In the past 2 decades, scholars have made a concerted effort to better understand the impact that selection procedures have on job candidates during and after the hiring process (e.g., Anderson et al. 2010; Chan and Schmitt 2004; Hausknecht et al. 2004; Ryan and Ployhart 2000). In short, the knowledge from this stream of literature suggests that job applicants form judgments about an organization’s selection procedure, and these perceptions affect subsequent job attitudes and behaviors (Ababneh et al. 2014; Ryan and Huth 2008).

Despite these advances, applicant reactions research has been criticized for lacking direct relevance to organizations (e.g., Ryan and Huth 2008). Some of this criticism stems from the belief that a self-serving bias appears to be driving research findings (Ababneh et al. 2014; Chan et al. 1998a; Fiske and Taylor 1991). A self-serving bias occurs when individuals attribute themselves credit for their successes based on their own internal or personal factors but assign blame for failures to outside forces that are beyond their personal influence (Miller and Ross 1975). Extant findings regarding this self-serving bias can be boiled down to a simple maxim: when applicants perform well on a selection procedure, they tend to like that selection procedure, and when applicants perform poorly on a selection procedure, they tend to dislike that selection procedure. Consequently, there is general agreement among scholars that selection procedure performance is a key driver of applicant reactions (e.g., Bauer et al. 1998; Chan et al. 1998a; Maertz et al. 2004).

In the present study, we challenge this view by taking a dispositional approach to the study of applicant reactions. Specifically, we argue that job applicants’ core self-evaluations (CSE) play an important role in explaining how they will react toward the organization and its selection procedures. CSEs are fundamental evaluations of the self (e.g., worthiness, competence, capabilities) that have the potential to affect how individuals perceive and interact with the world (Ferris et al. 2013; Johnson et al. 2008; Judge et al. 1997).The crux of our argument is that job applicants who evaluate themselves in a positive manner (i.e., high CSE) will evaluate the focal organization and its testing procedures in a similarly positive manner. Correspondingly, we expect job applicants with low CSE (i.e., negative evaluations of the self) to view the focal organization in a negative manner resulting in more negative reactions. Further, we believe the CSE-application reactions relationship will hold even after controlling for selection procedure performance, which we conceptualize as a measure of the self-serving bias.

This study makes several theoretical contributions to the literature. First, building on previous personality and applicant reactions research (e.g., Bretz and Judge 1994; Truxillo et al. 2006; Viswesvaran and Ones 2004), we heed the call of scholars to investigate applicant personality from a broader and more global perspective (e.g., Chan and Schmitt 2004; Ryan and Ployhart 2000). In short, by measuring disposition at a very wide scope, we aim to show that some job applicants enter the selection process with dispositional “baggage.” This baggage means that individuals are likely to walk away from the process with a negative view of the organization—even when these applicants perform well in the selection procedure and are offered a position. Second, we contribute to the understanding of CSE and its consequences by reexamining the relative importance of the self-serving bias. We argue that CSE should predict applicant reactions even after controlling for selection test performance—further supporting the notion that CSE is a stable personality construct that affects how behavioral intentions are formed. Finally, we build on justice research by showing that CSE impacts the well-established relationship between fairness perceptions (Gilliland 1993) and applicant reactions (e.g., Bell et al. 2006; Chapman et al. 2003; Smither et al. 1993).

Theoretical Background

Core Self-Evaluations

Judge and colleagues (2003) describe CSE as “a basic, fundamental appraisal of one’s worthiness, effectiveness and capability as a person” (Judge et al. 2003, p. 304). Because CSE reflects “a baseline appraisal that is implicit in all other beliefs and evaluations” (Chang et al. 2012, p. 83), it is not surprising that CSE has shown to be an important driver of how individuals perceive and interact with the world—especially the workplace (Judge et al. 2002, 2008). For example, studies have repeatedly linked CSE with a number of positive work (e.g., job satisfaction, performance) and life (e.g., life satisfaction) outcomes (e.g., Chang et al. 2012; Judge and Bono 2001; Judge et al. 2005; Kacmar et al. 2009).

From a theoretical perspective, CSE is thought to have an influence on outcomes both directly and indirectly (Ferris et al. 2011). Direct effects of CSE occur when positive evaluations of the self “spill over” into other appraisals (e.g., emotional generalization). Indirectly, CSE is thought to affect the appraisals people make (e.g., job characteristics) as well as the actions they engage in (e.g., working longer at tasks), which ultimately affect outcomes such as job satisfaction and job performance (Chang et al. 2012). Although two previous studies have examined CSE in an applicant reactions context (Anderson et al. 2012; Nikolaou and Judge 2007), these two works were mostly exploratory in nature and did not invoke CSE theory to explain how applicant CSE affects outcomes such as selection procedure performance or behavioral intentions toward an organization.

Applicant Reactions and Behavioral Intentions

Applicant reactions research attempts to understand “attitudes, affect, or cognitions an individual might have about the hiring process” (Ryan and Ployhart 2000, p. 566). This research is important because applicant reactions are responses that job candidates have about the selection systems that organizations implement. If selection systems create negative reactions, job candidates will likely form more negative reactions toward the organization (Hoang et al. 2012; Ryan and Ployhart 2000; Steiner and Gilliland 1996). Thus, they should be less likely to accept the job, reapply for the position, or encourage others to apply for the position (Anderson et al. 2010; Chan et al. 1998b; Hausknecht et al. 2004). Although research examining applicant-focused outcomes (e.g., fairness, likability) has advanced our understanding of the selection process, scholars have called for applicant reactions research to examine outcomes that are more relevant to the hiring organization (Chan and Schmitt 2004; Hülsheger and Anderson 2009; Ryan and Huth 2008).

Accordingly, we focus on applicant behavioral intentions in the present research, as intentions are important precursors to future behaviors in which applicants are likely to engage (Ajzen 2011; Hausknecht et al. 2004). Behavioral intentions are defined as “a person’s subjective probability that he or she will perform some behavior” (Fishbein and Ajzen 1975, p. 288). Whether or not an applicant intends to accept a job offer matters a great deal to organizations because a selection tool’s effectiveness and utility is greatly diminished if top candidates respond negatively to a selection procedure and turn down job offers (Becker et al. 2010; Gilliland 1994). Moreover, reapplication intentions (i.e., whether applicants will make multiple efforts to obtain a position within the organization) and recommendation intentions (i.e., the likelihood that the applicant will encourage other people to apply for a similar position with the organization) impact the quality and quantity of future applicant pools (Kluger and Rothstein 1993), which also have an effect on test utility.

Although previous research has examined applicant behavioral intentions (e.g., Bell et al. 2006; McCarthy et al. 2009; Walker et al. 2008), there is little research related to the role that disposition plays in determining applicant intentions (Chan and Schmitt 2004; Ryan and Ployhart 2000). Consequently, we utilize CSE theory (Judge et al. 1997, 1998) to advance understanding of the relationship between applicant personality and applicant behavioral intentions.

Hypotheses

CSE and Behavioral Intentions

Research has shown that self-beliefs are major drivers of behaviors and behavioral intentions (Ajzen 1991; 2011). CSE theory (Judge et al. 1997) indicates that decisions and behaviors are driven by evaluations of how individuals view their own lives (i.e., attitude toward the self) as well as how they view their own place in the world (i.e., their ability to control their own destiny). Individuals with high CSE tend to have more positive evaluations of objects (e.g., organizations) and actions (e.g., test performance) in general (Chang et al. 2012). Because positive attitudes represent an overall favorable mindset and lead to positive evaluations about one’s surroundings (Watson et al. 1988), applicants with high levels of CSE are likely to form positive opinions about the organizations with which they interact. Those positive evaluations of the organization should ultimately result in more positive behavioral intentions toward that organization (e.g., Judge et al. 2005). Conversely, individuals with low levels of CSE should have generally negative views of themselves and their surroundings (including organizations with which they interact) and are thus likely to express more negative intentions and behaviors.

Based on this logic, we expect that applicant CSE will be positively related to intentions to accept the job, reapplication intentions, and recommendation intentions. The link between CSE and intentions to accept the job can be explained indirectly by efficacy beliefs (Bandura 1977, 1982). According to Ajzen (2011), individuals are likely to have stronger behavioral intentions when they believe they can perform a behavior successfully. Because individuals with high levels of CSE have an ability to counter setbacks and seize opportunities more successfully while displaying a greater sense of self-determination (Judge and Hurst 2007), they should feel more strongly about their chances of success in a new position and form positive intentions regarding the potential to accept a position when offered. Conversely, low CSE applicants are more likely to doubt their skills and abilities (Judge et al. 2003) and question whether they are really good enough to do the job (if it is offered to them). As a result, these applicants should be less likely to intend to accept the job.

Regarding reapplication intentions, applicants with high CSE who are rejected during the selection process should be more likely to reapply to the organization for the same position because they believe that they have what it takes to succeed and they enjoy the challenge of shaping their own future by attempting to exert some control over it (Judge 2009). On the other hand, applicants with low levels of CSE should display a lack of self-confidence and less assurance of their own destiny when confronted with a negative selection outcome, and those beliefs should have a negative impact on their intentions to reapply for the same position.

Finally, in relation to recommendation intentions, high CSE applicants will likely exude positive expressions of their experience toward others (Judge et al. 1997) and willingly form recommendation intentions that reflect their good feelings about the process regardless of the actual outcome. Low CSE applicants—with a negative outlook on life—should be less willing to recommend the organization to other people because they are likely to pick up on more negative information during the selection process (Ferris et al. 2011). They are also more likely to think that others are incapable of performing the job and will not recommend the position to other people based on this negative outlook.

As noted above, the self-serving bias is thought to be a driver of applicant responses and intentions (e.g., Chan et al. 1998a) as research has shown that when applicants perform well on a selection procedure they tend to react more positively, and when they perform poorly on a selection procedure they tend to react more negatively (e.g., Bauer et al. 1998). However, our dispositional arguments suggest that the effects of a self-serving bias may not be as strong as previously thought as the degree of CSE held by job applicants should impact the formation of behavioral intentions. Because disposition remains relatively stable across situations (Judge et al. 2003), CSE is unlikely to change even after applicants find out about their test performance; therefore, the impact of CSE on behavioral intention formation should be present regardless of the favorability of the outcome (Ajzen 2005). Thus, stable, long-term dispositional factors should have a more meaningful impact on the formation of intentions (Ajzen 1991, 2005) than relatively short-term factors encapsulated by a transient self-serving motive.

Hypothesis 1

CSE will explain a significant amount of variance in job applicant behavioral intentions (i.e., intentions of job acceptance, reapplication intentions, and recommendation intentions) above and beyond the applicant’s test performance (i.e., the self-serving bias).

A Moderating Role for Core Self-Evaluations

Scholars have long known that a strong relationship exists between fairness perceptions and applicant reactions (Hausknecht et al. 2004; Ryan and Ployhart 2000). For example, studies have linked applicant fairness perceptions with test-taking self-efficacy (Bauer et al. 1998), organizational attractiveness (Lazar et al. 2007; Schinkel et al. 2013), recommendation intentions, acceptance intentions, and reapplication intentions (e.g., Bell et al. 2006; McCarthy et al. 2009; Ployhart and Ryan 1997). From a theoretical perspective, fairness perceptions are thought to have such a strong effect on subsequent applicant attitudes and behaviors because fairness communicates important information to applicants about how the organization views them (Tyler 1994).

However, not all job candidates are likely to react to the same justice information in the same way (Gilliland 1993). Indeed, the idea that some applicants are likely to be more sensitive to justice information than other applicants is consistent with the tenets of CSE theory. For example, Ferris et al. (2011) maintain that a core aspect of CSE involves sensitivity to positive and negative information. With this in mind, we draw from these previous works to argue that applicants’ CSE represents a boundary condition for understanding the fairness perceptions-applicant reactions relationship.

Because low CSE applicants doubt their own abilities to succeed, have an overall negative disposition, and perceive themselves as pawns to the whims of the outside world (Judge 2009), it is likely that they enter the selection process with the preconceived notion that they will be treated poorly. Thus, low CSE applicants should be “on the lookout” for unfair situations in a selection process, and the occurrence of justice violations should confirm and enhance their negative self- and world-views, and cause them to react even more negatively toward the organization. This line of reasoning is consistent with other research showing that the combination of negative personal and situational factors lead to more negative behavioral intentions (e.g., Armitage and Conner 2001; McEachan et al. 2011). Conversely, high CSE applicants should be less sensitive to the information that negative justice information generally communicates (Ferris et al. 2011; Judge et al. 1997). More generally, because high CSE applicants tend to hold positive attitudes and believe that they control their own destiny (Judge 2009), their behavioral intentions are less likely to be influenced by external situational factors in general (Ajzen 1991). In short, violations of justice rules in the selection process should have a stronger (negative) effect on low CSE applicants than high CSE applicants.

In a similar manner, we expect adherence to justice rules to also have a stronger effect on low CSE applicants. Because low CSE applicants are more influenced by external situations (e.g., Judge et al. 2003), they should be more susceptible to the positive information that is communicated by an organization’s testing processes. Specifically, a perceived fair selection process should signal to low CSE applicants that the external situation is playing into their favor and is likely to enhance their potential success. The end result should be that they form even more positive behavioral intentions toward the organization. On the other hand, because high CSE applicants have positive dispositions and more perceived control over their own behavior (Judge et al. 2008), they should be less sensitive to justice information. More generally speaking, we expect that justice information (negative or positive) will not matter as much to high CSE applicants.

Hypothesis 2

CSE moderates the positive relationship between justice perceptions and behavioral intentions (i.e., intentions of job acceptance, reapplication intentions, and recommendation intentions) such that this relationship is stronger for applicants with low levels of CSE.

A visual representation of both Hypothesis 1 and 2 and the importance of test performance is displayed in Fig. 1.

Fig. 1
figure 1

The relationship between fairness perceptions, core self-evaluations (CSE), and behavioral intentions. Solid lines indicate hypothesized relationships whereas dashed lines indicate relationships established in the literature

Full size image

Method

Sample

The sample consisted of job applicants for the position of police officer in a large Eastern city in the United States. The original sample size consisted of 742 individuals applying for the position. Due to the desire to measure CSE, fairness perceptions, and behavioral intentions at different points in time, only respondents who completed surveys at all the time points were included in this analysis. This resulted in final sample of 194 job applicants (a response rate of 26 %)—a decrease in initial sample size that is consistent with findings in the extant literature (e.g., Bell et al. 2006; Truxillo et al. 2002) and suitable for research of this type (Salgado 1998).

Procedure

Surveys and selection procedures were administered to the job applicants at four different points in time. We employed this multi-wave procedure for a number of different reasons. First, our aim was to better understand applicant behavioral intentions after the job candidates found out how they performed on the organization’s selection procedure—in effect, allowing us to measure and control for the self-serving bias. Second, scholars have recommended this approach (e.g., Hausknecht et al. 2004) as it has shown to increase predictive validity during applicant reactions research (e.g., Bauer et al. 1998). Finally, the use of multiple measures over multiple points in time is likely to diminish the influence of common method variance on our findings (e.g., Podsakoff et al. 2003; Richardson et al. 2009).

At measurement time 1, job applicants who signed up to take the police officer exam were mailed a survey that measured their level of CSE. Approximately one month later, at measurement time 2, job applicants completed the police officer exam in a high-stakes setting. Specifically, all applicants completed the exam at the same testing facility in the focal city. After the applicants completed the test, they were given a survey to measure fairness perceptions (time 3). Approximately 1 month later, all applicants were mailed information regarding their performance on the selection test. This information included both the raw score on the selection test as well as their overall test score ranking among the entire pool of applicants. It was at this point (time 4) that job candidates were surveyed regarding their behavioral intentions (i.e., job acceptance intentions, intentions to re-apply, and recommendation intentions). Thus, applicants responded to the surveys at time 4 after they knew their performance on the selection exam (and relative rank among the other applicants). Due to the nature of this procedure, we also tested for non-inclusion bias in our sample; we found that average test score did not significantly differ for those who did and did not respond to our survey at time 4. Similarly, there were no significant differences in average levels of CSE for responders and non-responders.

Measures

Core Self-Evaluations

CSE was measured at time 1 using the core self-evaluations scale (CSES) developed by Judge et al. (2003). This Likert-type scale is composed of 12 items and includes statements such as “Sometimes when I fail I feel worthless” (reverse scored) and “I am confident I get the success I deserve in life” (Judge et al. 2003). The scale ranges between 1 and 5 and descriptions are anchored from 1 = strongly disagree to 5 = strongly agree. The CSES in the current sample exhibited a coefficient alpha of .81. CSE scores were not included in the final civil service exam score given to the applicants.

Test Performance

The selection procedure was administered by the hiring organization at time 2. The civil service exam administered to police officer candidates was similar to exams studied in prior applicant reactions research and is common for civil service positions of this nature (e.g., Barrett et al. 1999; McCarthy et al. 2009; Truxillo et al. 2002). Specifically, the procedures were developed by the focal organization in collaboration with several industrial and organizational psychologists to measure a combination of cognitive ability, integrity, and situational judgment for the purpose of external selection. Although specific information regarding how the final scores were calculated is not available due to proprietary reasons, the psychometric properties, structure, and criterion validity are similar to those of other civil service exams. Moreover, to the point of this paper, job applicants were aware that their performance on the selection procedure (i.e., overall test score) would determine whether or not they would be offered the position. Thus, the selection procedure was administered in a setting that can be characterized as “high-stakes” (e.g., Levashina et al. 2012; Lievens and Patterson 2011).

Fairness Perceptions

The selection procedural justice scale (SPJS; Bauer et al. 2001) was used to measure applicants’ fairness perceptions of the selection process at time 3 (i.e., after the applicants completed the selection process). The SPJS contains Likert-type measures and is composed of items with statements such as “I was treated honestly and openly during the testing process” and “I was comfortable with the idea of expressing my concerns at the test site.” The scale has ranges between 1 and 5 and includes descriptions anchored from 1 = strongly disagree to 5 = strongly agree (Bauer et al. 2001). The alpha reliability for the SPJS in our sample was 0.89.

Behavioral Intentions

Three different behavioral intentions were measured at time 4 (i.e., after applicants learned of their test performance). Intentions of job acceptance reflect a desire to take the job if it is offered. A single item adapted from Macan et al. (1994) was used to assess this variable, “I intend to accept the job if it is offered.” Reapplication intentions relate to plans to pursue a position again in the future. A single item derived from Ployhart and Ryan (1997) that reflects this behavioral intention, “I would apply to this organization again,” was used to capture this notion. Recommendation intentions demonstrate a willingness to promote the organization as well as the selection process to others. A single item for this measure was adapted from Smither et al. (1993) and included the statement “Based on my experience with this process, I would encourage others to apply for employment with this organization.” For all three intentions, Likert scales were used ranging from 1 to 5 representing 1 = strongly disagree to 5 = strongly agree.

Analysis

Study hypotheses were assessed using hierarchical regression analyses. Test performance was entered in Step 1, and main effects of CSE and fairness perceptions at Step 2 and Step 3, respectively. Finally, the interaction term (CSE X fairness) was entered at Step 4. We also performed a simple slopes test for each regression to determine if moderation was statistically significant at different levels of CSE for the three behavioral intentions measured (Preacher et al. 2004).

Results

Table 1 provides the means, standard deviations, and correlations of the variables in this study. For Hypothesis 1, we examined the effect of CSE on behavioral intentions over and above that of test performance (i.e., the self-serving bias.) Specifically, after entering test performance in step 1 (see Table 2), our results indicate that CSE had significant incremental impact on acceptance intentions (β = .24, ∆R 2 = .05, p < .01), reapplication intentions (β = .35, ∆R 2 = .09, p < .001), and recommendation intentions (β = .27, ∆R 2 = .06, p < .01). Thus, Hypothesis 1 was supported. Although our data suggest that test performance may play a role in the way that applicants behave after experiencing an organization’s selection process (i.e., the self-serving bias), CSE appears to also be an important driver of applicant behavioral intentions.

Table 1 Means, standard deviations, correlations, and reliabilities
Full size table
Table 2 Results of hierarchical regression analysis for behavioral intentions
Full size table

We used moderated hierarchical regression analysis to test Hypothesis 2—which examined the extent to which CSE moderates the fairness perceptions-behavioral intentions relationship. A significant change in R 2 for each of the three dependent variables after entering the interaction term of CSE X fairness in the regression model at step 4 indicates a moderation effect. Our results indicate that the interaction term explains a significant level of variance as the R 2 change in each of the three models was shown to be significant. Results for acceptance intentions show a ΔR 2 value of .05 (p < .001) with the interaction term included. Reapplication intentions show a ΔR 2 value of .03 (p < .05) and recommendation intentions show a ΔR 2 value of .02 (p < .05) after including the interaction term. Thus, Hypothesis 2 was supported. Table 2 provides a complete breakdown of the regression models.

We also conducted a simple slopes analysis to determine if each interaction was significant at different levels of fairness perceptions and plotted the interaction at −1 SD, and +1 SD. CSE and fairness perceptions significantly interacted to impact acceptance intentions at −1 SD (t = 6.74, p = .00). However, at +1 SD the simple slope was not significant (t = 1.39, n.s.). The magnitude of the slope change was stronger for applicants with low levels of CSE (compared to those with high levels of CSE), and this supports Hypothesis 2. Reapplication intentions were plotted and examined in a similar manner—and significant differences were obtained at both −1 SD (t = 7.01, p = .00), and at +1 SD (t = 3.90, p = .00), again supporting Hypothesis 2. Finally, recommendation intentions were plotted and examined at both −1 SD (t = 7.81, p = .00) and at +1 SD (t = 5.62, p = .00), and again there were significant differences in the predicted direction. In all three cases the magnitude of the slope change (between justice perceptions and behavioral intentions) was stronger for low levels of CSE. Figures 2, 3 and 4 graphically depict these interaction effects (Dawson 2014).

Fig. 2
figure 2

The moderating effects of CSE on the fairness perceptions to acceptance intentions relationship. CSE is core self-evaluations, fairness perceptions are measured by selection procedural justice scale (SPJS)

Full size image
Fig. 3
figure 3

The moderating effects of CSE on the fairness perceptions to reapplication intentions relationship. CSE is core self-evaluations, fairness perceptions are measured by selection procedural justice scale (SPJS)

Full size image
Fig. 4
figure 4

The moderating effects of CSE on the fairness perceptions to recommendation intentions relationship. CSE is core self-evaluations, fairness perceptions are measured by selection procedural justice scale (SPJS)

Full size image

Discussion

Applicant reactions to selection procedures remain of vital interest to researchers and practitioners due to their potential to impact organizational outcomes (Anderson et al. 2010; Hausknecht et al. 2004; Ryan and Ployhart 2000). Although numerous studies have shown that circumstances (e.g., test performance) play a role in understanding applicant behavioral intentions, approaches examining the impact of personality on this phenomenon have been quite infrequent (Bretz and Judge 1994; Truxillo et al. 2006; Viswesvaran and Ones 2004). Our aim was to show that global disposition (i.e., CSE) matters when attempting to explain applicant intent—even when test performance and fairness perceptions are considered. Specifically, we found that CSE is significantly related to post-hiring-decision behavioral intentions (i.e., acceptance, reapplication, and recommendation intentions) even after controlling for selection test performance (i.e., the self-serving bias). Thus, CSE appears to play a significant role in explaining and understanding applicant reactions.

Additionally, we found support for the notion that CSE moderates the fairness perception-behavioral intentions relationship. The plots of these interactions (presented in Figs. 2, 3, 4) along with tests of simple slopes paint an even more detailed picture about the moderating role of CSE. In particular, our results suggest that job applicants with a negative disposition (i.e., low CSE) are the candidates most likely to be affected by (both positive and negative) justice information as the combination of low CSE and perceived unfairness most strongly impacts intentions. Overall, our results suggest that CSE is an important mechanism in understanding how people respond to organizations during a high-stakes selection setting.

Because high levels of CSE are associated with greater confidence in one’s own ability, previous research has found that individuals with high CSE generally outperform those who have low levels of CSE (e.g., Erez and Judge 2001; Kacmar et al. 2009; Judge and Hurst 2007). These previous findings led us to investigate whether or not CSE was related to applicants’ performance on the high-stakes civil service exam. Our hunch was confirmed by the significant and positive relationship (r = .48, p < .01) between CSE and selection test performance. These results suggest that having a positive self-concept also influences performance in high-stakes testing contexts—possibly because the same traits that enable people to cope with change effectively (e.g., low levels of neuroticism; high levels of self-efficacy) allow them to perform better in testing conditions that induce anxiety and stress (Judge et al. 1998). These supplementary results provide us with another useful piece of information about CSE that suggests the construct could be particularly useful in the selection process (Ferris et al. 2011).

Theoretical Implications

Our findings advance theoretical understanding in few different ways. First, we contribute to CSE theory by linking it with applicant behavior. Previously, researchers have suggested that incorporating disposition into applicant reactions should produce fruitful understanding of the mechanisms at play in the formation of behavioral intentions (Ryan and Ployhart 2000). We believe our arguments and results successfully demonstrate that CSE encapsulates elements of the personal factors that impact intention creation. Because beliefs about the self and beliefs about the ability to control one’s surroundings are derived from self-concept arguments invoked in the CSE literature, we argued for a relationship between CSE and behavioral intentions. Thus, our work reveals more information about how and why CSE impacts work outcomes such as behavioral intentions. Judge (2009) has suggested that the measurement of CSE would help organizations to better understand their employees. Our results enhance these arguments by extending this logic to job applicants.

Another contribution of our research comes from the fact that we conducted our study over the course of multiple time periods. By utilizing a longitudinal design, we were able to test the theory behind CSE and demonstrate in a high stakes setting that it has a significant effect on applicant reactions regardless of test performance. Moreover, by showing how CSE predicts multiple outcomes over multiple points in time, we further highlight its stability as a dispositional construct.

Finally, we make a contribution to the applicant reactions literature by showing when and how CSE provides a boundary condition in explaining the justice perceptions-behavioral intentions relationship. Specifically, our results show that low CSE applicants are more sensitive to justice information—both negative and positive—when forming their reactions toward the hiring organization. This effect is important because fairness perceptions are associated with higher levels of acceptance, reapplication and recommendation intentions (e.g., Bell et al. 2006; Chapman et al. 2003; Smither et al. 1993). By studying the combination of fairness and CSE, our research extends prior work and helps to explain other drivers of applicant responses to selection procedures. In particular, our findings have implications for research on perceived violations of justice rules. As previous research has shown, perceived violations of justice rules have a stronger impact on outcomes than adherence to justice rules (e.g., Bauer et al. 1998; Maertz et al. 2004). Our results show that CSE—especially low levels of CSE—are also important as a specific condition that impacts candidate intention formation.

Practical Implications

For organizational practitioners, one implication of our results is that job applicants likely enter the selection process with dispositional “baggage.” Ultimately, this means that applicants have some part of their judgments and reactions already in place—even before they experience the selection procedure and outcome. Clearly, the impact of CSE is not the sole or overriding issue in the formation of intentions, but our results demonstrate that it is significant and meaningful. Because test performance and intent seem to be partially predetermined—via levels of CSE—before applicants even experience the selection system, there is likely little that an organization can do to completely eliminate negative applicant reactions in the selection process. Thus, our advice is that disposition should be accounted for early on in the selection process, as it appears to be a major driver of applicant perceptions and behavioral intentions.

The personality factors that influence differences in applicant behavioral intentions are important for human resources professionals to understand because of the impact they have on the organization’s chances of obtaining quality employees, developing a positive reputation, or getting quality applicants to try again when a new position is open. For example, a one point increase in applicant CSE leads to an increase in behavioral intentions of approximately 7 % for each outcome we examined. Moving beyond a job applicant context, our results also suggest that CSE is likely to affect employee reactions to other aspects of organizational staffing decisions including promotion and termination. This is an important consideration for practitioners as the number of lawsuits filed against organizations in promotion and termination contexts is even higher than the number of lawsuits that are filed in application contexts (e.g., DiLorenzo 2010; Tomlinson and Bockanic 2009).

We also present evidence here that disposition might even play a role in determining the operational validity of the selection methods that are used. Specifically, our results showed that levels of CSE are related to test performance levels which mirror findings regarding CSE’s relationship with job performance and other work outcomes (e.g., Chang et al. 2012). While there could be some overlap between the integrity portion of the selection test and CSE, we do not believe this to be a significant issue as overt integrity is a distinct construct from personality variables. Practically speaking, the importance of CSE becomes more relevant when considering that a one point increase in applicant CSE results in approximately 7 % higher test performance scores. Thus, organizations may wish to test potential job applicants for levels of CSE early on in the selection process as low levels of CSE may be indicative of less successful employees. Not only would such testing save time, money, and effort in recruiting processes, research suggests that they are very few ethnic or gender group differences in CSE, which should minimize concerns of adverse-impact (Chang et al. 2012).

Limitations and Future Research

As with all empirical studies, our study has limitations that must be considered. First, our inability to measure applicant behavior beyond the selection process represents a limitation. However, because the purpose of the study was to measure the relationship between CSE and behavioral intentions, we do not see this limitation as severe considering our study context focuses on applicant responses to selection systems (Ryan and Ployhart 2000). Moreover, prior research suggests that the relationship between behavioral intentions and behaviors is very strong (e.g., Ajzen 2011), and our development of arguments for the CSE-behavioral intentions relationship suggests that the generalizability of our proposed model will be of merit to scholars.

A second limitation is our use of a single measure of personality to examine applicant reactions. However, we believe that this is not a severe issue as CSE is a personality variable that is wide in scope and measures disposition in the broadest sense possible (e.g., Judge 2009; Judge et al. 2003). From a more practical perspective, previous research suggests that using multiple measures of constructs in applied settings is extremely difficult (e.g., Wanous et al. 1997). Due to the high-stakes nature of applicant testing, organizational stakeholders are often unwilling to allow researchers to manipulate and lengthen selection tests associated with these samples due to increased applicant fatigue (Nguyen et al. 2003), limited space on applicant surveys (Wanous et al. 1997), or negative applicant reactions that result from completing repetitive measures (Ni and Hauenstein 1998; Ryan and Ployhart 2000). Still, we recommend that this research be replicated using different measures of personality in other high-stakes assessment settings. For example, both self and other ratings of the Big Five framework of personality have demonstrated that multiple measures of personality enhance the validity of predicting workplace performance outcomes (e.g., Kluemper et al. 2015; Oh et al. 2011)—similar results may be found for applicant reactions as well.

Third, we are cognizant of the fact that our dependent variables were measured using single items and could therefore be seen as less reliable than what might otherwise be desired. This may be especially relevant due to the relatively high mean scores we documented for each intention. Using multiple items could have created more variance in our outcome variables and reduced the negative skew in our sample’s distribution. Negative skew may not be a severe issue in our research because compression of a score scale does not always lead to loss of information in that region (Ho and Yu 2014), and this issue can be common in a setting where high scores are the norm (Koedel and Betts 2008, 2010). Additionally, in the measurement of intentions, the use of multi-item measures may not be as critical as it is for other constructs. For example, Wanous et al. (1997) detail multiple reasons why the use of a single-item measure can be appropriate. Practically speaking, single-item measures require less space and time to complete which reduces fatigue on the part of the respondent. In particular, behavioral intentions are straightforward in their content coverage (Hulin and Judge 2003) and asking a simple question multiple times could even exacerbate negative applicant responses. In this vein, Sackett and Larson (1990) recommend that single-item measures are appropriate when the construct of interest is unidimensional, clear to the respondent, and sufficiently narrow in scope (Wanous and Hudy 2001). We believe that our behavioral intention items meet these qualifications and our results provided adequate variance to draw meaningful conclusions.

We are also limited from the standpoint of generalizability of our results to different applicant populations. Due to the sensitive nature of the selection system in this high-stakes setting, the hiring organization did not allow us to collect data regarding age, sex, race and prior work experience. Therefore, these measures normally assessed as control variables were unavailable in our analysis. However, we do not believe this to be a significant limitation (see Becker 2005) because prior research has demonstrated that gender, age, and race play little part in predicting applicant reactions and therefore are not critical for use as control variables (Hausknecht et al. 2004).

A final limitation is our use of the applicant’s knowledge of their test score as a proxy for the self-serving bias. We are not as concerned with this potential limitation because all applicants were aware in advance that the test score would be used to determine whether or not they would receive a job offer. Applicants were also aware of their final ranking among all the candidates who underwent the selection process. Therefore, test score should act as a sufficient approximation of the applicants’ self-serving bias. From a theoretical perspective, one might also consider our findings as an alternative perspective to the self-serving bias approach that this research stream generally takes. For example, CSE could be viewed as another kind of self-serving bias from the standpoint that self-evaluations can be geared toward more positive outlooks about the self—which ultimately drive more positive applicant reactions and test performance in general. From this perspective, one could argue that CSE represents a different type of self-serving bias—one that is more person-focused than situation-focused—and that such a view should be considered when attempting to understand how applicants react in selection settings. Accordingly, previous research that conceptualized the self-serving bias in narrow terms (i.e., test outcome) may need to be broadened and integrated with a dispositional perspective, and future research should be conducted to explore this possibility.

Conclusion

As opposed to many studies that have taken a situational approach to understand how applicants react to the selection process, this study examines these issues from a dispositional perspective. Specifically, we found that CSE has a positive influence on behavioral intentions—even after applicants found out about their selection test performance and whether or not they would be offered the position. Additionally, we show that low levels of CSE strengthen the positive relationship between fairness perceptions and behavioral intentions. By theoretically linking disposition and behavioral intentions, this study adds to both the applicant reactions and CSE literatures and helps provide a better understanding of the impact that personality has in the hiring process.