SEARCH

SEARCH BY CITATION

Abstract

  1. Top of page
  2. Abstract
  3. Realistic Job Previews and Turnover: Possible Mediators
  4. Possible Moderators
  5. Method
  6. Results
  7. Discussion
  8. Conclusion
  9. REFERENCES

A meta-analytic path analysis with k = 52 studies and sample size of roughly 17,000 showed that enhanced perceptions of organizational honesty is the primary mechanism by which realistic job previews (RJPs) influence voluntary turnover. This suggests revisions to RJP theory to incorporate social exchange and the way RJPs lead individuals to feel about the organization. Results assessing several new potential moderators of relationships between RJPs and turnover found that the most effective RJP design may be an oral or written RJP delivered posthire and designed to signal organizational honesty. We discuss several key theoretical and practical implications of the results based on a novel signaling theory perspective on RJPs.

Organizations strive to attract, recruit, and retain qualified and productive employees, as human capital is essential for organizational effectiveness. Successfully accomplishing these goals, however, can be challenging. The recruitment and retention of key employees remain critical issues for organizations (Allen, Bryant, & Vardaman, 2010). Even when voluntary turnover rates drop because of unfavorable labor markets, there is evidence that high unemployment rates have little impact on the turnover of high-performing employees or those with in-demand skill sets (Trevor, 2001). Thus, organizations must take steps to ensure their productive employees remain on the job or risk potentially exorbitant financial costs (Ployhart, Schneider, & Schmitt, 2006).

Although often treated separately in practice and research, recruitment and turnover are to some extent inextricably linked. Recruitment practice and research tends to focus on attracting a sufficient quantity and quality of applicants to the organization and is largely grounded in signaling theory. Signaling theory is based on the idea that job seekers faced with incomplete information use whatever information is available as signals about job and organizational attributes (Rynes, 1991; Spence, 1973). This signaling occurs as organizations attempt to persuasively communicate information across three stages of the recruitment process: attracting applicants, maintaining applicants, and influencing job choice (Allen, Mahto, & Otondo, 2007; Barber, 1998). Turnover practice and research tends to focus on how job, organization, and environmental factors shape quit decisions and is largely grounded in the theory of organizational equilibrium: Individuals are motivated to continue participating as long as the inducements for participation outweigh the expected contributions (March & Simon, 1958). These two processes are linked to the extent that information, signals, and experiences during recruitment influence subsequent evaluations of jobs and organizations, desire to quit, and turnover decisions.

One primary linking mechanism between recruitment and turnover is the realistic job preview (RJP). RJPs are defined as programs, materials, and/or presentations that provide applicants with realistic and balanced (positive and negative) information about a job (Barber, 1998; Breaugh & Starke, 2000; Phillips, 1998). More than 20 years ago, two quantitative reviews reported similar results regarding the relationship between RJPs and turnover. McEvoy and Cascio (1985) reported a corrected correlation between RJPs and turnover of r =−.09 and concluded that given this relatively modest effect, turnover reduction efforts might be better targeted elsewhere. Alternatively, Premack and Wanous (1985) reported a corrected correlation of r =−.06 but concluded that given the relatively modest investment required to implement an RJP, providing RJPs is a useful turnover management strategy. Phillips (1998) updated these prior meta-analytic results and reported corrected correlations of r =−.06 with voluntary turnover and r =−.05 with aggregate turnover (i.e., voluntary, involuntary, and not specified). Thus, RJPs have been consistently associated with reduced turnover, albeit with modest effect sizes.

These prior reviews also presented mixed evidence regarding the presence and importance of potential moderators of the RJP-turnover relationship. McEvoy and Cascio (1985) concluded there was little evidence of significant unexplained variance to warrant a search for moderators. Premack and Wanous (1985), on the other hand, noted that job complexity might moderate the relationship such that the effect is stronger for more complex jobs (r =−.12) than for less complex jobs (r =−.02). Providing yet another perspective related to moderators, Phillips (1998) reported evidence that research setting, RJP timing, and RJP medium moderate the relationship such that the effect is stronger for field rather than lab studies, RJPs given right before hiring rather than very early in recruitment or after hiring, and verbal and written rather than videotaped RJPs. Phillips (1998) concluded that future research in the area should focus on the conditions under which RJPs are likely to be most effective.

In addition to identifying potential moderators, five primary mediating mechanisms (met expectations, role clarity, perceptions of honesty, self-selection, and perceptions of fit that influence attraction to the organization) have been proposed to explain how and why RJPs influence turnover (Breaugh and Starke, 2000). Despite scattered empirical support for each of these mechanisms, research to date has not assessed each mechanism in a single study, thus providing limited evidence concerning the relative importance of each. Learning more about the role of each mechanism has important conceptual implications for understanding how RJPs work and important practical implications for organizations desiring to implement an RJP.

We intend to make several important contributions to the literature. First, although previous research has shown a relationship between RJPs and turnover, there is no consensus on the mechanisms by which this occurs. In this study, meta-analytic path analysis was used to directly test the roles of several proposed mediating mechanisms. To our knowledge, this is the first test of this kind in the area and may provide new insights into how and why RJPs work. Second, although there has been some consideration of the boundary conditions related to the RJP–turnover relationship, there have been repeated calls for more attention to explaining the conditions under which RJPs are more or less likely to influence voluntary turnover. We assessed a variety of moderators including several that have not been included in previous meta-analytic reviews, providing new insights into when and under what conditions RJPs are likely to be most effective. In addition, we updated the most recent RJP meta-analysis (Phillips, 1998) with more than a decade of subsequent research incorporating an additional 30% effect sizes. This update is particularly important because the rapid increase in the availability of job and organization information due to the Internet could make the information signaling function of RJPs less important, yet no research has examined this possibility.

Realistic Job Previews and Turnover: Possible Mediators

  1. Top of page
  2. Abstract
  3. Realistic Job Previews and Turnover: Possible Mediators
  4. Possible Moderators
  5. Method
  6. Results
  7. Discussion
  8. Conclusion
  9. REFERENCES

The goal of an RJP is to present applicants with a balanced and accurate view of a particular job within an organization. As previously mentioned, evidence suggests that RJPs reduce subsequent turnover (both voluntary and aggregate). Thus, we expect to find the use of RJPs to be negatively associated with turnover. In addition to examining this simple relationship, we assess five mediating mechanisms identified from the literature including met expectations, role clarity, perceptions of honesty, self-selection (acceptance), and attraction to the organization. Because self-selection has been most often studied as group-level acceptance rates, and because individuals who self-select out of the process (and thus are not hired) cannot provide data related to subsequent turnover, this mechanism was included in the meta-analytic review but not the mediator analyses. Therefore, we developed hypotheses related to the extent to which four potential mechanisms partially mediate the RJP–voluntary turnover relationship. We propose partial mediation because there are multiple mechanisms and because other mechanisms yet to be identified may also play a role.

Job Acceptance

One of the primary roles of an RJP is to provide applicants with realistic information about the positive and negative aspects of a job and working environment. The realistic information allows applicants to make an informed decision about whether they should continue with the selection process (Bretz & Judge, 1998). This self-selection function influences voluntary turnover because applicants who perceive that they would have difficulty adjusting to the work environment may be more likely to opt out of the process. Realistic information provides applicants with the ability to compare job alternatives, which in theory should allow them to make more informed job acceptance decisions (Breaugh & Starke, 2000). In the absence of an RJP, these same applicants may find themselves hired into an inhospitable or incompatible job or work environment resulting in their desire to leave. Meta-analytic reviews report that RJPs have a negative relationship with acceptance (Phillips, 1998; Premack & Wanous, 1985). Thus, we expected the use of RJPs to be associated with lower acceptance. Job acceptance is typically measured and reported as a group-level index of the percentage or rate at which applicants accept offers of employment. Unfortunately, because only individuals who accept an employment offer can turnover, acceptance rates cannot be used to empirically evaluate the role of acceptance as a mediator of the RJP–turnover relationship.

Expectations

Another primary goal of RJPs is to lower unrealistically optimistic expectations on the part of new hires that could stem from organizations presenting overly positive information during the recruitment process (Breaugh & Starke, 2000). Meeting the expectations of applicants is important; Fedor, Buckley, and Davis (1997) argued that discrepancies between applicant expectations and what they experience on the job may result in negative attributions towards the organization. These negative feelings may lead individuals experiencing discrepancies to voluntarily turnover. Applicant expectations are commonly separated into initial expectations and met expectations. Initial expectations refer to level of applicant expectations after exposure to an RJP but before starting the job. Met expectations represent the accuracy or degree to which applicant expectations prior to starting the job match expectations experienced once on the job. We included both initial and met expectations in this meta-analysis. Results from previous empirical research and reviews (e.g., Horner, Mobley, & Meglino, 1979; Phillips, 1998; Premack & Wanous; 1985; Wanous, 1973) led us to predict that use of RJPs would be associated with lower initial expectations and greater met expectations. Further, expectations (both initial and met) were expected to partially mediate the relationship between RJPs and voluntary turnover.

Perceptions of Honesty

An ancillary consequence of providing an RJP may be that applicants appreciate the willingness of the organization to provide an open and honest assessment of the job and work environment. Organizations providing realistic, and possibly negative, information may be seen as trustworthy and honest. Presumably an organization perceived as being honest during recruitment will foster loyalty and encourage employees to remain on the job, and there is some evidence that organizational dishonesty may be associated with higher turnover intentions and turnover (Cialdini, Petrova, & Goldstein, 2004; Schwepker, 1999). When the realistic information about the job presented during recruitment matches experiences on the job, an employee may feel that the organization has fulfilled part of the psychological contract (Breaugh & Starke, 2000; Rousseau, 1995). Alternatively, a lack of information can be seen by applicants as an indicator of negative organizational characteristics such as a lack of honesty, professionalism, or interest in the candidate (Breaugh & Billings, 1988). Organizations that fail to display honesty may inadvertently promote feelings of mistrust or ambivalence for employees that could result in increased turnover. Therefore, we expected a positive relationship between RJPs and perceptions of organizational honesty and that perceptions of honesty would partially mediate the RJP-turnover relationship because those who view the organization as more honest are less likely to quit (Cialdini et al., 2004).

Role Clarity

Role clarity, the degree to which an individual understands his or her role on the job and in the organization, has been an observed consequence of RJPs for more than 2 decades (Saks & Cronshaw, 1990; Wanous, 1980). By providing an accurate depiction of the job and work environment, RJPs provide greater role clarity. RJPs decrease the chances of role ambiguity occurring on the job, which thereby decreases the negative effects associated with role ambiguity, such as turnover (Meglino, DeNisi, Youngblood, & Williams, 1988). Thus, we expected that RJPs would be associated with greater perceived role clarity and that role clarity would partially mediate the RJP–turnover relationship (Griffeth, Hom, & Gaertner, 2000).

Organizational Attraction

Another key role of an RJP is to provide applicants with a sense of how well they are likely to fit with the job and the work environment. Organizational fit refers to the degree of similarity between an organization and an individual's values and goals and/or the job requirements and an individual's skills and interests. Individuals who find they are not a good fit with their job or organization are more likely to be become unhappy and leave (Breaugh & Starke, 2000). The realistic information presented in RJPs aids applicants in determining how well they may or may not fit with the organization and the previewed position before they accept a job offer. Thus, RJPs may increase or decrease individual perceptions of fit. However, individuals who view an RJP and feel they fit well with the organization should be less likely to leave the organization. Despite its conceptual/theoretical importance, in practice, RJP researchers have not assessed fit, instead typically measuring perceptions of job and organizational attractiveness. Although RJPs could increase or decrease individual perceptions of attraction depending on how the RJP information affects assessments of fit, the more realistic the information being presented, the greater the number of applicants who may find the organization or job unattractive (Saks, Wiesner, & Summers, 1996). This reduced attraction to the organization/job may result from the negative aspects of the realistic information or an overestimation of the negative aspects of the job (Meglino, Ravlin, & DeNisi, 1997). Thus, we expected that attractiveness would partially mediate the relationship between RJPs and turnover, because those who still find the job attractive after receiving the RJP would be more likely to fit with the organization and less likely to quit (Allen et al., 2007).

Possible Moderators

  1. Top of page
  2. Abstract
  3. Realistic Job Previews and Turnover: Possible Mediators
  4. Possible Moderators
  5. Method
  6. Results
  7. Discussion
  8. Conclusion
  9. REFERENCES

RJPs have been used in many contexts and taken a variety of forms. Moderators are important both conceptually and practically because they identify boundary conditions under which RJPs are more or less effective. Previous research has focused on three moderators: study setting (i.e., field or lab), presentation medium, and timing in which the intervention is administered (Phillips, 1998; Premack & Wanous, 1985). Given the diversity of RJP applications, there is value in expanding this set of boundary conditions. Conceptually, previous moderator analysis has focused on two characteristics of the RJP itself (medium and timing) and one methodological consideration (research setting). Although RJP design is certainly important, RJPs are not presented in a vacuum. They are given in particular contexts to particular types of applicants, and it is likely that elements of both may affect responses to an RJP. That is, different types of applicants for different types of jobs may respond differently to the same RJP. Therefore, we considered nine moderators across four conceptual categories derived from results of previous meta-analytic studies, qualitative reviews, and logical arguments (Meglino, Ravlin, & DeNisi, 2000; Phillips, 1998; Premack & Wanous, 1985): RJP characteristics (presentation medium, timing, and duration of exposure); applicant characteristics (experience and education); methodological considerations (study setting and timing of measurement of mediators); and context (industry type and Internet accessibility). Of note, the majority of these moderators (six of nine) have not been reported in prior RJP meta-analyses.

RJP Characteristics

Characteristics of RJPs might be important largely because RJP design affects the information provided and the opportunity for recipients to self-select. For example, exposure may impose constraints on the amount of realistic information applicants can access. Longer exposure time may lead to greater amounts of realistic information being presented to applicants; more information may allow for greater impact on expectations, role clarity, perceptions of honesty, and subsequent turnover. Previous research has already demonstrated that both the medium and the timing of an RJP affect relationships with turnover. Medium affects the nature of information communication (Allen et al., 2007); both verbal and written RJPs have been found to have stronger effects on turnover. Verbal RJPs may provide greater opportunity for interaction, increasing perceptions of credibility and honesty; written communication, although lean, is effective at communicating specific information that may be used to manage expectations and role clarity (Allen et al., 2007). Prehire timing of an RJP has also been found to have stronger effects than posthire timing, perhaps because prehire timing allows recipients to self-select out more readily than does posthire timing. In sum, we proposed that RJPs that are longer, delivered prehire, and communicated in verbal or written form would be more strongly associated with outcomes.

Applicant Characteristics

Characteristics of applicants may also moderate the effects of RJPs because individuals bring different values, experiences, and expectations that may influence how they respond to RJPs. In particular, experience and education may influence a priori job knowledge and alternative job opportunities. For example, applicants with more job experience may already possess extensive knowledge and information about a particular job; thus, RJPs might not be expected to have as large an effect on expectations and role clarity. Applicants who have been exposed to relevant job experiences likely already have clear expectations and role information; thus, an RJP would be expected to have less impact on their self selection and subsequent turnover. There may also be a role for perceived mobility in that individuals who perceive themselves as more marketable, that is those with greater experience and education, may be more likely to self-select out to pursue alternative opportunities. In sum, we proposed that RJPs provided for employees with at least a college education rather than those with a high school education or less, and provided for entry level jobs compared with those requiring experience, will be more strongly associated with outcomes.

Methodological Considerations

Prior research has demonstrated stronger relationships between RJPs and voluntary turnover in field as compared to lab studies. Timing of when key variables are measured also likely plays a role in the effectiveness of RJP interventions. This may be important given the variance across studies in when variables such as attraction, honesty, expectations, and role clarity are measured. We expected RJPs provided posthire to have weaker relationships with outcomes because the employee would have already experienced exposure to job and organizational conditions, thus potentially limiting the need for a realistic preview. In sum, we proposed that RJP studies conducted in field rather than lab settings and those measuring mediator variables prehire will be more strongly associated with outcomes.

Context

The information provided by an RJP may be more or less important in some contexts than others. For example, industries characterized as “white collar” typically require more previous experience, education, and professional credentials than those characterized as “blue collar.” Applicants who have already been exposed to relevant job experiences likely already have clear expectations and role information; thus, an RJP would be expected to have less impact on their self selection and subsequent turnover. Applicants in industries that have fewer experience requirements are likely to be more influenced by the realistic information provided by an RJP. Further, we suggested earlier that widespread availability of job and organization information accessible via the Internet could have reduced the need for, and impact of, realistic previews. Because of this shift of accessibility and use of online information, applicants may be exposed to additional information about an organization and job beyond what is presented in an RJP. Therefore, RJPs may not be as important because of the increased exposure and information available to applicants through online sources. In sum, we proposed that RJPs provided in blue-collar as opposed to white-collar industries and those conducted prior to the widespread accessibility of the Internet would be more strongly associated with outcomes.

Method

  1. Top of page
  2. Abstract
  3. Realistic Job Previews and Turnover: Possible Mediators
  4. Possible Moderators
  5. Method
  6. Results
  7. Discussion
  8. Conclusion
  9. REFERENCES

This review incorporated both published and unpublished research, including journal articles, dissertations, theses, and technical reports. The inclusion of works from sources other than peer-reviewed journals was made to minimize the “file drawer problem.” Works were primarily identified and collected through the investigation of references used by previous reviews, an electronic search of the PsychInfo, PsychArticles, Business Source Premier, and Dissertations Abstracts databases, and searches of electronic conference programs from the Society of Industrial and Organizational Psychology and the Academy of Management. Authors of works not readily available were contacted to request copies of unpublished studies, conference submissions, and technical reports. Initial search results used “RJP” and “realistic job previews” as keywords and resulted in 82 studies conducted between 1956 and 2009. Studies were included in this review only if the study documented the implementation and evaluation of a realistic job preview, contained at least one comparison group, and reported sufficient data to calculate correlation coefficients.

After eliminating studies that did not provide necessary information, the review contained 52 studies: 36 published journal articles, 9 dissertations, 3 theses, and 4 technical reports or other unpublished works. Of these, 18 were included in Premack and Wanous (1985) and 34 were included in Phillips (1998). Thus, this meta-analysis included 18 unique studies. When present, multiple RJP interventions were identified within a given study. As a result, the 52 studies produced 75 unique effects. Identification of multiple RJP interventions within a given study was based on the use of different samples, with different forms of media, and under different experimental procedures. Two independent raters coded each study for RJP and study characteristics. Observed rater agreement was k = .86. All discrepancies were resolved through the review of studies in question and mutual agreement.

Each study in this review was coded for overall turnover, voluntary turnover, the proposed mediation mechanisms (acceptance, attraction, initial and met expectations, perceptions of honesty, and role clarity), and potential moderators (RJP timing, medium, setting, industry, exposure time, experience required by job, education, measurement timing of mediators, and Internet accessibility). Coding was only completed if the variable was explicitly stated in a particular study using the term or terms provided in this review.

Moderators

Moderators included in previous reviews were coded according to a modified scheme developed previously (Phillips, 1998). RJP setting refers to the location in which the RJP took place (laboratory or field study) with medium representing the form of media used to communicate RJP information (written, oral, video, online, audio, other, or combination) and timing identifying when in the selection process applicants were exposed to the RJP (prehire or posthire). RJP exposure is the time applicants had access to the RJP and was coded as lasting less than 1 hour or more than 1 hour. Industry type was based on the nature of the open position and coded as white collar, blue collar, military, education, or healthcare. Experience requirements for the open position were coded as entry level or managerial. Participant education was also coded using the mean participant level of education (high school diploma and equivalent or college level and beyond). Internet accessibility was coded based on the publication date of the study. Because there is no definitive date for when applicants would have had access to online resources, a date was chosen that fell around the time the Internet became relatively accessible and allowed for the retention of a reasonable number of studies in both pre- and post-Internet periods. Netscape Navigator and Internet Explorer were launched in 1994 and 1995, respectively, and approximately 14% of American adults (roughly 24 million) were online in 1995 (Pew Internet, 2003). Therefore, studies were coded as those with publication dates of 1994 and earlier as “pre-Internet” and studies with publication dates of 1995 and later as “post-Internet.” Finally, timing of measures was coded as occurring prehire (before job acceptance), posthire (after job acceptance but before job start), or on the job.

Acceptance

Acceptance refers to whether or not an applicant accepted an offer of employment or continued in the recruitment process. In previous RJP studies, individual choices are aggregated and reported as overall acceptance rates. This variable was either directly reported or calculated using dropout rate information provided within reviewed studies. Coding for this variable only occurred when a study explicitly stated the acceptance rate or rate of attrition from the recruitment process. Applicant attrition and acceptance rates were combined due to the small number of studies that explicitly reported applicant attrition.

Expectations

Applicant expectations refer to the degree to which applicants’ expectations were lower than those of other participants (initial expectations) or the accuracy of pre- and posthire expectations (met expectations). Across measures of expectations, items were generally measured using Likert-type items such as “I am going to be satisfied with this job” from Buckley, Fedor, Veres, Wiese, and Carraher (1998). Effect sizes were only calculated for studies that explicitly reported levels of participant/applicant expectations. Of note, initial expectations could not be included in the meta-analytic path analysis because of an insufficient number of valid cells in the associated correlation matrix.

Role Clarity

Role clarity represents the degree to which an employee's role or tasks are unambiguous. Studies reporting role clarity generally measured this variable using self-report survey items. Sample role clarity items are, “I have a very good idea of what the job entails” and “I am well aware of the duties required of the job” (Saks & Cronshaw, 1990).

Perceptions of Organizational Honesty

Perceptions of organizational honesty refer to applicant's perceptions that the organization is honest in the information it portrays and the level to which it can be trusted. Honesty information was typically collected through self-report surveys with items such as, “The organization was direct and open in dealing with me as a prospective employee” (Saks & Cronshaw, 1990).

Attraction

Attraction refers to an applicant's reported level of attraction to the organization. Typically, applicants expressed perceived organizational attractiveness through self-report measures involving Likert-type items. Coding for this variable occurred when studies explicitly reported the self-reported levels of perceived attraction of participants or applicants.

Overall Turnover

Overall turnover represents the number (or percentage) of individuals who leave an organization for any reason. The time frame for which turnover was reported varied across studies with all reported levels of turnover included within the calculated effect size. All turnover information, including data that were explicitly reported as voluntary turnover, was included within this variable effect size. Studies were coded for turnover if they reported employee turnover, voluntary turnover, or survival rates.

Voluntary Turnover

Voluntary turnover was typically assessed within studies by exit interviews or determined by the authors of the studies. Only information explicitly identified by the authors of a given study as voluntary turnover or “quits” was included. This variable did not include measures of overall turnover or survival rates.

Meta-Analytic Procedures

The Hunter-Schmidt Meta-Analysis Package (Schmidt & Le, 2004) was used to calculate meta-analytic estimates. Once effect sizes were generated, tests of homogeneity were applied to each effect size. We used three common indicators of homogeneity: residual standard deviation, variance accounted for by sampling error, and chi-square test or Q as suggested by Hunter, Schmidt, and Jackson (1982). Tests for moderator effects used a weighted least squares regression approach and were only conducted for effects that showed sufficient heterogeneity.

Path Analysis of Meta-Analytic Correlations

To evaluate the relationships specified in the proposed model (Figure 1), the calculated meta-analytic correlations were subjected to path analysis. The full correlation matrix used for the path analysis was created using studies included in this review, previous meta-analytic reviews, and studies containing correlations of target variables. First, correlations generated from this meta-analysis were used to fill all possible cells. These cells included all those in which RJP was a primary variable. Second, results from published meta-analytic reviews were used to fill as many remaining cells as possible. These included relationships between voluntary turnover and met expectations (Griffeth et al. 2000), attraction (Kristof-Brown, Zimmerman, & Johnson, 2005), and role clarity (Bauer, Bodner, Erdogan, Truxillo, & Tucker, 2007). Third, remaining correlations were identified from studies used in this review. Specifically, the correlation between perceptions of honesty and attraction was taken from Zuliani (1988) and the correlation between perceptions of honesty and role clarity was taken from Saks and Cronshaw (1990). Finally, for the remaining cells, published results from primary studies outside the RJP literature were used to generate meta-analytic estimates of the necessary coefficients. Specifically, the relationship between met expectations and attraction was based on results from two studies (Vigoda, 2000; Vigoda & Cohen, 2002), the relationship between met expectations and role clarity was based on results from five studies (Babakus, Cravens, Johnston, & Moncreif, 1999; Blau, 1988; Grant, Cravens, Low, & Moncreif, 2001; Naumann, Widmier, & Jackson, 2000; Saks, Wiesner, & Summers, 1994), the relationship between perceptions of organizational honesty and voluntary turnover was based on results from two studies (Hemdi & Nasurdin, 2006; Hicks & Klimoski, 1987), and the relationship between attractiveness and role clarity was based on a single study (Harris, Artis, Walters, & Licata, 2006). Sample size was handled by computing the harmonic mean of the respective cells used in the analysis (e.g., Burke & Landis, 2003; Colquitt, LePine, & Noe, 2000; Viswesvaran & Ones, 1995).

image

Figure 1. Maximum-Likelihood Parameter Estimates for the Proposed Multiple Mediation Model.

Note. Statistics are standardized path coefficients (with standard errors in parentheses); N = 339; *p < .05.

Download figure to PowerPoint

Given that the model specifying partial mediation depicted in Figure 1 would lead to a test with zero degrees of freedom, we chose to complete the analyses employing a product-of-coefficients approach for testing models involving multiple mediators (Preacher & Hayes, 2008). This approach involves computation of the indirect effect of each mediator in the presence of all other mediators. The resulting parameters are then used as input to a Sobel test by which the statistical significance of each mediator may be evaluated. All models were tested using EQS 6.1 and produced parameters based on maximum likelihood (ML) estimation techniques.

Results

  1. Top of page
  2. Abstract
  3. Realistic Job Previews and Turnover: Possible Mediators
  4. Possible Moderators
  5. Method
  6. Results
  7. Discussion
  8. Conclusion
  9. REFERENCES

Meta-analytic correlations and confidence intervals are reported in Table 1 along with total sample size and number of included effect sizes. Confidence intervals were used to interpret the strength of the reported effects of RJP on turnover and each mechanism. Use of RJPs was related to lower voluntary turnover (r =−.07), decreased overall turnover (r =−.04), increased acceptance (r = .02), lower levels of attraction (r =−.10), lower initial expectations (r =−.12), lower met expectations (r =−.01), greater perceptions of honesty (r = .11), and higher role clarity (r = .10). Of note, confidence intervals for all relationships did not include zero with the exception of met expectations. These meta-analytic results match closely with results found in previous meta-analytic reviews, with the exception of the relationships between RJPs and acceptance (r =−.03), and RJPs and perceptions of honesty (r = .05) reported by Phillips (1998).

Table 1.  Meta-Analytic Results for the Effects of RJPs
Outcome variablerkNLower confidence intervalUpper confidence intervalQdfp
  1. Note. *indicates that the confidence intervals do not contain zero for the observed effects.

Acceptance.02*246,601.01.0496.8423.01
Attraction−.10*151,585−.16−.0365.9914.01
Initial expectations−.12*162,538−.17−.08135.4215.01
Met expectations−.0192,084−.06.04124.868.01
Perceptions of honesty.11*101,554.05.1722.379.01
Role clarity.10*5871.02.2172.204.01
Turnover−.04*4817,230−.05−.02179.3847.01
Voluntary turnover−.07*154,924−.10−.0424.5914.04

Path Analysis

The correlation matrix and associated sample sizes and number of studies are presented in Table 2. Figure 1 includes maximum-likelihood parameter estimates associated with the test of the multiple mediation model (n = 339, the harmonic mean of the cells in Table 2). For simplicity, the intercorrelations between residuals of the mediator variables are not included in the figure. Given that all mediator covariances are not likely to be solely a function of a single antecedent and such residual covariances are likely substantial, intercorrelations between mediators residuals were fixed to zero following recommendations of Preacher & Hayes (2008).

Table 2.  Meta-Analytic Correlations Between Variables in Path Analyses
VariableRJPMet expectationsPerceptions of organizational honestyAttractivenessRole clarity
  1. Note.a indicates a meta-analytic r derived from this meta-analytic review; b indicates a meta-analytic r derived from a single meta-analytic review; c indicates meta-analytic r derived from studies included in this review; d indicates meta-analytic r derived from studies not included in this review.

Met expectations−.01a    
k, N9,2084    
Perceptions of organizational honesty.11a.01c   
k, N10,15542,272   
Attractiveness−.10a.44d.38c  
k, N15,15852,5941,233  
Role clarity.10a.30d.45c.44d 
k, N5,8715,6851,601,140 
Voluntary turnover−.07a−.15b−.38d−.14b−.11b
k, N15,49248,14862,52210,21574,315

The specific indirect effects observed for the tested model were a1b1= .002 (through met expectations), a2b2= .015 (through role clarity), a3b3=−.052 (through perceptions of honesty), and a4b4=−.009 (through attraction). Results of Sobel tests associated with each mediator revealed a significant effect for only perceptions of honesty (Z =−2.01, p = .04).

Moderator Analyses

Tests of homogeneity identified that effect sizes associated with all outcomes contained sufficient unaccounted variance to allow for moderator investigations. The variance accounted for by moderators for each observed effect is reported in Table 3. Moderators accounted for 76% of the variance observed in acceptance, 53% of the variance in attraction, 60% of the variance in initial expectations, 45% of the variance in perceptions of honesty, 30% of the variance in overall turnover, and 63% of the variance in the observed relationship between RJPs and voluntary turnover. All of the R2s were statistically significant (p < .05). Tables  4 through 9 report results of the moderators on the relationships between use of RJPs and acceptance, attraction, initial expectations, perceptions of honesty, total turnover, and voluntary turnover. An investigation of moderator effects was not conducted for met expectations and role clarity due to the relatively small number of effect sizes.

Table 3.  Variance Accounted for by Moderators at the RJP-Level Derived From Regression Analysis
Variables
ModeratorsAcceptanceAttractionInitial expectationsPerceptions of honestyTurnoverVoluntary turnover
  1. Note. Values represent percentages. *p < .05. Dashes indicate the moderator was not included in the analysis.

Setting14*12*6*1*1*6*
Medium31*19*8*3*10*35*
Timing9*4*1*5*7*
Exposure1*5*1*1*
Job industry8*5*9*8*4*
Experience17*9*1*
Education4*9*1*10*
Internet accessibility18*9*5*17*3*
Timing of measure4*1*5*
R276*53*60*45*30*63*
Table 4.  Meta-Analytic Results of Moderators of RJPs and Acceptance
ModeratorLevel of moderatorrkNLower 95% confidence intervalUpper 95% confidence interval
  1. Note. Only moderators are included for which k ≥ 2 per level.

Setting      
 Field.05124,905.03.09
 Laboratory−.06121,696−.09.01
Medium      
 Oral−.074490−.12.05
 Written.02143,262−.01.06
 Video.1131,628.01.21
 Combined−.0731,221−.15.09
Exposure      
 Less than 1 hour.0752,047.02.15
 More than 1 hour−.142221−.72.73
Job industry      
 White collar−.06183,433−.10−.02
 Blue collar.1121,277−.24.47
 Military.1121,493−.21.44
 Healthcare.012398−.63.65
Education      
 High school.0743,643.03.13
 College−.05131,848−.09.02
Internet      
 1900–1994.03196,362.01.05
 1995–present−.245239−.35−.13
Table 5.  Meta-Analytic Results of Moderators of RJPs and Attraction
ModeratorLevel of moderatorrknLower 95% confidence intervalUpper 95% confidence interval
  1. Note. Only moderators are included for which k ≥ 2 per level.

Setting      
 Field.052305−.69.78
 Laboratory−.12131,280−.19−.07
Medium      
 Written−.199686−.29−.12
 Online−.014624−.14.12
Timing      
 Prehire−.14121,165−.21−.08
 Posthire.082193−.73.83
Internet      
 1900–1994−.185550−.26−.10
 1995–present−.01101,035−.07.04
Timing of measure      
 Prehire−.097457−.17−.01
 Posthire−.373259−.47.25
Table 6.  Meta-Analytic Results of Moderators of RJPs and Initial Expectations
ModeratorLevel of moderatorrknLower 95% confidence intervalUpper 95% confidence interval
  1. Note. Only moderators are included for which k ≥ 2 per level.

Setting      
 Field−.09101,927−.15−.04
 Laboratory−.216611−.34−.12
Medium      
 Oral−.234602−.38−.11
 Written−.0781,509−.13−.01
 Combined−.132237−.97.70
Timing      
 Prehire−.14112,077−.20−.10
 Posthire−.035461−.16.11
Exposure      
 Less than 1 hour−.255561−.37−.13
 More than 1 hour−.122412−.75.51
Job industry      
 White collar−.19111,386−.26−.14
 Blue collar−.023680−.18.15
 Education−.072472−.66.52
Experience      
 Entry level−.07142,438−.11−.04
 Manager.522100−.73.95
Internet      
 1900–1994−.10111,620−.15−.06
 1995–present−.165918−.22−.09
Timing of measure      
 Prehire−.0861,172−.13−.03
 Posthire−.273321−.37.17
 On the job−.126933−.18−.07
Table 7.  Meta-Analytic Results of Moderators of RJPs and Perceptions of Organizational Honesty
ModeratorLevel of moderatorrknLower 95% confidence intervalUpper 95% confidence interval
  1. Note. Only moderators are included for which k ≥ 2 per level.

Setting      
 Field.1261,301.04.18
 Laboratory.104253−.11.30
Medium      
 Written.072261−.75.86
 Video.1131,020−.002.25
 Combined−.042139−.84.74
Timing      
 Prehire.127864.01.17
 Posthire.103690−.06.27
Job industry      
 White collar.184212−.04.41
 Blue collar.102477−.49.69
 Military.073723−.09.23
Experience      
 Entry level.0671,034−.01.14
 Manager.192500−.38.77
Internet      
 1900–1994.1081,300.05.15
 1995–present.152254.03.27
Timing of measure      
 Prehire.115445.02.20
 Posthire.192500−.11.28
 On the job.043609−.04.12
Table 8.  Meta-Analytic Results of Moderators of RJPs and Overall Turnover
ModeratorLevel of moderatorrkNLower 95% confidence intervalUpper 95% confidence interval
  1. Note. Only moderators are included for which k ≥ 2 per level.

Setting      
 Field−.044015,134−.06−.02
 Laboratory−.0282,096−.07.04
Medium      
 Oral−.155578−.27−.04
 Written−.05226,149−.07−.02
 Video.04128,002−.06−.01
 Other−.0261,642−.08.05
 Combined.053859−.10.20
Timing      
 Prehire−.023613,674−.04−.01
 Posthire−.08123,556−.12−.04
Exposure      
 Less than 1 hour−.03152,981−.07.01
 More than 1 hour−.0473,288−.08.01
Job industry      
 White collar−.04255,879−.07−.02
 Blue collar−.156923−.24−.07
 Military−.02109,433−.05.01
 Education.043298−.21.30
 Healthcare−.193308−.44.06
Experience      
 Entry level−.043214,984−.05−.01
 Manager−.163516−.35.03
Education      
 High school−.0398,365−.06−.01
 College−.04111,762−.09.01
Internet      
 1900–1994−.044215,794−.06−.02
 1995–present−.0261,436−.07.03
Table 9.  Meta-Analytic Results of Moderators of RJPs and Voluntary Turnover
ModeratorLevel of moderatorrknLower 95% confidence intervalUpper 95% confidence interval
  1. Note. Only moderators are included for which k ≥ 2 per level.

Setting      
 Field−.07123,828−.09−.02
 Lab−.0531,096−.13.03
Medium      
 Written−.0962,091−.14−.03
 Video−.093846−.24.06
 Combined.042741−.43.51
 Other−.052279−.82.71
Timing      
 Prehire−.06103,804−.09−.02
 Posthire−.1051,120−.18−.02
Exposure      
 Less than 1 hour−.0241,100−.11.08
 More than 1 hour−.063974−.20.07
Job industry      
 White collar−.0682,404−.10−.01
 Blue collar−.023495−.23.16
 Military−.0831,870−.18.02
Education      
 High school−.0721,636−.39.24
 College−.1151,386−.19−.04

On average, RJP characteristics accounted for approximately 25% of the variance explained by moderators across all outcomes. Longer RJP exposure tended to be associated with stronger effects (e.g., voluntary turnover with exposure more than one hour: r =−.06; voluntary turnover with exposure less than 1 hour: r =−.02), with the exception of initial expectations. However, in all cases confidence intervals overlapped. With regard to turnover, posthire RJPs tended to be associated with stronger effects (e.g., voluntary turnover with posthire RJP: r =−.10; voluntary turnover with prehire RJP: r =−.06). The confidence intervals did not overlap with respect to total turnover but did overlap with respect to voluntary turnover. With regard to other outcomes, prehire RJPs tended to be associated with stronger effects; however, in all cases confidence intervals overlapped. When there were enough studies to evaluate, oral RJPs tended to have stronger effects; however, confidence intervals overlapped in all cases.

Across outcomes, applicant characteristics accounted for approximately 10% of the variance explained by moderators. With regard to education, RJPs provided to college-educated samples tended to have stronger effects (e.g., voluntary turnover with college educated samples: r =−.11; voluntary turnover with high school educated samples: r =−.07). With regard to experience, RJPs with managerial samples tended to have stronger effects (e.g., total turnover with managerial samples: r =−.16; total turnover with entry level samples: r =−.04). In all cases for applicant characteristics as moderators, confidence intervals overlapped.

Across outcomes, methodological characteristics accounted for approximately 10% of the variance explained by moderators. With regard to research setting, field studies tended to have larger effects on turnover and honesty (e.g., voluntary turnover with field studies: r =−.07; voluntary turnover with lab studies: r =−.05). For other outcomes, lab studies tended to have stronger effects. Posthire measurement of outcomes tended to be associated with stronger effects (e.g., honesty with posthire measurement: r = .19; honesty with prehire measurement: r = .11). Despite any trends for methodological characteristics as moderators, however, all confidence intervals overlapped.

Across outcomes, context accounted for approximately 15% of the variance explained by moderators. With regard to Internet availability, effects were stronger for more recent studies with respect to acceptance, initial expectations, and honesty (e.g., honesty with more recent studies: r = .15; honesty with older studies: r = .10), but weaker for more recent studies with respect to attraction and overall turnover (e.g., overall turnover with more recent studies: r =−.02; overall turnover with older studies: r =−.04). With respect to industry type, RJP effects were stronger on voluntary turnover in white collar (r =−.06) as opposed to blue collar (r =−.02) contexts and the opposite was true for total turnover (r =−.04 and r =−.15 for white collar and blue collar, respectively) and most other outcomes. Confidence intervals for the context moderators overlapped for all but the effects of industry type on total turnover.

Discussion

  1. Top of page
  2. Abstract
  3. Realistic Job Previews and Turnover: Possible Mediators
  4. Possible Moderators
  5. Method
  6. Results
  7. Discussion
  8. Conclusion
  9. REFERENCES

This research makes a number of important contributions to the RJP literature, including using meta-analytically derived estimates in a path analysis to assess the relative impact of key theoretical mechanisms intended to explain the RJP–voluntary turnover relationship; updating previous meta-analyses of the RJP–turnover relationship with the inclusion of additional studies published since the last review; and expanding the search for moderators that further understanding of the conditions under which RJPs are more or less effective.

Review of Results

Overall, meta-analytic results for the RJP–turnover relationship (corrected r =−.07 for voluntary turnover) were quite consistent with previous reviews (i.e., r =−.09, McEvoy & Cascio, 1985; r =−.06, Premack & Wanous, 1985; r =−.06, Phillips, 1998), suggesting that RJPs have remained a low investment strategy for modestly influencing turnover despite evolving workplaces and employment systems and the growth in readily available information online. Meta-analytic results also showed that RJPs are related to each of the theoretical mediating mechanisms proposed in the literature; however, RJPs are most strongly related to perceptions of organizational honesty (r = .11), role clarity (r = .10), and attraction (r =−.10). These results generally follow those reported in previous reviews (Phillips, 1998; Premack & Wanous, 1985). When compared to the most recent review by Phillips (1998), all relationship directions were the same with the exception of acceptance, which was observed as r = .02 in this review and by Phillips as r =−.03. Observed relationship strengths were comparable across all other outcomes. Specifically, Phillips reported negative relationships between RJPs and initial expectations (r =−.18), met expectations (r =−.02), turnover (r =−.05), and voluntary turnover (r =−.06), and a positive relationship between RJPs and climate for honesty (r =−.05). Results from Phillips (1998) reveal a stronger negative relationship with initial expectations (r =−.18) than this study (r =−.12). Conversely, a stronger relationship was observed between perceptions of honesty and RJPs for this study (r = .11) than was observed by Philips (r = .05).

Results of the path analysis reveal that honesty was the only significant mediating mechanism between RJPs and voluntary turnover, and that this effect fully mediated this relationship. These findings suggest that perceptions of honesty play a far more important role in understanding the effects of RJPs than previously realized, as evidenced by the relatively few studies investigating this mechanism relative to most of the other proposed mediators. Previous conceptual work has suggested that the effectiveness of RJPs is based primarily on managing expectations, followed closely by variations of applicants self-selecting out of the process. These results suggest reevaluating these conceptualizations to more fully consider how RJPs influence individuals’ perceptions about the hiring organization. For example, extensive research on perceived organizational support (POS) shows that perceptions that an organization cares about and supports employees are related to important outcomes such as turnover (e.g., Allen, Shore, & Griffeth, 2003). It may be that perceptions of honesty engendered by RJPs are one way organizations signal employees about such caring and support. Similarly, research on psychological contract breach (Rousseau, 1995) suggests that honesty during recruitment would lead new employees to be less likely to see their psychological contract as being breached. Future research is needed on the role of these types of social exchanges in understanding RJPs.

Our findings also provide boundary conditions advancing understanding of when RJPs are more or less effective at influencing turnover, although overlapping confidence intervals suggest caution in interpreting these results. Consistent with prior research, effects on voluntary and total turnover were stronger in field setting than lab settings; it may be that lab settings are limited in the ability to recreate conditions similar enough to the complexities of quitting a job. Also consistent with prior research, oral RJPs were more effective than other presentation modes at influencing total turnover (Phillips, 1998), although there were not sufficient studies to assess the effects of oral presentation on voluntary turnover. Oral presentation modes may be effective because, according to research on media richness theory (Daft & Lengel, 1984), face-to-face communication is the richest form. Contrary to some prior research, our findings suggest that posthire RJPs were somewhat more effective than prehire ones at influencing voluntary and total turnover. This finding is consistent with the idea that honesty is a key mechanism linking RJPs and turnover; prehire RJPs would be more likely to affect self-selection mechanisms; perceptions of honesty could likely be influenced at any time. Consistent with prior research, longer RJPs were slightly more effective at influencing voluntary and involuntary turnover than shorter RJPs. We did not find strong support for the idea that greater Internet accessibility would make more recent RJPs less effective at influencing total turnover (not enough recent studies assessed voluntary turnover); thus, it appears that even with the greater job and organization information available through the Internet, RJPs still provide useful information to applicants.

Theoretical Contributions

Taken together, these findings suggest that a signaling-based theory of RJPs may be a particularly fruitful avenue for future research that offers the opportunity to extend understanding of the effects of RJPs beyond the traditional focus on met expectations and self-selection. Signaling theory (Spence, 1973) focuses on the communication of information to influence decision making under conditions of information asymmetry and suggests that signaling is particularly important when there are information asymmetries about quality and intent (Connelly, Certo, Ireland, & Reutzel, 2011). In the context of RJPs, job seekers have imperfect information about unobservable job and work environment attributes, such as job characteristics and organizational culture. During the recruitment process, organizations and their agents send numerous signals about these attributes. Because signals are most important when incentives for veracity are weak (Spence, 1973), RJPs may provide particularly key information. We think it is likely that job seekers may reasonably question the complete veracity of publically available information and recruitment communication because organizations have a vested interest in creating and maintaining a positive public image.

We argue that RJPs serve a signaling function in two primary ways: instrumental and symbolic (Highhouse, Thornbury, & Little, 2007). Instrumentally, the information included in an RJP provides signals about previously unobserved work and environmental characteristics that reduce uncertainty. This instrumental function aids job seekers in managing expectations, mobilizing coping resources, or self-selecting, and has been represented by the bulk of RJP research to date. Symbolically, the act of providing an RJP also signals something about unobservable organizational characteristics such as organizational honesty, organizational support, and care for employees. Unfortunately, this symbolic function has received less attention. This result suggests that the symbolic function could be as important, or even more important, than the instrumental function and should be the subject of future research.

This signaling perspective on RJPs suggests other interesting opportunities for future RJP research. For example, research is needed on precisely what different recruitment messages (such as RJPs) signal to job seekers in terms of unobservable organizational attributes such as honesty, support, caring, or social exchange. A signaling perspective may present a new way to view research on RJP delivery as well. Prior research on delivery method has tended to focus on media characteristics, but perhaps the delivery method itself also serves a signaling function. That is, orally presented RJPs may signal positive organizational attributes to job seekers as compared to written RJPs. Signaling may also provide a novel perspective on RJP timing. An expectations management perspective implies that RJPs must be given prehire; however, our results suggest this may not be the case. If signaling organizational honesty is a key mechanism underlying RJPs, then timing may not be critical. A signaling perspective may also suggest that RJPs given more than once may be more effective as the frequency and consistency of signals increases signaling effectiveness (Connelly et al., 2011). Perhaps providing multiple RJPs (e.g., both before and after hiring) would result in a stronger signaling effect.

Signaling theory also suggests that some receivers interpret signals differently than others (Connelly et al., 2011; Spence, 1973). Although previous RJP research has examined the moderating role of a few individual differences (such as work experience), a signaling perspective suggests a far more nuanced investigation into what types of individuals respond in what manner to which types of signals under which circumstances. A more fine-grained analysis of these boundary conditions may uncover populations or circumstances under which RJPs result in more substantial effect sizes than those that have typically been reported. When evaluating a job and organization, applicants make probability assessments about unobservable attributes based on observable signals. In part, these probability assessments are based on previous experiences, explaining why previous job market experiences may affect responses to RJPs. It may also be the case that previous exposure to RJPs in other job settings adjusts these probability assessments. The extent to which previous recruitment messages and RJPs have displayed what Connelly et al. (2011) call signal fit (i.e., the fit between signals and underlying attributes) may affect responses to an RJP and may be a promising avenue for future research.

One issue that previous research has largely ignored is the factors leading organizations to offer RJPs. There are costs associated with manipulating signals (e.g., providing an RJP). For example, Spence (1973) suggested there might be monetary costs, time investments, and physical costs associated with signaling; however, signaling costs may be a prerequisite to obtaining the most preferred applicants and to fitting the right applicants to the right jobs. Signaling theory suggests that frequency of being in the recruiting market will be positively related to investment in acquiring signaling reputations. Research is needed to assess whether frequency of recruiting is, in fact, related to offering an RJP. Further, organizations with prominent reputations may be able to invest less in signaling because job seekers will perceive that they have more information about quality. Thus, research is needed on possible differences in the marginal signaling costs for different organizations. A final interesting proposition associated with signaling theory is that if any particular signal is extremely effective, all organizations will invest in it, and hence it will cease to serve a useful signaling function. Thus, the utility of RJPs may depend on the extent of adoption in a particular industry, job type, or geographic region.

Finally, the application of signaling theory to RJPs represents an extension to research and thinking about signaling. Connelly et al. (2011) note that signaling research focuses mainly on the intentional communication of positive qualities and attributes. However, in the context of RJPs, signaling takes the form of intentionally communicating negative qualities and attributes, but with the expectation that communicating this negative information may provide positive instrumental and symbolic signals about the organization. Researchers may want to consider whether there are other contexts in which signaling theory might help explain why organizations voluntarily and intentionally communicate negative information and the results of doing so (e.g., voluntarily reporting an ethical violation or being transparent about poor financial performance).

Practical Implications

The results also provide practical implications for organizations considering using RJPs to manage turnover. First, RJPs are modestly related to voluntary and total turnover, despite the proliferation of widely available job and organizational information on the Internet. We agree with Premack and Wanous’ (1985) perspective that, despite the modest effect sizes, the low investment required to implement an RJP makes them a cost-effective tool for managing turnover. Second, RJPs are also related to increased role clarity, increased perceptions of organizational honesty, and reduced initial expectations that make it more likely the job and organization will meet individuals’ expectations. Role clarity and expectations are important variables related to a variety of positive outcomes in their own right; thus, organizations can use RJPs to manage these as well. The key role of honesty in this process also suggests that organizations should investigate other mechanisms for signaling honesty to applicants and employees. Third, RJPs reduce perceptions of organizational attractiveness. In situations where successfully recruiting a sufficient number of qualified applicants is more challenging than retaining employees, RJPs may not be an appropriate strategy. Finally, our signaling perspective suggests that individuals might respond differently to RJP signals. Similarly to research demonstrating the impact of customizing recruitment information in general (e.g., Dineen & Noe, 2009), RJPs that are customized to particular job seekers may result in stronger effects.

In terms of implementing RJPs, both oral and written presentation modes are effective at reducing turnover. RJPs presented posthire do not produce significantly different effects than those presented prehire. This is particularly important because it enables organizations to benefit from the turnover-reducing effects of an RJP while mitigating the reduction in attraction associated with prehire RJPs. In sum, the most effective RJP design appears to be an oral or written RJP delivered posthire and designed to signal organizational honesty. As noted earlier, a further recommendation is for organizations to consider offering multiple RJPs given that such an approach might increase signal strength and consistency.

Limitations

A limitation of any meta-analysis is the reliance on existing primary studies that may not provide complete or sufficient data. In this study, some of the reported results were based on relatively few effect sizes, and the interpretation of such results must be made more cautiously, particularly in cases with divergent findings. The limitations associated with few studies should provide a road map to future researchers interested in clarifying the boundary conditions under which RJPs are more or less effective. In a related vein, we were also forced to omit acceptance rates from the mediating mechanism analyses because we could find no studies linking acceptance rates to subsequent turnover; thus, we could not provide an empirically based estimate for path modeling. This is an important limitation because self-selection is a key theoretical mechanism potentially linking RJPs with subsequent turnover. Future research that incorporates acceptance rates and turnover associated with RJPs is warranted, as this will aid in evaluating the relative importance of self-selection vis-à-vis the key honesty mechanism uncovered in this study.

A similar limitation is that lack of meta-analytic evidence for some relationships in the correlation matrix used to estimate path analyses forced us to use estimates drawn from prior empirical studies, in some cases only a small number of studies. Perhaps this result can serve as a foundation upon which to design research that directly focuses on these particular variables and relationships. We were also limited in our assessment of fit as a mediating mechanism because, although fit has been proposed as a theoretical mechanism, in practice the empirical evidence has relied exclusively on assessments of attraction to the organization. Future research that explicitly incorporates the role of fit perceptions is warranted. It may also be the case that more extensive meditational chains explain the impact of RJPs on turnover. For example, it is plausible that RJPs could influence attractiveness and honesty, which in turn affect acceptance rates and ultimately turnover. Future research that is designed to assess longer causal chains would be valuable.

Conclusion

  1. Top of page
  2. Abstract
  3. Realistic Job Previews and Turnover: Possible Mediators
  4. Possible Moderators
  5. Method
  6. Results
  7. Discussion
  8. Conclusion
  9. REFERENCES

We sought to investigate the mechanisms underlying the relationship between RJPs and turnover, extend previous meta-analytic findings on RJPs, and investigate previously unexamined moderators. Through the use of meta-analytic path analysis procedures, we observed that only perceptions of organizational honesty appear to mediate the relationship when all mechanisms are included concurrently. This finding indicates the potential importance of the information and message presented by the organization in RJPs and the possible benefits of adopting a signaling theory perspective for directing future RJP research.

REFERENCES

  1. Top of page
  2. Abstract
  3. Realistic Job Previews and Turnover: Possible Mediators
  4. Possible Moderators
  5. Method
  6. Results
  7. Discussion
  8. Conclusion
  9. REFERENCES
  • References marked with one asterisk indicate studies included in the meta-analysis and marked with two asterisks indicate studies used to complete the meta-analytic correlation matrix used in path analysis.
  • Allen DG, Bryant PC, Vardaman JM. (2010). Retaining talent: Replacing misconceptions with evidence-based strategies. The Academy of Management Perspectives, 24, 4864.
  • Allen DG, Mahto R, Otondo RF. (2007). Web-based recruitment: Effects of information, organizational brand, and attitudes toward a web site on applicant attraction. Journal of Applied Psychology, 92, 16961708.
  • Allen DG, Shore LM, Griffeth RW. (2003). The role of perceived organizational support and supportive human practices in the turnover process. Journal of Management, 29, 99118.
  • *Avner BK, Guastello SJ, Aderman M. (1982). The effect of a realistic job preview on expectancy and voluntary versus involuntary turnover. Journal of Psychology, 111, 101107.
  • **Babakus E, Cravens DW, Johnston M, Moncrief WC. (1999). The role of emotional exhaustion in sales force attitude and behavior relationships. Journal of the Academy of Marketing Science, 27, 5870.
  • Barber AE. (1998). Recruiting employees: Individual and organizational perspectives. Thousand Oaks , CA : Sage.
  • *Barber AE, Hollenbeck JR, Tower SL, Phillips JM. (1994). The effects of interview focus on recruitment effectiveness: A field experiment. Journal of Applied Psychology, 79, 886896.
  • **Bauer TN, Bodner T, Erdogan B, Truxillo DM, Tucker JS. (2007). Newcomer adjustment during organizational socialization: A meta-analytic review of antecedents, outcomes, and methods. Journal of Applied Psychology, 92, 707721.
  • **Blau G. (1988). Testing the generalizability of a career commitment measure and its impact on employee turnover. Journal of Vocational Behavior, 35, 88103.
  • *Bourgeois N. (2002). Factors influencing the effects of realistic job previews on applicant judgments of organizational attractiveness. Unpublished master's thesis. Texas Christian University , TX .
  • Breaugh JA, Billings RS. (1988). The realistic job preview: Five key elements and their importance for research and practice. Journal of Business and Psychology, 2, 291305.
  • Breaugh JA, Starke M. (2000). Research on employee recruitment: So many studies, so many remaining questions. Journal of Management, 26(3), 405434.
  • Bretz RD, Judge TA. (1998). Realistic job previews: A test of the adverse self-selection hypothesis. Journal of Applied Psychology, 83, 330337.
  • *Buckley MR, Fedor DB, Veres JG, Wiese DS, Carraher SM. (1998). Investigating newcomer expectations and job-related outcomes. Journal of Applied Psychology, 83(3), 452461.
  • *Buckley MR, Mobbs TA, Mendoza JL, Novicevic MM, Carraher SM, Beu DS. (2002). Implementing realistic job previews and expectation lowering procedures: A field experiment. Journal of Vocational Behavior, 61(2), 263278.
  • *Buda R, Charnov BH. (2003). Message processing in realistic recruitment practices. Journal of Managerial Issues, 3, 32326.
  • Burke MJ, Landis RS. (2003). Methodological and conceptual challenges in conducting and interpreting meta-analyses. In Murphy KR (Ed.), Validity generalization: A critical review (pp. 287309). Mahwah , NJ : Erlbaum.
  • Cialdini RB, Petrova PK, Goldstein N. (2004). The hidden costs of organizational dishonesty. Sloan Management Review, 45, 6773.
  • *Colarelli SM. (1984). Methods of communication and mediating processes in realistic job previews. Journal of Applied Psychology, 69, 6168.
  • *Colarelli SM, Stumpf SS. (1990). Compatibility and conflict among outcomes of organizational entry strategies: Mechanistic and social systems perspectives. Behavioral Science, 35, 110.
  • Colquitt JA, LePine JA, Noe RA. (2000). Toward an integrative theory of training motivation: A meta-analytic path analysis of 20 years of research. Journal of Applied Psychology, 85, 678707.
  • Connelly BL, Certo ST, Ireland RD, Reutzel CR. (2011). Signaling theory: A review and assessment. Journal of Management, 37, 3967.
  • *Cronin BE. (forthcoming). The impact of realistic job information and information salience on applicant attraction in web-based recruitment. Unpublished work.
  • Daft RL, Lengel RH. (1984). Information richness: A new approach to manager information processing and organizational design. In Straw B, Cummings LL (Eds.), Research in Organizational Behavior. Greenwich , CT : JAI Press.
  • *Dean RA, Wanous JP. (1984). Effects of realistic job previews on hiring bank tellers. Journal of Applied Psychology, 69, 6168.
  • *Dilla BL. (1987). Descriptive versus prescriptive information in a realistic job preview. Journal of Vocational Review, 30, 3348.
  • Dineen B, Noe R. (2009). Effects of customization on application decisions and applicant pool characteristics in Web-based recruitment context. Journal of Applied Psychology, 94, 224234.
  • *Dugoni BL, Ilgen DR. (1981). Realistic job previews and the adjustment of new employees. Academy of Management Journal, 24, 579591.
  • *Erffmeyer RC, Erffmeyer ES. (1983). An application of a realistic job preview in an instructional setting. Psychological Reports, 53, 939944.
  • *Farr JL, O’Leary BS, Bartlett CJ. (1973). Effect of work sample test upon self-selection and turnover of job applicants. Journal of Applied Psychology, 58, 283285.
  • Fedor DB, Buckley MR, Davis WD. (1997). A model of the effects of realistic job previews. International Journal of Management, 14, 211221.
  • **Fisher CD, Gitelson R. (1983). A meta-analysis of the correlates of role conflict and ambiguity. Journal of Applied Psychology, 68, 320333.
  • *Ganzach Y, Pazy A, Ohayun Y, Brainin E. (2002). Social exchange and organizational commitment: Decision-making training for job choice as an alternative to the realistic job preview. Personnel Psychology, 55, 613637.
  • *Gaugler BB, Thornton GC, III. (1990). Matching job previews to individual applicants’ needs. Psychological Reports, 66, 643652.
  • **Grant K, Cravens DW, Low GS, Moncrief WC. (2001). The role of satisfaction with territory design on the motivation, attitudes, and work outcomes of salespeople. Academy of Marketing Science Journal, 29, 165178.
  • **Griffeth RW, Hom PW, Gaertner S. (2000). A meta-analysis of antecedents and correlates of employee turnover: Update, moderator tests, and research implications for the next millennium. Journal of Management, 26, 463488.
  • *Haccoun RR. (1978). The effects of realistic job previews and their position within the selection sequence on telephone operator behavior and attitudes. Unpublished manuscript, Behavioral Science Group, Bell Telephone of Canada, Montreal .
  • **Harris EG, Artis AB, Walters JH, Licata JW. (2006). Role stressors, service worker job resourcefulness, and job outcomes: An empirical analysis. Journal of Business Research, 59, 407415.
  • *Hedley HH. (1985). The relationship of job preview to absenteeism, turnover, and job satisfaction of public school teachers. Unpublished doctoral dissertation, College of William and Mary, Williamsburg , VA .
  • **Hemdi MA, Nasurdin AM. (2006). Predicting turnover intentions of hotel employees: The influence of employee development human resource management practices and trust in organization. Gadjah Mada International Journal of Business, 8, 2142.
  • *Hicks WD, Klimoski RJ. (1987). Entry into training programs and its effects on training outcomes: A field experiment. Academy of Management Journal, 30, 542552.
  • Highhouse S, Thornbury EE, Little IS. (2007). Social-identity functions of attraction to organizations. Organizational Behavior and Human Decision Processes, 103, 134146.
  • *Hom PW, Griffeth RW, Palich LE, Bracker JS. (1998). An exploratory investigation into theoretical mechanisms underlying realistic job previews. Personnel Psychology, 51, 412451.
  • *Horner SO, Mobley WH, Meglino BM. (1979). An experimental evaluation of the effects of a realistic job preview on Marine recruit affect, intentions and behavior. Unpublished technical report, University of South Carolina , Columbia .
  • Hunter JE, Schmidt FL, Jackson GB. (1982). Meta-analysis. Cumulating research findings across studies. Beverly Hills , CA : Sage.
  • *Ilgen DR, Seeley W. (1974). Realistic expectations as an aid reducing involuntary resignations. Journal of Applied Psychology, 59, 452455.
  • *Kraus E. (1992). The moderating effects of realism of course expectations on motivation, satisfaction, performance, and attrition (realistic course preview). Unpublished doctoral dissertation, Hofstra University .
  • *Krausz M, Fox S. (1981). Two dimensions of a realistic preview and their impact upon initial expectations, expectation fulfillment, satisfaction and intentions to quit. Journal of Occupational Behavior, 2, 211216.
  • **Kristof-Brown AL, Zimmerman RD, Johnson EC. (2005). Consequences of individuals’ fit at work: A meta-analysis or person-job, person-organization, person-group, and person-supervisor fit. Personnel Psychology, 58, 281342.
  • *Lent RW, Nota L, Soresi S, Ferrari L. (2007). Realistic major previews in the school-to-college- transition of Italian high school students. The Career Development Quarterly, 56, 183191.
  • *Lockman RF. (1980). The realistic job preview experiments at Great Lakes and San Diego. Paper no. 80–1885, Center for Naval Analysis, Alexandria , VA .
  • *Macedonia RM. (1969). Expectations, press, and survival. Unpublished doctoral dissertation, New York University.
  • March JG, Simon HA. (1958). Organizations. New York, NY : John Wiley.
  • *Martin AD. (1996). Exploring the upper limits of realistic job preview effectiveness for performance of new employees. Unpublished doctoral dissertation, Wayne State University.
  • McEvoy GM, Cascio WF. (1985). Strategies for reducing employee turnover: A meta-analysis. Journal of Applied Psychology, 70, 342353.
  • *Meglino BM, DeNisi AS, Ravlin EC. (1993). Effects of previous job exposure and subsequent job status on the functioning of a realistic job preview. Personnel Psychology, 46, 803822.
  • *Meglino BM, DeNisi AS, Youngblood SA, Williams KJ. (1988). Effects of realistic job previews: A comparison using an enhancement and a reduction preview. Journal of Applied Psychology, 73, 259266.
  • Meglino BM, Ravlin EC, DeNisi AS. (1997). When does it hurt to tell the truth? The effects of realistic job reviews on employee recruiting. Public Personnel Management, 26, 414421.
  • Meglino BM, Ravlin EC, DeNisi AS. (2000). A meta-analytic examination of realistic job preview effectiveness: A test of three counterintuitive propositions. Human Resource Management Review, 10(4), 407433.
  • *Miceli MP. (1985). The effects of realistic job previews on newcomer behavior: A laboratory study. Journal of Vocational Behavior, 26, 277289.
  • **Naumann E, Widmier SM, Jackson DW. (2000). Examining the relationship between work attitudes and propensity to leave among expatriate salespeople. Journal of Personal Selling and Sales Management, 20, 227241.
  • *Parmerlee M. (1982). A laboratory investigation of realistic job previews. Unpublished doctoral dissertation, Indiana University , Bloomington .
  • Pew Internet. (2003, December 22). Web use and communication activities (America's online pursuits). Retrieved from http://www.pewinternet.org/Reports/2003/Americas-Online-Pursuits/4-Web-Use-and-Communication-Activities/1-Close-to-twothirds-of-Americans-now-go-online-to-access-the-Internet.aspx.
  • Phillips JM. (1998). Effects of realistic job previews on multiple organizational outcomes: A meta-analysis. Academy of Management Journal, 41, 673690.
  • Ployhart RE, Schneider B, Schmitt N. (2006). Staffing organizations: Contemporary practice and theory (3rd ed). Mahwah , NJ : Erlbaum.
  • Preacher KJ, Hayes AF. (2008). Asymptotic and resampling strategies for assessing and comparing indirect effects in multiple mediator models. Behavior Research Methods, 40, 879891.
  • Premack SL, Wanous JP. (1985). A meta-analysis of realistic job preview experiments. Journal of Applied Psychology, 70(4), 706719.
  • *Raphael MA. (1975). Work previews can reduce turnover and improve performance. Personnel Journal, 54, 9798.
  • *Reilly RR, Tenopyr ML, Sperling SM. (1979). Effects of job previews on job acceptance and survival of telephone operator candidates. Journal of Applied Psychology, 64, 218220.
  • *Reilly RR, Brown B, Blood MR, Malatesta CZ. (1981). The effects of realistic previews: A study and discussion of the literature. Personnel Psychology, 34, 823834.
  • *Rosenburg MM. (1991). The effects of individual expectations and realistic job previews on organizational commitment. Unpublished doctoral dissertation, Claremont Graduate School , Claremont , CA .
  • Rousseau DM. (1995). Psychological contracts in organizations: Understanding written and unwritten agreements. Thousand Oaks , CA : Sage.
  • Rynes SL. (1991). Recruitment, job choice, and post-hire consequences: a call for new research directions. In Dunnette MD, Hough LM(Eds.), Handbook of industrial and organizational psychology (2nd ed, pp. 399444). Palo Alto , CA : Consulting Psychologists Press, Inc.
  • *Saks AM. (1989). An examination of the combined effects of realistic job previews, job attractiveness and recruiter affect on job acceptance decisions. Applied Psychology: An International Review, 38, 145163.
  • *Saks AM, Cronshaw SF. (1990). A process investigation of realistic job previews: Mediating variables and channels of communication. Journal of Organizational Behavior, 11, 221236.
  • *Saks AM, Wiesner WH, Summers RJ. (1994). Effects of job previews on self-selection and job choice. Journal of Vocational Behavior, 44, 297316.
  • *Saks AM, Wiesner WH, Summers RJ. (1996). Effects of job previews and compensation policy on applicant attraction and job choice. Journal of Vocational Behavior, 49, 6885.
  • Schmidt FL, Le H. (2004). Software for the Hunter-Schmidt meta-analysis methods (p. 42242). Iowa City , IA : Department of Management Organizations, University of Iowa.
  • Schwepker CH. (1999). An exploratory investigation of the relationship between ethical conflict and salesperson performance. Journal of Business and Industrial Marketing, 18, 435446.
  • Spence AM. (1973). Job market signaling. Quarterly Journal of Economics, 87, 355374.
  • *Suszko MK, Breaugh JA. (1986). The effects of realistic job previews on applicant self-selection and employee turnover, satisfaction, and coping ability. Journal of Management Review, 9, 116.
  • *Thornsteinson TJ, Palmer EM, Wilff C, Anderson A. (2004). Too good to be true? Using realism to enhance applicant attraction. Journal of Business and Psychology, 19, 125137.
  • *Travagline AM. (2001). Online recruiting: Implementing Internet-based realistic job previews. Unpublished doctoral dissertation, Temple University .
  • Trevor CO. (2001). Interactions among actual ease-of-movement determinants and job satisfaction in the prediction of voluntary turnover. Academy of Management Journal, 44, 621639.
  • **Vigoda E. (2000). Internal politics in public administration systems: An empirical examination of its relationship with job congruence, organizational citizenship behavior, and in-role performance. Public Personnel Management, 29, 185210.
  • **Vigoda E, Cohen A. (2002). Influence tactics and perceptions of organizational politics: A longitudinal study. Journal of Business Research, 55, 311324.
  • Viswesvaran C, Ones DS. (1995). Theory testing: Combining psychometric meta-analysis and structural equations modeling. Personnel Psychology, 48, 865885.
  • *Wanous JP. (1973). Effects of a realistic job preview on job acceptance, job attitudes, and job survival. Journal of Applied Psychology, 58, 327332.
  • Wanous JP. (1980). Organizational entry: Recruitment, selection and socialization of newcomers. Reading MA : Addison-Wesley.
  • **Wanous JP, Poland TD, Premack SL, Davis KS. (1992). The effects of met expectations on newcomer attitudes and behaviors: A review and meta-analysis. Journal of Applied Psychology, 77, 288297.
  • *Weitz J. (1956). Job expectancy and survival. Journal of Applied Psychology, 40, 245247.
  • *Wiesner WH, Saks AM, Summers RJ. (1991). Job alternatives and job choice. Journal of Vocational Behavior, 38, 198207.
  • *Youngberg CF. (1963). An experimental study of ``job satisfaction'' and turnover in relation to job expectations and self-expectations. Unpublished doctoral dissertation, New York University.
  • *Zaharia ES, Baumeister AA. (1981). Job preview effects during the critical initial employment period. Journal of Applied Psychology, 66, 1922.
  • *Zuliani RA. (1988). The realistic organizational preview and the joining decision process. Unpublished doctoral dissertation, University of Minnesota, Minneapolis .