Outcome and process accountability in negotiation: a motivated information-processing approach

Past research indicate that negotiators under outcome accountability, compared to non-accountable ones, are more prone to competitive behavior which leads to suboptimal agreements, even when there is the possibility of obtaining higher joint gain. However, recent research showed that negotiators under process accountability made more accurate estimates of the other party’s preferences and interests and obtained higher joint gain than the non-accountable counterparts. Moreover, there’s some evidence that equality in gain sharing may be moderated by social motives. The current study with professional negotiators (N = 88) focus on the effects of both outcome and process accountability on the negotiation processes in a prosocial climate. Results indicate that accountable negotiators tend to maximize the agreement’s value, thus suggesting a positive inﬂuence of the interaction of these two variables on the negotiation’s outcomes and processes. Non-accountable negotiators and negotiators held accountable only for outcome tend to get lower gains than those obtained by the negotiators under process accountability, although they are prone to divide gains more equitably. Theoretical implications of these results as well as its consequences for the negotiation practice in organizations are discussed. performance? Some signiﬁcant research on decision making suggests that accountability may be one of those factors.

behave throughout the negotiation, as if they were purely rational agents. This view was challenged by bounded rationality theory (Simon, 1955) which inspired a more descriptive approach of negotiation focused on the actual behavior and cognitions of negotiators. In fact, observations of negotiators' behavior show that they systematically violate the fundamental principles of the normative-rational theories. The descriptive approaches of negotiation indicate that even when they reach agreement, negotiators often arrive to inefficient outcomes, wasting opportunities for enlarging joint gain (Raiffa, 1982).
Most negotiations are mixed-motive situations, such that parties are motivated to cooperate with one another to reach agreement, but also to compete to claim resources (Walton & McKersie, 1965). In these situations, different patterns of priorities yield integrative potential, i.e., there's a set of possible combinations of preferences and interests which allow obtaining optimized gains for both parties. Unfortunately, negotiators often don't recognize these possibilities of maximizing joint gain and tend to reach suboptimal agreements. This difficulty has been attributed to the lack of accuracy of judgments about the other party's priorities and interests. These inaccuracies prevent them from expanding the "pie" of available resources (e.g., Thompson & Hastie, 1990, Neale & Bazerman, 1991. In fact, past research has pointed the negotiators' tendency to base their judgments on inappropriate information and faulty beliefs as the main cause of unsuccessful negotiations. For instance, even when negotiators recognize the value of mutual agreement, frequently, they fall prey to a "fixed pie assumption", that is, the biased perception that parties' interests are completely opposed (Bazerman & Neale, 1986;Thompson & Hastie, 1990) and, in consequence, they fail to agree on a valuable solution for both parties.
Although these biases are robust and pervasive, some social context factors may have a de-biasing effect in negotiators' cognition, leading them to recognize some compatible interests. So, a critical question is: what factors may have a de-biasing effect which, in turn, would allow negotiators to escape from the systematic cognitive biases that usually hinder performance? Some significant research on decision making suggests that accountability may be one of those factors.

Accountability in negotiation
In human interdependent activities it is difficult to escape the evaluative scrutiny of others. Accountability is a permanent feature of organizational life, leading individuals to act in line with prevailing expectations from significant others and to anticipate justifications for deviant behaviors (Tetlock, 1998;1999). In the broadest sense, accountability occurs whenever the individual has an expectation, implicit or explicit, to be called upon to justify their actions, beliefs or feelings before others. However, in a finer approach, it appears that this is a multidimensional phenomenon and can be activated in different ways (Lerner & Tetlock, 1999). For example, it occurs when someone has the expectation that her performance is being observed by others or simply when she expects to be evaluated by unknown or known others. The fact that individuals have the expectation of being personally identified with their actions also constitutes a support to accountability.
As an almost ubiquitous context variable, accountability would seem to have an obvious relevance in organizational negotiation: for logistic reasons, in organizations, negotiators act on behalf of other people (supervisors, shareholders) and are accountable to them. However, cognitive paradigm, witch shaped most of negotiation research in the last decades (Bazerman et al., 2000), has relegated the study of accountability effects to a secondary plan. Ironically, the typical design of those studies about social cognition in negotiation is close to the framework of economic models based on games theory: "strangers who meet once, interact strategically and then they go their separate ways" (Kramer et al., 1993, p. 639). Accountability, being treated as an undesirable interfering variable in the experimental plan, was annulled deliberately and overlooked in data analysis.
Diverging from this orientation, recent literature points to a more systematic attention to the effects of social context on cognitive processes in negotiation.
However, interconnectedness of these contextual factors make difficult to understand their specific effects. Therefore, the inherent methodological difficulties recommend researchers to focus on specific variables which occur more visibly and frequently in the social context of the negotiations. Accountability undoubtedly meets this condition: in organizational contexts, rarely negotiators act on behalf of their own interests. On the contrary, negotiations are embedded in social networks which evoke, explicitly or implicitly, the way significant others judge their results and actions.
Past research indicate that negotiators who act on behalf of others, and must justify the agreements they obtain, tend to make fewer concessions, be more competitive, and get sub-optimal agreements more often than the non-accountable ones (Rubin & Brown, 1975, Pruitt & Carnevale, 1993. The tendency to exacerbate this competitive orientation appears to be directly proportional to the social pressure degree of the constituents (Ben-Yoav & Pruitt 1984b).
The most usual explanation of these results has been based on the effects that the representative role could exert on the behavior of negotiators (Enzle, Harvey, & Wright, 1992). For instance, Blake and Mouton (1961) suggested that the implicit obligation imposed by the constituent and representative role on the one hand, the pressures associated to explicit social or financial rewards, on the other hand, will inevitably lead accountable negotiators to use inflexible and competitive negotiation strategies. As representative, the negotiator would assume an implicit contract to defend the interests of the constituency, and this assumption would be enough to evoke a strong win-lose orientation (Druckman, Solomon & Zechmeister, 1972), even in the absence of other external pressures. The accountable negotiator would be much more competitive the higher the solidity of the link with a significant in-group (Breaugh & Klimoski, 1977), the lower the confidence of the constituents (Wall, 1975) and the lower the social status of the negotiator (Kogan, Lamm & Trommsdorff, 1972). Another explanation is related to the negotiator's perception that he or she is been evaluated. Because accountability makes individual identity salient, accountable negotiators would seek out strategies which allow them to save face through behaviors which are deemed appropriate to guarantee a positive judgment on the part of the constituents, even when those are unknown (Gruder, 1971).
This approach seems to suggest an inevitable association between accountability and competitiveness in negotiation, and it leaves little room for the way in which accountable individuals perceive their relationship with the constituents, as well as the nature of subsequent expectations about consequences of this relationship. On the other hand, in organizational adhocratic contexts (Mintzberg, 1979), or in which there are weak hierarchical links and diminished power distance, accountability relationships tend to occur in a more fuzzy and less formal way than assumed in the literature, thus suggesting the need of a greater attention to how accountable individuals cognitively construct the situation.

Outcome and process accountability
Decision making approaches (e.g., Tetlock, 1999;Lerner & Tetlock, 1999) posit a major distinction in accountability condition: individuals may be held accountable for the outcomes of their decisions (what they've got) or/ and they may be under scrutiny for the process they use to make decisions (how they've decided).
Research suggests that individuals under process accountability (individuals who believe they must justify the way they make decisions) are resistant to ordinary biases because they analyze the available information more carefully, and are more self-critical concerning the decision-making process. Research on negotiation also showed that process accountability contributes to enhance epistemic motivation (Kruglanski & Webster, 1996) and individuals are more likely to engage in systematic information processing. This enhanced cognitive effort and reduces some common biases on the negotiator judgment such as the fixed-pie perception (the inclination to believe that own and other's interests and outcomes are diametrically opposed). As a result, accountable negotiators reach more integrative agreements than the non-accountable ones (De Dreu, Koole & Steinel, 2000;Simões, 2005).
In organizational decision making, Simonson and Staw (1992) showed that outcome accountability, compared with process accountability, enhances the decision maker's propensity to engage in escalation of commitment. The authors argue that outcome accountability induces the need for self-justification causing a tendency to defend past lines of action. Instead, accountability process is associated with a greater propensity to weigh alternatives carefully, decreasing the importance of self-justification, once the expectation of evaluation by others is centered in decisions' strategy choice to the detriment of the effectiveness of the decision in itself. Another line of distinction between the effects of process and outcome accountability is based on the differentiation of the motivational potential of each of these conditions. For example, Siegel-Jacobs and Yates (1996) argue that accountability process contains the tacit suggestion that the improvement of strategies will be positively reflected on performance. Therefore, it is a stronger incentive than outcome accountability which, despite the anticipation of reward, involves a high level of uncertainty.

Information processing and social motivation
Integrative negotiation is a cognitively demanding activity in terms of information analysis and problem-solving skills. To the extent that it fosters openness and information exchange, cooperation may facilitate integrative agreements. But, accordingly to dual concern model (Pruitt & Rubin, 1986) this can only happen when negotiators have a problem-solving orientation, which imply a combination of high concern for one's own outcomes with a high concern for the other party's outcomes.
Negotiator's cognitions seem to be influenced by social motivation directly.
Individuals tend to seek, encode and retrieve information which is consistent with their social motivation. Specifically, negotiators tend to look for confirmatory information, consistent with their motivational orientation: prosocial individuals select information associated with equality, consensus and joint gain, whereas people with a selfish orientation are prone to confirm their egocentric beliefs, (e.g., De Dreu & Boles, 1998).

A study on the effects of accountability on negotiating performance
The association between accountability and the difficulty in obtaining integrative agreements assumed by classical research does not appear to verify when accountability focuses on the process. On the other hand, the double concern model suggests that outcome accountability has opposite effects depending on the nature of social motivation of the negotiators. In fact, it interacts positively with prosocial orientation in achieving high joint gain, contrary to what happen with accountable negotiators who are oriented only to self-interest. In this view, the propensity to build integrative agreements in negotiators with prosocial motivation and held accountable for outcomes is interpreted as reflecting the interaction between the level of aspiration, induced by high accountability, and encouragement of a problem solving orientation promoted by the interests of the other party. So, resistance to yielding (associated with a high level of aspiration) would be a crucial variable to explain the increase of joint gain when social motivation is high. However, the dual concern model does not mention anything about the consequences of this cognitive process. Despite the absence of direct empirical evidence, De Dreu and Carnevale (2003) suggested that high levels of aspiration may be associated with the strengthening of epistemic motivation. On the other hand, other studies (De Dreu et al., 2001;Simões, 2005) established that the positive influence of process accountability on joint gain is mediated by the increase of accuracy in information processing. This influence appears to be independent of the effects of social motivation. But the association between social motivation and process accountability moderately favors the discovery of integrative potential in the negotiations. Hence, we assume that the effect of outcome accountability may be similar to accountability process, but only when negotiators have a prosocial orientation. Since there is no specific research on this issue, to date, and as far as we know, it is possible to hypothesize that the effects of two types of accountability can be merged together, enlarging mutual gains as a result of an orientation that could be named "intelligent cooperation" 2 .
Past research has also shown that negotiators, who simultaneously display high social motivation and are under process accountability, have a propensity to divide the gains more evenly than accountable negotiators who have low social motivation (Simões, 2005). Those findings suggest that the same may happen with outcome accountability. On the other hand, when negotiators share a social identity, and are accountable for the outcomes of the other party, they tend to conclude egalitarian agreements (Kramer, Newton, &.Pommerenke, 1993). Assuming that the effects of the salience of common social identity and social motivation may be identical, we believe that the negotiators who are accountable for outcomes and urged by a third party to cooperate with the opponent may also engage in an equitable gains distribution.
The simultaneous influence of the two types of accountability may also have organizational relevance. The intricate web of interdependence in today's organizations and increased concern about future interaction with the other party (today's opponents may be tomorrow's partners and vice-versa) lead more and more to situations in which negotiators attach high utility to a cooperative climate, whilst feeling strongly accountable not only for outcomes but also for the negotiation process. Additionally, volatility of organizational functions and roles generates in negotiator's constituents a more explicit concern about the negotiation process ("how") and not just about the outcomes ("what" or "how many").

Objective and hypotheses
This study aims to examine the specific effects of the interaction between process and outcome accountability on the negotiation performance when negotiators have a prosocial orientation. Considerations of the previous section lead us to formulate the following hypotheses: Hypothesis 1: When having high social motivation, negotiators accountable for both process and outcome will get higher joint gain than negotiators held accountable for outcomes only; Hypothesis 2: When having high social motivation, negotiators held accountable for outcomes show more equal gain sharing than those who are simultaneously under process and outcome.

Experimental plan
To test these hypotheses we built a 2 factor (outcome accountability high vs. low) x 2 (process accountability: high vs. low) plan. Since we intended to examine the effects of two types of accountability in a prosocial climate, and in order to ensure the generalization of this condition, all participants were induced to cooperate with the other party during the interaction. To this end we used two procedures cumulatively: participants were informed that they would work together on a task of mutual interest the next day (induction of expected future cooperation) and, at the same time, they were instructed to collaborate with the other party during negotiations.
To check the effectiveness of the procedure, after negotiation participants answered a short questionnaire consisting of four questions that seek to assess 1) the extent to which participants believed in future cooperation with the other party and to what extent the negotiator had contributed 2) to help the other party reach her goals, 3) to attain a good performance during the negotiation, and 4) felt satisfied with the agreement. These last two indicators are similar to those used by emphasized instructions (e.g,, Giebels, De Dreu, & van de Vliert, 2003).
The questions were answered using a seven-point scale (1 -strongly disagree to 7 -strongly agree). These last four indicators were aggregated and averaged into a single index ( = .63).

Participants
Eighty-eight professionals 3 (42% were women), from the service and industry sectors, and who were usually involved in negotiations, participated in the study. Forty-four dyads were formed and reached an agreement within the assigned time.

Task and procedure
Each dyad negotiated face-to-face for up to 20 minutes, without any restriction in the communication process, except for the interdiction of accessing the specific written information of the other party, with the aim of reaching an agreement on negotiating a setting of five items. This one is an adaptation of a task used in previous studies (O'Connor, 1997;Thompson, 1990) concerning a labor negotiation and involving two parties: a candidate (already selected) to a job and an employer's representative 4 . Each item had five alternative proposals. Each alternative corresponds to a certain point value which determines the relative preferences of each party. Participants were informed that they should negotiate by using only the 25 alternatives (5 items x 5 alternatives) included in the tables provided. Roles were randomly assigned and the 3 Initially, the number of participants was 94. However, after performing the negotiation task, three dyads had displayed a value of social motivation significantly lower than the others. Since the pro-social environment is treated as a general condition in this study, these three dyads were excluded from the sample.
4 The task involves organizational roles which may interfere with the understanding of the structure of negotiation items and priorities. So, we've taken an additional procedure in order to neutralize some possible effects of social role, particularly those that might arise from the perception of power asymmetry. At this end, before reading the initial information about the task, as described, participants received the following information: "On the part of the company, the recruitment of this element is complete. In turn, the candidate is only interest in this company. So, the negotiation is intended solely to settle the conditions of their employment". This information was reinforced in the written instructions. task was orally presented and explained in general terms. Then, participants received a set of written information, including the task description, the payoff tables corresponding to each role as well as information to manipulate the independent variables. All doubts were cleared within the limits of the experimental conditions.

Process accountability
Two levels of process accountability (high and low) were implemented by using standard experimental manipulation from the relevant literature. Participants were (not) informed that a group of experts would examine how they would lead negotiation, and would examine the procedures they followed, the decisions they made and the strategies they adopted.

Outcome Accountability
This variable was manipulated (not) informing the participants that the quality of the agreement would be analyzed by a group of experts. Adopting a procedure used by De Dreu et al. (2000), the manipulation of both types of accountability has been strengthened by informing participants that they would be interviewed the next day by a negotiation expert called for that purpose. So they could take personal notes both during and after negotiation, in view of the "interview". At that end, they were delivered a sheet of paper entitled "Notes for the interview," including one or two fields, depending on the conditions: "Notes on my outcomes"and / or "Notes on my way of negotiating"

Judgment accuracy
Judgment accuracy (the first dependent variable) was measured by calculating, for each party, the difference between her estimate of the other party's preferences and the other party's actual preferences. At the end of the negotiation, participants were given a table similar to that indicated the same preferences that they used in negotiating, with the difference that this table did not indicate the points assigned to each alternative. Then they were asked to complete this "empty" table, assigning the estimate of the number of points in the table of the opponent. They were also informed about the maximum and minimum values mentioned in both tables (400 and 0 points, respectively). Following the suggestions of Thompson and Hastie (1990), the measure of judgment accuracy took only into account the participant's estimates of the two integrative items and the common value item. The measure was constructed by calculating, first, the sum of absolute differences between the estimates of each alternative and the actual values, and this sum was then divided by the maximum possible inaccuracy value, measured in points (2240), which allowed to obtain a measure of imprecision. Finally, that value was subtracted from the unit, allowing a more accurate indicator expressed as a value ranging from 0 (complete inaccuracy) and 1 (full accuracy).

Negotiation performance
Concerning the negotiation performance (the second dependent variable) two main measures were used: joint gain, i.e., the sum of points scored by each party in the final agreement; and gain distribution for each dyad, the ratio of the partial gains (highest value/ lowest value). The complete equality corresponds to 1, and, therefore, the less egalitarian the agreement the more the value of this indicator is departing from 1.
To allow detailed analysis of the results were still negotiating collected other measures of performance: level of aspiration, measured by value of the first offer, and resistance to yielding, measured by the difference between the values of the first offer and the individual value of agreement.

Results
Negotiation is an interdependent activity, implying that the results of each participant depend on the results of the other party and vice-versa. Therefore, in treatment of the data dyad was used as the unit of analysis. The measures used in this study meet a typical pattern of dyadic variables, that is to say, they vary within each dyad and between dyads (Kenney, 2001). Thus, following an established procedure in research on dyadic negotiations (e.g., Gelfand & Realo, 1999), we averaged the scores of both negotiators to create a dyad measure. Exceptions to this procedure are the measures that are dyadic, by definition, that is, the result of a specific relation (sum, difference) involving the individual values, as described in the preceding section. PSYCHOLOGICA, 2011, 55 Manipulation check Measured effects of the emphasized instructions and the induction of future cooperation in order to obtain a prosocial orientation in every cell of the experimental design are in line with the intended, as can be verified by the mean indicator of prosocial (minimum 1, maximum 7) in Table1. Post hoc tests (p <.05) indicate no significant differences between dyads of the four experimental cells. The analysis of participants' responses to manipulation checks questions indicate that it was positive for both independent variables. Process accountable participants displayed a belief of being scrutinized for the way they negotiate (M=4.2) significantly stronger (F (1,84) = 38.2, p <.01) than the non-accountable (M=2.04).

361
Also, participants who were told that only the outcomes of the agreement would be scrutinized differ significantly (F (1,84) = 86.9, p <.001, M = 5.1) from those who did not receive this information (M = 1.89).

Hypothesis 1
The hypothesis that negotiators who were accountable for both process and outcome would obtain higher joint gain than those who had only been held accountable for outcomes received support. Univariate ANOVAs revealed significant effects of process accountability (F (1,42) = 47.7, p <.001) and outcome accountability (F (1,42) = 49.8, p <.03) on joint gain . Comparing the effects of both variables on joint gain through Tukey's post hoc tests (p <.05) confirmed that the dyads who were accountable only for the process obtained higher joint gain (M = 1150) than (M = 968) the non-accountable negotiators (M = 916). The same applies to the difference between these two groups of dyads and those accountable for both process and outcome (M = 1326). There are no significant differences between the gains obtained by non accountable dyads and outcome accountability dyads. The analysis of joint gain revealed an interaction effect between the two independent variables (F (1, 40) = 5.37; p <.03), indicating that negotiators held accountable for the process maximize joint gain when they are simultaneously accountable for outcome. This result seems to support our main hypothesis. However, taking into account that negotiators who were only held accountable for outcome obtained join gains which were similar to those obtained by non-accountable dyads and lower than process accountable negotiators, this interaction effect suggest that these two types accountability have different influences on joint gain. Apparently, outcome accountability has only significant influence on the achievement of joint gain when coupled with process accountability.

Accuracy of judgment
The accuracy of judgment does not seem to be influenced by outcome accountability, F (1,42) =. 29, p <.59, ns. Rather, confirming the results of previous studies (e.g., Simões, 2005), we found a significant effect of process accountability F (1, 42) = 28.08, p <.001), indicating that accountable negotiators made more accurate judgments of the other party's preferences (M =. 64) than the non-accountable counterparts (M =. 39). Since those previous results highlighted the mediating role of the accuracy in judging of the other party's preferences on the achievement of joint gain, we also analyzed this possibility in the present study, following the steps prescribed by Baron and Kenney (1986). The first regression showed that the accountability process is correlated with joint gain ( =. 62, p <.01) and, separately, with the accuracy ( =. 54, p <.01). The integration of independent variables and the hypothetical mediator in the equation led to the breakdown of significance of the effect of accountability ( =. 14, p =. 30, ns), confirming that accuracy of the judgments about the preferences of the other party mediates the achievement of joint gain in negotiations in which individuals are under process accountability. The mediation test on outcome accountability has been discarded because there is no significant correlation between this independent variable and judgment accuracy, the hypothesized mediator.

Hypothesis 2
Descriptive statistics points to a general tendency to equitable outcomes in all cells of the experimental design, regardless of the joint gain level. However, analysis show a main effect of outcome accountability on gain distribution indicator, F (1,42) = 4.09, p <.03. In line with our prediction, the dyads in this condition have divided gains more e equitably (M = 1.04, SD =. 6) than those who did not feel accountable (M = 1.26, SD =.5). A similar analysis on process accountability has not proved significant (F (1,42) = 09.01, p =. 43, ns), and the same is true regarding the an interaction effect between these two variables, F (1,40) = 2.08, p =. 15, ns.

Discussion
First of all, results confirm that process accountability enhances the discovery of integrative agreements via an increase in accuracy in estimating the preferences of the other party. This effect is associated with higher joint gain as noted in previous studies (De Dreu et al, 2000;Simões, 2005). Secondly, when having a high social motivation, participants simultaneously held accountable for process and outcome obtained higher joint gain, suggesting the existence of a positive interaction between the two variables in creating value in negotiation.
Concerning gain distribution, results indicate that non-accountable dyads tend to be more egalitarian, although they present lower joint gains, which is in line with what was observed in past research. The most salient finding, in our view, refers to the fact that the association of prosocial climate and outcome accountability can also lead to a tendency toward equal division of the gain, but departing from a lower level of joint gains comparing to the negotiators who are also under process accountability. On the other hand, the most intriguing result, and not anticipated in our hypotheses, refers to the low value of the joint gain achieved by dyads solely accountable for the outcomes, which apparently contradicts classic research which predicts high joint gain in this condition when social motivation is high. However, in those studies, outcome accountability was manipulated through instructions that emphasized the representative role and, on the other hand, evoked explicit normative control of outcomes on the part of the constituents. In a seminal study in this field (Ben-Yoav & Pruitt, 1984a), highly accountable participants acted on behalf of a company and the points they obtained in the agreement were transformed into real money which would be divided between participants and the company, according to the evaluative judgment of the company's owner. In the high accountability condition, participants were told that the criteria for evaluating the effectiveness depended exclusively on the owner's view and that he would write a brief assessment of their effec-tiveness in negotiation. The effects of the combination of social motivation and accountability were, in this study, primarily explained as "creative impact" in the negotiators' reaction to the conflict between the needs of their constituents and responding to the needs of the opposing side " (Ben-Yoav & Pruitt, 1984a, p. 293).
Differently, in our study, accountability refers to unknown constituents and does not evoke norms involving explicit and direct consequences for negotiators. So, the divergence of results concerning joint gain may correspond to a difference in the nature of accountability pressures. In fact, according to the model of social contingency (Tetlock, 1999), accountability to constituents whose orientations are unknown tend to activate preventive self-criticism and stimulates epistemic motivation. Conversely, when the constituents' views are known, there is the tendency of individuals to adopt positions and choices which they believe that are in accordance to the expectations of the constituents. Thus, in the same way as the competitive orientation of accountable negotiators (indicated by classic research) may result from the assumption that constituents expect from them a "hard" negotiation behavior, participants of our study, who were held accountable for outcomes, might have assumed that the expectation of unknown constituents pointed to the equal division of resources. We suspect that the instructions of collaboration with the other party might have been retrospectively interpreted as tacit expectations of equality. Participants will have possibly focused on the outcome, as gain equilibrium, than on quality of the agreement as a whole. The fact that this egalitarian tendency has proved to be much less significant, and in tandem with the achievement of higher joint gain, in dyads accountable for the way they've negotiated seems to indicate that focusing in the process of negotiating, rather than on outcomes, serves to maximize the benefits of a prosocial climate in negotiation.

Limitations of study
The participation of professional negotiators in experimental studies can be seen as an essential desideratum to increase the ecological validity of results which is not negligible in a field with high importance is the organizational and negotiation. However, the risk of possible interference of factors not subject to experimental control, such as personal beliefs and "mental models" prevailing in the organizations of origin, advises caution in generalizing these results. However, it should be noted that the clearest results are consistent with those obtained in previous studies with students, namely the confirmation that the process accountability positively influences the accuracy in information processing, which, in turn, is conducive to achieving higher joint gain.

Conclusions and practical implications
This study confirms that process accountability affects how negotiators process information, leading them to analyze it more accurately. Process accountability is a factor in the social context of negotiations of that influences the epistemic motivation, similarly to what literature on individual decision-making has suggested.
As in previous studies (e.g., Simões, 2005) we found no significant indications that the level of resistance to yielding is associated with epistemic motivation, unlike the prediction of the double concern model. Since the measure of resistance to yielding referred to a level of aspiration spontaneously displayed by the participants, future research, in which resistance to yielding being manipulated by (not) establishing high levels of aspiration (e.g., Ben-Yoav & Pruitt 1984b), may shed new light on of this issue. Outcome accountability has not the same effect when negotiators interact in a prosocial climate. It may constitute an ambiguous construct referring to a situation which essence may change contingently with the negotiators' perception about what they believe to be the specific expectations of the constituents, and so this is an issue that requires further investigation.
Apparently, when held accountable for both outcome and process, negotiators with high social motivation tend to coordinate satisfaction of self-interest with maximizing joint gain, taking into account the interests of the opponent. And they get it while considering information on the structure of the items and the preferences of both parties. Put it another way, the prosocial orientation seems to favor the maximization of joint gain but only when epistemic motivation is high.
Prescribing some generic relational skills development to foster cooperation between negotiators, as is currently frequent in training, may be insufficient to ensure the preparation of professional negotiators to enhance the quality of agreements. Whilst the skills to promote a positive interpersonal relationship may be a factor for facilitating the exchange of information, further analysis of this, particularly in structurally complex negotiations, depends on a set of contextual factors. Among these, some may be deliberately changed. Our results suggest that accountability for the way they negotiate lead practitioners to consider the scope a greater number of alternatives by deepening the processing of information available, either in situ or by using data processing systems to support decision making.
Contrary to outcome accountability which refers almost exclusively to the sphere of relationships between management and negotiator, and focus on the control of objectives, it is our understanding that process accountability fall within the broader context of organizational cultural values and standards. For example, in an environment that promotes organizational learning, i.e., where knowledge sharing is valued regularly and collectively, it should be normal for individuals to be under high accountability not only for what they get but also for the ways they succeed. However, as mentioned above, the construction of accountability links is contingent on how individuals perceive the relationship with the constituents. Hence, in the current state of research, it would be unwise to set a strict organizational policy that does not take into account the characteristics of the specific micro-context. But, as a general rule, accountability seems to strongly foster negotiation outcomes and processes, especially in a cooperative environment.