Motivating raters through work design: Applying the job characteristics model to the performance appraisal context

Abstract In general, raters are responsible for evaluating their subordinates’ performance and delivering their final performance ratings. As most organizations prefer accurate ratings, past research has focused on encouraging raters to produce more accurate appraisals. Extant research on rater motivations has highlighted the various factors affecting rating accuracy, but the job characteristics of appraisals represent an unexplored area. The current study attempts to extend the rater motivation research by introducing a work design perspective of appraisal and exploring the ways in which the job characteristics of appraisals can be used to motivate raters. First, rater motivation theory and Hackman and Oldham’s job characteristics model (JCM) are reviewed. Then, the current study illustrates how the JCM can be applied to the performance appraisal context.


Introduction
Performance appraisal is one of the core tasks of a manager with supervisory status. In general, supervisors are responsible for evaluating their subordinates' performance and delivering their final performance ratings. Performance ratings and appraisal information from the appraisal process are commonly used for administrative decisions, including promotions, salary increases, performancebased pay, transfers, and training (Campbell, Campbell, & Chia, 1998;Cleveland, Murphy, & Williams, 1989;Coens & Jenkins, 2000;Meyer, 1975). Furthermore, contemporary research has begun to emphasize the link between individual-level performance management based on performance appraisals and organizational-/firm-level performance (DeNisi & Smith, 2014). Consequently,

PUBLIC INTEREST STATEMENT
Most organizations prefer accurate performance ratings over inaccurate ones. For that reason, motivating performance raters to produce more accurate appraisal has been a popular research topic. Such research is often called as the rater motivation literature. The current study attempts to extend the rater motivation research by introducing a work design perspective of appraisal. The basic assumption is that the job characteristics of performance appraisal as a task of a supervisor can be designed to provide incentives for raters to rate their subordinates more accurately.
The current study attempts to extend rater motivation research by introducing the work design perspective of performance appraisals and exploring the ways in which the job characteristics of appraisals can be used to motivate raters. First, the current paper provides a brief history of rater motivation theory to summarize past research trends. Second, Hackman and Oldham (1976)'s job characteristics model (JCM), one of the most widely recognized and cited work design models in the literature, is reviewed. The current study then illustrates how JCM can be applied to the performance appraisal context. Lastly, theoretical and organizational implications and avenues for future research are provided.

Rater motivation theory
Scholarly interest in rater motivation has been a recent phenomenon over the past two decades. From the 1950s to the late 1970s, early research on rating accuracy primarily focused on issues such as rating format and scale until the work of Landy and Farr (1980) and Decotiis and Petit (1978) changed the paradigm of rating accuracy research (Arvey & Murphy, 1998;Ferris, Judge, Rowland, & Fitzgibbons, 1994;Murphy & Cleveland, 1995). Landy and Farr (1980) suggested that researchers shift their attention from rating scales and rating formats (e.g. forced choice rating and ranking) to the rater's cognition and information processing (Arvey & Murphy, 1998). The body of research following Landy and Farr (1980) produced a number of topics such as the rater's cognitive process (DeNisi, 1996;DeNisi & Peters, 1996), rater memory (Murphy & Balzer, 1986), and rater training (Bernardin & Buckley, 1981;Roch, Woehr, Mishra, & Kieszczynska, 2012). Despite the contributions of the early research and subsequent studies following Landy and Farr (1980), research continued to assert that rating errors and inaccurate appraisals were not the result of deliberate actions but the result of invalid rating instruments or the raters' cognitive limitations (Lahuis & Avis, 2007;Murphy & Cleveland, 1995;Wang et al., 2010).
The work of Decotiis and Petit (1978) and the stream of research following it challenged the premise that raters are willing to do their best to produce accurate appraisals. Decotiis and Petit (1978, p. 636), stressed that "consideration must be given to what energizes rater behavior, what determines the direction of such behavior and how this behavior is sustained" and introduced the concept of rater motivation. According to rater motivation theory, raters make intentional decisions based on individual motives that eventually affect performance rating accuracy (Levy & Williams, 2004). Thus, the common theme of rater motivation studies has concerned identifying the motivational factors (e.g. accountability and rating consequences) of appraisals and to use those factors to minimize rating distortions and to maximize rating accuracy. Popular topics include comprehensive rater motivation models (RMMs) (Harris, 1994;Murphy & Cleveland, 1991), rater goals (Murphy, Cleveland, Skattebo, & Kinney, 2004;Wang et al., 2010), rater accountability (Mero & Motowidlo, 1995;Mero, Motowidlo, & Anna, 2003;Roch, 2005), organizational politics in appraisals (Longenecker, Sims, & Gioia, 1987), the effect of rating discomfort on leniency (Villanova, Bernardin, Dahmus, & Sims, 1993), and rating inflation to increase unit (subordinate) performance (Ilgen, Mitchell, & Fredrickson, 1981). However, most research has explained rater motivation by examining a limited number of factors and their effects on rating accuracy. Few scholars, such as Cleveland (1991, 1995) and Harris (1994) have proposed comprehensive RMMs that depict various motivating factors in the appraisal process (Park, 2014).
First, Harris (1994, p. 739) constructed an integrated RMM using the five situational antecedents (accountability, organizational human resource management (HRM) strategy, task/outcome dependence, trust, and forms) and three personal antecedents (amount of information, self-efficacy, and mood) of motivational factors (rewards, negative consequences, and impression management) that result in performance appraisal behaviors (observation, storage, retrieval, integration, rating, and feedback). However, the complex relationships between the eight antecedents and three motivational factors were insufficiently addressed. In addition, the dependent variable is unclear, as all six stages of rating (from observation to feedback) were separately set as dependent variables. Additionally, even if all the appraisal behavior variables were accepted as dependent variables, questions still remain about whether the three motivational factors repeatedly affect each stage of rating behavior. Consequently, despite its contribution, Harris' RMM (1994) presents a number of limitations. Cleveland's RMM (1991, 1995, p. 243) argues that rating behaviors are determined by combining the two types of rater motivations: (1) the motivation to rate accurately and (2) the motivation to distort ratings. The authors used Vroom's expectancy theory (1964) to explain the motivational variation in both types. According to Murphy and Cleveland (1991), raters will be motivated to rate accurately if rewards exist for accurate ratings and if the possibility of receiving such rewards is correlated with rating accuracy. Rewards for accurate ratings, although rarely provided in practice, include "salary increases, promotions, and assignment to choice positions," which partly depend on the rater's accurate performance appraisals (Murphy & Cleveland, 1995, p. 259). Raters are also motivated to inflate (distort) ratings if assigning accurate ratings has negative consequences and if the possibility of facing such consequences is high. For example, negative consequences include discouraging employees who get negative ratings, damaging interpersonal relationships with subordinates, losing rewards (e.g. salary increases and promotions) based on the rater's unit performance (i.e. subordinate performance), and inducing negative reactions from subordinates during appraisal feedback (Murphy & Cleveland, 1995). Murphy and Cleveland (1995, p. 242) stress that although raters are capable of making accurate appraisals, they willingly alter the accuracy of ratings to attain rewards and to avoid negative consequences as there are "surprisingly few good reasons to give accurate ratings." Therefore, the authors assume that rating distortions can be prevented by reducing the negative consequences of providing accurate ratings, which will decrease the raters' willingness to distort ratings. Consequently, discussions about how to encourage raters to make accurate appraisals and to build an appraisal system that positively affects raters have been scarce.
Past studies related to rater motivation, even the comprehensive ones, have not considered the approach of defining performance appraisal as the supervisor's job and using such job characteristics to investigate rater motivation. Several studies have indicated that, for most raters, conducting performance appraisals is a mandatory but occasional and insignificant task; thus, raters often invest limited effort, rigor, and time in preparing for and performing appraisals (Murphy & Cleveland, 1995;Roberts, 1998;Tziner, Murphy, Cleveland, Beaudin, & Marchand, 1998). Nevertheless, these studies do not provide a framework for identifying the job characteristics of appraisals that may affect the raters or possible solutions to raters' negative reactions to appraisals as a task. Accordingly, establishing an effective job design model that can motivate raters to put more effort into the appraisal process can serve as a new extension of the rater motivation literature. In the following sections, the JCM and its application in the performance appraisal context are introduced.

Job characteristics model
Since its introduction almost four decades ago, Hackman and Oldham's (1976) JCM has continued to have a strong influence on the job design theory literature (George, 2009). Hackman and Oldham's (1976) JCM explains the job characteristics that strengthen employees' internal work motivations, implying that, "rather than being motivated by, say, the promise of rewards or the prospect of receiving (avoiding) supervisory attention, people would try to perform well simply because it felt good when they did-and it felt bad when they did not" (Oldham & Hackman, 2010, p. 464). The JCM was originally introduced to "specify the conditions under which individuals will become internally motivated to perform their jobs" (Hackman & Oldham, 1976, p. 250). According to the authors, the core logic of the JCM is that an employee's job characteristics ultimately affect work outcomes and that his or her psychological states act as a mediator between the job characteristics and work outcomes.
In the original work of Hackman and Oldham (1976, p.256), the first part of the model explains how the five perceived "core job dimensions" (skill variety, task identity, task significance, autonomy, and feedback) determine the three "critical psychological states" as follows: (1) skill variety, task identity, and task significance result in "experienced meaningfulness of the work"; (2) autonomy contributes to "experienced responsibility for outcomes of the work"; and (3) feedback leads to "knowledge of the actual results of the work activities." Next, the three critical psychological states collectively affect the four personal and work outcomes (internal work motivation, quality of work performance, work satisfaction, and low absenteeism and turnover) (Parker, Wall, & Cordery, 2001). Lastly, "employee growth need strength" acts as a moderator: (1) at the link between core job dimensions and critical psychological states and (2) at the link between critical psychological states and outcome variables. In sum, the assumption implied in the JCM is that the job characteristics influence the employees to "consciously think and feel certain things about their jobs (e.g. feel accountable or responsible for their work, know how well they are doing, and think the work is meaningful), and it is these conscious thoughts and feelings that ultimately lead to personal and work outcomes (e.g. job satisfaction and job performance)" (George, 2009, p. 3).
Since the introduction of the JCM, research has sought to verify the model's assumptions. In one of first extensive reviews and meta-analyses of the JCM literature, Fried and Ferris (1987) found general support for the links among job characteristics, psychological states, and work outcomes in the JCM model (Johns, Xie, & Fang, 1992;Tiegs, Tetrick, & Yitzhak, 1992). In a more recent metaanalysis of 677 articles, Humphrey, Nahrgang, and Morgeson (2007) replicated and expanded the work of Fried and Ferris (1987) to include recent research on the JCM. Humphrey et al. (2007) found strong support for the positive relationship between the five core job dimensions and work outcomes (job satisfaction, growth satisfaction, internal work motivation, and job performance), with absenteeism as the only exception. Next, the authors validated the simultaneous mediating role of the three critical psychological states between the four core job dimensions (skill variety, task identity, task significance, and autonomy) and work outcomes. However, employees' psychological states did not mediate the relationship between feedback and work outcomes.
The moderating effect of the growth need strength (GNS) has been a rather diminished area of interest in the JCM literature. A series of studies found weak empirical support for the moderating role of the GNS in the JCM (Fried & Ferris, 1987;Johns et al., 1992;O'brien, 1982;Tiegs et al., 1992). Thus, the JCM model was often empirically tested and found to be valid without including GNS as a moderator (Graen, Scandura, & Graen, 1986;Renn & Vandenberg, 1995). In sum, the JCM and its components have found strong support, and, except for GNS, they continue to be widely used in the literature.
The current paper, based on Hackman and Oldham's original work (1976), applies the core job characteristics, critical psychological states, and work outcomes of the JCM to the context of performance appraisal. The JCM in the appraisal context is depicted in Figure 1. Although the JCM has been repeatedly tested and modified since its development, the model has not been applied to performance appraisal context. Thus, the current study uses the original JCM to break new ground using work design to internally motivate raters to make accurate appraisals.

Experienced meaningfulness of work
According to Hackman and Oldham (1976, p. 256), experienced meaningfulness of work is "the degree to which the individual experiences the job as one which is generally meaningful, valuable, and worthwhile." The job characteristics of skill variety, task identity, and task significance determine the meaningfulness of work, as explained below. Hackman and Oldham (1976, p. 257) define skill variety as "the degree to which a job requires a variety of different activities in carrying out the work, which involve the use of a number of different skills and talents of the person." The authors argue that if employees find that they need to challenge or stretch their skills and abilities, they will find their jobs meaningful. In contemporary organizations, performance appraisals are not only used for employee evaluations but also for employee development, reward distribution, and the integration of employee performance and organizational performance for better performance management (Fletcher, 2001). Thus, raters are expected to account for the various purposes of appraisals, and such expectations make the job itself challenging, resulting in high skill variety. The level of perceived skill variety will depend on the number of skills and talents that raters use for conducting performance appraisals. Such skills and talents, including the collection, storage, and retrieval of information on employee performance, require the raters to utilize various observational skills, information processing skills, and their cognitive abilities to promote rating accuracy (DeNisi & Peters, 1996;Harris, 1994;Landy & Farr, 1980;Woehr & Huffcutt, 1994). Second, raters have to use both objective and subjective performance measures, which are usually not interchangeable, meaning that they have to understand and use both types of measures to evaluate a single employee (Arvey & Murphy, 1998;Bommer, Johnson, Rich, Podsakoff, & Mackenzie, 1995). Third, raters are often responsible for managing employee participation during the appraisal process by setting performance standards and goals (e.g. management-by-objectives) with their subordinates (McConkie, 1979;Roberts, 2003). Lastly, interpersonal skills and communication are necessary, as the appraisal process is affected by the social context in organizations (Levy & Williams, 2004). In short, raters who extensively use their skills and abilities may perceive the task of appraisal to be meaningful, according to the logic of Hackman and Oldham (1976, p. 257), who claim that "when a job draws upon several skills of an employee, that individual may find the job to be of enormous personal meaning-even if, in any absolute sense, it is not of great significant or importance." Hackman and Oldham (1976, p. 257) define task identity as "the degree to which the job requires completion of a 'whole' and identifiable piece of work; that is, doing a job from beginning to end with a visible outcome." The authors also argue that individuals who are responsible for a larger part of a "whole" task will consider that work more meaningful than the individuals with smaller parts. The rating task provides high task identity to raters who are involved in the performance appraisal from the beginning to the end. In general, according to Storey and Sisson (1993) and Brown and Benson (2005), the performance appraisal process includes three stages: objective setting, monitoring (feedback), and evaluation. In the objective setting stage, raters are responsible for starting the appraisal process, either by imposing objectives or setting goals with the employees' participation. In the second stage, raters monitor whether the employees are doing well in achieving the goals set in the previous stage and provide formal or informal mid-term progress feedback to the employees. Lastly, in the third stage, raters finally evaluate their employees' performances during the appraisal cycle and assign performance ratings. As raters inevitably participate in all appraisal processes, the level of perceived task identity will be linked to the absolute amount of rater participation. For instance, raters who invest more time and effort in objective or goal setting and in mid-term progress feedback will have higher levels of task identity than others who spend less time on these tasks.

Task significance
Task significance is "the degree to which the job has a substantial impact on the lives or work of other people, whether in the immediate organization or in the external environment" (Hackman & Oldham, 1976, p. 257). Thus, the authors also stress that "when an individual understands that the results of his work may have a significant effect on the well-being of other people, the meaningfulness of that work usually is enhanced." As a task, the performance appraisal, which usually results in performance ratings, has a substantial impact on both the employees and the organization. First, with regard to the impact on employees, the most evident purpose of performance appraisals is to inform administrative decisions, such as promotions, salary increases, disciplinary action (e.g. termination), pay-for-performance, employee transfers, and employee development and training (Bretz, Milkovich, & Read, 1992;Cleveland et al., 1989;Murphy & Cleveland, 1995). Next, many contemporary organizations include performance appraisals in the broad frame of performance management, as one of the goals of performance appraisals is to improve employee performance (Gruman & Saks, 2011;Spence & Keeping, 2011). Accordingly, performance appraisals are used to manage employee and organizational performances and to integrate these two performance types (Fletcher, 2001). Therefore, depending on the extent of appraisal usage within their organizations, raters can easily recognize the importance of appraisals and perceive higher levels of task significance.

Experienced responsibility for work outcome
Experienced responsibility for work outcome refers to "the degree to which the individuals feels personally accountable and responsible for the results of the work he or she does" (Hackman & Oldham, 1976, p. 256). According to the authors, an individual's autonomy in his or her job is the single determinant of experienced responsibility.

Autonomy
Autonomy is "the degree to which the job provides substantial freedom, independence, and discretion to the individual in scheduling the work and in determining the procedures to be used in carrying it out" (Hackman & Oldham, 1976, p. 258). To the authors, a high level of autonomy means that the task's outcome is primarily determined by the individual's effort and decisions rather than the supervisor's guidance or job procedure manuals. In the appraisal process, the immediate supervisor (i.e. the rater) is mainly responsible for assessing the employee's performance during the evaluation period and for assigning final performance ratings. However, a rater's autonomy in the appraisal may decrease if other parties (e.g. subordinates or the rater's supervisor) participate in the appraisal process. Under a 360-degree appraisal or a multi-source rating system, in addition to the rater's appraisal, his or her subordinates, supervisors, and peers can participate in the appraisal process (Conway & Huffcutt, 1997;van der Heijden & Nijhof, 2004). Nevertheless, the 360-degree appraisal system is usually used for developmental purposes and has a limited influence on final performance ratings, as its effectiveness has been contested (DeNisi & Kluger, 2000;DeNisi & Pritchard, 2006). Also, depending on the appraisal system settings, in some organizations, the rater's direct supervisor may review and validate the final ratings. In such case, although rare in practice, the rater's supervisor is likely to provide some form of guidance or comment (Harris, 1994).

Knowledge of results
Knowledge of results is "the degree to which the individual knows and understands, on a continuous basis, how effectively he or she is performing the job" (Hackman & Oldham, 1976, p. 257). The authors propose feedback as a job characteristic that fosters knowledge of results. Hackman and Oldham (1976, p. 258) define feedback as "the degree to which carrying out the work activities required by the job results in the individual obtaining direct and clear information about the effectiveness of his or her performance." Recently, in an updated study on the JCM, Oldham and Hackman (2010) stressed that although the JCM includes some social dimensions of work in the feedback characteristics, these social aspects have been neglected in the JCM literature. According to Oldham and Hackman (2010, p. 468), because the social dimensions of work can contribute to "the motivation, performance, and well-being of jobholders," future research may want to consider incorporating social aspects into the JCM. In the current study, following the recommendation of Oldham and Hackman (2010), the social dimension of performance appraisals as a task has been incorporated into the feedback characteristics by adapting the theories of rater accountability to the JCM. In this section, rater accountability is only used to discuss the relationship between feedback and knowledge of results in the appraisal context. The effect of accountability on rating performance is addressed in the next section.

Feedback
In general, organizational accountability involves "an actor or agent in a social context who potentially is subject to observation and evaluation by some audiences, including oneself" (Frink & Klimoski, 2004, p. 3). The audience (e.g. supervisors, subordinates, or customers) and the employees held accountable form a social system founded on shared expectations and rules that that guide work decisions and actions (Breaux, Munyon, Hochwarter, & Ferris, 2009;Frink & Klimoski, 1998. For that reason, accountability in organizations is assumed to affect the work performance of those held accountable (Antonioni, 1994;Hochwarter et al., 2007;Tetlock, 1985Tetlock, , 1992. In the performance appraisal context, accountability is operationalized as rater accountability, which is defined as the "perceived potential to be evaluated by someone and being held responsible for one's decision or behaviors" (Levy & Williams, 2004, p. 887). The potential audiences in the performance appraisal process usually consist of two players: the supervisor and the subordinate. How, then, do the raters receive feedback on their performance of appraisals as a task?
First, the rater's supervisor may question the quality of performance ratings (Ferris, Munyon, Basik, & Buckley, 2008). For example, in organizations in which the raters' supervisors review and approve the performance ratings, raters can receive information about their effectiveness in the appraisal process. Second, employees' direct feedback on the raters' effectiveness can occur on two occasions: (1) when reacting to the feedback session or interview and (2) when engaging in the appeal process. Before explaining direct feedback, it is worth mentioning that the employee's attitudes and behaviors due to such appraisals, such as job dissatisfaction, work motivation, trust in the appraisal system, and absenteeism, are not directly and clearly provided to the rater; they are instead an indirect indication of the rater's appraisal quality. Accordingly, following Hackman and Oldham (1976)'s definition of feedback, the employee's indirect feedback is excluded from discussion.
Past research has indicated that appraisal reactions are important criteria when assessing appraisal effectiveness (Williams & Levy, 2000). Overall, compared with poor performers, highly rated employees tend to show more accuracy in self-appraisals (Mitchell, Green, & Wood, 1981;Murphy & Cleveland, 1995). Consequently, the rater and the high-performing employee are likely to agree on the assigned performance rating, resulting in the employee's favorable reactions during the appraisal session (Dorfman, Stephan, & Loveland, 1986). Thus, the raters receive feedback from high performers about their appraisal effectiveness in the form of positive reactions. On the contrary, poorly rated employees tend to rate themselves higher than ratings by their supervisors and react negatively or defensively during their appraisal sessions (Facteau & Craig, 2001;Harris & Schaubroeck, 1988;Mount, 1984). In such cases, direct feedback from the employees is provided in the form of http://dx.doi.org/10.1080/23311908.2017.1287320 negative reactions as they perceive that the raters showed poor performance in conducting appraisal by assigning lower ratings than expected. The next type of feedback is the formal appeal process, which allows employees to have a voice during the appraisal process (Korsgaard & Roberson, 1995). If employees believe that their performance ratings are incorrect or unfair, they can go through formal appeal process (Thurston & McNall, 2010). Hence, if employees decide to appeal their ratings to higher authorities, such as human resource department and the performance evaluation committee, they provide direct feedback to their rater about his or her appraisal quality. Hackman and Oldham's (1976) original JCM offers internal motivation, work performance, work satisfaction, and absenteeism and turnover as the final personal outcomes. In the first meta-analysis of the JCM literature, Fried and Ferris (1987) categorized the outcomes into overall job satisfaction, growth satisfaction, internal work motivation, job performance, and absenteeism. In the next and most recent meta-analysis, Humphrey et al. (2007) re-sorted the four work outcomes into behavioral outcomes (work performance, absenteeism, and turnover intentions) and attitudinal outcomes (job satisfaction and internal work motivation). A common thread ties this research together: the core psychological states are seen to affect the internal motivation of employees, resulting in higher quality work outcomes and greater overall job satisfaction (Oldham & Hackman, 2010). In the current paper, only work performance is discussed as a work outcome because the goal of providing raters with internal motivation is to encourage them to make accurate appraisals. Outcomes such as absenteeism and turnover intentions are excluded because individuals are unlikely to leave their organizations because of the job characteristics of appraisals, which are commonly the task of supervisors. In addition, job satisfaction is also excluded, as performance appraisals only constitute a small fraction of supervisors' tasks. According to Mintzberg (1975, p. 55), supervisors perform various interpersonal (figurehead, leader, and liaison), informational (monitor, disseminator, and spokesperson), and decision-related (entrepreneur, disturbance handler, resource allocator, and negotiator) roles. As such, the effect of appraisals on supervisors' job satisfaction is limited.

Work outcomes
In the appraisal context, work performance is represented by the level of performance appraisal accuracy. All five job characteristics, mediated by psychological states, internally motivate the raters and influence the accuracy of performance appraisals. First, conducting appraisals requires a variety of skills, and such requirements can make the job more challenging and meaningful, leading to increased rating accuracy. Second, task identity is represented by the raters' responsibility through the entire performance appraisal process, which has a visible outcome in the form of performance ratings. As previously noted, if individuals perceive that they are performing a task that is whole and complete, the work matters more to them; thus, higher work performance is achieved (Hackman & Oldham, 1980;Rosso, Dekas, & Wrzesniewski, 2010). Third, the task significance of appraisals is reflected in performance ratings' significant impact on personnel decisions and employee development. The more ratings are used in various HRM practices, the more raters are motivated to rate accurately. Fourth, raters, depending on their level of autonomy in rating their employees' performances, are likely to feel responsible and attempt to invest more effort to make better appraisals.
Lastly, receiving direct feedback from their supervisors and subordinates, raters acquire knowledge about their task results and eventually attempt to improve their work performance by increasing their accuracy. Notwithstanding, depending on the institutional settings or rating systems, raters may sometimes distort ratings in response to such feedback. We can assume that raters in almost all organizations are expected to administer accurate appraisals. If raters do not meet these expectations, they may be required to justify their rating decisions to their audience (e.g. supervisors or subordinates); thus, rater accountability functions as a motivating factor in appraisals (Mero & Motowidlo, 1995;Mero, Guidice, & Brownlee, 2007).
When raters are held accountable to their supervisors, they tend to conduct accurate appraisals by preparing to justify their ratings with the most relevant performance information and by processing the information in a consistent manner (Mero & Motowidlo, 1995). However, when evaluating subordinates using a rating appraisal system with absolute standards, raters often inflate ratings to avoid confrontations with subordinates, to manage subordinates' impressions, and to maintain subordinates' motivation (Murphy & Cleveland, 1995;Spence & Keeping, 2011;Wayne & Kacmar, 1991). Therefore, raters tend to become more accurate in their appraisals when justifications are provided to their superiors (supervisors) compared with when justifications are provided to their subordinates (Ferris et al., 2008).
By contrast, under a forced distribution rating system (FDRS), raters are motivated to be as accurate as possible in their appraisals. The FDRS is an appraisal system in which the rater is obligated to assign performance ratings based on a predetermined distribution (e.g. normal distribution); the raters are thus forced to avoid lenient appraisals (Schleicher, Bull, & Green, 2009). Accordingly, the FDRS encourages raters to distinguish between high performers and low performers, resulting in a high-performance and high-talent culture in which poor performances are penalized and honest appraisal feedback is expected (Guralnik, Rozmarin, & So, 2004;Meisler, 2003;Schleicher et al., 2009;Scullen, Bergey, & Aiman-Smith, 2005). As low performers are likely to respond negatively to appraisal results and require justification for their ratings, raters must be as accurate as possible (Facteau & Craig, 2001;Harris & Schaubroeck, 1988;Mount, 1984). In short, in the case of subordinate feedback, raters' attempts to make accurate appraisals are affected by whether the organization uses an absolute standard or fixed distribution rating system.

Conclusion
The current research applies Hackman and Oldham's (1976) JCM to the performance appraisal context to reveal new perspectives on improving rater accuracy. To date, extant research on rater motivation has highlighted various motivational factors that affect rating accuracy, but the job characteristics of appraisals have constituted an unexplored area. The current paper offers a first step in considering the work design aspect of appraisals and illustrates how raters can be internally motivated to perform more accurate appraisals.

Implications for research
The current study examines appraisals as the task of supervisors and explores the ways in which raters are internally motivated to increase appraisal accuracy. Several interesting research questions may be addressed in future research. First and foremost, future research must acknowledge that raters respond to performance appraisals as a job that has its own characteristics. Past research on rater accuracy focused on issues such as rating instruments, rater error, raters' cognitive processes, and rater motivation, which assumed that raters had good intentions in the appraisal process and simply fell short of expectations for other reasons. Therefore, traditional concerns about rater accuracy neglected the work design aspect of appraisals. Second, the significance of appraisals as a task must be reexamined. Although conducting appraisals is a typical task for supervisors, little research has addressed the importance of the appraisal task to raters. Few studies have found that raters consider the task appraisal an insignificant job (Murphy & Cleveland, 1995;Roberts, 1998;Tziner et al., 1998). However, as performance appraisal is one of the crucial aspects of the organizational performance management process, determining how to increase raters' perceived significance of appraisals is certainly important.
Third, future research may want to reassess the adoption of 360-degree appraisal system, which is often used to increase employees' rating acceptance and to provide multiple source feedback on employee performance (Antonioni & Park, 2001;DeNisi & Kluger, 2000;Latham & Wexley, 1982). Through the lens of the JCM, the 360-degree appraisal system seemingly deteriorates the rater's autonomy in making appraisal decisions, resulting in decreased responsibility for rating accuracy (work outcome). Accordingly, the relationship between a rater's autonomy and multi-source ratings can be further investigated. Fourth, the current study proposes that the various skills (e.g. interpersonal and information processing skills) that raters are required to use for appraisals and ways to enhance those skills should be identified. According to the JCM, individuals respond differently to work designs; thus, work outcomes are affected by individual-level differences (George, 2009). Similarly, in the http://dx.doi.org/10.1080/23311908.2017.1287320 appraisal context, the required skills for conducting appraisals may be similar but the outcomes depend on the raters' individual differences. For instance, although almost all raters engage in performance measurement and feedback, some raters may be good at measuring performance but not so good at inducing employees' acceptance of ratings. Accordingly, developing a framework to identify a common set of required skills for appraisals and to resolve individual variations that may discourage accuracy may yield valuable insights into managing and improving rater skills.

Implications for practice
The current study offers several recommendations for HRM practices. First, rater training, including programs and contents about the significance of appraisals, can be developed and used to internally motivate raters. Although popular rater training methods, such as the frame-of-reference (FOR) method, are useful in increasing rater accuracy (London, Mone, & Scott, 2004;Roch et al., 2012), they do not show the rater why appraisal is important for the organization and what role appraisal plays in performance management. Thus, organizations can train raters by providing information about how individual-level performance affects overall organizational performance and strategic management to emphasize the task significance of appraisals.
Second, organizations may consider using the FDRS over the absolute rating system. If the organization's goal is to encourage accurate and honest appraisals, the FDRS is the better option because it controls the rater's desire to inflate ratings. Some scholars argue that the FDRS has positive effects, such as motivating high performers, eliminating or penalizing low performers, and installing a climate of meritocracy and performance-oriented culture in organizations (Blume, Baldwin, & Rubin, 2009;Scullen et al., 2005;Welch, 2001). However, some studies claim that the FDRS undermines productivity, discourages collaboration, results in unfair appraisals, and forces high-performing employees to underperform (Pfeffer & Sutton, 2000;Scullen et al., 2005). Therefore, organizations should carefully think about their ultimate goals before adopting the FDRS. Third, additional work is required to develop organizational policies that provide direct feedback to the raters regarding their rating accuracy. As previously noted in the "feedback" section, raters only receive direct feedback from their supervisors and subordinates after the final rating decisions have been made. Such feedback only allows the raters to incorporate these comments into the next performance evaluation period, which usually occurs only once or twice a year. Hence, more frequent feedback (e.g. midterm evaluations or review sessions between evaluation periods) can be used to provide raters with direct feedback about their appraisal decisions. Based on the feedback received from their supervisors and subordinates, raters can thus correct potential problems prior to making their final rating decisions at the end of the evaluation period.