Research has consistently shown that one of the best ways to improve the overall performance of an organization is to hire strong performers (Hunter & Hunter, 1984; MacLane & Walmsley, 2010; Schmidt & Hunter, 1998). However, in recent years, it has become increasingly difficult to attract, hire, and retain high-performing employees (Aguinis et al., 2012). One way to facilitate the retention of high-performing employees is to recognize and reward their accomplishments via a well-designed performance management system (Trank et al., 2002). Moreover, researchers have found that the use of certain types of performance management systems, such as those that employ relative ratings, can contribute to the overall improvement of performance in the workplace. One of the way the implementation of relative ratings can improve organization-wide performance through inciting lower-performing employees to voluntarily or involuntarily leave the company based on their performance ratings (Berger et al., 2013; Scullen et al., 2005).

Broadly speaking, relative ratings, sometimes referred to as social-comparative ratings, are a type of performance rating that leverage social comparisons to facilitate the evaluation of employees (Olson et al., 2007). In this type of rating system, an employee’s performance is compared to that of a carefully chosen reference group, or in some cases other specific employees (Goffin & Olson, 2011; Olson et al., 2007). Social-comparative rating systems have demonstrated several advantages beyond traditional absolute rating systems (e.g., rating systems that use pre-determined standards for performance) including increased criterion-related validity, accuracy, and rater agreement, as well as reduced leniency in single ratee and multi-source performance ratings (Carver et al., 2021; Feeney et al., 2018; Feeney et al., 2016; Feeney et al., 2023; Freund & Kasten, 2012; Goffin et al., 1996; Goffin et al., 2009; Mabe & West, 1982; Wagner & Goffin, 1997). Moreover, compared to traditional (i.e., absolute) rating systems, social-comparative rating systems are better at differentiating amongst employees based on their performance (Goffin et al., 1996; Guralnik et al., 2004). For example, social-comparative rating methods such as the Relative Percentile Method (RPM) discourages raters from assigning identical scores to ratees, limiting the likelihood that multiple employees will be assigned identical performance scores when their performance is not, in fact, equal (Goffin et al., 1996). As a result, it is easier for managers to identify top performers (Blume et al., 2013).

Researchers have found that job seekers pay attention to the various human-resource management features of an organization during the recruitment process, including how performance will be evaluated (e.g., absolute vs social-comparative ratings) and how rewards will be distributed based as a function of evaluations being used (Gerhart & Milkovich, 1990; Wayne & Casper, 2012). Employees, especially top performers, may seek out organizations that allow them to stand out as well as facilitate and reward their successful performance (Bretz & Judge, 1998; Cadsby et al., 2007; Moon et al., 2017; Turban & Keon, 1993). Moreover, employees demonstrate preferences for certain features of performance appraisal systems such as their purpose (i.e., developmental or administratively focused), degree of formality, and frequency (Gosselin et al., 1997). There is also evidence suggesting individuals have preferences for different styles of performance rating scales (Blume, 2013).

Organizations may be able to take strategic advantage of individual differences associated with having specific preferences for social-comparative, or traditional absolute, performance rating systems to attract and ultimately retain top talent as a function of increased person-organization fit (Barrick & Parks-Leduc, 2019; Kristof, 1996). However, little is known about the individual differences associated with those who may be interested or disinterested in working for a company that uses social-comparative performance evaluations in order for organizations to be able to do so. At present, approximately 83% of organizations use absolute performance ratings (Gorman et al., 2017), and organizations implicitly assume that most employees would benefit from and prefer to receive absolute performance ratings. However, no extant research has examined this assumption and evaluated the potential consequences and correlates of employee preferences for either absolute or social-comparative performance evaluations. Specifically, if an organization were to advertise a position and it were known that this organization places an emphasis on rewarding top performers by using a social-comparative performance rating system, would this change the profile of the psychological characteristics of individuals who are attracted to the organization? Likewise, would this impact an applicant’s interest in working for an organization that used this type of performance evaluation processes? Accordingly, the purpose of the present study is to identify whether applicants demonstrate a preference for social-comparative performance ratings over traditional absolute performance ratings, and to identify key psychological characteristics of the individuals who may demonstrate such a preference. To that end we ask the following research question:

  • RQ1: What proportion of employees will indicate a preference for social-comparative versus absolute performance ratings?

Individual difference characteristics and preferences for social-comparative rating systems

Social comparisons are interpersonal comparisons made to evaluate oneself in comparison to others, or evaluate others relative to others (Buunk & Gibbons, 2007; Festinger, 1954). According to social comparison theory, all humans engage in social comparisons in one form or another and do so to evaluate how individuals compare to one another on a wide range of behaviours, skills, attitudes, and more (Festinger, 1954; Goffin & Olson, 2011). Some researchers believe that the frequency with which individuals make use of social comparisons may reflect individual differences and preferences for this type of information and feedback (Buunk & Gibbons, 2007). Across a series of studies, Gibbons and Buunk (1999) developed a measure of Social Comparison Orientation (SCO) to assess this individual difference in having an affinity for seeking out and making social comparisons. Buunk and Gibbons (2007) argued that individuals high on SCO prefer to make judgements based on social comparisons. Given that individuals high on SCO should demonstrate a preference for social comparison-based information over non-social comparison information, in the context of the workplace, this affinity for social-comparisons should manifest as a preference for social-comparative performance ratings. Thus, we hypothesize that:

  • Hypothesis 1: Greater levels of SCO will be positively associated with a preference for social-comparative rating systems.

Research has also consistently shown that cognitive ability is the best predictor of individual job performance and that hiring individuals high on cognitive ability is one of the best ways to improve performance within an organization (MacLane & Walmsley, 2010; Schmidt & Hunter, 1998). In workplace settings, individuals high on cognitive ability tend to prefer performance rating systems that reward and recognize individual performance (Trank et al., 2002), and feature high reward differentiation from other employees (Blume et al, 2009), as these types of performance management systems allow those high in cognitive ability to maximize potential rewards. Consequently, Blume et al. (2013) found that individuals high in cognitive ability prefer forced-distribution performance management systems over traditional absolute performance management systems since forced-distribution rating systems, like other social-comparative performance rating systems, facilitate high reward differentiation which is contextually signaled to employees due to increased competition for a limited reward pool (Garcia et al., 2013; Goffin et al., 1996). Accordingly, given that other well-designed social-comparative rating systems will also feature high reward differentiation, we hypothesize that:

  • Hypothesis 2:Greater levels of cognitive ability will be positively associated with a preference for social-comparative performance rating systems.

Those high in self-efficacy may also demonstrate a preference for social-comparative rating systems. Individuals with high levels of self-efficacy are said to be confident in their cognitive skills, as well as other abilities to succeed in a given situation (Chen et al., 2001; Judge et al., 1997). Previous research has found that self-efficacy is positively related to performance across a variety of domains, including that of job performance (Stajkovic & Luthans, 1998). However, it has also been found that those high in self-efficacy tend to over-estimate their performance abilities relative to their actual performance (Stone, 1994). Consequently, this may facilitate those high in self-efficacy to believe that they will be strong performers, regardless of their actual levels of performance. In turn, those who believe they will be strong performers should demonstrate a preference for a rating and reward system that is best able to differentiate between and reward strong performers, such as a social-comparative rating system (Goffin et al., 1996; Trank et al., 2002). Thus, we hypothesize that:

  • Hypothesis 3: Greater levels of Self-Efficacy will be positively associated with a preference for social-comparative performance rating.

Meta-analytic evidence has consistently demonstrated that Conscientiousness is the strongest personality predictor of cognitive ability as well as work-performance behaviours (Barrick & Mount, 1991; Dudley et al., 2006; Hurtz & Donovan, 2000; Schmidt & Hunter, 1998). Consistent with the idea that high achievers and strong performers will prefer rating and reward systems that recognize and facilitate their individual success (Trank et al., 2002), we broadly expect those high on Conscientiousness to prefer social-comparative rating and reward system. However, narrow personality traits tend to be more useful for predicting behaviours due to their specific definitions and less expansive content domains (Paunonen & Ashton, 2001). Consequently, we argue that this effect may be driven by two of the narrow trait subscales embedded within Conscientiousness: Organization and Diligence.

In the HEXACO model, the four narrow traits that form Conscientiousness are: organization, perfectionism, prudence, and diligence (Lee & Ashton, 2004). Individuals high in organization have a strong desire for order and structure. Moreover, those who have a high need for structure seek to understand the world by simplifying and processing complex information as efficiently as possible (Neuberg & Newsom, 1993). This includes social as well as non-social information such as performance ratings and feedback. Compared to absolute ratings and feedback, social-comparative ratings and feedback are arguably less cognitively complex and more cognitively efficient. It has been argued that social comparisons are a necessary and pervasive part of everyday life, and that human’s proclivity for social comparisons may have become hard-wired as result of evolutionary processes (Goffin & Olson, 2011). Accordingly, our ability to make social comparisons may have been developed over generations to become a natural and efficient cognitive process for humans. Absolute ratings are cognitively inefficient as they do not provide the extra information about the performance of others that social-comparative ratings provide (Farh & Dobbins, 1989). For a ratee to obtain this information, they must expend additional cognitive, and perhaps even social resources, to identify how they performed compared to others. On the other hand, social-comparative ratings package more complex information into a single rating, making them more cognitively efficient as they provide the recipient with information about their own level of performance and those of others at the same time (Goffin & Olson, 2011; Goffin et al., 2009). Moreover, because it is easier to process social-comparative rating information, it may be the case that this type of rating information is easier to make use of and understand thereby providing a more efficient pathway towards improving future performance. Therefore, we hypothesize that:

  • Hypothesis 4a: Greater levels of organization will be positively associated with a preference for social-comparative performance rating systems.

One of the primary characteristics of individuals high on diligence is their need for achievement (Lee & Ashton, 2004). Individuals characterized by a high need for achievement (i.e., diligence) have a strong work ethic, enjoy hard work, are ambitious, and take their jobs seriously (Bluen et al., 1990). Individuals who are motivated by achievement also prefer merit-based rewards (Turban & Keon, 1993). Combined with their desire to succeed and work hard, a preference for merit-based rewards may enable a preference for social-comparative performance ratings as this type of evaluation and reward system will better facilitate individual success. Therefore, we hypothesize that:

  • Hypothesis 4b: Greater levels of diligence will be positively associated with a preference for social-comparative performance rating systems.

Methods

Participants

In total, 867 participants were recruited for this study online through Amazon’s online survey and data collection system, Mechanical Turk (MTurk). To qualify for this study, participants had to be employed, either part-time or full time. Due to extensive concerns about ongoing issues with low quality responses and bots on the Mturk platform, as reported in both the academic literature (Aguinis et al, 2021) and industry literature (Dreyfuss., 2018; Moss & Litman, 2018a, b), we elected to engage in extensive data cleaning and only retain responses from participants who passed all available careless responding checks. In addition to standard directed response careless responding checks (Meade & Craig, 2012), we asked a series of content validation questions to ensure that participants did indeed read the materials presented in our study. Requiring participants to successfully complete multiple responding and content validation checks resulted in a substantial loss in sample size, however, we feel strongly that we were left with high quality responses as a result.

Accordingly, after removing careless responders who failed to correctly respond to the directed response questions (n = 147), those who subsequently indicated we should not use their data (n = 5), and those who failed to correctly answer content questions (n = 570), 145 participants remained. Additional details about the nature of the careless responding and content questions can be found in the Measures section. Importantly, we conducted an a-priori power analysis for ordinal logistic regression analyses which was the primary analysis used to evaluate our hypotheses. According to Chen et al. (2010), a small effect size for odds ratios is 1.68 (i.e., the equivalence of cohen’s d = 0.2) and a medium effect for odds ratios is 3.47 (i.e., the equivalence of d = 0.5). We used an odds ratio value of 2.5 (i.e., small to medium effect) to calculate the number of participants needed when power = 80%, alpha = 0.05 with equal probabilities for the ordinal response categories (i.e., preference for social-comparative feedback). The results of this analysis indicated a sample size of 120 was needed to obtain sufficient power, therefore our final sample size of 147 is sufficient to test the hypotheses in our study.

Participant ages ranged from 21 to 66 (M = 40.00, SD = 10.56). All were employed (85% full-time) and most reported identifying as female (66%). Additionally, 78% indicated that they have provided another employee with feedback on their job performance at some point during their career, and 94% of participants have had their performance evaluated formally. All participants were compensated for their time and responded to all items using a 5-point Likert-type scale ranging from 1 (Strongly Disagree) to 5 (Strongly Agree) unless otherwise specified. The data collected for this study are available via the OSF: https://osf.io/kh6vj/. The study was approved by [University name redacted] Non-Medical Research Ethics Board (#114,686).

Measures

Preferences for performance feedback

Aguinis and Bradley (2014) recommended the use of vignettes to study phenomena that would be unethical or impossible to manipulate or study in real world situations. Therefore, to capture participants’ preferences for performance feedback, they were presented with a vignette and asked to respond to a single question pertaining to its content. The vignette first provides participants with a job description for a retail sales manager, and a list of important skills and tasks that typical of an individual employed in this role. Next, participants were asked to imagine they are searching for a job as a retail sales manager and have completed a final round of interviews for two very similar jobs at different companies they are interested in and qualified for (see supplemental materials). They were informed that although the two companies are similar in most regards (e.g., both are very large companies in the same industry), the two companies use very different systems to evaluate and reward their employees (social-comparative versus absolute).

Participants were then presented with a chart explaining the two evaluation and reward systems, as well as sample ratings from each system. The two performance rating scales that were described to participants and used as examples were the RPM and Graphic Rating Scales (GRS). The RPM is a well-established social-comparative rating method and communicates performance ratings using percentile scores ranging from 0 to 100, where the performance of the target is compared to that of relevant others (see Goffin & Olson, 2011 for additional details). Similarly, the GRS is among the most commonly used absolute rating methods (Gorman et al., 2017) and communicates performance ratings using visual markers on scales that typically range from poor to excellent (see Catano et al., 2019 for additional details).

Last, participants were told to consider all the information they were presented with and were asked “Which company would you accept an offer of employment from?” and responded using a four-point scale: 1 (Strongly prefer to work for the company that uses the relative rating system), 2 (Slightly prefer to work for the company that uses the relative rating system), 3 (Slightly prefer to work for the company that uses the absolute rating system), and 4 (Strongly prefer to work for the company that uses the absolute rating system).

Cognitive ability

Next, participants were asked to complete a timed 23-item self-report version of the International Cognitive Ability Resource (ICAR; Condon & Revelle, 2014) in order to measure cognitive ability. This measure of cognitive ability contains four item types: Verbal Reasoning items (9 items), Letter and Number Series items (6 items), Matrix Reasoning items (4 items), and Three-Dimensional Rotation items (4 items). The authors of the measure reported the internal consistency reliability across all four subscales to be 0.81 and provided evidence for its validity (Condon & Revelle, 2014). Participants were given six minutes to answers these questions, which is commensurate with the amount of time provided for other well-established cognitive ability tests (McKelvie, 1994; Wright & Laing, 1943).

Social comparison orientation

Participants completed a modified 6-item version of the ability subscale from the Social Comparison Orientation Scale (Gibbons & Buunk, 1999). The items from the ability subscale were modified to fit the context of our study and were rewritten to reference social comparisons in the workplace. For example, a sample item reads: “I often compare how I am performing at work to the performance of my co-workers”. The internal consistency reliability for the original scale was between 0.77 and 0.85. In addition, the authors provided evidence of the original measure’s validity (Gibbons & Buunk, 1999).

Personality

Participants completed the 100-item self-report HEXACO personality inventory (Lee & Ashton, 2004; Lee & Ashton, 2018) and the 8-item General Self Efficacy Scale (Chen et al., 2001) The HEXACO uses six 16-item subscales to measure Openness to Experience, Conscientiousness, Extraversion, Agreeableness, Emotionality, and Honesty-Humility and 4-item subscales to measure each of the narrow traits. The authors of the HEXACO reported that the internal consistency reliability for the narrow traits was between 0.59 and 0.89 and provided evidence in support of the measure’s validity (Ashton & Lee, 2007; Lee & Ashton, 2004; Lee & Ashton, 2018). Similarly, the internal consistency reliability for the General Self-Efficacy scale is between 0.86 and 0.90 and evidence has been provided in support of the measure’s validity (Chen et al., 2001).

Careless responding

Careless responders were identified using best practice recommendations (Meade & Craig, 2012). Participants were asked to complete five directed-response items (e.g., “Please respond strongly disagree to this item”) embedded throughout the questionnaire portion of the study. Additionally, at the end of the study and before participants were debriefed, participants were asked to respond “yes” or “no” to the question “In your honest opinion, should we use your data in our analyses in this study?”. Participants who responded “no” to this question, and participants who incorrectly answered any of the aforementioned directed-response items were removed from all analyses.

Content questions

As part of identifying careless responders, we also asked participants to answer four content question to ensure only participants who read the study materials and paid sufficient attention were included in our analyses. Participants were asked “In the job description you were provided with, which was listed as an important skill?”, “What type of job were you told to imagine you were looking for?”, “Which of the following was a key feature of the Relative Rating System?”, and lastly “Which of the following was a key feature of the Absolute Rating System?”. Participants who incorrectly answered any of these questions were removed from our analyses.

Procedure

Participants were recruited online and were asked to take on the role of a Retail Sales Manager who had been presented with job offers from two very similar companies after completing a series of interviews. Next, participants were told they would need to review the two different methods these companies used to evaluate the performance of their employees (i.e., a social-comparative or absolute performance management system) and subsequently indicate which company they would prefer to work for. To limit potential order effects, we randomized which company and job offer they read about first. After reading and responding to the vignette participants were given six minutes to complete the ICAR cognitive ability test, followed by content questions about the vignette they read to identify careless responders. Next, participants completed the measures of personality and social comparison orientation presented in a randomized order and without a time limit. Last, participants were asked the remaining questions used to help further mitigate careless responding. Participants were compensated $0.50 USD for their participation.

Results

Check for order effects

First, we evaluated whether the order participants read about the different reward systems impacted which reward system they preferred. No significant differences were found between responses to the vignette from the two different orders of presentation t(126) = -0.99, p = 0.320. Therefore, we found no evidence that the order materials were presented in impacted which reward system participants preferred (M1 = 2.54, SD1 = 1.20; M2 = 2.73, SD2 = 1.09).

Main findings

Means, standard deviations, correlations, and Cronbach’s alphas for the main variables are reported in Table 1.Footnote 1 Before testing our hypotheses, we evaluated the proportion of participants who indicated a preference for absolute and social-comparative performance evaluations in order to answer our RQ1. We found that the majority of participants (54%, n = 79) either ‘slightly preferred’ (n = 33) or ‘strongly preferred’ (n = 46) absolute feedback, whereas only 46% (n = 66) of participants either ‘slightly preferred’ (n = 31) or ‘strongly preferred’ (n = 35) the social-comparative feedback.

Table 1 Means, standard deviations, and correlations
Full size table

Hypothesis tests

Ordinal logistic regression was used to evaluate our hypotheses and research question given the ordered nature of the preference for performance feedback variable. Hypothesis 1 predicted that greater levels of SCO would be positively associated with a preference for the social-comparative rating system. A significant positive association was found between SCO and preference for performance feedback (b = -0.54, 95% CI [-0.89, -0.18], OR = 0.58, p = 0.003), indicating that individuals higher on SCO were more likely to prefer the social-comparative performance options. Therefore, Hypothesis 1 was supported.

Hypothesis 2 predicted that greater levels of cognitive ability would be positively associated with a preference for the social-comparative rating system. No significant association was found between cognitive ability and preference for performance feedback (b = 1.48, 95% CI [-0.54, 3.50], OR = 4.38, p = 0.151). Therefore, Hypothesis 2 was not supported.

Hypothesis 3 predicted that greater levels of self-efficacy would be positively associated with a preference for the social-comparative rating system. A significant positive association was found between self-efficacy and preference for performance feedback (b = -0.58, 95% CI [-1.05, -0.11], OR = 0.56, p = 0.016), indicating those higher on Self-Efficacy were more likely to prefer the social-comparative performance options. Therefore, Hypothesis 3 was supported.

Hypothesis 4a predicted that greater levels of organization would be associated with a preference for social-comparative performance rating systems. Likewise, Hypothesis 4b maintained that greater levels of diligence would also be positively associated with preference for social-comparative performance rating systems. To test these hypotheses, the four narrow traits that comprise conscientiousness were entered into the same ordinal logistic regression model. We found that organization (b = -0.56, 95% CI [-1.01, -0.13], OR = 0.57, p = 0.011) and, diligence (b = -0.52,95% CI [-1.03, -0.01], OR = 0.59, p = 0.046) were significant and positive predictors of a preference for social-comparative performance feedback, prudence (b = 0.67, 95% CI [0.19, 1.14], OR = 1.94, p = 0.005) was a significant predictor of a positive preference for absolute performance feedback, and perfectionism (b = -0.01, 95% CI [-0.48, 0.45], OR = 0.98, p = 0.950) was not found to be a significant predictor. Therefore, Hypotheses 4a and 4b were supported.

Next, we conducted exploratory analyses to address whether the narrow traits that underlie the remaining HEXACO personality traits (honesty-humility, emotionality, extraversion, agreeableness, and openness), might be associated with a preference for social-comparative or absolute performance rating systems. To evaluate this research question, we tested five separate ordinal logistic regression models and used the four relevant narrow traits as independent variables. None of the narrow traits were significant predictors in the logistic regression models for honesty-humility, extraversion, agreeableness, and openness to experience (See Table 2 for additional details). However, in the logistic regression model for emotionality, we found that fearfulness (b = 0.52, 95% CI [0.05, 0.99], OR = 1.68, p = 0.029) was a significant positive predictor of preference for absolute performance feedback, whereas anxiety (b = -0.12, 95% CI [-0.53, 0.28], OR = 0.88, p = 0.553), dependance (b = 0.29, 95% CI [-0.14, 0.71], OR = 1.33, p = 0.189), and sentimentality (b = -0.35, 95% CI [-0.81, 0.11], OR = 0.70, p = 0.134) were not found to be significant predictors.

Table 2 Results of ordinal logistic regression for HEXACO narrow traits
Full size table

Discussion

The purpose of the present study was to identify whether job applicants demonstrate a preference for social-comparative performance ratings over traditional absolute performance ratings. We also sought to identify whether any key psychological characteristics and individual differences were associated with having a preference between the two performance rating systems. The results of the study provide empirical evidence that some individuals may hold preferences for social-comparative performance evaluations over traditional, absolute performance evaluations. In the current sample, nearly half of the participants expressed such a preference. Moreover, support was found for the notion that certain individual differences are associated with having a preference for either social-comparative or absolute performance ratings.

One of the most important and interesting findings produced by this study came about through our research question: what proportion of employees will indicate a preference for social-comparative versus absolute performance ratings? We found that, when given a choice, some participants indicated a preference for social-comparative performance evaluations over absolute performance evaluations. The proportion of participants who indicated a preference for social comparative performance evaluations was unexpectedly high (46%) given that only 17% of organizations use social-comparative performance evaluations (Gorman et al., 2017). Moreover, this finding is important because most organizations exclusively use absolute performance evaluations. The results of this study provide the first empirical evidence that some individuals may in fact prefer social-comparative performance evaluations over absolute evaluations. This opens the doors for future researchers to thoroughly investigate the important related question of why they might prefer social-comparative performance evaluations. As discussed in greater detail below, one possible antecedent for this preference is individual differences in personality.

As predicted, we found that greater levels of SCO were associated with having a preference for social-comparative performance rating systems. This finding is theoretically consistent with our hypotheses, as well as those made by Gibbons and Bunk (1999), who indicated that individuals high on SCO have a proclivity for social-comparative based information. One implication of this finding is that it may help facilitate the development of specific feedback for employees. Individuals differ in their reaction to performance ratings and feedback (Keeping & Levy, 2000), and knowing in advance how to frame or present an individual’s performance ratings and feedback may facilitate the acceptance of and/or use of the feedback by the employee. That is, if a human resource manager is aware that an employee is particularly interested in receiving a certain type of feedback, the employee may be more receptive to the feedback if it is presented in a format or style that they are receptive to (Jawahar, 2006). In the future, researchers may wish to explore the affective, behavioral, and cognitive consequences of providing employees with the same feedback using different rating formats (i.e., relative or absolute). Importantly, this finding also demonstrates that individual differences may play a role in determining people’s preferences for social-comparative or absolute rating systems.

On the other hand, no support was found for the notion that greater levels of cognitive ability would be associated with a preference for the social-comparative rating system. This was unexpected, as extant empirical research and theory suggest there should be a relationship between cognitive ability and preference for social-comparative rating systems. A study conducted by Blume et al. (2013) found that cognitive ability was associated with individuals’ attraction to an organization that used forced-distribution ratings, a type of social-comparative rating. This is supported by the argument advanced by Trank et al. (2002), who posited that strong performers tend to have an interest in comparative performance. One possible explanation for our unexpected findings is that the type of social-comparative rating used in the present study differs rather substantially from the type of social-comparative ratings used by Blume et al. (2013). Researchers have noted that the various social-comparative rating formats are not all the same and can differ from one another in important ways (Roch et al., 2007). That is, it may be the case that the social-comparative rating format used in the study by Blume et al. (2013) differs in important ways from the social-comparative rating format used in the present study (the RPM). For example, the RPM involves comparing a ratee to a predetermined and relevant referent group, whereas other social-comparative methods, such as the ranked comparison or forced distribution methods, involve direct comparisons of individuals to other individuals. It is possible that such differences across the various social comparison-based methods may have contributed to limiting the generalizability of previous findings.

As predicted, we found that greater levels of self-efficacy were associated with preference for social-comparative ratings, thereby providing support for Hypothesis 3. Our hypothesis was predicated on the notion that individuals high in self-efficacy may accurately or inaccurately believe that they are stronger performers (Stajkovic & Luthans, 1998; Stone, 1994), and deserve to be rewarded accordingly. It is, therefore, plausible that they would demonstrate a preference for a social-comparative rating and reward system as these types of rating systems feature high reward differentiation for top performers (Goffin et al., 1996).

In support of Hypotheses 4a and 4b, greater levels of organization and diligence were associated with a preference for social-comparative feedback. The finding that individuals with a high need for structure and order (organization) prefer social-comparative feedback is consistent with conclusions drawn by Neuberg and Newsom (1993), who found that these types of individuals aim to process information as efficiently as possible. Compared to traditional absolute ratings, social-comparative performance ratings communicate more information in a condensed fashion making, which is why they may appeal to individuals high on organization. Moreover, humans are very accustomed to interpreting and making use of social comparative information, as social comparisons are used throughout everyday life to help understand how an individual measures up in comparison to others (Goffin & Olson, 2011). Accordingly, it may be the case that in the context of the workplace, those high on organization prefer a social-comparative rating system, as it provides them with information that they regularly interpret and implement to help facilitate efforts to improve their work performance. With regards to diligence, we found that greater levels of diligence were associated with a preference for social comparative rating systems. Given that social-comparative rating systems feature high reward differentiation (Goffin et al., 1996), this finding is consistent with previous research that has found that achievement-oriented individuals (i.e., those high on diligence) prefer merit-based rewards (Turban & Keon, 1993).

As mentioned, meta-analytic evidence suggests that those high in conscientiousness overall tend to be strong employees (Barrick & Mount, 1991; Dudley et al., 2006; Hurtz & Donovan, 2000; Schmidt & Hunter, 1998). In this study, individuals high on two of the components of conscientiousness (i.e., organization and diligence) demonstrated a preference for a social-comparative rating systems. This finding is consistent with Trank et al. (2002) and their argument that strong performers are more likely to be interested in comparative performance and reward systems.

This finding is also consistent with Moon et al. (2017) theory that introducing social-comparative ratings to an organization may induce a sorting effect. A sorting effect reflects the simultaneous process of organizational attraction and attrition due to a change in an element of organizational design. More specifically, a change in organizational design would result in higher performers being attracted to an organization while simultaneously motivating low performers to exit the organization (Cadsby et al., 2007; Gerhart & Fang, 2014; Trevor et al., 2012). Accordingly, the results of the present study suggest that if an organization were to introduce social-comparative ratings into its human resources and management systems (e.g., recruitment and selection), those who may be strong performers (i.e., high on two elements of conscientiousness) are more likely to demonstrate a preference for social-comparative ratings.

Exploratory analyses

Lastly, we would like to draw attention to some additional correlational findings. First, it was found that three of the four narrow facets of Openness to Experience (Creativity, Inquisitiveness, and Aesthetic Appreciation) were associated with a preference for social-comparative feedback. One possible explanation for this finding is that, in the context of our study, individuals high on openness view the possibility of working for a company that provides relative ratings as an opportunity to not only try something new, but to apply a new (to them) solution to a problem they have been faced with: the performance rating process. Relative ratings may be viewed as novel and alternative solution to problems they may have experienced with traditional absolute performance rating systems (Adler et al., 2016; Aguinis et al., 2012; Smith et al., 2000), thus facilitating a preference for relative performance ratings. Similarly, forgivingness, a facet of Agreeableness, was associated with a preference for social-comparative feedback. Those high on forgivingness can be characterized by a tendency towards accepting and trusting others who have previously treated them poorly. The performance appraisal process adversely affects the social relationship between the provider and recipient of workplace feedback (Pearce & Porter, 1986), this may be especially true of social-comparative feedback as it communicates implications regarding the social standing, and performance of, the recipient in the context of their own workplace which can be a negative experience (Goffin & Olson, 2011; Levy & Williams, 2004). However, those high on forgivingness may be more willing to look past the potentially more negative social information that may accompany this process. Next, fearfulness, a facet of Extraversion, was associated with a preference for absolute feedback. Those high on fearfulness can be characterized by a tendency towards harm avoidance (Lee & Ashton, 2004). Perhaps it is the case that these individuals perceive alternative performance rating systems as detrimental towards their ability to succeed, and therefore demonstrated a preference for the performance management system that they were more familiar with, which are absolute performance ratings (Gorman et al., 2017).

Finally, we would like to point out is the negative relationship found between the four facets of Honesty-Humility (Modesty, Greed Avoidance, Sincerity, and Fairness) and SCO. This is interesting because the four facets of Honesty-Humility were not found to be associated with a preference for absolute or social-comparative feedback. This implies there is variance in SCO that explains its relationship with preference for feedback type that is independent of a desire to attain status, power, and manipulate others for self-serving interests (Lee et al., 2013). Future researchers may wish to explore this distinction further.Footnote 2 However, it must be mentioned these correlational findings must be interpreted with caution due to the ordered categorical nature of the preference for absolute vs social-comparative feedback variable (Bürkner & Vuorre, 2019).

Practical implications

The results of our study supported the notion that individual differences predict preferences for social-comparative performance ratings. This has practical implications for organizations that either already use social-comparative performance ratings or are seeking to implement them. Previous research on social comparison has found that the use of social-comparative performance evaluations may encourage competitiveness amongst employees which could impact the culture of the organization (Chun et al., 2023; Garcia et al., 2013). Nevertheless, organizations can take strategic advantage of an individual’s interest in social-comparative performance evaluations and improve the ways in which their recruitment and selection systems are designed as a function of person-organization fit. For example, previous research has found that when employees experience poor person-organization fit, they are more likely to experience low levels of organizational commitment (Meyer et al., 2002) and subsequently engage in turnover (Judge, 1994). However, when job applicants are given extensive and realistic information about a job or organization in advance of being hired, they are less likely to engage in turnover (Bretz & Judge, 1998). To take advantage of this finding, organizations that make use of social-comparative ratings could consider targeting individuals high on SCO during recruitment and selection processes, however, this should not be done without consideration for the finding that SCO can be negatively correlated with Honesty-Humility (see Table 1). This in turn may improve person-organization fit for those who enter the organization and may reduce their long-term intentions to turnover.

Another way organizations using social-comparative ratings can implement the findings of the present study is to continue selecting for individual differences that may be related to job performance. This includes individuals high on organization, diligence, and self-efficacy. Selecting individuals high on organization and diligence can be achieved via selection for individuals high on conscientiousness (Lee & Ashton, 2004). As discussed in the context of the findings relating to organization and diligence, organizations that use social-comparative ratings may be able to attract stronger performers to the organization as individuals high on conscientiousness (i.e., those who would also be high on diligence and organization) are not only more likely to prefer social-comparative ratings, but also tend to be strong performers at work (Barrick & Mount, 1991). Likewise, previous meta-analytic evidence points to a strong positive relationship between self-efficacy and job performance (Stajkovic & Luthans, 1998). Therefore, by informing prospective applicants that social-comparative ratings are used within an organization it is possible to improve overall organizational performance, as stronger performing employees may be more attracted to, and subsequently more inclined to enter the organization.

Limitations

As is the case with all studies, our research is not without its limitations. The primary limitation of this study was our use of vignettes as their generalizability is limited (Aguinis & Bradley, 2014). In developing our study, we followed best practices in our design and use of vignettes. According to Aguinis and Bradley (2014), the use of vignettes is ideal for studies where it would be unethical or impractical to manipulate certain elements of a situation. For example, it is not practical to manipulate the type of performance rating system within a real-world organization. Future researchers interested in evaluating and building upon the generalizability this work should consider moving beyond vignettes and conduct a field test using a simulated or actual recruitment scenario. However, the results of lab and field studies typically converge (Anderson et al., 1999). Similarly, in the vignettes from the present study, we asked participants to imagine that they were looking for jobs as retail sales managers. Consequently, due to our use of vignettes, we were unable to examine how our findings may have generalized across industries or job types. Accordingly, another future direction that researchers may wish to explore is whether findings from the present study generalize across industries and job types. For instance, it may be prudent to examine jobs that vary in the extent to which their tasks are cognitively demanding, which may in turn allow future researchers to revisit the relationship between cognitive ability and preferences for social-comparative performance feedback.

Conclusion

In conclusion, the study presents evidence that individual differences may indeed play a role in individuals’ preferences for different types of performance rating systems. These findings have both theoretical and practical implications for organizational performance that should be considered when designing organizational recruitment and selection systems.