The effects of the leader–member exchange relationship on rater accountability: A conceptual approach

Abstract In contemporary organizational psychology research, the effect of the leader–membership exchange (LMX) relationship on rater accountability has received little attention. The current study seeks to fill this gap in the literature by explaining how the quality of the LMX relationship between the rater and the audience of appraisal affects rater accountability. Moreover, this paper examines how the upward strategies of both the rater and the ratee influence LMX relationships in the appraisal context. Employee resistance in performance appraisal is assumed to moderate the relationship between LMX and rater accountability. A conceptual model depicting the relationship between upward influence tactics, LMX relationship quality, rater accountability and performance rating is presented.


PUBLIC INTEREST STATEMENT
In organizational psychology research, the dyadic relationship between a leader and followers in an organization is called the leader-membership exchange (LMX) relationship. The LMX can be divided into high-and low-quality relationships. Followers are incentivized to engage in high-quality LMX relationships with their supervisor because the members of this relationship enjoy positive features such as trust, liking and a favourable position towards each other. The LMX can influence supervisors' evaluation of subordinates to assign performance ratings. Specifically, LMX can affect the rater accountability of supervisors because raters tend to evaluate ratees who are engaged in highquality LMX relationships with them. Therefore, the subordinates, as ratees, attempt to construct highquality relationships by exercising upward influence tactics such as ingratiation and self-promotion. The current study examines the relationship between ratees' upward influence behaviours, the LMX relationship and rater accountability.
potentially is subject to observation and evaluation by some audience(s), including one's self" (Frink & Klimoski, 2004, p. 3;Hall et al., 2003). Thus far, the main interest of accountability research has been to examine the role perceived accountability plays in determining individual behaviour and performance (Mero, Guidice, & Werner, 2014). In the performance appraisal context, in which felt accountability is regarded as rater accountability, the accountability of raters to the audiences (e.g. ratee, rater's supervisor) in the performance appraisal process affects the accuracy of the appraisal (Park, 2016).
Ample evidence in the literature shows that raters who feel more accountable tend to comply with the audience's view in making appraisal decisions by either following the audience's view or preparing to justify the accuracy of the performance rating to the audience (Mero & Motowidlo, 1995;Mero, Motowidlo, & Anna, 2003). As a result, although studies on the relationship between rater accountability and performance appraisal have flourished, research on the antecedents of rater accountability have received relatively little attention. Only a few scholars, such as Mero et al. (2014), have conducted studies on the antecedents of perceived accountability, including managerial monitoring behaviour. For that reason, in the rater accountability literature, there is no specific stream of research that addresses the antecedents of rater accountability because the literature assumes that accountability is formed as a reaction to the view of the audience. This assumption fails to consider the possibility of the quality of the relationship between the rater and the audiences of appraisal. Although raters may respond to the view of the audience, their level of perceived accountability may vary depending on the level of quality (e.g. high or low) of the relationship with each audience.
The current study seeks to fill this gap in the literature by proposing the quality of relationship between the rater and audiences as the antecedent of rater accountability. The audience of the appraisal process is defined as the ratee and rater's supervisor. The rater's supervisor is one type of audience because the supervisor is ultimately responsible for the appraisal decisions made by the rater. Other parties such as colleagues are excluded, as they do not have a direct interest in the appraisal process and are not allowed to attain information about the ratee's appraisal results and ratings.
The current study is conducted in the following order. First, the current study provides a brief literature review of rater accountability and leader-member exchange (LMX) theory. Next, it presents a conceptual model representing the relationship between the LMX relationship and rater accountability. In this process, it also illustrates how the upward tactics of both the rater and the ratee influence LMX relationships in the appraisal context. Lastly, theoretical contributions and organizational implications are suggested.

Rater accountability
Rater accountability is a construct that depicts felt accountability in the performance appraisal context. Past organizational psychology research has defined felt accountability as "an implicit or explicit expectation that one's decisions or actions will be subject to evaluation by some salient audience(s) (including oneself), with the belief that there exists the potential for either rewards or sanctions based on this expected evaluation" (Frink & Klimoski, 2004;Hall et al., 2006, p. 88;Tetlock, 1985Tetlock, , 1992. Therefore, contemporary organizational accountability research (Breaux, Munyon, Hochwarter, & Ferris, 2009;Frink & Klimoski, 2004;Mero, Guidice, & Brownlee, 2007) has shown that the felt accountability of employees is rooted in the expectations of the audiences to whom the individuals are potentially held accountable. The audiences and the employees held accountable form a social system that sets shared expectations between the audiences and accountable employees (Frink & Klimoski, 1998;Tetlock, 1985Tetlock, , 1992. These shared expectations are crucial for maintaining the social system; thus, organizations establish formal mechanisms, such as a formal supervision relationship, performance appraisal, and a financial incentive system, to manage employee accountability (Frink & Klimoski, 2004).
Rater accountability is similar to the definitions and concepts of felt accountability, but its application is limited to the performance appraisal context. Levy and Williams (2004, p. 887) define rater accountability as "the extent to which a rater is held answerable to someone else for his or her ratings of another employee." Raters prepare for the potential possibility of being held accountable for the ratings and tend to make appraisal decisions that are most acceptable to the audiences, including the rater's subordinate (ratee) and supervisor (Klimoski & Inks, 1990;Park, 2014). The rater's response in the appraisal process differs depending on the audience to which they may be held accountable. The ratee is the audience with the highest interest in the performance appraisal outcome. Raters, as supervisors, have the duty to provide annual or other periodic feedback to the ratees regarding their performance appraisal results. In the feedback process, whether the feedback takes the form of a face-to-face discussion or a written document, justification or rationale for the performance rating is required. Consequently, without exception, raters tend to feel accountable to ratees. Research indicates that raters are motivated to adjust the performance ratings of poor performers when raters must confront poor performers about their low performance (Klimoski & Inks, 1990;Levy & Williams, 2004;Shore & Tashchian, 2002). Therefore, if ratees are the audience, facing the possibility of being held accountable by the ratee results in raters providing more positive ratings to reduce the chances of having to justify their appraisal to the ratees and to increase the acceptance of the rating by the ratee (Katz & Kahn, 1978;Mero et al., 2007;Murphy & Cleveland, 1995;Park, 2016).
If the rater's supervisor is the audience, raters provide more accurate ratings when they know their supervisors prefer accurate ratings and more positive ratings when the supervisors express favourability for higher ratings (Mero & Motowidlo, 1995;Mero et al., 2007). However, some scholars contend that even though raters' supervisors are ultimately responsible for the final ratings assigned by the raters, the supervisors rarely object to or disagree with the ratings (Bernardin & Villanova, 1986;Harris, 1994).
Despite the abundant work on the audiences of appraisal, the majority of extant research has conducted experimental studies using student populations (non-working samples) to observe rater behaviour (Park, 2016). For that reason, the effect of the task environment in the appraisal context has been neglected because students have been instructed to respond to a simulated setting in which they were asked to justify their decisions to others (Mero et al., 2014). In those simulated settings, participants who perform the role of raters do not share work relationships with the ratees as they have never actually worked together. In practice, the performance rater is likely to form unique relationships with each member of the audience, and this relationship can affect the task environment represented by the performance appraisal process. Furthermore, past research has not yet considered the relationship between audiences as a determinant of rater accountability. Few scholars, such as Harris (1994), have noted that the rater's supervisor may have a particular position (e.g. positive or negative) about the ratee that may decrease accuracy, but these studies have not advanced to discuss the relationship between the rater's supervisor and the ratee.
In sum, until now, rater accountability research has not addressed the influence of the relationship between the rater and audiences with regard to accountability. From the above discussion, the current study draws on the basic premises for rater accountability before presenting the conceptual model. First, rater accountability is a mediator between relationships with the audience and the performance rating. For audiences (the ratee and the rater's supervisor) to hold the rater accountable, the actual result of the performance rating is required for the audience to judge the accuracy and for the raters to justify their rating decision. Second, both good and poor performers are defined as audiences because all employees receive performance ratings from their supervisor and may request justification for their ratings. Lastly, accountable raters may intend to please the audiences rather than merely attempting to be as objective as possible in performance appraisals.
The current study seeks to expand the rater accountability literature by introducing the LMX theory to propose that the quality of the relationships among the rater, the ratee, and the rater's supervisor may affect rater accountability.

The leader-member exchange theory (LMX) of leadership
The basic premise of the LMX theory is that the relationship between the supervisor and the subordinate is dyadic and that each supervisor-subordinate dyadic exchange relationship is unique (Elicker, Levy, & Hall, 2006;Graen & Uhl-Bien, 1995). Supervisors construct unique relationships with each subordinate by differentiating their interactions with each subordinate (Dulebohn, Bommer, Liden, Brouer, & Ferris, 2011). The unique exchange relationship between the supervisor and the subordinate varies depending on the relationship quality and can be divided into two categories of relationship groups: high-quality and low-quality LMX relationships (Graen & Cashman, 1975;Graen, Novak, & Sommerkamp, 1982). According to the social network perspective on LMX, high-quality LMX relationships represent "commitment to one another, a willingness to support one another, and confidence in the other's loyalty" (Goodwin, Bowler, & Whittington, 2009, p. 957). Supervisors and subordinates in low-quality LMX relationships tend to have fewer positive features of the high-quality LMX relationship. The supervisor and subordinate who share a high-quality LMX relationship experience "mutual trust, liking, loyalty, professional respect, and reciprocal behaviours that benefit their supervisor" (Rosen, Harris, & Kacmar, 2011, p. 822). Subordinates in the high-quality LMX group have more positive interactions with their supervisors (leaders) and receive more attention, information and resources than do low-quality LMX members (Elicker et al., 2006;Graen et al., 1982). For this reason, low-quality LMX members tend to experience more negative consequences and interactions than do high-quality LMX members (Rosen et al., 2011).
One topic that has gained increasing interest in the current LMX literature is how LMX relationship quality affects the performance appraisal process between rater and ratee pairs across all levels of the organization. Elicker et al. (2006, pp. 533-534) proposed several focal points regarding the comparison between high-and low-quality LMX relationships in the appraisal context. First, ratees in high-quality LMX relationships participate in performance appraisal sessions with a high level of trust and a positive attitude due to receiving more work-related resources and attention from their supervisor (Liden, Wayne, & Stilwell, 1993). Second, high-quality LMX relationship ratees have more voice, which means "the opportunity to present one's point of view," in the appraisal process and can influence the final appraisal decision using that voice. Lastly, high-quality LMX ratees possess more relevant information about the performance appraisal session based on frequent prior communication with the supervisor and can receive better treatment in the appraisal process.
Based on the discussion above and accepting the existence of high-and low-quality LMX relationships, the current study focuses on the determinants of the quality of LMX relationships. Specifically, upward influence behaviour, a type of employee political skill that is well known as an antecedent of LMX, is used in the performance appraisal context (Dulebohn et al., 2011). Political skill in organizations is often defined as "the ability to effectively understand others at work and to use such knowledge to influence others to act in ways that enhance one's personal and/or organizational objectives" (Ferris et al., 2007, p. 291). Ferris et al. (2007, p. 306) proposed an interpersonal process approach to conceptualize the effect of political skill on others at work. They suggested the concept of upward appeal, which "involves obtaining the support of individuals higher up in the organization hierarchy." Similarly, LMX research views such upward appeal as "upward influence behaviour," which illustrates the purposeful behaviour of followers to make leaders form positive perceptions of them and to enhance the quality of LMX relationships (Dienesch & Liden, 1986;Dulebohn et al., 2011, p. 10;Lord, 1985;Schriesheim, Castro, & Yammarino, 2000).
To conduct influence behaviour, one must use interpersonal influence tactics to effectively exert influence on persons with authority. The categorization of influence tactics began with the work of Kipnis, Schmidt, and Wilkinson (1980), which included eight influence tactics. This categorization was refined by Kipnis and Schmidt (1988) to six tactics (Higgins, Judge, & Ferris, 2003). According to Higgins et al. (2003, p. 91), these tactics include the following: (1) assertiveness: using a forceful manner to get what one wants, (2) ingratiation: using behaviours designed to increase the target's liking of oneself or to make oneself appear friendly in order to get what one wants, (3) rationality: using data and information to make a logical argument supporting one' request, (4) exchange: making an explicit offer to do something for another in exchange for their doing what one wants, (5) upward appeals: relying on the chain of command, calling in superiors to help get one's way, (6) coalitions: mobilizing others to help in persuading the target individual.
The early work on influence tactics has been revised for application to the upward influence behaviour context by Dulebohn et al. (2011, p. 10). These authors offer three upward influence tactics: (1) ingratiation: efforts to attain liking through increasing friendliness; (2) self-promotion: efforts to be regarded as competent by emphasizing one's accomplishments; and (3) assertiveness: aggressive and persistent efforts in making requests to attain what one wants. More specifically, Yukl and Falbe (1990) define ingratiation tactics as the attempt of an individual to put others in a good mood or to produce a favourable perception towards the individual before asking others to do something. If these influence tactics are successful, the result is that leaders' perception of followers becomes more positive, and supervisors, whether consciously or subconsciously, remember the ratee's proactive behaviours and use this information to evaluate subordinates' behaviours, such as task performance (Dulebohn et al., 2011, p. 10;Lam, Huang, & Snape, 2007;Liden & Mitchell, 1988).
Building on the early work on LMX (Graen & Cashman, 1975;Graen et al., 1982), the social network view of LMX, and the upward influence tactics theory (Dulebohn et al., 2011), the current study assumes that high-quality LMX ratees have a greater possibility of receiving more favourable performance appraisals than do low-quality LMX ratees. The rationale is that more interaction and liking and higher reciprocal dependency between the high-quality LMX rater and the ratee will make the rater feel more accountable and favourable to the high-quality LMX ratee.
Based on the discussion of rater accountability and LMX theory in the previous sections, the current study proposes a conceptual model and propositions that depicts the relationship between LMX relationships and rater accountability.

The conceptual model
The conceptual model of the current study is composed of four relationships. First is the relationship between upward influence behaviour and LMX relationships. Rater-ratee, rater's supervisor-ratee and rater's supervisor-rater LMX relationships are discussed because the audiences of the appraisal process are the ratees and the rater's supervisor. Second, after the LMX relationships are formed, they affect rater accountability. High-quality LMX is assumed to positively affect rater accountability, whereas low-quality LMX is considered to have a weaker or no effect on rater accountability. Third is the relationship between rater accountability and performance rating. The final rating is assumed to directly reflect the level of rater accountability. Last, the performance rating generates feedback to the rater, ratee and rater's supervisor in assessing their LMX relationship quality. Figure 1 presents the comprehensive model of the current study.

Upward influence tactics and the quality of LMX relationships
In the conceptual model of the current study, Dulebohn et al.'s (2011) three types of upward influence tactics are used to explain the effect of the upward influence behaviour of raters and ratees on the quality of LMX relationships. The basic assumption is that both raters and ratees aim to attain high-quality LMX with their supervisor and that they will use upward influence tactics in the process. Although followers' ingratiation and self-promotion tactics positively affect LMX relationships, assertive tactics by followers tend to form low-quality LMX relationships because assertive tactics make leaders feel uncomfortable and lose trust in the follower (Colella & Varma, 2001;Dulebohn et al., 2011;Krishnan, 2004;Lam et al., 2007). Therefore, the current study uses ingratiation and selfpromotion tactics to explain the behaviour of raters and ratees. Despite almost no research investigating the influence of ingratiation and self-promotion tactics on LMX in the performance appraisal context, a recent meta-analysis by Dulebohn et al. (2011) provides some empirical support. According to Dulebohn et al. (2011), ingratiation tactics (21 studies, N = 3,749, supervisor and subordinate reported data combined) and self-promotion tactics (4 studies, N = 1,132) showed a positive relationship with LMX quality.
For that reason, the current study assumes that such relationship can be applied to the appraisal context.
Ratees form LMX relationships with the rater and the rater's supervisor with the clear goal of attaining a positive appraisal, which can lead to higher performance ratings. Ratees attempt to attain the rater's favour towards the ratee and emphasize their achievements during the period of performance evaluation. Ratees also invest effort in improving the relationship with the rater's supervisor. As the rater's supervisor is the person who eventually approves the rater's evaluation of the ratee, ratees may want to form high-quality LMX relationships with the rater's supervisor as well. Raters also form LMX relationships with their supervisor. Thus, raters follow the same pattern of subordinate-to-supervisor upward influence tactics. One key difference is that whereas ratees aim to receive positive ratings, raters care more about managing their work-unit performance because it composes a large portion of the rater's performance review by the rater's superior (Murphy & Cleveland, 1995).
Proposition 1: The ratee's ingratiation tactics positively affect the quality of LMX relationships with the rater and the rater's supervisor.
Proposition 2: The ratee's self-promotion tactics positively affect the quality of LMX relationships with the rater and the rater's supervisor Proposition 3: The rater's ingratiation tactics positively affect the quality of the LMX relationship with the rater's supervisor.
Proposition 4: The rater's self-promotion tactics positively affect the quality of the LMX relationship with the rater's supervisor.

LMX relationships and rater accountability
The upward influence tactics performed by the rater and the ratee result in two types of outcomes: high-quality LMX relationships and low-quality LMX relationships. Raters and ratees who perform successful upward influence tactics construct high-quality LMX relationships, whereas less successful attempts are likely to result in low-quality LMX relationships. Although no empirical study has directly tested the relationship between LMX relationship and rater accountability, past findings indicate that LMX relationship does have a significant impact on the performance appraisal process. For example, Elicker et al. (2006), using the survey data of 188 employees, found that the quality of the rater-ratee LMX relationship is positively associated with the opportunity to express the ratee's view and provide relevant information to the rater during the appraisal session.
In the current study, the quality of the LMX relationship is expected to affect rater accountability on several occasions. First, ratees engaged in high-quality LMX relationships with the rater and the rater's supervisor will receive more precise or favourable performance ratings than low-quality LMX ratees. Second, ratees in high-quality LMX relationships have advantages over low-quality LMX members in both the objective and subjective dimensions of performance appraisal. Ratees with high-quality LMX have more interactions with raters and rater's supervisors and can attain more information about their objective achievements. Thus, if raters fail to recognize some portion of the ratee's achievements, high-quality LMX ratees, in their frequent interactions with their rater, may fill in the gap of unrecognized objective achievements, whereas low-quality LMX members have lower chances of doing so. For the subjective dimension, supervisors and subordinates in high-quality LMX relationships share the emotion of liking and have mutual dependence on each other, which generates increased affect between them (Chi & Lo, 2003;Dulebohn et al., 2011;Ferris et al., 2009). Consequently, raters perceive higher rater accountability to high-quality LMX ratees than to lowquality LMX ratees.
Third, the effect of the rater and the rater's supervisor LMX relationship on rater accountability also differs according to the relationship quality. In high-quality LMX relationships, the rater is likely to follow the preference of the supervisor. Based on frequent communication in the past and reciprocal liking, the rater's motivation is to comply with the supervisor's view of the appraisal. If the supervisor has a stake in the performance appraisal, such as affection for a specific ratee, the rater is likely to perceive accountability to assign a more positive rating. If the supervisor prefers an accurate appraisal, the rater is expected to conduct an accurate appraisal to prepare for possible justification of the rating. For low-quality LMX relationships, the rater and the rater's supervisor are likely to have had little interaction, meaning that the rater may not be sure of the supervisor's position towards the appraisal. In this case, the rater is likely to engage in "pre-emptive self-criticism" and to conduct accurate appraisals to prepare for audiences (supervisors) with unknown views (Lerner & Tetlock, 1999;Park, 2016). Drawing from the arguments above, raters in high-quality LMX relationships with their supervisors are expected to incorporate the supervisor's interest in the performance appraisal more than in the case of low-quality LMX relationships.
The current study assumes that ratees' resistance can moderate the relationship between LMX relationship quality and rater accountability. During the performance appraisal process, ratees may refuse to accept the rater's negative evaluation of their performance. Tepper et al. (2006) argues that ratees would either refuse to accept the performance evaluation or attempt to negotiate with the rater to receive a better evaluation. Past research has revealed that raters rated lower quality LMX ratees more strictly regardless of the form of resistance, whereas higher quality LMX ratees were rated more favourably if the ratees used negotiation for resistance (Tepper et al., 2006). Thus, LMX quality was assumed to moderate the relationship between ratee resistance and the rater's performance evaluation. However, the current study proposes that employee resistance moderates the relationship between LMX quality and rater accountability because, based on the notion of rater accountability, the quality of the LMX relationship is a fixed condition prior to perceiving accountability, and employee resistance occurs between the appraisal session held with ratees already engaged in certain types of LMX relationships with the rater. Furthermore, employee resistance towards performance appraisal results can be interpreted as the opinion of the audiences of appraisal, which affects rater accountability.
Proposition 5: The rater and ratee LMX relationship quality positively affects rater accountability.
Proposition 6: The rater's supervisor and the ratee's LMX relationship quality positively affects rater accountability.
Proposition 7: The rater's supervisor and the rater's LMX relationship quality positively affects rater accountability.
Proposition 8: Ratee resistance will moderate the relationship between LMX relationship quality and rater accountability.

Performance rating and feedback
Performance rating is generated as the final outcome of the appraisal process. Ratings typically consist of a five-point scale of "outstanding," "above average," "average," "below average" and "poor," but they vary by the appraisal system in each organization. The performance rating generates feedback in the form of information provision to the ratee and the rater's supervisor, and these audiences evaluate the rater's justification for the result of performance appraisals based on the performance ratings. The performance rating provides information to the audiences of the appraisal about whether their LMX relationships with the rater are of high or low quality. If the ratee believes that the performance rating is acceptable and satisfactory, the ratee will acknowledge that the LMX relationship with the rater is high quality. In the case of the rater's supervisor, depending on the supervisor's preference (e.g. accurate rating or lenient rating), the perceived LMX relationship quality of the supervisor can change based on how satisfied the rater's supervisor is with the rater's decision to incorporate the supervisor's preference. Raters are likely to evaluate the LMX relationship quality through their interaction with the audiences after the appraisal process and assess whether they met the expectations of the audience to maintain or enhance the quality of the LMX. In sum, the feedback produced by providing performance ratings reorganizes the LMX relationship quality between the rater and the audiences of appraisal and provides the basis for another round of the appraisal process that begins with the participants of the appraisal process investing effort to build high-quality relationships with their superiors.

Conclusion
This article has potential theoretical contributions for the organizational psychology research in several aspects. First, this paper provides a preliminary step in assessing the effects of LMX relationship quality on rater accountability in the performance appraisal context. Thus far, past research has merely assumed that each audience affects rater accountability without consider the relationship between the audiences. Second, until now, no research has considered the upward influence tactics of both the ratee and rater during the appraisal process. Contrary to the traditional approach of accounting for the supervisor-subordinate relationship in the appraisal process, the current study suggests three dyadic pairs (ratee-rater, ratee-rater's supervisor rater-rater's supervisor) of relationships for the first time in the literature. Lastly, the current study provides a new approach of identifying LMX relationship quality in appraisal using the performance rating. The idea is that the audiences of the appraisal process can evaluate their LMX relationship quality with the rater using the objective measure of the performance rating assigned to the ratees.

Implications for research and practice
The current study explores how upward influence tactics form LMX relationships in the appraisal context and describes how these relationships affect rater accountability. The current study offers several implications for future research and practice.
First, almost all organizations are aware that performance ratings are the most commonly used measure of employee performance and contribution to the organization (Rosen, Kacmar, Harris, Gavin, & Hochwarter, 2017). Based on the propositions of the current study, human resource managers and management may consider that supervisors who conduct performance appraisals face many factors, such as appraisal audiences, which deter them from making the most accurate appraisal. In this sense, future research may want to account for the existence of relationships among the audiences of appraisal. Simply framing accountability as a construct determined by various audiences that independently affect the rater neglects the possibility of the relationship dynamics among the audiences. Although the current study addressed only the ratee and the rater's supervisor, more audiences, such as colleagues and human resource departments, can take part in the appraisal process. Second, studies related to conscious distortion in performance appraisal may include LMX relationship quality as a determinant of rating distortion because raters consciously favour ratees in high-quality LMX relationships. One of the most recent and extensive works in this stream of research was the work by Spence and Keeping (2011), in which the authors presented politics and impression management as independent variables that affect rater bias in appraisal. Thus, including LMX theory as one of the determinants of rating bias and seeking solutions to mitigate those biases may be a topic of interest for the expansion of rating distortion studies.
Third, the current study can be a valuable addition to the rater motivation theory (RMT) literature (Harris, 1994;Murphy & Cleveland, 1995). In the RMT literature, the rater perceives the incentives and disincentives of making an appraisal and manipulates the accuracy of the appraisal to gain positive results (e.g. work-unit performance increase) and avoid negative consequences (e.g. dissatisfaction from delivering negative feedback) (Park, 2014). However, no rater motivation literature has directly acknowledged maintaining the quality of LMX relationships as an independent variable that can encourage raters to adjust ratings.
Fourth, the current study has implications for research in the rater training literature and practice. Contemporary rater training research can be represented by the frame-of-reference (FOR) training approach (Roch, Woehr, Mishra, & Kieszczynska, 2012). According to Roch et al. (2012, p. 2), FOR training increases rating accuracy by "helping raters understand what behaviours constitute specific levels of performance on specific dimensions" and "establishing performance prototypes that allow raters to counteract normal information loss (i.e. forgetting) by categorizing ratee performance." Nevertheless, FOR training does not provide information or standard procedures to prepare for ratees' upward influence tactics presented in the current paper. Including such counteraction methods for these tactics in the FOR may enable raters to establish standards for performance and to prepare for ratees' behaviours to influence the raters to provide higher ratings.
Fifth, research in performance management by employee engagement can benefit from the current study. In the contemporary literature on employee engagement performance management, employee engagement is stimulated by effective job design, coaching and social support, training, and trust and justice during performance appraisal (Gruman & Saks, 2011). In these approaches, LMX relationship quality is offered as a way for leaders to increase employee engagement. However, the possibility of followers using a high-quality LMX relationship to alter the appraisal outcome and advance their interest has not been addressed. Future research in employee engagement is encouraged to consider that employees may engage in performance management-related activities not only to improve organizational performance but also to fulfil their self-interest of attaining high ratings by forming high-quality LMX relationships with their superiors.
Lastly, organizations interested in improving appraisal accuracy may benefit from the conceptual model and propositions in the current study. Management and human resource managers should recognize that an employee's performance rating is affected by interactions prior to the appraisal period because ratees constantly exercise upward influence tactics and attempt to build high-quality LMX relationships. Therefore, future research and organizations may consider developing more objective measures of performance appraisal that would result in similar appraisal results regardless of successful upward influence tactics and high-quality LMX relationships.