or possible for a negotiation, how canemail negotiators converge on meaning, shared understanding and agreement? In otherwords, given that email is a sub-optimal medium for convergence processes (Denniset al. 2008), are there particular behaviors that we can identify that make email nego-tiations more or less successful? In turn, we focus on medium management, relation-ship building, emotion transmission, and negotiation strategies as important behaviorsshaping success in email negotiation.

3 Behaviors of Interest

3.1 Medium Management

The characteristics and costs of the email medium may require that email negotiatorsmanage the medium for greater success. The role of the email user may be centralto success in email negotiations. For example, recent research (Geiger and Parlamis2014) has found that those with greater email comfort—an individual’s subjectiveattitude toward email—do better in email negotiations. Additionally, this research

123

J. D. Parlamis, I. Geiger

suggests that media richness theory (Daft and Lengel 1986) as applied to negotia-tions is limited in that it does not take into account the interaction of the individualwith the medium as an important moderator of the richness. Indeed, channel expan-sion theory (Carlson and Zmud 1999) suggests that the interaction of the person andthe communication medium is critical for understanding media richness perceptions.As mentioned above, media synchronicity theory (Dennis et al. 2008) has expandedthe discussion on media richness by suggesting that a mode is not inherently betteror worse for communication, rather different communication media enable differenttasks. For example, email, a text-based asynchronous medium, is good for transmis-sion of information (conveyance), whereas, telephone or face-to-face communication,synchronous media, would facilitate shared understanding (convergence) because itallows for mutual and simultaneous dialogue. MST also postulates that familiarity ofthe user with the medium may soften each medium’s appropriateness for conveyanceand convergence processes (Dennis et al. 2008). Thus, success in email negotiationsneeds to take the complex interplay of user and medium into account; those whoare more competent with the medium (understanding the attributes and costs of themedium) will perform better. It could be argued that in order to best employ theadvantages (good for communicating information) and to best manage the deficits(not as good for converging on meaning or understanding) of the email medium,active endeavors need to be undertaken. We call these actions ‘medium management’.They may include identifying and clarifying misunderstandings, suggesting best timesto communicate and conveying content as appropriate to better position oneself to suc-ceed.

We propose that medium management can take on two distinct styles: proactiveor reactive. Proactive and reactive behavior patterns have been used to describe man-agement styles (e.g., Larson et al. 2007) and seem very appropriate differentiators ofsuccess in email negotiation.

Proactive management of the medium describes behaviors where individuals thinkahead and initiate action to prevent problems and find ways to maintain a flow ofinformation that can lead to mutual understanding. It includes such acts as suggestinghow to manage time or language differences, rephrasing information that is unclear, ormaking practical, explicit suggestions for response timing and expectations. This typeof medium management facilitates convergence processes, ultimately fostering mutualunderstanding and agreement. Since understanding mutual differences in priorities isa prerequisite for achieving joint gain in negotiation (Thompson 2012), proactivemedium management can be expected to foster such outcomes.

Reactive medium management is characterized by behaviors that do not createopportunities for understanding but rather retrospectively try to ameliorate problemsor misunderstandings or mitigate deficits that have already been caused by the medium(such as mention missed emails or lack of response and identifying misunderstand-ings). It can be characterized as a repair mechanism when mutual misunderstandingshave surfaced. In sum, we propose:

Hypothesis 1 Proactive medium management will positively impact joint gainwhereas reactive medium management will negatively impact joint gain.

123

Mind the Medium

3.2 Relationship Building

Building a relationship may help negotiators persist when convergence is needed, butdifficult to achieve, in the email interaction. In a study investigating communicationmedium richness and negotiation outcomes, results indicated that those negotiating ina richer environment (face-to-face as compared to telephone or email) showed morecollaborative behavior, greater satisfaction and more interest in future negotiations(Purdy et al. 2000). It has been suggested that to make up for the lack of richness, indi-viduals should make sincere attempts to build rapport prior to negotiations or engagein self-disclosure (Drolet and Morris 2000). Alternatively, Walther (1995) argues thatdifferences in relational communication between the face-to-face and electronicallymediated mode fade away if enough time is available for communication. Morris et al.(2002) found that a phone call in advance of negotiations, where counterparts had theopportunity to establish a relationship, yielded greater rapport and increased settle-ment success. The critical role of relationship was underscored in research conductedby Pesendorfer and Koeszegi (2007). They found that in severe conflicts prior rela-tionship was significantly positively related to settlement success. Other theorists havesuggested that intensification or escalation of conflict can be moderated by relation-ship dynamics (e.g., Friedman and Currall 2003). Friedman and Currall (2003) profferthat stronger social bonds will interfere with the negative spiral of process problemsin the email negotiation.

A theoretical rationale for why relationships impact negotiation outcomes positivelyis offered by the literature on in-group bias and out-group discrimination and socialidentity theory (see Brewer 1979; Sherif et al. 1961; Tajfel and Billig 1974; Tajfel andTurner 1979). In the classic minimal group paradigm experiments, researchers dividedparticipants into over- and under-estimators and then asked them to allocate resources.Choice of allocation appeared to coincide with the arbitrary groups to which they hadbeen assigned. Tajfel and others argued that group identity plays a significant role inself-image. By allocating resources to the group to which we belong (our friends), weare maximizing self-interest and potentially invoking a reciprocity norm that will beof benefit in the long run. If individuals in a negotiation begin to form a relationship,it is a way to transform the “out-group” into an “in-group”. If the counterpart in anegotiation is part of the in-group, they are more likely to share goals and be moreinterested in satisfying the other party’s interests.

Extrapolating from the findings that compare different communication media innegotiation (e.g., Morris et al. 2002), from media richness theory (Daft and Lengel1986; Trevino et al. 1987), the dispute-exacerbating model of email (Friedman andCurrall 2003) and social identity theory (e.g., Tajfel and Turner 1979) it would followthat the email dyads that spend more time on relationship building would performbetter in the negotiation than those who do not spend time on establishing a relation-ship. Indeed, Pesendorfer and Koeszegi (2007) assert that “personal relationship helpsnegotiators to better manage the communication challenges” (p. 403). Thus,

Hypothesis 2 The more relationship building in an email negotiation the more suc-cessful the joint outcome.

123

J. D. Parlamis, I. Geiger

3.3 Emotion Transmission

We further propose that explicit emotion transmission potentially has an influence onjoint outcomes in email negotiation. In an email communication environment, whichlacks non-verbals such as voice tone and posture, more explicit emotion commu-nication conveys not only important content but also a meta-message: interest andinvestment in the negotiation. In other words, this higher emotional content could befunctioning as non-verbal information that communicates, “I care about this negotia-tion and I’m here to work to find a settlement”.

Indeed, the functional perspective of emotions is gaining traction in the literature(e.g., see Frijda 1986; Morris and Keltner 2000; Van Kleef 2010). It is suggestedthat emotions and emotional expression serve a social function that communicatesinformation such as intentions, needs, and expectations to others and to the self. Aparticular emotion can signal to a counterpart a particular intention that can lead tosome predictable outcomes, inferences and behaviors. Van Kleef (2010) describesthe emotions as social information (EASI) model as an approach to emotion thatemphasizes the impact emotional expression has on influencing another’s inferencesand affective behavior. For example if a negotiator expresses anger or happiness at aproposal, this signals the next course of action, either offering a counterproposal orpush to close the deal, respectively. Also, expression of emotion can offer incentive orpunishment that in turn reinforces or suppresses certain behaviors in others. The EASIfurther proposes that especially in competitive situations and when the receiving partyis engaged in deep information processing, drawing inferences from the transmittedemotion is accentuated (Van Kleef 2010).

While in most face-to-face environments individuals are getting this emotion-related information from many sources (facial, gestural or speech cues) and, therefore,can somewhat reliably know the feelings of the other, in the email context emotionalexpression is only communicated via text and therefore may be less knowable.

Emotions, whether genuine or tactical (Schroth 2008) have been shown to signifi-cantly impact the outcomes in negotiations although findings are somewhat equivocal.Seminal research (Allred et al. 1997) found that compassion experienced during anegotiation increased joint gains while anger did the opposite. Furthermore, Kopel-man et al. (2006) found that strategic display of positive emotion resulted in greaterconcessions from a negotiation partner. This research suggests that positive emotionsare beneficial for negotiations (i.e., lead to better outcomes) and negative emotionsare detrimental for negotiations (i.e., lead to worse outcomes). However, a numberof other studies (e.g., Overbeck et al. 2010; Sinaceur and Tiedens 2006; Van Kleefet al. 2004a, 2004b) have found that expressing positive emotions (e.g., happiness)was less effective at obtaining concessions from a counterpart than negative emotions(e.g., anger). The impact of emotions in negotiation is complex and it is importantto note that, much of the research on emotions in negotiation examined negotiationswithout reference to the communication mode (in-person or via technology). Whenexamining e-negotiations specifically, Hine et al. (2009), using Linguistic Inquiryand Word Count software to search for specific emotion words appearing in the textof e-negotiations, found significantly more positive emotion words in successful e-negotiations but found similar levels of negative emotion words across groups when

123

Mind the Medium

e-negotiations were not divided into phases. Other research on electronic negotiationshas found that both positive and negative emotions are present in successful and failednegotiation (Griessmair and Koeszegi 2009). Griessmair and Koeszegi (2009) contentanalyzed statements made in electronic negotiation experiments and found that suc-cessful negotiators did show negative emotions but were able to steer the negotiationto a more positive emotional state and were able to break a cycle of negative emotionsfound in the failed negotiations. Recent research (Parlamis and Ames 2010) found thatthose negotiating face-to-face rated both positive (e.g., happy, confident, hopeful) andnegative (e.g., nervous, challenged, guilty) emotions higher than those in the emailcondition where settlement success was lower. They suggest that it is not the valenceof the emotion that matters but, rather, the mere presence of it that is critical; greateremotional engagement (positive and negative) may be required for success in emailnegotiation.

We are interested in investigating the impact that positive and negative emotionshave on the relative success of email negotiations. Positive and negative affect havebeen identified as the two major emotional dimensions in previous research (seeWatson et al. 1988). While the bipolarity of the dimensions have been questioned,positive activation (PA) and negative activation (NA) are considered distinct factors ina broader structure of emotions (Watson et al. 1999). Research has shown that positiveor negative emotions can co-occur or be separately experienced across many differentcontexts (see Larson et al. 2001).

Previous research shows that both positive and negative emotional communicationcan be found in successful negotiation. For example, Hine et al. (2009) found thatboth “positive and negative emotion had a significant relationship with e-negotiationsuccess” (p. 207) and research conducted by Griessmair and Koeszegi (2009) indi-cated that “even successful negotiations contain negative emotional communication”(p. 227). Moreover, the theoretical foundation provided by the EASI and the social-functional perspective suggest that successful email negotiations will show moreexplicit expression of emotion than less successful email negotiations. Since emo-tion tends to cluster along two distinct dimensions (positive and negative activation),we thus propose:

Hypothesis 3 More (positive and negative) emotion communication in email negoti-ation leads to more successful joint outcomes.

3.4 Strategies

We propose that medium management, relationship building, and emotion transmis-sion are critical elements of the negotiation that take a different, more explicit form inemail as compared to face-to-face negotiation. The strategic positioning of the nego-tiator is another critical element of a negotiation that can shape negotiation outcomes.In particular, two negotiation strategies or orientations have been discussed in the lit-erature: distributive and integrative (e.g. Lewicki et al. 2010; Thompson 2012). Theformer strategy refers to bargaining where competition over limited resources shapes a“zero-sum” frame with a focus on individual needs. The latter strategy refers to a more

123

J. D. Parlamis, I. Geiger

cooperative, problem-solving orientation focusing on creative solutions and collectiveneeds.

Depending on the negotiation strategy or orientation, different types of informationcould be shared. For instance, distributive information can be characterized by a focuson facts and positions and integrative information can be characterized as a focus onpriorities and interests. These orientations have related behaviors that correspond toclaiming value, e.g., substantiation, threats, and creating value, e.g., packaging offers,log-rolling (Pesendorfer and Koeszegi 2007; Weingart et al. 2004).

Most complex negotiations can be described as having elements of both distribu-tive and integrative orientations (i.e., mixed-motive). In other words, negotiators aremotivated to “claim value” or compete to benefit one’s own position and interestsand “create value” or cooperate with their counterpart to reach agreement (Lax andSebenius 1986; Walton and McKersie 1965; Fisher et al. 1991). In addition, mostnegotiation situations are not a “fixed pie” (Bazerman and Neale 1992; Thompson andHastie 1990) but rather have the potential to “expand the pie” to benefit both sides.

While a combination of distributive and integrative strategies can be found in suc-cessful negotiations—research has shown that a balance between claiming and cre-ating value will lead to mutual benefit for both parties (e.g., maximum joint gain;Lewicki et al. 2010)—other research has shown that more integrative behaviors (e.g.,multiple equivalent offers, asking questions about interests and priorities, cooperativeattitudes and perspective-taking) and less competitive attitudes and distributive behav-iors leads to greater settlement success and more profitable outcomes in face-to-facenegotiations (e.g., Moore et al. 1999; Medvec and Galinsky 2005; Thompson 2012).Extant research shows that a focus on distributive behaviors—which are prevalentover integrative behaviors in many negotiations anyway (Zachariassen 2008)—leadsnegotiators to focus their cognitive capacities on distributing the available resources,instead of trying to find creative ways to increase the total negotiation pie (e.g., Hyderet al. 2000), so that joint value creation suffers. We thus propose:

Hypothesis 4 More integrative information and creating behaviors lead to higher jointgain, while more distributive information and claiming behaviors lead to lower jointgain.

4 Subjective Value

Besides economic outcomes, we are also interested in how these different behaviorsimpact social psychological outcomes such as satisfaction with the negotiation processand interaction. Curhan et al. (2006) developed a widely used measure of satisfaction:the subjective value (SV) inventory. The global SV measure comprises four sub-scalespertaining to satisfaction with the instrumental outcomes (e.g., satisfaction with theeconomic settlement), with the self (e.g., feelings of competence and satisfactionwith one’s negotiation behaviors), the process (e.g., the perception of an efficient andjust process), and the relationship between negotiators (e.g., the perception of trustand liking). Curhan et al. (2006) describe SV as an important predictor of futureoutcomes. In a recent study conducted by Curhan et al. (2010), findings of a multi-

123

Mind the Medium

round negotiation indicated that SV predicted economic gains in later rounds as wellas interest in negotiating in the future with the same party.

SV has been shown to “increase negotiators’ commitment and sheer enduranceto meet the challenge of reaching an effective integrative settlement” (Curhan et al.2010, p. 694). This is particularly important for email negotiations where the deficits,coupled with the inherent challenges of the medium (such as being less suitable toconvergences processes), might require greater persistence, optimism and confidence.

Research has shown that in negotiations behaviors tend to get reciprocated (Rubinet al. 1994) and if these behaviors are negative (e.g., aggressive tactics) this can leadto negative perceptions toward the counterpart (Friedman and Currall 2003) and thenegotiation (Geiger 2012). Negative emotional expressions may on the one hand trans-mit implicit information so that the other party can draw inferences from it. On theother hand, it may trigger an even stronger (negative) affective reaction (Van Kleef2010). Reactive medium management may be perceived in a similar vein, because itis aimed at fixing some deficiency that has occurred due to the characteristics of themedium.

Conversely, if the behaviors are positive (e.g., sincere attempts at building a relation-ship, expressing positive emotions, using integrative and creative strategies, and proac-tive management of the medium) these too will be reciprocated and lead to a more pos-itive evaluation of the negotiation. Email tends to weaken social bonds (Friedman andCurrall 2003) so those dyads that spend more time strengthening the social bonds andengaging in the other positive behaviors outlined above, should positively impact SV.

Of the behaviors (relationship building, emotion expression, strategies, and mediummanagement) that we predict will have a significant relationship with joint gain, wepropose that:

Hypothesis 5 Proactive medium management, relationship building, positive emo-tional expression, integrative information, and creating value behaviors lead to higherglobal SV while reactive medium management, negative emotion transmission, dis-tributive information, and claiming value behaviors lead to lower global SV.

5 Method

5.1 Participants

Our study was conducted in an intercultural and intercontinental setting, a realisticsetting with regard to email negotiations. Participants were master’s students at a USand a German university. The negotiation exercise was a required assignment withinthe respective classes. One hundred and four subjects (52 dyads) participated in theexercise. The average age for the participants from the German university was 24with 39 % male and 61 % female. The average age for the participants from the USuniversity was 28 with 46 % male and 54 % female. English language proficiency wasnaturally higher among US students where mean English proficiency was 4.8 (on ascale from 1 to 5) while mean English proficiency for German students was 3.9.

123

J. D. Parlamis, I. Geiger

Table 1 Payoff matrix of thenegotiation exercise

Negotiation item Options Benefit pointsseller

Cost pointsbuyer

1. Price Option 1.1 20 25Option 1.2 15 20

Option 1.3 10 15

Option 1.4 0 0

2. Schooling Option 2.1 8 36

Option 2.2 6 12

Option 2.3 4 8

Option 2.4 2 4

Option 2.5 0 0

3. Financing Option 3.1 8 15

Option 3.2 7 10

Option 3.3 6 5

Option 3.4 −5 0

4. Assembly Option 4.1 −2 0

Option 4.2 0 0

Option 4.3 2 1

Option 4.4 4 3

5. Warranty Option 5.1 20 5

Option 5.2 16 4

Option 5.3 12 3

Option 5.4 8 2

Option 5.5 4 1

6. Maintenance Option 6.1 4 0

Option 6.2 5 5

Option 6.3 8 15

Option 6.4 12 25

5.2 Task and Procedure

The negotiation simulation entailed a German manufacturer selling a special purposemachine to a high efficiency pump company in America. Per the storyline, Germanstudents played the role of seller while US students played the role of buyer. Thiswas a mixed-motive negotiation simulation where six items of differential worth andpriority were up for negotiation: price, schooling, financing, assembly, warranty, andmaintenance. The payoff matrix is given in Table 1.

Prior to the negotiation, participants were asked to answer several demographic(e.g., age, gender) and intercultural (e.g., English proficiency, intercultural experi-ences) questions. After answering pre-negotiation questions, dyads were paired andthen each subject was given an information package containing general and role-specific information about the negotiation exercise, their partner’s name and emailaddress as well as an instruction sheet containing all rules for the negotiation. Before

123

Mind the Medium

negotiations started, we asked participants to indicate an individual negotiation targetgain. Participants were given ten days to conduct the negotiation. Each email had tobe copy sent to a general email account maintained by the researchers. This allowedexperimenters to keep a record of email transcripts and informed on the progress ineach dyad. Agreement was reached when a final contract was proposed by one partyand confirmed in a subsequent email by the other. Parties could also agree to disagree.Upon conclusion of the negotiation, subjects took an online post-negotiation surveythat included items on SV and intercultural differences. Participants were debriefedin class and negotiation results were discussed.

5.3 Measures and Coding Scheme

The email transcripts from the experiment were coded and analyzed to determinebehaviors and strategies used during the email negotiation (Weingart et al. 2004).Weingart et al. (2004) outline several choices researchers must make prior to quali-tative coding of negotiations. First, the level of coding and the extent of tactics andstrategies to be coded need to be determined. We decided to use a coding schemethat originated from negotiation theory (see Pruit 1981; Pruit and Lewis 1975). Wemodified and reduced the codes described in Weingart et al. (2007) to accommodatespecific constraints of the email framework and goals of this research. For full cat-egory coding used in this research see Table 2. We refined the coding scheme usingtwo independent raters and an iterative process of coding and testing for agreement.Second, we used semantic instead of syntactic unitizing where each thought unit wasone unit to be coded. Thus, we allowed different units per email; also a particularcode could appear more than once in one email. Three independent, trained codersengaged in a first round of unitizing and coding on their own. Inter-rater reliabilityfor unitizing as measured by Guetzkow’s U was found to be 96.6 % and coding reli-ability measured by Cohen’s Kappa was found to be 66.4 %, which indicates fair togood inter-coder reliability (Weingart et al. 2004; Banerjee et al. 1999; Bakeman andGottman 1986). Ratings were compared and discrepancies discussed. A second roundof coding occurred independently. Finally, coders assessed remaining inconsistenciesand resolved all discrepancies through discussion. Third, all behaviors were aggre-gated to the level of the dyad. Fourth, we divided the number that each code appearedin each negotiation by the total number of codes per negotiation to arrive at a relativefrequency of each specific behavior. For analysis we used the relative frequency ofeach behavior.

Joint gain as the first dependent variable (DV), was measured as the sum of theseller’s and the buyer’s individual gain, i.e. the surplus benefit points for the sellerand the surplus cost points saved for the buyer. The subjective satisfaction ratings ofthe negotiation was measured by the subjective value inventory (SV, Curhan et al.2006), using the post-negotiation questionnaire. All dimensions of the SV displayedsimilar values for internal consistency and item loadings as the original scale validationby Curhan et al. (2006). Internal consistency was at α = 0.819 for SV instrumental(loadings from 0.706 to 0.921), at α = 0.612 for SV self (loadings from 0.513 to 0.803),at α = 0.892 for SV process (loadings from 0.814 to 0.897), and at α = 0.953 for SV

123

J. D. Parlamis, I. Geiger

Table 2 Content analytical coding system

No. Behavioral category Single behavior included in behavioral category

1 Integrative Information States issue preferencesStates issue priorities

Asks questions about preferences/priorities

Open and honest communication interests

2 Create value Makes multi-issue offer

Shows insight

Notes general differences and similarities

Makes positive comments

Suggests compromise

Suggests package trade-offs

3 Distributive information Facts

Asks for bottom line

Asks about others’ substantiation (attack arguments)

Asks miscellaneous task related questions

Notes differences

Negative reactions

Suggests discuss one issue

4 Claim value Makes single issue offer

Refers to bottom line

Substantiates position

Refers to mutual interest to influence other party

Makes threats or uses aggressive tactics

Refers to power

Suggests creative solutions to meet own interests

5 Push to closure Time checks

Notes similarities

Exploding offers

6 Process management Suggests using reciprocity

Suggests vote

Suggests to move on and/or how to do so

7 Relationship building Personal opening statements

Off-task comments

Discloses personal information

Asks friendly questions of the other party

8 Reactive medium management Clarifies misunderstandings caused by the medium

Mentions missed emails or lack of response

Manages miscommunication

relationship (loadings from 0.920 to 0.952). The sub-scale scores (instrumental, self,process, and relationship) were computed by summing the individual items. We alsocalculated a global score of subjective value by adding the four sub-scale scores.

123

Mind the Medium

Table 2 continued

No. Behavioral category Single behavior included in behavioral category

9 Proactive medium management Suggests how to manage time difference

Describes/suggests best times to communicate

Addresses all questions/issues asked

Manages complexity of the medium with practical solutions

Rephrases offers as necessary

Recaps information conveyed

Suggests how to manage language difficulties

10 Negative emotion communication Communicates anger, frustration, disappointment, regret etc.

Uses emoticons or punctuation to emphasize negative feelings

11 Positive emotion communication Communicates happiness, eagerness, warmth, etc…

Expresses apology

Uses emoticons or punctuation to emphasize positive feelings

12 Other Failed negotiation/Impasse

Stalled process/no response

Filler statements

13 Finish/Confirm Settlement Confirms settlement

Make closing remarks

Wrap-up

Note: Coding categories 1–7, 11 adapted from Weingart et al. 2007

5.4 Control Variables

When conducting analyses on joint gain, we controlled for “total amount of codes” thusminimizing the possible confounding effects of length and number of emails. In linewith theoretical considerations (e.g., Geiger 2012), we also controlled for individualgain and individual target gain (or joint gain for SV self) when analyzing subjectivevalue to remove the impact greater success might have on the overall satisfaction withthe e-negotiation.

6 Results

Table 3 shows the means, standard deviations, and correlations of independent anddependent variables in the following equations. As can be seen, joint gain is signifi-cantly correlated with ‘reactive medium management’ (r = −.319, p < .05), ‘totalamount of codes’ (r = .326, p < .05) and marginally correlated with ‘create value’(r = .249, p < .1). Global subjective value is significantly correlated with ‘claimvalue’ (r = −.302, p < .05), ‘proactive medium management’ (r = .309, p < .05),and marginally correlated with ‘distributive information’ (r = −.242, p < .1) and‘negative emotion transmission’ (r = −.248, p < .1). This descriptive analysis pro-vides first evidence that medium management, emotion transmission, and negotiation

123

J. D. Parlamis, I. Geiger

Tabl

e3

Mea

ns,s

tand

ard

devi

atio

ns,a

ndco

rrel

atio

ns

MS

DC

orre

latio

ns

12

34

56

78

910

1112

1314

1516

1Jo

intp

rofit

21.6

27.

391

2SV

inst

rum

enta

l21

.35

5.55

.551

∗∗1

3SV

self

22.2

23.

24.3

56∗∗

.599

∗∗1

4SV

proc

ess

20.3

45.

71.3

57∗∗

.757

∗∗.5

21∗∗

1

5SV

rela

tions

hip

20.0

56.

24.3

51∗

.685

∗∗.4

75∗∗

.916

∗∗1

6SV

tota

l83

.96

17.4

1.4

63∗∗

.883

∗∗.6

77∗∗

.949

∗∗.9

22∗∗

1

7In

tegr

ativ

ein

form

atio

n.0

51.0

50−.

158

−.19

7−.

124

−.03

2.0

12−.

088

1

8C

reat

eva

lue

.207

.084

.249

+.1

49.0

45.0

43.1

33.1

13−.

308∗

1

9D

istr

ibut

ive

info

rmat

ion

.041

.043

0−.

112

.021

−.26

7+−.

370∗

∗−.

242+

−.00

3−.

131

10C

laim

valu

e.1

39.0

87−.

024

−.19

4.0

3−.

357∗

∗−.

394∗

∗−.

302∗

−.19

8−.

111

.152

1

11R

elat

ions

hip

build

ing

.150

.081

−.17

7−.

158

−.27

9*−.

063

−.01

−.11

8.2

4+−.

073

−.23

4+−.

292*

1

12R

eact

ive

med

ium

man

agem

ent

.031

.039

−.31

9∗−.

169

−.22

5−.

169

−.18

5−.

206

.2.0

11−.

011

−.28

4*.0

81

13Pr

oact

ive

med

ium

man

agem

ent

.052

.050

.003

.17

.339

*.3

19∗

.294

∗.3

09∗

.173

−.30

1∗−.

257+

−.25

9+−.

054

−.26

71

14N

egat

ive

emot

ion

tran

smis

sion

.012

.023

.076

−.05

7−.

148

−.24

6+−.

379∗

∗−.

248+

−.11

5−.

125

.195

.305

∗−.

167

.025

−.14

21

15Po

sitiv

eem

otio

ntr

ansm

issi

on.1

51.1

06.1

42.1

95.0

28.1

82.2

14.1

96−.

375∗

∗−.

178

−.16

9−.

212

−.36

0∗∗

−.16

1.0

83−.

044

1

16To

tala

mou

ntof

code

s35

.35

20.3

7.3

26∗

−.09

8.0

02−.

26+

−.25

1+−.

198

.034

−.18

5.1

55.1

08−.

278∗

−.02

−.06

8.2

19.3

03∗

1

∗∗p

<.0

1,∗

p<

.05,

+p

<.1

(2-t

aile

d)

123

Mind the Medium

Table 4 OLS regression analysis for Joint Gain

Predictor Variable Std. beta t p

Intercept 0 2.102 .041Integrative information −.007 −.045 ns

Create value .339 2.290 .027

Distributive information −.017 −.118 ns

Relationship building −.015 −.106 ns

Reactive medium management −.299 −2.181 .035

Proactive medium management .054 .351 ns

Negative emotion communication .051 .379 ns

Total amount of codes .373 2.691 .010

Dependent variable: joint gain

strategies are related to economic and social-psychological outcomes in email nego-tiations.

For testing our hypotheses about joint gain we used ordinary least squares (OLS)regression. Because all theoretical predictors are relative frequencies of behaviorstaken from a single coding scheme, we first assessed the degree of multicollinearityof all regressors. Claim value and positive emotion transmission displayed varianceinflation factors of 3.18 and 3.72, respectively. These values are greater than the criticalthreshold of 2.86 (corresponding to a tolerance value of 0.35; Diamantopoulos andSiguaw 2006) so that we excluded the respective behaviors as regressors in the equationfor joint gain. Table 4 displays the OLS regression results.

The overall model was significant (F(8) = 2.407; p < .03) and accounted for30.9 % of the DV’s variance. We find that the total amount of codes, create value,and reactive medium management significantly predicted joint gain (see Table 4).Relationship building and emotion communication were not significant predictors ofjoint gain; so Hypotheses 2 and 3 were not confirmed. However, Hypotheses 1 and 4were both partially confirmed. Creating value behaviors positively impacted joint gainwhile reactive medium management negatively predicted joint gain. Total amount ofcoded behaviors also positively influenced joint gain.

For testing our predictions about SV in negotiation we used hierarchical linearmodeling (e.g., Peugh and Enders 2005), because each individual’s SV judgments arenested within dyads (e.g., Turel 2010). We performed the analyses for global SV aswell as for the SV dimensions for a supplementary, more fine-grained analysis. Therespective models take the following form:

SVijk = μ + αi + βj · (dyad level predictorj)

+ γk · (individual level predictork) + εijk,

with αi ∼ N(0, σα), εijk ∼ N(0, σε), subscript i denoting the dyad (i = 1…52), jdenoting the hypothesized dyad level predictors, and k denoting the individual levelcontrol variables. SVijk is a linear combination of the global mean μ, the dyad’srandom effect αi, the dyad level predictors’ fixed effects βj, the individual level fixed

123

J. D. Parlamis, I. Geiger

effect γk (individual gain and individual target gain a controls) and the individuallevel error term εijk. Significant βj coefficients in the model signify that the dyad levelpredictors had an effect on the DV, while the γk coefficients indicate the individuallevel control variables’ influence. For computations, all independent individual levelvariables were group mean centered and all dyad level independent variables weregrand mean centered (Enders and Tofighi 2007). With these centering options one caninterpret the resulting coefficients in the same manner as unstandardized coefficientsin OLS regression (Nezlek 2001).

We applied this analytical technique in the following manner: First we computed anunconditional base model for each DV containing only μ, αi, and εijk as a reference forall following models (Nezlek 2001). By using Akaike’s information criterion (AIC) wethen checked whether including theoretical control variables on the individual level(individual gain, target individual gain) or on the dyad level (joint gain) producedbetter model fit (Gelman and Hill 2007). For all DVs except SV self individual levelvariables showed to be more suitable as controls. We then computed a model in whichwe included all theoretical predictors. Finally, we computed purified models in whichonly the significant predictors remained, in line with, e.g., Godes and Mayzlin (2009).We display the results of the base models and the final models in Table 5. As can be seen,for all models the variance of level 2 random effect αi decreases substantially whenlevel 2 predictors are added to the model, compared to the unconditional base model.For all dependent variables except SV self the same can be said about the individuallevel error term εijk which was decreased through the inclusion of individual levelfixed effects (control variables).

In line with Hypothesis 5, distributive behaviors (distributive information, claimvalue) and reactive medium management decrease negotiators’ global SV apprecia-tion. However, we did not find an influence of integrative information, create value,proactive medium management, or negative and positive emotion transmission onglobal SV. Relationship building, supposed to have a positive impact, even seems tohamper global SV, albeit only on a marginal level of significance (p < .1). Hypothesis5 can thus only be partially substantiated.

These results are also found for SV process and SV relationship and partially forSV instrumental. Additionally, SV relationship is influenced in the predicted negativedirection by negative emotion transmission. SV self is not predicted by these behaviors,but only by proactive medium management. Overall, we find mixed evidence for ourhypothesis relating behaviors in email negotiation to negotiators’ SV.

7 Discussion, Implications and Future Research

7.1 Discussion

While some researchers see email as a medium less suitable for the difficult taskof negotiation than other media (e.g. face-to-face communication) this paper exam-ines particular, relevant skills related to success in email negotiation demonstratingthat email does not always disadvantage the e-negotiator. We hypothesized that morerelationship building, positive and negative emotion transmission, proactive medium

123

Mind the Medium

Tabl

e5

Hie

rarc

hica

llin

ear

mod

els

for

SV

Dep

ende

ntva

riab

les

Glo

balS

VSV

inst

rum

enta

lSV

self

SVpr

oces

sSV

rela

tions

hip

Para

met

ers

[ind

epen

dent

vari

able

]B

ase

mod

elFi

nalm

odel

Bas

em

odel

Fina

lmod

elB

ase

mod

elFi

nalm

odel

Bas

em

odel

Fina

lmod

elB

ase

mod

elFi

nalm

odel

μ[i

nter

cept

]83

.96∗

∗∗84

.24∗

∗∗21

.35∗

∗∗21

.48∗

∗∗22

.22∗

∗∗22

.27∗

∗∗20

.34∗

∗∗20

.36∗

∗∗20

.05∗

∗∗19

.91∗

∗∗β

1[I

nteg

rativ

ein

form

atio

n]

β2

[Cre

ate

valu

e]

β3

[Dis

trib

utiv

ein

form

atio

n]−1

07.5

3∗∗

−31.

23∗

−36.

79∗∗

β4

[Cla

imva

lue]

−82.

52∗∗

∗−1

4.19

+−2

5.29

∗∗∗

−25.

07∗∗

∗β

5[R

elat

ions

hip

build

ing]

−44.

37+

β6

[Rea

ctiv

em

ediu

mm

anag

emen

t]−1

26.5

2∗∗

−37.

08∗

−38.

37∗∗

β7

[Pro

activ

em

ediu

mm

anag

emen

t]17

.47∗

∗β

8[N

egat

ive

emot

ion

tran

smis

sion

]−5

8.64

∗β

9[P

ositi

veem

otio

ntr

ansm

issi

on]

β10

[joi

ntga

in]

0.12

∗∗γ

1[i

ndiv

idua

lgai

n]0.

76∗∗

∗0.

24∗∗

0.33

∗∗∗

0.32

∗∗∗

γ2

[ind

ivid

ualt

arge

tgai

n]−0

.34∗

−0.1

4∗−0

.10+

−0.1

1+V

ar(ε

ijk)

[ind

ivid

uall

evel

erro

r]10

4.38

∗∗∗

82.7

6∗∗∗

12.0

5∗∗∗

11.2

2∗∗∗

7.72

∗∗∗

7.72

∗∗∗

15.7

9∗∗∗

13.5

6∗∗∗

19.3

6∗∗∗

17.5

5+V

ar(α

i)[d

yad

leve

lran

dom

effe

ct]

195.

83∗∗

∗10

6.73

∗∗∗

18.4

6∗∗∗

12.0

6∗∗

2.70

+1.

12ns

16.5

0∗∗∗

8.26

∗19

.27∗

∗∗5.

25ns

AIC

865.

5875

4.00

632.

9356

0.80

541.

2953

0.97

646.

7755

4.84

666.

2657

0.48

BIC

873.

5277

6.79

640.

8657

6.00

549.

2354

4.20

654.

7057

5.10

674.

1959

3.27

∗∗∗

p<

.001

;∗∗

p<

.01;

∗p

<.0

5;+

p<

.1;t

wo-

taile

dte

sts.

Mod

els

are

estim

ated

with

max

imum

likel

ihoo

d

123

J. D. Parlamis, I. Geiger

management, and instrumental integrative behaviors as well as less reactive mediummanagement and less instrumental distributive behaviors would lead to higher jointgain. We also proposed that the same behaviors would have a similar effect on thenegotiators’ global perception of SV in negotiation, except for negative emotionswhich we assumed to have a negative effect. The hypotheses were tested on simulatedbuyer-seller email negotiations between subjects from a US and German university.

In line with our predictions, value creating behaviors (such as making multi-issueoffers, suggesting compromises or creative solutions, packaging trade-offs, and show-ing insight by noting similarities and differences in interests and needs) increased andreactive medium management decreased joint gain. Interestingly, the total amount ofcommunication also had a positive influence on joint gain. In other words, the morecommunication in general, and the more integrative behaviors and the less behaviorsmotivated by the inadequacies of the medium specifically, the greater joint success inthe negotiation. Similar to previous research (Parlamis and Ames 2010) it appears thatactive engagement of negotiators is critical for negotiation success. These findings arealso in line with both existing research on the influence of negotiation strategies onjoint gain (e.g., Olekalns and Smith 2000) as well as reasoning about the challengesof low-synchronicity media such as email (Dennis et al. 2008). More messages andno neglect of the peculiarities of the medium are needed to reach higher joint gain.

Reactive medium management also decreases global SV, as well as the SV sub-dimensions that relate to the negotiation process and the negotiators’ relationship. Thisfinding confirms the damaging role that “letting the medium go” in email negotiationshas. In line with previous findings (Geiger 2012), instrumental distributive behaviors(distributive information, claim value) also had a detrimental effect on global sub-jective value. Moreover, negative emotion transmission decreased SV about the rela-tionship. These findings about subjective value indicate that it’s not the positive typesof behavior (relationship building, proactive medium management, positive emotiontransmission, integrative instrumental behaviors) that increase but rather the negativeones that decrease negotiators’ overall perceptions of the negotiation. Only with regardto SV about the self did proactive medium management have a positive influence. Inother words, negotiators are mostly more sensitive towards negative deviations fromthe average email negotiation behavior than towards positive ones.

Contrary to our predictions, relationship building significantly predicted neitherjoint gain nor global subjective value. The same can be said of emotion communi-cation, which had no influence on joint gain or global subjective value. Distributiveinstrumental behaviors did not damage joint gain, and integrative instrumental behav-iors did not influence global SV.1

The first puzzling yet interesting finding pertains to relationship building. First, inthis research relationship building does not significantly influence joint gain whichstands in contrast to previous research that demonstrates the importance of relation-ships to negotiation success (e.g., Moore et al. 1999; Morris et al. 2002; Pesendorfer

1 It is important to note that for multicollinearity reasons we did not include claim value and positiveemotion transmission in the regression equation for joint gain. Thus, a strict formal test for these is missing.However, because those two variables’ correlations with joint gain are also insignificant (see Table 3), weare confident that they do not predict joint gain in our data.

123

Mind the Medium

and Koeszegi 2007). We suspect that the one-shot nature of the negotiation could bepartially responsible. The negotiation occurred across continents with two individ-uals who certainly had no long-term relationship potential and no past relationship.The one-shot, instrumental nature of the task could have dampened the relationshipbuilding aspect of the negotiation.

A second possible explanation for why relationship building did not impact jointgain could be found in research on social information processing and computer medi-ated communication (e.g., Walther 1995). This research argues that relational commu-nication differences between face-to-face and electronically mediated communicationmodes fade away if enough time is available for communication. In other words, “CMCgroups have been less social and more task-oriented relative to FtF, [but] when groupsinteract over time, warmer relational tone and friendship development is frequentlyseen” (Walther 1995, p. 198). It could be that our research did not allow enough timefor the relational information to make a difference in the outcome.

Importantly, this type of negotiation, one time only with no lasting relationshipties and requiring only a short amount of interaction time, is becoming an increas-ingly commonplace experience for e-negotiations. Consider online car negotiationsor negotiations for selling off items on Craigslist. In these cases no past or futurerelationship is expected and very little time is allotted to the interaction. Our researchis first evidence to suggest that relationship will not necessarily play a significant rolein this context.

Further, our research could have meaningful application here. Specifically, in theabsence of future negotiation potential, this research suggests that it is most beneficialto spend energy on creating value behaviors and to pay attention to the deficiencies ofthe medium so as to not require repair behaviors, rather than relationship building forone-shot negotiations.

A second unexpected finding concerns emotion transmission. We had hypothe-sized that both positive and negative emotion transmission could serve as cues for thereceiving person who, according to the EASI model, could draw inferences about thesender’s intent and behavior (Van Kleef 2010). While past research has acknowledgedthat emotion may be transmitted in emails despite the medium’s leanness (Griessmairand Koeszegi 2009), our study shows that these emotions did not have an effect oneconomic outcomes. Only with regard to SV relationship did negative emotions showany influence. A possible explanation for these findings may reside in the strength ofemotions: While we identified them as explicit behaviors in the emails, negotiatorsmay not have given them too much attention because they were more focused on thesubject matter of the negotiation. In face-to-face negotiation, emotion transmissioncan happen through many more channels such as visual and para-verbal cues andemotions may thus be a lot more salient. Hence, our research puts the applicability ofthe EASI model (Van Kleef 2010) in email negotiations partly in doubt.

Lastly, the non-existing influence of distributive behaviors (distributive informa-tion, claim value) on joint gain, where we had expected a negative one, warrantssome attention. Possibly, in contrast to in-person negotiations, particular characteris-tics of the email environment make it more likely that distributive elements are notas damaging in this particular setting. Specifically, email does not have the advan-tage of co-presence, immediate responses and feedback, or context clues that exist

123

J. D. Parlamis, I. Geiger

in face-to-face interactions. Therefore, those communicating via email have dimin-ished capacity to converge on the meaning of messages and arrive at mutual under-standing (Dennis et al. 2008; Friedman and Currall 2003). Consequently, individualsnegotiating over email may be more explicit in their communication, noting facts,bottom lines and taking time to substantiate positions in an effort to ensure mutualunderstanding.

In addition, it could be that distributive-type behavior in an email context is lesscounter-normative than in a face-to-face environment and therefore less damaging.Individuals who have used the email medium for years or are native email usersmay be accustomed to the “flame” of email (negative, aggressive or impolite state-ments). Researchers over the past 30 years have documented these negative, uninhib-ited communication behaviors in computer mediated communication such as email(e.g., Sproull and Kiesler 1986). Our research might indicate that an individual’sexpectations about communication may differ across mode and therefore responsesassociated with those behaviors may also diverge.

7.2 Implications and Future Research

As noted by many researchers and theorists (e.g., Daft and Lengel 1984; Purdy et al.2000; Friedman and Currall 2003) the attributes of the email medium (e.g., lack ofco-presence, immediate feedback, or non-verbal cues) can hinder outcomes in negoti-ations. Our research suggests that negotiators should not let the detached and discon-nected aspects of email drive the tenor of the interaction; rather, negotiators need tobe engaged, active and connected during the email interaction, developing and testingsettlements, creating value and taking control of the email medium so medium-relatedissues do not require repair. As MST and other recent research suggests (Dennis et al.2008; Geiger and Parlamis 2014), it is the familiarity and comfort of the user withthe medium that is critical to maximizing the utility of the medium for complex taskslike negotiation. Further, this research proposes that increased communication, useof strategies and control of the medium, may influence perceived interest and com-mitment to the negotiation. Future research should look at perceived interest as amediating variable in email negotiations.

Since e-negotiation is a ubiquitous part of our business world today, determiningpractical, empirically based suggestions for increasing success is greatly needed. Ourresearch suggests that e-negotiators should engage in value creation behaviors andavoid behaviors that emphasize the inadequacies of the medium. These findings furthersuggest a very active approach to the negotiation yields higher joint gain. For greaterpost-negotiation satisfaction, limiting distributive negotiation strategies and keepinga grip on the medium appears to be critical. The detachment that email provides us(Daft and Lengel 1986) is seductive but we should not allow the medium to dictate thelevel of engagement or to simply react to the deficits and costs of the medium. Rather,one needs to commit to the email negotiation process and be cautious that the deficitsof the email medium do not create barriers to communication that lead to suboptimaloutcomes. While proactive medium management did not predict joint outcome or

123

Mind the Medium

SV, it might lay the necessary groundwork for the communication behaviors that arerequired for complex tasks.

References

Adair WL, Brett JM (2005) The negotiation dance: time, culture, and behavioral sequences in negotiation.Organ Sci 16(1):33–51

Allred KG, Mallozzi JS, Matsui F, Raia CP (1997) The influence of anger and compassion on negotiationperformance. Organ Behav Hum Dec Process 70(3):175–187

Ambrose E, Marshall D, Fynes B, Lynch D (2008) Communication media selection in buyer-supplierrelationships. Int J Organ Prod Manage 28(4):360–379

Banerjee M, Capozzoli M, McSweeney L, Sinha D (1999) Beyond kappa: a review of interrater agreementmeasures. can J Stat 27(1):3–23

Bakeman R, Gottman JM (1986) Observing interaction. Cambridge Universtiy Press, CambridgeBazerman MH, Neale MA (1992) Negotiating rationally. Free Press, New YorkBrewer MB (1979) In-group bias in the minimal intergroup situation: a cognitive-motivational analysis.

Psychol Bull 86(2):307–324Carlson JR, Zmud RW (1999) Channel expansion theory and the experiential nature of media richness

perceptions. Acad Manage J 42(2):153–170Clark HH, Brennan SE (1991) Grounding in Communication. In: Resnick LB, Levin JM, Teasley SD

(eds) Perspectives on socially shared cognition. American Psychological Association, Washington D.C.,pp127–149

Croson RT (1999) Look at me when you say that: an electronic negotiation simulation. Simul Gaming30(1):23–37

Curhan JR, Elfenbein HA, Eisenkraft N (2010) The objective value of subjective value: a multi-roundnegotiation study. J Appl Soc Psychol 40(3):690–709

Curhan JR, Elfenbein HA, Xu H (2006) What do people value when they negotiate? Mapping the domainof sujective value in negotiation. J Pers Soc Psychol 91:493–512

Daft RL, Lengel RH (1984) Information richness: a new approach to managerial behavior and organizationdesign. Res Organ Behav 6:191–233

Daft RL, Lengel RH (1986) Organization information requirements, media richness, and structural design.Manage Sci 32(5):554–571

Dennis AR, Fuller RM, Valacich JS (2008) Media, tasks, and communication processes: A theory of mediasynchronicity. MIS Q 32(3):575–600

Diamantopoulos A, Siguaw JA (2006) Formative versus reflective indicators in organizational measuredevelopment: a comparison and empirical illustration. Br J Manage 17(4):263–282. doi:10.1111/j.1467-8551.2006.00500.x

Drolet AL, Morris MW (2000) Rapport in conflict resolution: accounting for how face-to-face contactfosters mutual cooperation in mixed-motive conflicts. J Exp Soc Psychol 36(1):26–50

Enders CK, Tofighi D (2007) Centering predictor variables in cross-sectional multilevel models: a new lookat an old issue. Psychol Method 12(2):121–138. doi:10.1037/1082-989x.12.2.121

Fisher R, Ury W, Patton B (1991) Getting to yes: negotiating agreement without giving in, 2nd edn. PenguinBooks, New York

Frijda NH (1986) The emotions: studies in emotion and social interaction. Cambridge University Press,Cambridge

Friedman RA, Currall SC (2003) E-mail escalation: dispute exacerbating elements of e-mail communication.Hum Relat 56:1325–1348

Galin A, Gross M, Gosalker G (2007) E-negotiation versus face-to-face negotiation what has changed—ifanything? Comput Hum Behav 23:787–797

Geiger I (2012) Media effects on the formation of negotiator satisfaction: the example of face-to-face and text based electronically mediated negotiations. Group Decis Negot 26:2012. doi:10.1007/s10726-012-9317-3

Geiger I, Parlamis J (2014) Is there more to email negotiation than email? The role of email affinity. ComputHum Behav 32:67–78. doi:10.1016/j.chb.2013.11.016

Gelman A, Hill J (2007) Data analysis using regression and multilevel, hierarchical models. CambridgeUniversity Press, Cambridge

123

J. D. Parlamis, I. Geiger

Godes D, Mayzlin D (2009) Firm-created word-of-mouth communication: evidence from a field test. MarkSci 28(4):721–739. doi:10.1287/mksc.1080.0444

Griessmair M, Koeszegi ST (2009) Exploring the cognitive-emotional fugue in electronic negotiations.Group Decis Negot 18:213–234

Hine MJ, Murphy SA, Weber M, Kersten G (2009) The role of emotion and language in dyadice-negotiations. Group Decis Negot 18:193–211

Hollingshead AB, McGrath JE, O’Connor K (1993) Group task performance and communication tech-nology: a longitudinal study of computer-mediated versus face-to-face work groups. Small Group Res24(3):307–333

Hyder EB, Prietula MJ, Weingart LR (2000) Getting to best: efficiency versus optimality in negotiation.Cogn Sci 24(2):169–204. doi:10.1207/s15516709cog2402_1

Kersten G, Zhang G (2003) Mining inspire data for determinants of successful internet negotiations. CentEur J Oper Res 11(3):297–316

Koeszegi ST, Pesendorfer E, Vetschera R (2011) Data-driven phase analysis of e-negotiations: an exemplarystudy of synchronous and asynchronous negotiations. Group Decis Negot 20(4):385–410

Kopelman S, Rosette AS, Thompson L (2006) The three faces of eve: strategic displays of positive, negative,and neutral emotions in negotiations. Organ Behav Hum Dec Process 99(1):81–101

Lax D, Sebenius J (1986) The manager as negotiator: bargaining for cooperation and competitive gain. FreePress, New York

Larson LL, Bussom RS, Vicars W, Jauch L (2007) Proactive versus reactive manager: Is the dichotomyrealistic? J Mgt Stud 23(4):385–400

Larson JT, McGraw AP, Cacioppo JT (2001) Can people feel happy and sad at the same time? J Soc Psychol81(4):684–696

Lewicki RJ, Barry B, Saunders DM (2010) Negotiation, 6th edn. McGraw-Hill, New YorkMedvec VH, Galinsky AD (2005) Putting more on the table: how making multiple offers can increase the

final value of the deal. HBS Negot Newsl 8(4):4–6Moore DA, Kurtzberg TR, Thompson LL, Morris MW (1999) Long and short routes to success in electron-

ically mediated negotiations: group affilitations and good vibrations. Organ Behav Hum Dec Process77(1):22–43

Morris MW, Keltner D (2000) How emotions work: the social functions of emotional expressions in nego-tiations. Res Organ Behav 22:1–50

Morris M, Nadler J, Kurtzberg T, Thompson L (2002) Schmooze or lose: social friction and lubrication ine-mail negotiations. Group Dyn Theory Res Pract 6(1):89–100

Nadler J, Shestowsky D (2006) Negotiation, information technology, and the problem of the faceless other.In: Thompson LL (ed) Negotiation Theory and Research. Tyalor and Francis Group, New York, pp145–172

Naquin CE, Kurtzberg TR, Belkin LY (2010) The finer points of lying online: e-mail versus pen and paper.J Appl Psychol 95(2):387–394

Naquin CE, Paulsen GD (2003) Online bargaining and interpersonal trust. J Appl Psychol 88(1):113–120Nezlek JB (2001) Multilevel random coefficient analyses of event- and interval-contingent data in

social and personality psychology research. Pers Soc Psychol Bull 27(7):771–785. doi:10.1177/0146167201277001

Olekalns M, Smith PL (2000) Understanding optimal outcomes. Hum Commun Res 24(4):528–556. doi:10.1111/j.1468-2958.2000.tb00768.x

Overbeck JR, Neale MA, Govan CL (2010) I feel, therefore you act: intrapersonal and interpersonal effectsof emotion on negotiation as a function of social power. Organ Behav Hum Dec Process 112(2):126–139

Parlamis J, Ames D (2010) Face-to-face and email negotiations: a comparison of emotions, perceptions andoutcomes. IACM 23rd annual conference paper. Available at SSRN: http://ssrn.com/abstract=1612871

Pesendorfer EM, Koeszegi S (2006) Hot versus cool behavioural styles in electronic negotiation: the impactof communication mode. Group Decis Negot 15(2):141–155. doi:10.1007/s10726-006-9025-y

Pesendorfer EM, Koeszegi S (2007) Social embeddedness in electronic negotiations. Group Decis Negot16(4):399–415

Peugh JL, Enders CK (2005) Using the SPSS mixed procedure to fit cross-sectional and longitudinalmultilevel models. Educ Psychol Meas 65(5):717–741. doi:10.1177/0013164405278558

Pruit DG (1981) Negotiation behavior. Academic Press, New YorkPruit DG, Lewis SA (1975) Development of integrative solutions in bilateral negotiation. J Pers Soc Psychol

31(4):621–633

123

Mind the Medium

Purdy JM, Nye P, Balakrishnan PV (2000) The impact of communication media on negotiation outcome.Int J Confl Manage 11(2):162–187

Rubin JZ, Pruitt DG, Kim SH (1994) Social conflict: escalation,stalemate and settlement, 2nd edn. McGraw–Hill, New York

Schroth HA (2008) Some like it hot: teaching strategies for managing tactical versus genuine anger innegotiations. Negot Confl Manage Res 1(4):315–332

Sherif M, Harvey OJ, While BJ, Hood WR, Sherif CW (1961) Intergroup conflict and cooperation: therobbers cave experiment. University Book Exchange, Norman

Sinaceur M, Tiedens LZ (2006) Get mad and get more than even: when and why anger expression is effectivein negotiations. J Exp Soc Psychol 42:314–322

Sproull L, Kiesler S (1986) Reducing social context cues: electronic mail in organizational communication.Manage Sci 32:1492–1512

Stuhlmacher AF, Citera M (2005) Hostile behavior and profit in virtual negotiation: a meta-analysis. J BusPsychol 20(1):69–93

Tajfel H, Billig M (1974) Familiarity and categorization in intergroup behavior. J Exp Soc Psychol 10:159–170

Tajfel H, Turner J (1979) An integrative theory of intergroup conflikct. In: Austin WG, Worchel S (eds)The social psychology of intergroup relations. Brooks/Cole, Montery, pp 33–47

Thompson L (2012) The mind and heart of the negotiator, 5th edn. Pearson, Upper Saddle RiverThompson L, Hastie R (1990) Judgment tasks and biases in negotiation. In: Sheppard BH, Bazerman MH,

Lewicki RJ (eds) Research on negotiation in organizations 1. JAI Press, Greenwich, pp 31–54Thompson L, Nadler J (2002) Negotiating via information technology: theory and application. J Soc Issues

58(1):109–124Trevino LK, Lengel RH, Daft RL (1987) Media symbolism, media richness, and media choice in organi-

zations: a symbolic interactionist perspective. Commun Res 14(5):553–574Turel O (2010) Interdependence issues in analyzing negotiation data. Group Decis Negot 19(2):111–125.

doi:10.1007/s10726-008-9118-xUtz S (2007) Media use in long-distance friendships. Inf Commun Soc 10:693–712Van Kleef GA (2010) The emerging view of emotion as social information. Soc Pers Psychol Compass

4(5):331–343Van Kleef GA, De Dreu CKW, Manstead ASR (2004a) The interpersonal effects of emotions in negotiations:

a motivated information processing approach. J Pers Soc Psychol 87(4):510–528Van Kleef GA, De Dreu CKW, Manstead ASR (2004b) The interpersonal effects of anger and happiness in

negotiations. J Pers Soc Psychol 86(1):57–76Walther JB (1995) Relational aspects of computer-mediated communication: experimental observations

over time. Organ Sci 6(2):186–203Walton RE, McKersie RB (1965) A behavioral theory of labor negotiations: an analysis of a social interaction

system. McGraw-Hill, New YorkWatson D, Clark LA, Tellegen A (1988) Development and validation of brief measures os positive and

negative affect: the PANAS scales. Soc Psychol 54(6):1063–1070Watson D, Wiese D, Vaidya J, Tellegen A (1999) The two general activation systems of affect: structural

findings, evolutionary considerations, and psychobiological evidence. Soc Psychol 76(5):820–838Weingart LR, Brett JM, Olekalns M, Smith PL (2007) Conflicting social motives in negotiating groups.

J Pers Soc Psychol 93(6):994–1010Weingart LR, Olekalns M, Smith PL (2004) Quantitative coding of negotiation behavior. Int Negot 9(Special

Issue on Research Methods in Negotiation and Social Conflict): 441–455.Zachariassen F (2008) Negotiation strategies in supply chain management. Int J Phys Distrib Logist Manage

38(10):764–781. doi:10.1108/0960003081091484

123