An Initial Test of the Effects of Cue Patterns on Behavior and Attributions in a Purchasing Negotiation

ABSTRACT - This paper describes an experiment which provided an initial test of the effects of varying degrees of behavioral consistency on attributions and behavior in a simulated purchasing negotiation. Subjects faced with behaviorally consistent negotiation partners perceived their partners to be more trustworthy, conceded more, and were more likely to reach an agreement in the negotiation than subjects who negotiated with inconsistent partners.

Citation:

Randall L. Rose and Peter R. Dickson (1988) ,"An Initial Test of the Effects of Cue Patterns on Behavior and Attributions in a Purchasing Negotiation", in NA - Advances in Consumer Research Volume 15, eds. Micheal J. Houston, Provo, UT : Association for Consumer Research, Pages: 101-107.

Advances in Consumer Research Volume 15, 1988      Pages 101-107

AN INITIAL TEST OF THE EFFECTS OF CUE PATTERNS ON BEHAVIOR AND ATTRIBUTIONS IN A PURCHASING NEGOTIATION

Randall L. Rose, University of South Carolina

Peter R. Dickson, The Ohio State University

ABSTRACT -

This paper describes an experiment which provided an initial test of the effects of varying degrees of behavioral consistency on attributions and behavior in a simulated purchasing negotiation. Subjects faced with behaviorally consistent negotiation partners perceived their partners to be more trustworthy, conceded more, and were more likely to reach an agreement in the negotiation than subjects who negotiated with inconsistent partners.

INTRODUCTION

Several researchers have proposed that the impact of one negotiator's behavior on the behavior of another negotiator is mediated, at least in part, by attributions made to account for that behavior (Siegel and Fouraker 1960; Rubin and Brown 1975; Chertkoff and Esser 1976). From this perspective, bargainers are motivated to understand the causes of their partners' negotiating behavior because of their mutual dependence on one another for the attainment of goals. Unfortunately, there are usually several plausible explanations for behavior as complex as bargaining. For example, a seller's tough negotiating behavior could be attributed to dispositional factors such as greed or competitiveness, situational factors such as cost constraints or instructions from superiors, or to both. Thus, negotiators can become mired in a dilemma of trying to choose among several plausible causes for the observed behavior. Recently, Rose and Dickson (1987) proposed that negotiators help resolve this dilemma by judging the degree of consistency exhibited by the other party's behavioral cues. Such cues include the size and pattern of concessions, body language, and verbal and written communications.

This experiment was intended to serve three key purposes: (1) to evaluate subjects' ability to observe and report accurately the patterns in the behavioral cues provided by negotiators, (2) to determine the impact of cue consistency on subjects' bargaining behavior and the impressions they form of their partners, and (3) to examine the strength of the relationship between these impressions and bargaining behavior.

BACKGROUND AND HYPOTHESES

Consumer researchers and marketing scholars have long recognized the importance of negotiation as a means of exchange for a wide variety of products in consumer as well as in industrial markets (Alderson 1957). Only recently has that perception begun to manifest itself in a growing body of published research designed to improve our understanding of negotiated exchanges and to develop prescriptions for the practice of buyer-seller negotiation. This burgeoning of interest in the study of negotiation has led to a greater depth and breadth of scholarly investigation. Consumer researchers have studied such diverse aspects of negotiation as learning across interactions (WaLker 19-71), power relationships (Dwyer 1984; Dwyer and Walker 1981; Johnston and Bonoma 1984, McAlister et al. 1986), perceived similarity between parties to the exchange (Mathews et al. 1972), constituent monitoring (Clopton 1984), and expectations of trust and toughness (Schurr and Ozanne }985). Most recently researchers have demonstrated increasing concern for the role of negotiation in business relationships (as opposed to single interactions) (Dwyer, Schurr, and Oh 1987) and for the development of heuristic models of the negotiation process (Evans and Beltramini 1987).

The work of consumer researchers in this area draws heavily on and also has begun to extend the results of a very large body of research on the topic of negotiation and conflict resolution in disciplines such as social psychology. (See Lewicki and Litterer 1985; Pruitt 1981; Rubin and Brown 1975.) The extent to which bargaining is characterized by mutual behavioral responsiveness has long been a topic of interest to bargaining scholars. However, most research in this area has examined the effects of individual behavioral cues such as a pattern of concessions on bargaining behavior. This approach maximizes experimental control, but also preempts any opportunity to study such interesting issues as the interactions of behavioral cues, the weights assigned to each cue in forming an impression of an actor, or the nature of the relationships among the cues.

In the model of attributional processing in negotiations proposed by Rose and Dickson (1987), the relationships among a bargainer's behavioral cues are characterized by three-aspects of consistency: (1) temporal consistency or the extent to which a cue exhibits a pattern over time, (2) directional consistency (consistency over modality to use Kelley's (1967) terminology) which is the consistency between the patterns formed by the cues, and (3) point consistency or the consistency between two or more cues at any one point in time (i.e., during any bid). The basic assumptions underlying the model are that cues which are directionally patterned over time (i.e., those which move consistently from soft to tough or from tough to soft) provide more information about the negotiator's motivations and intentions than those which are not patterned and that the most plausible cause for a negotiator's behavior is the one which is most consistently signalled by all of the behavioral cues available in the interaction.^ To use an example from Rose and Dickson (1987): "If at least one cue provides information that is inconsistent with the other available cues, the credibility of the bargainer will be damaged. For example, if a bargainer is inconsistent in his verbal communications, sometimes tough and other times soft, the degree of consistency between the message behavior and any particular concession will be high when the concession is small (large) and the message is tough (conciliatory) and low when the concession is large (small) and the message is tough (conciliatory)."

A question which must be answered concerns the factors, if any, which mediate the effects of behavioral consistency on negotiating behavior. One factor which has received considerable attention in the bargaining literature is interpersonal trust (Deutsch 1958; Loomis 1959; Kimmel et al. 1980; Pruitt 19813. In general, high levels of mutual trust are associated with joint problem solving and integrative bargaining behavior. While the development of an atmosphere of mutual trust in a negotiation does not guarantee joint problem-solving, the deleterious effects of a lack of trust are well established. Low levels of trust are likely to produce negotiations which are distributive in nature, characterized by aggressive behavior and high levels of conflict (Lewicki and Litterer 1985). Recently, Schurr and Ozanne (1985) reported the results of an experiment in buyer-seller negotiations in which subjects' expectations of the trustworthiness and toughness of their hypothetical suppliers were manipulated. Expectations of the seller's trustworthiness were shown to moderate the effects of the seller's tough bargaining behavior on subjects' concessions and the likelihood of reaching agreement. Their findings suggest that a tough bargaining stance is likely to benefit the negotiator only to the extent to which his partner perceives him to be trustworthy.

Bargainers who are consistent in their behavior should be viewed as being more trustworthy than inconsistent bargainers; and, therefore, the credibility of the signal conveyed by their behavior should be increased. This effect of cue consistency on perceptions of trustworthiness is based partly on the fact that consistent bargainers behave in a manner which is congruent with their verbal communications while inconsistent bargainers do not. A negotiator who claims to be unable to make further concessions on one bid, but then makes a concession on the next bid suffers from diminished trustworthiness. This leads directly to the following hypotheses:

H1: Bargainers exposed to behavioral cues from their negotiating partners which consistently signal a soft-to-tough direction to their behavior will perceive the seller to be more trustworthy than those facing inconsistent negotiators.

H2: Bargainers who perceive the seller to be more trustworthy will concede more than those who perceive the seller to be less trustworthy.

H3: Bargainers exposed to behavioral cues from their negotiating partners which consistently signal a soft-to-tough direction to their behavior will concede more than those facing inconsistent negotiators.

METHOD

Experimental Design: The between-subjects design employed in this study was a simple one with only one factor, relative cue consistency, operationalized at two levels, low and high. In the low cue consistency condition (LOWCON), the programmed seller took a random amount of time to make random concessions and send messages of random toughness. A pattern of decreasing concessions, soft-to-tough messages, and increasing deliberation time comprised the high consistency treatment (HIGHCON).

Three behavioral cues were manipulated to construct the consistency conditions of interest in this . research: (1) the pattern of concessions, (2) the deliberation time between bids, and (3) the toughness of the messages used. The selection of the most theoretically interesting levels of the independent variables was a difficult task. At this early stage of our investigation of the effects of cue patterns on bargaining behavior and impressions of the actor, it was deemed advisable to test patterns representing opposite extremes of consistency. Thus, our choice of very consistent and very inconsistent patterns was intended to establish the endpoints of possible inductions of behavioral consistency. The question of how much inconsistency in behavior is enough to undermine the effects hypothesized above is an interesting one worthy of future research.

The extent to which any two of the cues were consistent with one another is an empirical issue which is not addressed in this research. Most problematic is the relationship between the deliberation time pattern and the other two cues. Arguments could be made supporting each of three positions regarding this cue. First, one could argue, as we do, that increasing deliberation time is consistent with soft-to-tough concession and message patterns because it signals an creasing reluctance to concede. On the other hand, it is also possible to interpret decreases in deliberation time as more consistent with a hardening of bargaining stance. This is plausible because a bargainer who has reached his/her limit no longer need deliberate over the next concession since no concession is possible. Finally, it is also plausible that deliberation time patterns, in themselves, provide no information concerning changes in a bargainer's stance. From this perspective, the time cue would be irrelevant to a bargainer's assessment of his partner's behavioral consistency. In general, little is known about what negotiators' infer from their partners' behavioral signals. Although in this study we were only concerned with inducing differences in bargainers' perceptions of their partners' overall behavioral consistency, the information value of cue patterns is also a topic worthy of future research.

Concession Pattern: With these goals in mind, two different types of concession patterns were chosen for pretesting: (1) random concessions and (2) a pattern of non-linear, decelerating concessions (similar to Druckman et al. 1972). The concession patterns used are described in Table 1. A very important issue concerns the contingent or non-contingent nature of the programmed concessions. One approach would be to make each concession in all conditions contingent on a previous concession on the same turn by the subject bargainer. It could be argued that this restriction is necessary in order to avoid unilateral forfeiture of profits by the programmed opponent. On the other hand, requiring all programmed concessions to be contingent may drastically reduce the effectiveness of the experimental manipulations by making the behavior patterns difficult or even impossible to perceive. Given the goals of this research, the latter outcome would be disastrous. The question of how subjects respond to concession patterns is an empirical one. There was no a priori reason to assume that a substantial number of subjects would simply sit back and watch the opponent concede without reciprocating. For these reasons a non-contingent concession pattern was used.

The programmed seller opened the negotiation with the same tough initial bid in each condition. This initial bid (52400) is tough because it is $400 above the buyer's break-even price. The non-linear, decelerating concession pattern was derived by decreasing the size of the concession by one third on each turn. The nonlinear, decelerating pattern starts with a large initial concession then decreases the size of further concessions at an increasing rate (soft to tough). This strategy should clearly signal a settlement point arrived at asymptotically through the pattern of concessions. In the other concession condition, the size of the concessions was deliberately constructed so as to avoid signalling a settlement point with a pattern.

TABLE 1

CONCESSION PATTERNS

In order to avoid confounding the concession patterns with the total amount conceded, the total amount conceded was held constant in all conditions. One way to do this is to end each negotiation after the same number of bids (Yukl 1974a,b). The number of bids should be sufficiently large to permit the subjects to discern the pattern of cues presented yet small enough to ensure that all sessions end simultaneously with either an acceptance of the subject's latest offer by the programmed opponent or a rejection of that offer for an external alternative. An alternative to this unrealistic constraint would be to allow the bargainers to finish the exercise subject to some time or number of bids constraint (Smith et al. 1982). The hypotheses concerning the effects of consistent behavioral cues on bargaining behavior and outcomes could then be tested using data up to the last turn on which all bargainers are still involved in the interaction. This method would permit examination of the effects of the various behavior patterns on the failure to reach agreement and actual outcomes. However, it could also introduce an element of time pressure to all conditions, the importance of which would depend on the length of time allowed for the negotiation. Under conditions of high perceived time pressure, bargainers have been found to make larger and more frequent concessions as the deadline nears both in laboratory experiments and in actual negotiations (e.g. labor negotiations) (Pruitt & Drews 1969; Chertkoff & Esser 1976; Raiffa 1982). This effect is moderated by the importance of reaching an agreement (i.e. the availability of alternatives to agreement within the negotiation).

As a compromise, the programmed seller made a maximum of 15 predetermined bids. If agreement was not reached after the fifteenth bid, the negotiation ended with the programmed seller indicating to the bargainer that it was accepting an alternative offer outside the negotiation. When the seller withdrew from the negotiation, subjects were compensated as if they had accepted their outside alternative as well. The 15 bid maximum, while admittedly lengthy, was selected in order to ensure that the negotiators would have sufficient time to observe and to interpret their partners' behavioral signals. Subjects were not told prior to the negotiation that there was a 15 bid limit.

The outside alternative provided to the bargainers was designed to be acceptable, but not very attractive. It was expected that subjects would perceive a moderate degree of pressure to reach agreement within the negotiation. The effects of this pressure might vary systematically across conditions if the amount conceded by the programmed seller were not held constant. Obviously, tough strategies are made more effective when one's bargaining partner feels compelled to reach an agreement. Since the amount conceded by the program was held constant in these studies, the only systematic effect on subject concessions should be related to the nature and consistency of the behavioral cues provided in the interaction.

Deliberation Time Pattern: The amount of time used by the programmed opponent between bids was varied in two levels: (1) an increasing pattern and (2) a no-pattern, randomized condition. Deliberation time was measured from the time the bargainer received a bid until the instant the counter-bid was made. The increasing pattern began with a 15 second deliberation time and ended with a delay of 119 seconds. The total deliberation time was held constant across conditions. Table 2 lists the deliberation time patterns.

Message Pattern: Two levels of communications were manipulated. In one condition the toughness of the messages sent by the programmed bargainer was increased on each turn. In the second condition the communications were randomly ordered in order to avoid creating a pattern. The message patterns used by the programmed seller are shown in Table 3. This manipulation required exploratory research in order to ensure that actual differences in the toughness of the messages were observable from one bid to the next in the increasing-toughness condition. Preliminary evidence was gathered by asking subjects to assume the role of a bargainer who has just received a particular message from a bargaining partner. Subjects were asked to judge the toughness of each message and the likelihood that a bargainer sending such a message would make further concessions on subsequent bids. The ordering of the messages suggested by this admittedly imprecise method was used to construct the initial message patterns which were later validated in the context of an actual negotiation.

TABLE 2

DELIBERATION TIME PATTERNS

TABLE 3

MESSAGE PATTERNS

The Bargaining Simulation: Subjects assumed the role of buyer in a purchasing negotiation. The seller's role was played by a programmed opponent using an interactive software program on micro-computers. (See Clopton 1984 and Schurr and Ozanne 1985 for similar approaches.) Subjects were told that they would be bargaining against each other over the price of a commodity industrial chemical called "Dexene". Use of a commodity is important to the external validity of the exercise since subjects only bargain over price (the quantity to be purchased is assumed to be fixed). In addition, it was vital to the internal validity of the experiment that the subjects believe that they were actually bargaining with a live opponent. For this reason every effort was made to enhance the mundane realism of the negotiation and its setting.

The negotiations were held in several microcomputer labs with approximately ten subjects participating in each session. The number of participants in a given session was deliberately kept small in order to ensure that subjects could be isolated physically from each other. Subjects were told that they were participating in a test of negotiating still. The most important objective of the instructions provided to the subjects was to convince them that they would actually be negotiating with a real opponent. Subjects were allowed to send written, one-line messages of their own to the seller during each turn but were not allowed to pose questions. These messages were chosen by the subjects from a list of fifteen messages that were identical (except for references to buyer or seller) to those sent by the programmed seller.

Subjects were given two incentives to maximize their outcomes in the negotiation. All subjects were told that they would receive a variable amount of course credit for participating dependent on performance in the negotiation. (In fact, all subjects received the same amount of credit.) A monetary reward was allocated to subjects also based on their performance in terms of the settlement price. To avoid deadlocks arising because the bargainers become overly zealous in pursuit of a reward, subjects were given an alternative price available from a vendor external to the negotiation which was automatically accepted if the negotiation ended without a settlement This price was not a very favorable one for the buyer ($1800), and thus if accepted, minimized the reward the bargainer could earn. The price also provided the bargainers with a point of comparison against which to evaluate the seller's offers. Subjects were given no information concerning the opponent's payoff schedule or reservation price.

After extensive pretesting and adjustment of the manipulations, this experimental procedure provided the subjects with a stimulating and realistic bargaining exercise. While a lot of time and effort was invested in enhancing the mundane realism of the experimental setting, the rewards in the form of more motivated subjects and more valid responses made these efforts worthwhile.

RESULTS

Forty-six senior, undergraduate business students participated in the negotiation. One subject in the consistent cue condition and two subjects in the inconsistent cue condition had to be dropped from most of the analyses, one due to her error (accidentally withdrawing from the negotiation by pressing the wrong key) and two due to a bug in the bargaining program (detected in the first session) which caused a premature ending of their negotiation. Final cell sizes were 21 and 22 for HIGHCON and LOWCON, respectively.

Manipulation Checks. Evidence of the successful operationalization of the cue patterns is provided by subjects' responses to several scales designed to measure subjects' perceptions of the cue patterns. Based on each cue (i.e. concessions, messages, and deliberation time), subjects were asked the extent of their agreement with statements describing the seller as becoming more reluctant to concede and tougher as the negotiation proceeded on a 7-point scale anchored by strongly disagree/strongly agree. In addition, subjects were asked the extent to which they agreed that the seller had taken more time to deliberate late in the negotiation than early on the same 7-point scale. Table 4 summarizes the treatment means for subjects' responses to each of these seven point scales anchored by strongly disagree/strongly agree. All of the contrasts between HIGHCON and LOWCON are significant at p<.01, and the mean differences are in the expected directions. As expected, subjects exposed to a consistent seller perceived the seller to be tougher, more reluctant to concede, and to deliberate longer as the negotiation progressed. No mean in the inconsistent cue condition is higher than the scale midpoint of four, nor is a mean in the consistent cue condition lower than the scale midpoint.

TABLE 4

PATTERN RECOGNITION SCALE MEANS

Further evidence concerning the validity of the consistency manipulation was provided by subjects' responses to the following item. Subjects were asked, "How consistent was the seller's behavior? In other words, were all aspects of the seller's bargaining behavior such as the pattern of concessions, the time needed to make an offer, and the messages sent consistent with each other?' Subjects' responses were recorded on a 7-point scale anchored by "not at all consistent/very consistent". Analysis of variance revealed that subjects in the two conditions perceived a clear difference in the over-all consistency of the cues provided by the seller as measured by the consistency scale. HIGHCON subjects perceived the seller to be more consistent than LOWCON subjects (m=5.42 vs. 3.82, p<.01).

Beliefs About the Seller: It was expected that bargainers in the consistent cue condition would view the seller as more trustworthy than bargainers exposed to inconsistent behaviors. Fourteen, 7-point semantic differential scales were used to evaluate subjects' beliefs about the seller. Subjects were asked to rate the seller on the following dimensions: cooperative/ competitive, deceptive/frank, strong/weak, skilled/unskilled, thoughtful/thoughtless, wise/foolish, honest/dishonest, reasonable/unreasonable, intense/mild, generous/selfish, tough/easy, professional/amateurish, trustworthy/untrustworthy, and good/bad.

Table 5 summarizes the results of a principal components analysis of these measures. Two components were identified which accounted for approximately 60% of the variance in subjects' responses. Items with loadings of at least .50 were interpreted as indicators of a latent factor. The loading of the good/bad item did not load heavily on either factor, suggesting the presence of a separate evaluative dimension. Cooperation, strength, skill, wisdom, intensity, generosity, toughness, and professionalism loaded on the first component. Deception, thoughtfulness, honesty, reasonableness, and trustworthiness loaded on the second component. The former was interpreted as toughness, and the latter was interpreted as trustworthiness. Responses to the items comprising each component were then summed. Reliability and correlation coefficients were computed on these indices. Cronbach's alpha was .74 and .91 for the trustworthiness and toughness scales, respectively. The two components were not significantly correlated (r=.115, p=.44). Differences between treatments on these summated scores were tested using analysis of variance. As expected, cue consistency had a significant effect on 'subjects' perceptions of the seller's trustworthiness (F=7.39, p<.01). The consistent seller was seen to be more trustworthy (mean=16.42) than the inconsistent seller (mean=19.77), with a lower score indicating greater trustworthiness. No differences in subjects' perceptions of the seller's toughness were expected or found (F=.63, p=.43) The means for the consistent and the inconsistent conditions were 32.13 and 34.27, respectively, with a higher score indicating a greater toughness.

TABLE 5

ROTATED FACTOR PATTERN FOR EVALUATIVE SCALES

Trust and Yielding: Bargainers who attributed greater trustworthiness to their partner were expected to yield more in the negotiation. Only limited support for this proposition was found by examining the correlation between subjects' perceptions of the seller's trustworthiness and their total concessions (r=-.26, p<.09). Further evidence concerning the strength and direction of the relationship between trust and yielding was obtained by dividing the sample into two groups based on a mean split of the trust scale. Subjects below the mean were categorized as high trusters while those above the mean were categorized as low trusters. This categorical variable was then used as a classification variable in analysis of variable. The results of this analysis indicated that high trusters conceded significantly more on average than low trusters ($549 vs. $372, F=4.14, p<.05). While no inferences concerning causality are appropriate based on these analyses, the conclusion that trust and yielding are positively related received some support. (See Schurr and Ozanne 1985 for a direct manipulation of trust.).

Effects on Bargaining Behavior: It was anticipated that consistent cues would be associated with greater yielding and fewer deadlocked negotiations. The results of this experiment provided support for these propositions. Subjects in HIGHCON made a higher final offer than subjects in LOWCON: $1471 vs. $1319 (F=3.04, p<.05, one-tailed test). In addition, bargainers faced with consistent cues were more likely to reach an agreement with the seller in the negotiation than subjects whose sellers' were inconsistent. Sixty-seven percent of the consistent cue group reached agreement while only 41%-of the inconsistent cue subjects settled within the negotiation. This difference was significant (t -1.72, p<.05, one-tailed test). This is an important result since no bargaining strategy is effective if it greatly decreases the likelihood of reaching an agreement.

DISCUSSION

One important outcome of this experiment was the clear demonstration that subjects are able to recognize accurately the cue patterns manipulated in this research. This was a concern because cue consistency theory places heavy demands on the ability and motivation of negotiators to perceive and process the complex cues provided by the behavior of their partners. This study provided strong evidence that even inexperienced negotiators are able to discern these patterns. The study also demonstrated that subjects were able to use these cue patterns to form an accurate impression of the seller's behavioral consistency as required by the theory.

As expected, consistent bargainers were perceived to be more trustworthy than behaviorally inconsistent bargainers. Subjects who believed the seller to be more trustworthy also tended to concede more. While no direct causal link between attributions of trustworthiness to the seller and bargaining behavior was demonstrated in this research, these results are certainly supportive of the proposition that attributions mediate the effect of one bargainer's behavior on the behavior of his bargaining partner. The findings also support the conclusions of Schurr and Ozanne (1985) that beliefs about the seller's trustworthiness mediate the impact of a seller's tough behavior on the behavior of a buyer. The prediction that consistent behavior would affect subjects' bargaining behavior was supported as well. In particular, it was found that subjects faced with a consistent seller tended to reach an agreement more often than subjects faced with an inconsistent seller. Similarly, the final price offered by subjects differed significantly as a function of the seller's behavioral consistency. If a key to skilled, competitive bargaining is awareness of the other party's strategy and behavioral signals, then it is reassuring that the subjects passed this test. What is not so reassuring is that some subjects also displayed the capacity to be deceived by a skilled partner who used a planned response pattern to create impressions of trustworthiness and reasonableness.

IMPLICATIONS FOR CONSUMER RESEARCH

Recently, consumer researchers have demonstrated a growing interest in studying influence in sales interactions from an information processing perspective (e.g., Sujan, Bettman, and Sujan 1986). This research should be considered one part of this broader investigation. This initial test of the effects of cue consistency on bargaining attributions and behavior suggests that behavioral consistency may be an important vehicle for influence in buyer-seller negotiations. In the future, consumer researchers should check the robustness of these effects by varying; (1) the types of behavioral cues provided in the interaction, (2) the context of the negotiation, and (3) the length of the negotiation. The importance of behavioral consistency as input to attributional processes in bargaining should be further examined by testing the effects of behavioral consistency on the latent structure of attributions (i.e., causal dimensions such as locus of causality, stability, or controllability) which were proposed in the paper by Rose and Dickson (1987). This research has demonstrated main effects of cue consistency on impressions of trustworthiness and bargaining behavior. Future research should be directed toward answering the more complex and, perhaps, more interesting question of when such effects are likely to occur. Individual difference variables such as interpersonal orientation, motivational orientation, and negotiating experience should all contribute to a bargainer's sensitivity to the behavior of his partner and to his susceptibility to interpersonal influence. Our knowledge of the roles played by behavioral consistency and attribution in determining negotiation behavior will remain rudimentary until the interactions of these variables with differences in individuals and situations are better understood.

REFERENCES

Alderson, Wroe (1957), Marketing Behavior and Executive Action, (Homewood, 111: Irwin).

Chertkoff, Jerome M. and James K. Esser (1976), "A Review of Experiments in Explicit Bargaining," Journal of Experimental Social Psychology 12, 464-486.

Clopton, Stephen (1984), "Seller and Buying Firm Factors Affecting Industrial Buyers' Negotiation Behavior and Outcomes," Journal of Marketing Research 21, 39-53.

Deutsch, Morton (1958), "Trust and Suspicion," Journal of Conflict Resolution 2, 265-279.

Druckman, Daniel, K. Zechmeister, and D. Solomon (1972), "Determinants of Behavior in a Bilateral Monopoly Situation," Behavioral Science 17, 514-531.

Dwyer, F. Robert (1984), "Are Two Better Than One? Bargaining Behavior and Outcomes in an Asymmetrical Power Relationship," Journal of Consumer Research 11, 685-683.

Dwyer, F. Robert and Orville C. Walker, Jr. (1981), "Bargaining in an Asymmetrical Power Structure," Journal of Marketing 45, 104-15.

Dwyer, F. Robert, Paul H. Schurr, and Sejo Oh (1987), "Developing Buyer-Seller Relationships," Journal of Marketing, 51 (April), 11-27.

Evans, Kenneth R. and Richard F. Beltramini (1987), "A Theoretical Model of Consumer Negotiated Pricing: An Orientation Perspective," Journal of Marketing, 51 (April), 58-73.

Johnston, Wesley J. and Thomas V. Bonoma (1984), "The Effect of Power Differences on the Outcome of Consumer Bargaining Situations," in Advances in Consumer Research 11, Thomas C. Kinnear, cd., (Ann Arbor, MI: Association for Consumer Research), 170-174.

Kelley, Harold H. (1967), "Attribution Theory in Social Psychology," in Nebraska Symposium on Motivation, cd. David Levine, (Lincoln, NB: University of Nebraska Press).

Kimmel, M. I., D. G. Pruitt, J. M. Magenau, E. Konar-Goldband, and P. J. D. Carnevale (1980), "Effects of Trust, Aspiration, and Gender on Negotiation Tactics," Journal of Personality and Social Psychology 38, 9-22.

Lewicki, Roy J. and Joseph A. Litterer (1985), Negotiation, (Homewood, Ill.: Irwin).

Loomis, James L. (1959), "Communication, the Development of Trust, and Cooperative Behavior," Human Relations 12, 305-315.

Mathews, H. Lee, David T. Wilson, and John F. Monoky, Jr. (1972), "Bargaining Behavior in a Buyer-Seller Dyad," Journal of Marketing Research, 9(February), 103-105.

McAlister, Leigh, Max H. Bazerman, and Peter Fader (1986), "Power and Goal Selling in Channel Negotiations," Journal of Marketing Research, 23(August), 228-236.

Pruitt, Dean G. (1981), Negotiation Behavior, (New York: Academic Press).

Pruitt, Dean G. and J. L. Drews (1969), "The Effect of Time Pressure, Time Elapsed, and the Opponent's Concession Rate On Behavior in Negotiation," Journal of Experimental Social Psychology 5, 43-60.

Raiffa, Howard (1982), The Art and Science of Negotiation, (Cambridge, MA: Harvard University Press).

Rose, Randall L. and Peter R. Dickson (1987), "He Says No, But Does He Really Mean It? Bargaining Behavior, Cue Consistency, and Attribution," in Advances in Consumer Research 14, P. Anderson and M. Wallendorf, eds., Provo, UT: Association for Consumer Research).

Rubin, Jeffrey Z. and Bert R. Brown (1975), The Social Psychology of Bargaining and Negotiation, (New York: Academic Press).

Schurr, Paul H. and Julie L. Ozanne (1985), "Influences on Exchange Processes: Buyers' Preconceptions of a Seller's Trustworthiness and Bargaining Toughness," Journal of Consumer Research 11, 939-955.

Siegel, Sidney and Lawrence Fouraker (1960), Bargaining and Group Decision Making, (New York: McGraw-Hill).

Smith, D. Leasel, Dean Pruitt, and Peter Carnevale (1982), "Matching and Mismatching: The Effect of Own Limit, Other's Toughness, and Time Pressure on Concession Rate in Negotiation," Journal of Personality and Social Psychology 42, 876-883.

Sujan, Mita, James R. Bettman, and Harish Sujan (1986), "Effects of Consumer Expectations on Information Processing in Selling Encounters," Journal of Marketing Research 23, 316-353.

Walker, Orville, C., Jr. (1971), "The Effects of Learning on Bargaining Behavior," in Combined Proceedings, cd. F. C. Allvine, (Chicago: American Marketing Association).

Yukl, Gary A. (1974a), "The Effects Of Situational Variables and Opponent Concessions on a Bargainer's Perceptions, Aspirations, and Concessions," Journal of Personality and Social Psychology 29, 227-236.

Yukl, Gary A. (1974b), "Effects of Opponent's Initial Offer, Concession Magnitude, and Concession Frequency on Bargaining Behavior," Journal of Personality and Social Psychology 30, 323-335.

----------------------------------------