Academy of Management Journal 2009, Vol. 52, No. 3, 527–544.

CHANGE IN NEWCOMERS’ SUPERVISOR SUPPORT AND SOCIALIZATION OUTCOMES AFTER ORGANIZATIONAL ENTRY
MARKKU JOKISAARI Finnish Institute of Occupational Health JARI-ERIK NURMI University of Jyväskylä
Using a four-wave longitudinal research design and a latent growth modeling approach, we modeled change in newcomers’ perceived supervisor support and socialization outcomes (role clarity, work mastery, job satisfaction, and salary). Further, the role of perceived supervisor support in socialization outcomes was examined. The results showed that, on average, newcomers’ perceived supervisor support declined during the period 6–21 months after organizational entry. The results showed further that the steeper the decline in perceived supervisor support, the greater the rate of decrease in role clarity and job satisfaction, and the slower the increase in salary over time.

Organizational socialization is an important process for both newcomers and organizations. How newcomers “learn the ropes” and assimilate to an organization during this socialization process presumably has long-lasting effects on their job attitudes and behavior (e.g., Schein, 1978; Wanous, 1992). A pivotal assumption in the organizational socialization literature is that interaction between newcomers and organizational insiders, such as supervisors, plays an important role in newcomers’ socialization and related adjustment to work (e.g., Graen, 1976; Louis, 1990; Reichers, 1987). Despite this assumption, few previous studies have examined how supervisor support for newcomers changes after organizational entry and the role that such change in supervisor support plays in socialization and related adjustment to work. Consequently, the purpose of the present study was to examine how newcomers’ perceived supervisor support changes during the 6 –21 months following organizational entry, and how change in perceived supervisor support contributes to change in socialization outcomes (role clarity, work mastery, job satisfaction, salary).

NEWCOMERS’ ORGANIZATIONAL SOCIALIZATION Organizational socialization is asserted to be a process by which newcomers’ adjustment to work and assimilation into an organization change as a consequence of their gaining the new knowledge and learning the behavioral patterns expected from a member of the particular organization (e.g., Feldman, 1981; Schein, 1978). Scholars of organizational socialization have identified tasks and challenges that are critical to newcomers’ success in their transition into an organization (Chao, O’Leary-Kelly, Wolf, Klein, & Gardner, 1994; Feldman, 1981; Morrison, 1993; Ostroff & Kozlowski, 1992; Saks & Ashforth, 1997; Schein, 1978). Specifically, one critical task for newcomers is to acquire an understanding of the responsibilities and goals of their jobs. This learning is then reflected in role clarity (Feldman, 1981; Graen, 1976; Morrison, 1993; Schein, 1978; Wanous, 1992). The role theory literature emphasizes that it is of high importance that newcomers attain role clarity in their jobs, because many organizational outcomes, such as job performance, depend on to what extent employees know what is expected from them in their work roles (e.g., Graen, 1976). For instance, low role clarity is likely to be related to low job performance since it indicates lack of knowledge about job goals and the behaviors required to accomplish them (e.g., Tubre & Collins, 2000). Another critical task for newcomers is learning the skills and routines required to perform their jobs. Such learning is reflected in employees’ work and task mastery
527

We thank Associate Editor Kenneth Law and four anonymous reviewers for helpful comments and suggestions on drafts of this article. This study was supported by the Finnish Cultural Foundation, Jenny and Antti Wihuri Foundation, and Academy of Finland (#207421; #124294). The bulk of this study was done while Markku Jokisaari was at the University of Jyvaskyla. ¨ ¨

Copyright of the Academy of Management, all rights reserved. Contents may not be copied, emailed, posted to a listserv, or otherwise transmitted without the copyright holder’s express written permission. Users may print, download or email articles for individual use only.

528

Academy of Management Journal

June

(Feldman, 1981; Morrison, 1993; Schein, 1978). When newcomers have confidence in their mastery of a job, they are likely to make more effort and show perseverance in their tasks and in their cooperation with their coworkers (Feldman, 1981). The reason for this is, according to the social cognitive theory, that perceived mastery enhances newcomers’ beliefs that their own action produces effects (e.g., Bandura, 2001). Previous research has also shown that work mastery is related to job performance (Bauer & Green, 1998) and innovative cooperation (Feldman, 1976). Research on socialization literature has proposed not just that newcomer learning of important tasks has a central role, but also that evaluations and signals of success are important in socialization (Fisher, 1986; Graen, 1976; Schein, 1978; Wanous, 1992). First, communication of organizational acceptance in a form such as salary growth is important feedback to a newcomer on how she or he is valued in an organization (e.g., Fisher, 1986; Schein, 1978). Along similar lines, organizational support theory suggests that organizational feedback, such as receiving rewards, is important for how newcomers perceive and attach to their organization (e.g., Rhodes & Eisenberger, 2002). In other words, organizational feedback and rewards presumably play a key role in how newcomers are assimilated to organizations (Fisher, 1986; Graen, 1976; Schein, 1978). Previous research has also shown that organizational rewards, such as salary growth, are related to the turnover (job leaving) process (e.g., Trevor, Gerhart, & Boudreau 1997). For instance, low performance-reward contingency increases the likelihood of turnover cognitions and thus the risk of actual turnover (Griffeth, Hom & Gaertner, 2000). In line with these suggestions, earlier research has shown that salary increase is an important factor for retaining high performers in particular (Trevor et al., 1997). Second, in the socialization literature it has been argued that how newcomers affectively adapt to their current jobs, which is reflected in newcomers’ job satisfaction, is an important indicator of success in socialization (Feldman, 1981; Reichers, 1987; Saks & Ashforth, 1997; Wanous, 1992). Indeed, as Harrison, Newman, and Roth succinctly put it, “A general, positive, job attitude leads individuals to contribute rather than withhold desirable inputs from their work roles” (2006: 320). The theory of reasoned action (Fishbein & Ajzen, 1975) also suggests that employees whose attitudes to their jobs are positive will show enhanced motivation and behavioral “intentions to perform,” which then presumably facilitates their job performance. Furthermore, employees who are dissatisfied with their jobs can be

assumed to be more likely to show withdrawal cognitions and behaviors, such as intentions to quit and search for a new job, which then increase the likelihood of turnover (e.g., Hanisch & Hulin, 1991). Earlier research has also shown that job satisfaction is related to job performance and job turnover (see reviews by Griffeth, Hom, and Gaertner [2000] and Judge, Thoresen, Bono, and Patton [2001]). Following this line of argument, the present study focuses on newcomer learning—that is, newcomers’ achieving role clarity and work mastery— and how success in socialization is signaled and evaluated in terms of salary growth and job satisfaction as socialization outcomes. We further reasoned that newcomers’ socialization and related adjustment to work extend from the first year in an organization to the end of the second year, which continues to be an important time for learning key features of a job and achieving success in socialization (e.g., Chao et al., 1994; Schein, 1978). Previous studies have also shown that evaluations of the period during which newcomers have learned their jobs or have achieved integration into their organizations span the first two years after entry (Lee & Allen, 1982; Moore, 1974). Many critical events, such as change in work assignments and supervisor-subordinate conflict, are also more likely to occur during the second year in an organization (Gundry & Rousseau, 1994; see also Kammeyer-Mueller, Wanberg, Glomb, & Ahlberg, 2005). Furthermore, newcomers’ expectations of their employers’ obligations, such as an increase in salary, tend to rise toward the end of the second year after organizational entry (Robinson, Kraatz, & Rousseau, 1994). However, previous research in the field (e.g., Bauer & Green, 1998; Chan & Schmitt, 2000; Chen & Klimoski, 2003; Kammeyer-Mueller & Wanberg, 2003; Lance, Vandenberg, & Self, 2000; Major et al., 1995; Morrison, 1993, 2002; Ostroff & Kozlowski, 1992) has mainly focused on the first year of socialization. Consequently, the present study focuses on socialization into an organization toward the end of the second year after organizational entry. Finally, many scholars in the field (Graen, 1976; Jablin, 2001; Louis, 1990; Reichers, 1987) have argued that the interaction between newcomers and organizational insiders, such as supervisors, is a main channel for newcomer socialization. For example, Graen’s (1976) interpersonal role-making model suggests that supervisors play an important role in how newcomers’ socialization proceeds. The reason for this is that supervisors are an important source of feedback contributing to newcomer’s job learning, and they have formal authority in newcomers’ role negotiations and an opportunity

2009

Jokisaari and Nurmi

529

to influence organizational feedback to newcomers, such as rewards (Graen, 1976). However, although the interaction between newcomers and organizational insiders is argued to be one of the main channels for newcomer socialization, little research has been conducted on how and to what extent the role of supervisors in newcomer socialization changes after organizational entry. In consequence, researchers in the field also have limited knowledge about to what extent change in the contribution of supervisors to newcomers’ socialization is related to the rate of newcomer adjustment to work. Accordingly, the aim of the present study was to extend the interactionist perspective on newcomers’ socialization (Graen, 1976; Jablin, 2001; Louis, 1990; Reichers, 1987) by proposing that the contribution of supervisors to the socialization of newcomers is a process in which change in supervisor support over time affects newcomers’ adjustment at work. We base our theoretical arguments for change in supervisor support on theories of relationship development (Fichman & Levinthal, 1991; Gabarro, 1987) and the “liability of newness” argument in relationship development (Burt, 2000, 2002), which are outlined below. Specifically, theories of relationship development and the liability of newness argument offer a perspective from which to delineate the interpersonal and organizational processes through which supervisors’ contribution to newcomers’ socialization changes over time. In addition, we propose that newcomers’ adjustment to work and attainment of organizational acceptance over time is related to changes in the support they receive from their supervisors. Change in Supervisor Support after Organizational Entry The theory of relationship development suggests that, after organizational entry, a “honeymoon period” protects the initial relationship between a newcomer and organizational insiders, such as supervisors, from conflict and negative outcomes (Fichman & Levinthal, 1991). On the one side, recent graduates may enter an organization with goodwill and high expectations after years of investment in education. On the other side, organizational insiders may also show goodwill to newcomers. For example, they may not expect newcomers to show as much proficiency as they expect from oldtimers (e.g., Jablin, 2001). These initial assets, such as goodwill, commitment, and investments, contribute to the development of a relationship by creating the so-called honeymoon period in which the relationship is shielded from setbacks and conflicts (Fichman & Levinthal, 1991).

The literature on relationship development shows no agreement about how long the honeymoon effect typically lasts after organizational entry, beyond agreement that that this duration may depend on what kind of initial assets the actors have (Bruderl, 2000; Fichman & Levinthal, 1991). ¨ However, research on job turnover suggests that the honeymoon period lasts about three to six months (Bruderl, 2000; Farber, 1994). The rate of newcom¨ ers’ job turnover is initially low after organizational entry, reaching a peak after three to six months, and declining thereafter. In other words, the longer the honeymoon period, the later the peak in job turnover (Fichman & Levinthal, 1991). Overall, theory and empirical findings suggest that, on average, newcomers’ perceived supervisor support is also likely to be high after the first six months in an organization. However, previous work on the development of social networks and working relationships suggests that, after the honeymoon period and initiation of the relationship between newcomer and supervisor, perceived supervisor support typically declines over time. It has been argued, first, that there is a liability of newness in the development of social relations over time (Burt, 2000, 2002). This means that relationships prompted by exogenous factors, such as work relations, tend to weaken with time and that the decay is more likely among new relationships than old ones. Furthermore, it is argued that relationships that span work roles and differentials in status, such as superior-subordinate ones, decay more rapidly than relations between people similar in role and status (Burt, 2002; McPherson, Smith-Lovin, & Cook, 2001). One reason for this greater rapidity is that interaction between people with differences in social characteristics such as status and role requires more effort than interaction between people whose social characteristics are similar (e.g., McPherson et al., 2001). For example, there is typically an asymmetry in power between supervisor and newcomer, and this may produce social distance and constrain the development of trust and ease of communication (e.g., Gabarro, 1987). Second, the literature on the development of working relationships suggests that in the case of the supervisor-subordinate relationships, this process takes a long time. For example, the first 6 to 12 months in such a relationship constitute a period of testing, during which the relationship is based on contract rather than liking and is affected by organizational constraints (Boyd & Taylor, 1998; Gabarro, 1987). For example, the ability of supervisors to provide high-level support over time might be limited by organizational constraints such as high workload and large numbers

530

Academy of Management Journal

June

of subordinates (Allen, Poteet, & Burroughs, 1997; Green, Anderson, & Shivers, 1996). Furthermore, being a highly supportive supervisor over an extended period of time may be costly in terms of the supervisor’s own time and performance (Graen & Uhl-Bien, 1995; Sparrowe & Liden, 1997) as well as reputation, if others see it as favoritism (Allen et al., 1997). Finally, research on the liability of newness in relationship development suggests that relationships tend to weaken during their first two years (Burt, 2000, 2002). For example, Burt (2002) reported a tapering off in the decay of work relations in an organization after the first two years. Thus, we hypothesized: Hypothesis 1. Newcomers’ perceived supervisor support declines during the 6 to 21 months after organizational entry. Role of Supervisor Support in Socialization Outcomes The interactionist perspective on organizational socialization suggests that supervisor-newcomer relationships are important channels through which the organizational socialization of newcomers is negotiated and implemented (Graen, 1976; Jablin, 2001; Louis, 1990; Reichers, 1987). In fact, in Graen’s interpersonal role-making model supervisors play a critical role in the assimilation of newcomers to organizations: “The supervisor legitimately can mediate various organizational outcomes for the focal person” (Graen, 1976: 1209). Earlier research has also provided evidence for the importance of supervisors in the socialization process and its outcomes (Bauer & Green, 1998; Louis, Posner, & Powell, 1983; Ostroff & Kozlowski, 1992; Major, Kozlowski, Chao, & Gardner, 1995). It has been suggested previously, that, among organizational insiders, it is supervisors in particular who make the major contribution to newcomers’ role learning (e.g., Graen, 1976; Schein, 1978). Supervisors have also formal authority in newcomers’ role negotiations (Graen, 1976) and therefore have an opportunity to influence newcomers’ work assignments and goals. In addition, supervisors play a central role in providing knowledge and feedback through which role expectations are learned (e.g., Ostroff & Kozlowski, 1992). In these ways supervisors are able to contribute to newcomers’ role clarity in their work (Bauer & Green, 1998; Morrison, 2002). The organizational socialization literature (Katz, 1980; Schein, 1978; Wanous, 1992) also emphasizes that a critical aspect in the socialization of a newcomer is the shift from initial, and often provisional, work assignments to new work

assignments during the second half of the first year in an organization and during the second year. For newcomers these new assignments typically bring new challenges and create new role responsibilities (Katz, 1980; Schein, 1978; Wanous, 1992). Moreover, as Fisher noted: “It [a job] must change, since the role behavior expected of newcomers is not the same as that expected of experienced insiders. Newcomers may be given simple tasks first and not held to the same time or quality standards as experienced incumbents. Gradually, more is expected” (1986: 120). As a consequence, continuous supervisor support is often needed during the socialization process to clarify and give feedback in relation to these new role expectations. Thus, it seems reasonable to assume that a decline in supervisor support to newcomers will also decrease newcomers’ knowledge of what is expected from them in their work and roles. Hypothesis 2. The greater the decrease in perceived supervisor support, the greater the decrease in role clarity during the 6 to 21 months after organizational entry. A further central task for newcomers after organizational entry is to achieve the skills and knowledge required to perform their jobs (Feldman, 1981; Morrison, 1993; Schein, 1978; Wanous, 1992). It is argued that this learning of skills and knowledge is reflected in work and task mastery (e.g., Feldman, 1981). We reasoned that supervisor support and feedback over time should play an important part in the development of work mastery among newcomers during the 6 –21 months after organizational entry, since supervisors are an important source of information and feedback pertaining to how newcomers perform their jobs (Bauer & Green, 1998; Morrison, 1993; Ostroff & Kozlowski, 1992). Furthermore, as noted above, newcomers often face new work assignments and expectations after their first provisional assignments, and these new work assignments are often the first challenging tasks that allow newcomers to show their proficiency and obtain feedback on their performance in their organization (Katz, 1980; Schein, 1978; Wanous, 1992). As Schein stated, “It [a first regular assignment] still left the new employee after six to twelve months with a feeling of being untested and untried. . . . It was only after additional months in a department doing ‘meaningful’ work that they began to get data about their own abilities and acceptability” (1978: 116). Furthermore, it has been reported that the expectations that supervisors and coworkers have of the performance of newcomers are related to newcomers’ work experience (Chen & Klimoski, 2003). Consequently, as newcomers face

2009

Jokisaari and Nurmi

531

new job responsibilities and new expectations regarding their performance over time, they often need ongoing support and feedback from supervisors in order to develop mastery of their work. Consequently, we hypothesized that: Hypothesis 3. The greater the decrease in perceived supervisor support, the greater the decrease in work mastery during the 6 to 21 months after organizational entry. The organizational socialization literature suggests that an increase in salary is an important signal of organizational acceptance for a newcomer (Fisher, 1986; Graen, 1976; Schein, 1978). Furthermore, the organizational reward system is also a mechanism by which newcomers are assimilated to an organization (Fisher, 1986; Graen, 1976). Reward inequity has often been reported as a negative critical incident among newcomers after organizational entry (Gundry & Rousseau, 1994). A supervisor can influence a newcomer’s salary progress in different ways. The supervisor often has positional power that directly contributes to the negotiations over the newcomer’s salary progress, particularly in a performance-based pay system (e.g., Ferris & Judge, 1991; Graen, 1976). The supervisor is also often involved in the evaluation of how the newcomer is performing in the job, and these evaluations may also contribute to salary progress. Earlier research has also underlined the influential position of organizational insiders in the outcome of newcomers’ salary negotiations (Seidel, Polzer, & Stewart, 2000). Conversely, a decline in supervisor support over time may also mean that the supervisor is contributing less to evaluations of progress in the newcomer’s socialization, thereby hindering subsequent progress in remuneration. Thus, we hypothesized that: Hypothesis 4. The greater the decrease in perceived supervisor support, the lower the growth in salary during the 6 to 21 months after organizational entry. A further important indicator of progress in the socialization process is job satisfaction (Feldman, 1981; Katz, 1980; Reichers, 1987; Wanous, 1992). Again, supervisor support should play a role in how newcomers’ job satisfaction changes over time. As Graen’s (1976) interpersonal role-making model suggests, a supervisor typically has the positional power to channel resources, tasks, and opportunities in the workplace. These aspects can be assumed to contribute to newcomers’ job enrichment and progress in socialization, and, consequently, they can be assumed to lead to changes in job satisfaction over time. For example, Chen and

Klimoski (2003) found that good interaction between organizational insiders and newcomers was related to the empowerment of newcomers in workplaces. Previous research has also shown that supervisor behavior is related to job satisfaction in newcomers (Bauer & Green, 1998; Major et al., 1995). In this study, we reasoned that a decrease in supervisor support may lead to a decrease in job enrichment and progress in socialization over time, and consequently to a decrease in newcomers’ job satisfaction. Hypothesis 5. The greater the decrease in perceived supervisor support, the greater the decrease in job satisfaction during the 6 to 21 months after organizational entry. METHODS Participants and Procedures The participants were graduates from four polytechnic schools in Finland. The participants represented three different occupational domains: technology (occupations such as software designer, telecommunications engineer, system manager), business and management (e.g., marketing manager, marketing assistant, sales manager), and health care (e.g., nurse, midwife, physiotherapist). Recruitment was arranged in cooperation with the schools’ personnel, who informed students about the study. From the original list of names provided by the school administrations, 422 students (80%) agreed to participate in the study. These students were first contacted during the last term of their last school year. They were asked to fill in a questionnaire including questions related to social ties and well-being. Thirteen students returned questionnaires that were not appropriately completed, and consequently they were excluded from the study. The remaining 409 participants were included in the study. In the present study we focused on four time points: (1) Half a year after their graduation (time 1), we asked the participants about their life situations—that is, whether they were currently employed, unemployed, studying, on maternity/ paternity leave, or in some other situation. Participants were then instructed to fill in those parts of the questionnaire that were in accordance with their life situation. For example, only those participants who were employed were asked to fill in measurements related to supervisor relations, work adjustment, and salary. This time the questionnaire was mailed to

532

Academy of Management Journal

June

the participants and returned by mail as well. A follow-up note was sent by e-mail or mobile phone message two weeks later encouraging participation. We also sent a further round of questionnaires to nonrespondents. This procedure was also used at all the remaining time points of the study. Of the 343 participants who returned their questionnaires (a response rate of 84 percent), 273 were employed. (2) A year after graduation (time 2), the participants were again asked to fill in a questionnaire following instructions analogous to those at time 1. Of the 321 participants who returned their questionnaires (response rate 78%), 272 were at work. (3) One and a half years after graduation (time 3), the participants were again asked to fill in a questionnaire following instructions analogous to those at times 1 and 2; however, the item on salary was excluded (see below). Of the 300 participants who returned their questionnaires (response rate 73%), 245 were working. (4) Two years after graduation (time 4) the participants were again asked to fill in a questionnaire following instructions analogous to those used on the first two occasions. Of the 292 participants who returned their questionnaires (response rate 71%), 236 were at work. The number of participants who responded and were employed at times 1, 2, 3, and 4 ranged from 273 to 236. Forty-eight participants who reported at time 2 or at time 4 that they had changed to a new employer during the past 12 months were excluded. Also excluded were 5 participants who indicated at time 1 that they had already been in the same job with the same organization for some years. With respect to the 137 respondents who had complete and usable responses for all four data collection waves, and who had continued with the same employer during the study period, the average age was 26 years (s.d. 6), 58 percent were women, and median job tenure was 6 months (s.d. 4) at time 1 and 21 months (s.d. 9) at time 4. With respect to occupational domain, 49 percent of the respondents represented health care; 14 percent, business and management; and 37 percent, technology. Measures Appendix A presents the items of the scales used as measures in our study, and Table 1 gives the related reliability coefficients. Perceived supervisor support. Supervisor support was measured by a three-item scale (Feij,

Whitely, Peiro, & Taris, 1995). We dropped one item (“To what extent has your supervisor made changes in assignments to improve your skills and knowledge?”) from the original supervisor support scale as it overlapped with another item on the scale. Adjustment to work. Adjustment to work was measured in terms of job satisfaction (Cammann, Fichman, Jenkins, Jenkins, & Klesh, 1983), role clarity (Dallner et al., 2000), and work mastery (Dallner et al., 2000). One item was eliminated from the work mastery scale (“Are you content with your ability to maintain a good relationship with your co-workers at work?”) as in the factor analysis reported by Dallner and others (2000) this item showed very low communality, and in addition it focuses not on work tasks but on interaction with coworkers. A confirmatory factor analysis conducted for supervisor support and the work adjustment scales at time 1 indicated an acceptable level of fit for the four-factor model; that is, distinct factors emerged for supervisor support and each of the work adjustment scales ( 2[48] 66.57; CFI .96; TLI .95; RMSEA .06; SRMR .06). Salary. At times 1, 2, and 4, we asked the participants to indicate their monthly salary. Since measurement at three points is typically the minimum needed to estimate latent growth curves (e.g., Bollen & Curran, 2006), we decided to examine salary only at three time points. Control variables. Social integration with coworkers was assessed by three items (“I feel comfortable around my coworkers,” “My coworkers seem to accept me as one of them,” and “I get along with the people I work with very well”) from Morrison (1993); reliability (coefficient alpha) ranged from .82 to .89. Occupational domain was categorized by two dummy-coded variables: technology (1 “yes,” 0 “other”) and management (1 “yes,” 0 “other”). Work experience was measured by asking the participants to indicate in years and months how much work experience they had in their current occupational domain. Job tenure was measured in months by asking the participants to indicate how long they had worked in their current jobs. Latent Growth Modeling Latent growth modeling (also known as latent growth curve analysis) provides a tool for examining the role of change in the level of a focal variable over time and whether there are interindividual differences in this change (see, e.g., Bollen & Curran, 2006; Chan & Schmitt, 2000). In latent growth

2009

Jokisaari and Nurmi

533

FIGURE 1 Exemplary Path Diagram of a Multivariate Multiple-Indicator Latent Growth Model for Perceived Supervisor Support and Socialization Outcomea

a

Covariances between same-item residuals were omitted for clarity.

modeling (LGM) within a structural equation modeling (SEM) framework, the observed variables measured at the different time points are represented as factors indicating intraindividual change and interindividual differences in intraindividual change over time. For example, in a two-factor LGM, the intercept factor typically indicates the initial level of a particular variable at the given time point, and the slope factor indicates the growth rate of that variable over time. LGM in an SEM framework also enables variables observed at different time points to be modeled as latent variables, a procedure described as multiple-indicator or second-order LGM. This procedure allows measurement invariance to be estimated over time and permits correlated residual variables for same-item measures over time. In LGM, both the mean structure (initial level and average growth in the sample) and the covariance structure (interindividual differences in initial level and in growth rate) can be estimated. For example, the mean of the slope factor indicates

whether there is significant change in the focal variable over time and whether the focal variable shows an increasing or decreasing rate of growth over time. With respect to our study, we were interested in whether perceived supervisor support declined over time; that is, we investigated whether the mean of the slope factor, shown as s in Figure 1, was negative and significant. The variance estimate of the slope factor, 2s, would indicate whether there was individual variation in change over time. In such a case, it is meaningful to investigate possible covariates of change. These estimates of the mean and variance of the slope factor are key modeling results in LGM. In addition, in LGM the slope and intercept factors are allowed to covary to indicate to what extent change in a variable of interest is related to its initial status. Finally, LGM also allows examination of the covariates of change. With respect to our study, for example, Figure 1 shows a multiple-indicator latent growth model for both perceived supervisor support and the socialization outcome variables

534

Academy of Management Journal

June

and the paths between the growth components. By using such a multivariate latent growth model, one can examine whether changes (slope factors) in focal variables over time are related to each other.1 For example, in the present study, Hypotheses 2–5 concerned to what extent the slope factors between supervisor support and the focal socialization outcome would be related to each other as shown in Figure 1. Analyses The analyses were carried out in three successive phases: (1) the longitudinal measurement models and measurement invariance of the perceived supervisor support and work adjustment variables at the four measurement points were tested first. (2) The growth components (linear/nonlinear) of intraindividual change over time for the perceived supervisor support and socialization outcomes variables were then modeled by means of LGM. In these analyses, we used multiple-indicator latent growth models (second-order LGM) to model the growth components related to the perceived supervisor support and work adjustment variables. (3) Finally, we examined the associations between the growth factors for perceived supervisor support and socialization outcomes by using multivariate LGM. Measurement invariance. We performed a series of model comparisons to evaluate measurement invariance separately for perceived supervisor support and each work adjustment variable. The meaningful interpretation of growth trajectories requires the assumption of measurement invariance over time (e.g., Bollen & Curran, 2006; Lance et al., 2000). The results indicated that the assumption of measurement invariance was met for the perceived supervisor support and all work adjustment variables. That is, the same single factor among items held over time and both factor loadings and intercepts for the same items remained the same over time points. The assumption that both factor loadings and intercepts for the same indicator remain the same over time was met for the perceived supervisor support and all work adjustment variables. Modeling growth rates. To investigate the form of growth trajectories related to the perceived supervisor support and socialization outcomes, we estimated linear and nonlinear models. First, we

constructed a model with two growth factor components—that is, the intercept growth factor and the linear growth rate factor (slope)— estimating a simple linear growth trajectory over four measurement points (linear model). The model was constructed by defining the intercept factor as the initial status of supervisor support and socialization variables by setting the loadings of variables from time 1 to time 4 to 1 on the intercept factor and to 0, 1, 2, 3 on the slope factor. Because salary was measured at three time points, the loadings of variable were 0, 1, 3 on the slope factor.2 Second, we examined the general assumption that the error variances for repeated latent variables were homoskedastic over time (e.g., Lance et al., 2000). The assumption of homoskedastic error variances was tenable in our models. All of the latent growth models and supplementary analyses reported in this study were tested by using a maximum likelihood estimation for models with partial missing data (e.g., Schafer & Graham, 2002). In this procedure, observations are sorted into missing patterns, and parameters are estimated using all the available data (n 201 for perceived supervisor support; n 204 for work mastery and job satisfaction; n 205 for role clarity; n 197 for salary). All models reported here were performed with the Mplus statistical program, version 3 (Muthen & ´ Muthen, 2005). Model fit was assessed with chi´ square ( 2, df), the comparative fit index (CFI), Tucker-Lewis Index (TLI), root-mean-square error of approximation (RMSEA), and standardized rootmean-square residual (SRMR) (see Hu & Bentler, 1999). RESULTS Table 1 shows descriptive statistics and correlations for the study variables. For example, the observed means for perceived supervisor support, role clarity, and job satisfaction decreased, whereas the observed means for salary increased over time. The effect size estimates ( 2) for these mean changes were .10 for perceived supervisor support,
We examined nonlinear growth rates by using unspecified (freed-loading) models (e.g., Bollen & Curran, 2006). The unspecified model was constructed by relaxing the constraints on the slope factor by estimating the loadings on the slope factor. The results of the nonlinear models are shown in Table 2. On the basis of model comparisons and parsimony, the linear models were selected as the final models describing individual change in perceived supervisor support and socialization outcomes.
2

The term “multivariate latent growth model” is typically used when a model combines two or more univariate LGMs (e.g., Bollen & Curran, 2006).

1

2009

Jokisaari and Nurmi

535

TABLE 1 Correlations, Descriptive Statistics, and Reliability Coefficients of the Major Study Variablesa
Variables Mean s.d. 1 .83 .60 .82 .51 .61 .83 .50 .59 .67 .83 .40 .29 .29 .22 .29 .26 .20 .14 .18 .21 .22 .26 .29 .33 .28 .26 .30 .28 .27 .18 .38 .26 .23 .17 .16 .23 .04 .18 .13 .10 .22 .20 .38 .17 .20 .30 .31 .30 .18 .14 .17 .15 .08 .19 .13 .28 .32 .34 .44 .28 .35 .33 .42 .10 .09 .08 .15 .17 .24 .29 .78 .60 .55 .48 .30 .18 .14 .10 .02 .02 .05 .32 .27 .25 .27 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19

1. Perceived supervisor 3.16 1.04 support, time 1 2. Perceived supervisor 3.06 0.98 support, time 2 3. Perceived supervisor 2.94 0.88 support, time 3 4. Perceived supervisor 2.85 0.91 support, time 4 5. Job satisfaction, time 1 4.21 0.61 6. Job satisfaction, time 2 4.09 0.66 7. Job satisfaction, time 3 4.02 0.72 8. Job satisfaction, time 4 3.89 0.85 9. Role clarity, time 1 4.21 0.63 10. Role clarity, time 2 4.20 0.63 11. Role clarity, time 3 4.11 0.66 12. Role clarity, time 4 4.05 0.70 13. Salary, time 1 1,709 416 14. Salary, time 2 1,828 460 15. Salary, time 4 2,001 513 16. Work mastery, time 1 3.94 0.47 17. Work mastery, time 2 3.98 0.44 18. Work mastery, time 3 3.89 0.52 19. Work mastery, time 4 3.88 0.52 a .81 .63 .55 .37 .31 .18 .21 .15 .17 .20 .28 .37 .27 .25

.80 .56 .31 .27 .31 .24 .14 .16 .21 .21 .21 .32 .22

.85 .34 .32 .29 .41 .11 .12 .16 .19 .18 .23 .29

.80 .50 .59 .53 .02 .06 .03 .32 .19 .31 .25

.80 .67 .65 .07 .06 .18 .22 .29 .31 .22

.79 .73 .02 .02 .05 .19 .22 .28 .16

.82 .01 .05 .05 .19 .17 .22 .20

.78 .71 .09 .02 .01 .01

.80 .11 .02 .02 .02

.09 .10 .10 .08

.63 .55 .64 .53 .54 .69 .42 .42 .60 .74

Correlations greater than .17 are significant at p

.05. Reliability coefficients (alphas) are shown boldface along the diagonal.

.09 (role clarity), .14 (job satisfaction), and .31 for salary. Growth Rates of Perceived Supervisor Support and Socialization Outcomes To examine mean level changes and interindividual variation in the growth components of the perceived supervisor support and the socialization outcomes variables, we carried out separate univariate LGM analyses for perceived supervisor support, job satisfaction, role clarity, work mastery, and salary. In these analyses, measurement equivalence and homoskedastic error structure constraints were used for different measurement points in the multiple-indicator LGM models. Furthermore, covariances between sameitem residuals were set to be equal over time in the multiple-indicator LGM models (Bollen & Curran, 2006). First, models containing the growth components for initial level and change (i.e., the intercept and slope factors) were constructed for perceived supervisor support. The fit indexes suggested that the linear model (model 1) fitted the data on perceived supervisor support adequately, as shown in Table 2. The growth factors explained a substantial proportion of the variability in perceived supervisor support at each time point (R2s .64 –.70). As

shown in Table 3, the estimated rate of growth in perceived supervisor support decreased per halfyear interval, and the decrease was statistically significant (t 3.12), supporting Hypothesis 1. In addition, the results showed interindividual variation in both the initial level of and linear change in perceived supervisor support, as shown in Table 3. The relationship between initial status and growth rate was statistically significant; that is, the higher the level of initial perceived supervisor support, the more perceived supervisor support declined over time. The linear models (model 1) showed good fit with the data on role clarity, work mastery, job satisfaction, and salary, as shown in Table 2. However, in testing the model on salary, we found negative residual variance related to salary at time 4. Because this residual variance was not statistically significant, it was set at zero (Chen et al., 2001). The growth factors explained a substantial proportion of the variability in role clarity (R2s .79 –.81), work mastery (R2s .79 –.82), and job satisfaction (R2s .58 –.68) at each time point. The growth factors explained also a substantial proportion of variability in salary at time 1 and time 2 (R2s .69 –.85; time 4 residual variance was set to 0). As shown in Table 3, the estimated growth rates in role clarity and job satisfaction decreased per

536

Academy of Management Journal

June

TABLE 2 Tests of Alternative Latent Growth Model Specificationsa
Variable Perceived supervisor support Model 1 (linear) Model 2 (nonlinear) Model 1 vs. model 2 Job satisfaction Model 1 (linear) Model 2 (nonlinear) Model 1 vs. model 2 Salary Model 1 (linear) Model 2 (nonlinear) Model 1 vs. model 2 Role clarity Model 1 (linear) Model 2 (nonlinear) Model 1 vs. model 2 Work mastery Model 1 (linear) Model 2 (nonlinear)b Model 1 vs. model 2 df
2

CFI

TLI

RMSEA

SRMR

2

df

65 63

88.35 83.35

0.98 0.98

0.98 0.98

.04 .04

.06 .06 5.00 2

65 63

70.12 64.26

0.99 0.99

0.99 0.99

.02 .01

.08 .07 5.86 2

2 1

1.01 0.09

1.00 1.00

1.00 1.00

.00 .00

.02 .01 0.98 1

65 63

79.16 78.27

0.99 0.98

0.99 0.98

.03 .03

.08 .07 0.89 2

65

70.27

0.99

.99

.02

.11

a n 201 for perceived supervisor support; n 204 for job satisfaction; n mastery. b Model failed to converge to a proper solution.

197 for salary; n

205 for role clarity; n

204 for work

half-year interval, and the decreases were statistically significant (t’s 2.12, 4.25). However, the mean growth rate in work mastery did not show a statistically significant change over time (t 1.02). The mean growth rate in salary showed a statistically significant increase over time (t 8.33). In addition, the results indicated interindividual variation in both the initial level of and change in role clarity, work mastery, job satisfaction (p .10), and salary. The relationships between initial statuses and growth rates were not statistically significant.

Change in Perceived Supervisor Support and Socialization Outcomes We conducted multivariate latent growth model analyses to examine Hypotheses 2, 3, 4, and 5—that is, to find out what extent the growth factors for perceived supervisor support and socialization outcomes were associated. Each of these models consisted of a pair comprising perceived supervisor support and one of the socialization outcome variables, as shown in Figure 1. In these analyses, occupational domain, work experience, and tenure

TABLE 3 Parameter Estimates of Latent Growth Models for Perceived Supervisor Support, Salary, Job Satisfaction, Role Clarity, and Work Mastery
Perceived Supervisor Support 3.45* 0.47* 0.07* 0.04* 0.06* Job Satisfaction 3.92* 0.25* 0.09* 0.02 0.00 Role Clarity 4.04* 0.30* 0.03* 0.01* 0.01 Work Mastery 3.99* 0.17* 0.01 0.01* 0.02

Parameter Mean initial status Variance in initial status Mean growth rate Variance in growth rate Covariance (initial status with growth rate) *p .05 or greater.

Salary 1,633.50* 1,364.20* 98.20* 97.90* 0.35

2009

Jokisaari and Nurmi

537

TABLE 4 Correlations between Growth Factors in Multivariate Latent Growth Analyses
Initial Status of Perceived Change in Supervisor Perceived Support Supervisor Support Change Change Change Change *p in in in in role clarity work mastery salary job satisfaction .31 .07 .26* .08 .77* .35 .31* .64*

.05 or greater.

were included in the models as covariates. These variables were allowed to correlate with the growth factors. First, we examined the associations between perceived supervisor support and role clarity. The multivariate model showed moderate fit with the data ( 2[355] 582.01; CFI .91; TLI .91; RMSEA .06; SRMR .07). The results indicated, as shown in Table 4, that change (slope factor) in perceived supervisor support was related to change in role clarity; that is, the steeper the decreasing trend in perceived supervisor support, the steeper the declining trend in role clarity over time ( .77, p .05). This result supports Hypothesis 2. Second, we examined the associations between perceived supervisor support and work mastery. The multivariate model provided moderate fit with the data ( 2[355] 508.79; CFI .92; TLI .92; RMSEA .05; SRMR .09). The results, as shown in Table 4, did not support Hypothesis 3, in that change in perceived supervisor support was not related to rate of change in work mastery ( .35, p .05). Next, we examined the associations between perceived supervisor support and salary. The multivariate model provided moderate fit with the data ( 2[148] 233.79; CFI .95; TLI .94; RMSEA .06; SRMR .08). The results showed, in line with Hypothesis 4, that the steeper the decreasing trend in perceived supervisor support, the less salary increased over time ( .31, p .05). Finally, we examined the associations between perceived supervisor support and job satisfaction. The multivariate model provided moderate fit with the data ( 2[355] 574.88; CFI .91; TLI .90; RMSEA .06; SRMR .08). As shown in Table 4, change in perceived supervisor support was related to change in job satisfaction; that is, the steeper the decreasing trend in perceived supervisor support, the steeper the decline in job satisfaction over time ( .64, p .05). This result supports Hypothesis 5.

We also performed each of the multivariate latent growth models reported above by including social integration with coworkers in the models to estimate the role of coworkers in the socialization process. Change in social integration with coworkers was first modeled by univariate LGM. The results showed that the linear model fitted the data on social integration adequately ( 2[62] 88.91; CFI .98; TLI .98; RMSEA .05; SRMR .11).3 In multivariate latent growth models, the growth factors for social integration and perceived supervisor support and the focal socialization outcome variable were intercorrelated. All of the multivariate latent growth models showed adequate fit with the data. The results showed that change in social integration was not related to changes in perceived supervisor support ( .36, p .05), role clarity ( .37, p .05), job satisfaction ( .39, p .05), or salary ( .08, p .05). However, the rate of growth in social integration was related to the growth rate in work mastery over time ( .68, p .05); that is, the higher was social integration with coworkers, the higher was work mastery. The results showed further that, with social integration included in the models, the relations between perceived supervisor support and socialization outcomes were analogous to those reported previously. (See Appendix B for additional analyses concerning the role of individual attributes and job characteristics in change in supervisor support and socialization outcomes.) We also performed an additional multivariate latent growth model analysis to examine whether perceived supervisor support was related to the socialization outcomes when all of the work adjustment variables were included in the same model, as reported previously. In this model, the growth factors for all the work adjustment variables and perceived supervisor support were intercorrelated. The results showed that the relations between perceived supervisor support and socialization outcomes were the same as reported above. Finally, we also conducted supplementary analyses concerning the cross-lagged relationships between supervisor support and each socialization outcome variable, if their slope factors were shown to be related to each other in the LGM analyses; role
The nested model comparisons showed that the equal residual variance of the repeated latent variables resulted in a significant decrease in fit when this model was compared to the heterogeneous residual variance 2 model ( df 3, 8.73, p .05). Consequently, for the social integration variable, the residual variances of the latent variables were allowed to be heterogeneous over time.
3

538

Academy of Management Journal

June

clarity, job satisfaction, and salary were thus examined. Analyses related to the supervisor support and attitudinal socialization outcome variables were done via use of four-wave cross-lagged models (see, e.g., Bollen & Curran, 2006). That is, socialization outcome at time t 1 was an additive combination of the focal socialization outcome variable at time t and supervisor support at time t, and supervisor support at time t 1, in turn, was an additive combination of supervisor support at time t and the focal socialization outcome variable at time t. In addition, occupational domain, work experience, and job tenure were control variables in these analyses (that is, supervisor support and the focal socialization outcome variables were regressed on these variables). In these analyses, measurement equivalence constraints were used and residual terms of identical measures were allowed to covary for different measurement points. The fit indexes suggested that the models fitted the data adequately (supervisor support and job satisfaction: 2[313] 490.81; CFI .92; TLI .91; RMSEA .05; SRMR .08; supervisor support and role clarity: 2[313] 501.06; CFI .92; TLI .91; RMSEA .05; SRMR .07). The results showed that neither role clarity ( s .12, .05, .12; t’s 1.52, 0.51, 1.44) nor job satisfaction ( s .04, .05, .09; t’s 0.44, 0.54, 1.00) at time t predicted perceived supervisor support at time t 1. Instead, perceived supervisor support at time 2 predicted job satisfaction at time 3 ( .18; t 2.25), and perceived supervisor support at time 3 predicted role clarity at time 4 ( .14; t 1.90; p .06). We used a three-wave cross-lagged model to further examine the relationship between supervisor support and salary, since salary was measured only thrice, at times 1, 2, and 4. Otherwise the procedure was similar as reported above for the four-wave cross-lagged models. The fit indexes suggested that the cross-lagged model fitted the data adequately ( 2[80] 129.74; CFI .96; TLI .94; RMSEA .06; SRMR .05). The results showed that salary level did not predict perceived supervisor support later on ( s .07, .07; t’s 0.91, 0.86), but perceived supervisor support at time 2 predicted salary level at time 4 ( .13; t 2.46). These results further support the importance of perceived supervisor support in newcomer socialization. DISCUSSION A major assumption in the organizational socialization literature is that the interaction between newcomers and organizational insiders such as supervisors is the main channel influencing rates of socialization and related work adjustment among

newcomers. The results of the present study showed that, on average, perceived supervisor support declined during the period 6 –21 months after organizational entry, and that the greater the decrease in perceived supervisor support, the greater the decrease in newcomers’ role clarity and job satisfaction, and the slower the increase in salary over time. However, change in perceived supervisor support was not related to rate of change in newcomers’ work mastery over time. Change in Perceived Supervisor Support and Socialization Outcomes Our findings provided support for three out of the four hypotheses concerning the relations between changes in perceived supervisor support and socialization outcomes. The results showed that the greater the decrease in perceived supervisor support, the greater the decrease in newcomers’ role clarity and job satisfaction during the 6 –21 months after organizational entry, supporting Hypotheses 2 and 5. In addition, the results showed that decreasing perceived supervisor support was related to a slower increase in newcomers’ salary, in line with our Hypothesis 4. These results are consistent with those in the previous organizational socialization literature, emphasizing the importance of supervisors in organizational socialization (e.g., Bauer & Green, 1998; Major et al., 1995; Morrison, 2002). Furthermore, many scholars in the field have argued that the interaction between newcomers and organizational insiders such as supervisors is an important channel through which the organizational socialization of newcomers is negotiated and implemented (Graen, 1976; Louis, 1990; Reichers, 1987). Our study extends the earlier research by showing that change in newcomers’ perceived supervisor support is also related to change in newcomers’ adjustment to work and salary progress over time. In addition, the results of the present study add to the literature by showing that the relations between perceived supervisor support and socialization outcomes extend beyond the first year in an organization. However, the results showed that the association between change in perceived supervisor support and work mastery failed to reach statistical significance. It is possible that supervisors’ role in the development of newcomers’ work mastery is at its most important at the beginning of organizational entry and that work mastery stabilizes thereafter. In line with this explanation and earlier research (e.g., Morrison, 2002), our results showed that initial perceived supervisor support was related to newcomers’ initial work mastery. One further possibil-

2009

Jokisaari and Nurmi

539

ity is that coworkers play a role in the development of work mastery over time (e.g., Chen & Klimoski, 2003). For example, coworkers may provide important feedback and information once a newcomer is accepted as a member of a work group (Feldman, 1981). In line with this explanation, our additional analyses showed that change in newcomers’ social integration with coworkers was related to rate of change in work mastery over time. Overall, the results of the present study support the argument that a key process in organizational socialization is connecting with relevant social contacts to access and mobilize resources (Graen, 1976; Louis, 1990; Reichers, 1987; Schein, 1978; Wanous, 1992). A supervisor typically has positional power to channel resources, tasks, and opportunities in a workplace (e.g., Graen, 1976). He or she may also be an important source of information and feedback (e.g., Ostroff & Kozlowski, 1992) as well as a formal authority in negotiations over a newcomer’s role and salary (e.g., Graen, 1976). Our study adds to earlier socialization research by showing how changes in perceptions of supervisor support over time are related to changes in key outcomes of organizational socialization. The present results suggest that support and feedback from a newcomer’s supervisor is needed over an extended period of time following organizational entry. In other words, a high level of adjustment to work and salary progress over time in newcomers is related to their perceptions that they have continuous supervisor support. Perceived Supervisor Support Over Time In accordance with Hypothesis 1, the results of the present study showed that, on average, after the honeymoon period and the initiation of the relationship between newcomer and supervisor (Fichman & Levinthal, 1991), perceived supervisor support declines during the 6 –21 months after organizational entry. Our finding is in line with the argument for a liability of newness in relationship development: relationships started mainly because of exogenous factors, such as work relations, tend to weaken over time, and this decay is more rapid among new relationships than old ones (Burt, 2000). Furthermore, it has been suggested that “heterophilous” relationships related to work roles and status, such as superior-subordinate ones, tend to weaken more rapidly over time than relations among people whose status and roles are similar (Burt, 2002; McPherson et al., 2001). Many organizational, interpersonal, and resource constraints may account for the decline in perceived supervisor support over time. For example, organizational

constraints such as workload and number of subordinates limit the feasibility of high supervisor support over an extended period of time (e.g., Green et al., 1996). Furthermore, it has been found that the critical issue most often reported by newcomers after organizational entry was conflict between them and their supervisors; moreover, negative experiences tended to increase over time (Gundry & Rousseau, 1994). The present study adds to earlier research on organizational socialization by showing that changes occur not only in newcomers’ adjustments to work but also in their perceptions of supervisor support during the 6 –21 months after organizational entry. Moreover, the amount of individual variation in the change in perceived supervisor support over time was considerable. This result supports the idea that there is selectivity in the development of the relationship between supervisor and newcomer over time, possibly owing to organizational, interpersonal, and resource opportunities and constraints. Practical Recommendations The results of the present study have some practical implications. First, perceived supervisor support was shown to change during the study period, and this change had particular implications for newcomer adjustment to work and rise in salary. This finding suggests that organizations should provide supervisors with the resources (e.g., Schein, 1978) to better channel socialization of newcomers after organizational entry. When supervisors have a long-term opportunity to invest in newcomers, the returns for organizations in terms of high newcomer adjustment to work appear to be high. Conversely, newcomers may benefit from being informed that, on average, a high level of supervisor support is not feasible over an extended period of time after organizational entry and thus they should capitalize on the supervisor support they receive on entering. Furthermore, newcomers could be encouraged to develop informal ties in the organization and not rely only on their immediate supervisors as a source of resources, since supervisor support seems to decline over time. The social network approach also suggests, along the same lines, that newcomers are well advised to develop informal ties rather than trust only in a supervisor relationship that is liable to weaken over time (Burt, 1992). However, initially a newcomer may need sponsorship from organizational insiders to gain the legitimacy required to access informal networks (Burt, 1992), since typically his or her own social ties will be few after organizational entry.

540

Academy of Management Journal

June

Supervisors could use their position and networks to introduce newcomers to others in a workplace (Sparrowe & Liden, 1997; see also Kozlowski et al., 1996). For example, a supervisor might extend a newcomer’s ties by sponsoring his or her participation in different projects. Such participation might enhance integration into informal networks and thus may help the newcomer acquire knowledge and other resources, as those taking a social network approach have argued. Furthermore, these informal ties are likely to advance newcomers’ rates of socialization and work adjustment (Morrison, 2002) by providing additional resources and compensating for decreasing supervisor support over time. For example, our additional analyses showed that change in newcomers’ social integration in the workplace was related to rate of change in their work mastery. Limitations and Future Directions Several limitations have to be taken into account when making generalizations based on the present study. First, our measurements were done at roughly half-year intervals. Similar time lines have been used in earlier research (e.g., Feij et al, 1995). However, the socialization literature offers no clear theoretical arguments about when and how often measurements should be conducted (Jablin, 2001; Saks & Ashforth, 1997). The key issue to adequately modeling socialization dynamics is the length of the intervals between measurements and the number of measurement occasions (e.g., Zaheer, Albert, & Zaheer, 1999). In future studies, researchers need to deploy different kinds of designs and measurement intervals to more fully capture the dynamics of socialization and related adjustment. Second, we used only self-reports for the work adjustment and other measures. Common method variance bias in the information obtained may be a result. Future research should collect information from other sources as well. For example, research could benefit from the use of archival data and multiple sources of ratings, such as coworkers and supervisors. Another problem with self-reports is the possibility of a ceiling effect. The present attitudinal indicators of work adjustment showed high levels of adjustment after organizational entry. However, in contrast to a possible ceiling effect, the work adjustment indicators did not show statistically significant negative covariance between the initial status of the variables and the rate of change in them (e.g., Rogosa, 1995). Other factors may also explain the negative covariance between the initial status and the rate of change in relation to perceived supervisor support found in this study. The

negative covariance may be a product of regression to the mean (e.g., Rogosa, 1995). That is, initial extreme values in the focal variable may partially be a result of measurement error, and these values are likely to grow closer to the sample mean over time. Third, our sample consisted of recent graduates, individuals facing the transition from school to work. It is therefore possible that their adjustment required a longer time period to level off and stabilize than would be the case for employees with more work experience (e.g., Saks & Ashforth, 1997). Furthermore, as this study was drawn from Finnish employees, and about half of our sample consisted of employees in the Finnish health care domain, caution is accordingly advised in generalizing the results presented here. Consequently, there is need to examine the socialization process using samples of different kinds. Fourth, in the present study we were not able to ensure that newcomers had the same supervisors throughout the study period. However, in our opinion, in cases in which newcomers possibly faced new supervisors, decline in perceived supervisor support over time likely would be underestimated. Both the liability of newness and the honeymoon period in social relationship development suggest that new relationships start with an initial period of exploration during which they are protected from decline (Burt, 2002; Fichman & Levinthal, 1991). Thus, when newcomers had new supervisors during our study period, one would expect a temporary leveling out in decline in the perceived supervisor support over time. We tested this assumption in proxy by comparing the mean levels of perceived supervisor support at time 3 between participants who stayed with the same employers throughout the study period—and presumably, stayed with the same supervisors—and participants who changed employers between time 1 and time 2 (that is, had new supervisors). The results showed that those participants who had new employers, and thus new supervisors, reported higher supervisor support at time 3 than those who had stayed with the same employers (t 1.79). However, the differences in mean levels of supervisor support had leveled out at time 4. Finally, it would be important in future studies also to explore the socialization of newcomers from the perspective of supervisors. For example, to what extent are supervisors’ organizational possibilities and constraints, such as number of subordinates and organizational support, related to the development of the relationships between supervisors and newcomers over time? Moreover, as supervisors’ and employees’ perceptions of employees’ organizational behavior seem to differ (e.g., Harris

2009

Jokisaari and Nurmi

541

& Schaubroeck, 1988), it would be important to examine supervisor’s perceptions of newcomers’ rates of adjustment to work and job proficiency. This analysis would also offer an advantage over newcomers’ self-reports by reducing possible common method variance. Furthermore, different dimensions of supervisor-newcomer interactions as well as the relational context surrounding their relationships should be examined. For example, measures related to leader-member exchange and social network research could be included. Conclusions Entry into a new job is accompanied by interaction between the newcomer and organizational insiders that may then have consequences for individual socialization and related adjustment to work. The present study extends understanding of organizational socialization by showing that perceived supervisor support decreased over time during a study period of 6 –21 months after organizational entry and that the greater the decrease in perceived supervisor support, the greater the decrease in newcomers’ adjustment to work, and the slower the increase in salary over time. REFERENCES
Allen, T. D., Poteet, M. L., & Burroughs, S. M. 1997. The mentor’s perspective: A qualitative inquiry and future research agenda. Journal of Vocational Behavior, 51: 70 – 89. Ashford, S. J., & Black, J. S. 1996. Proactivity during organizational entry: The role of desire for control. Journal of Applied Psychology, 81: 199 –214. Bandura, A. 2001. Social cognitive theory: An agentic perspective. In S. Fiske, D. Schacter, & C. ZahnWaxler (Eds.), Annual review of psychology, 52: 1–26. Palo Alto, CA: Annual Reviews. Bauer, T. N., & Green, S. G. 1998. Testing the combined effects of newcomer information seeking and manager behavior on socialization. Journal of Applied Psychology, 83: 72– 83. Bollen, K. A., & Curran, P. J. 2006. Latent curve models: A structural equation perspective. Hoboken, NJ: Wiley. Boyd, N. G., & Taylor, R. R. 1998. A developmental approach to the examination of friendship in leaderfollower relationships. Leadership Quarterly, 9: 1–25. Bruderl, J. 2000. The dissolution of matches: Theoretical ¨ and empirical investigations. In J. Weesie & W. Raub (Eds.), Theoretical models and empirical studies of households and organizations: 129 –130. Amsterdam: Thela Thesis.

Burt, R. S. 1992. Structural holes. Cambridge, MA: Harvard University Press. Burt, R. S. 2000. Decay functions. Social Networks, 22: 1–28. Burt, R. S. 2002. Bridge decay. Social Networks, 24: 333–363. Cammann, C., Fichman, M., Jenkins, G. D., Jenkins, G. D., Jr., & Klesh, J. R. 1983. Assessing the attitudes and perceptions of organizational members. In S. E. Seashore, E. E. Lawler III, P. H. Mirvis, & C. Cammann (Eds.), Assessing organizational change: A guide to methods, measures, and practices: 71–138. Wiley: New York. Chan, D., & Schmitt, N. 2000. Interindividual differences in intraindividual changes in proactivity during organizational entry: A latent growth modeling approach to understand newcomer adaptation. Journal of Applied Psychology, 85: 190 –210. Chao, G. T., O’Leary-Kelly, A. M., Wolf, S., Klein, H. J., & Gardner, P. D. 1994. Organizational socialization: Its content and consequences. Journal of Applied Psychology, 79: 730 –743. Chen, F., Bollen, K. A., Paxton, P., Curran, P. J., & Kirby, J. B. 2001. Improper solutions in structural equation models. Sociological Methods & Research, 29: 468 – 508. Chen, G., & Klimoski, R. J. 2003. The impact of expectations on newcomer performance in teams as mediated by work characteristics, social exchanges, and empowerment. Academy of Management Journal, 46: 591– 607. Dallner, M., Elo A.-L., Gamberale, F., Hottinen, V., Knardahl, S., Lindstron, K., Skogstad, A., & Orhede, E. ¨ 2000. Validation of the general Nordic questionnaire (QPSNordic) for psychological and social factors at work. Copenhagen: Nordic Council of Ministers. Farber, H. S. 1994. The analysis of interfirm worker mobility. Journal of Labor Economics, 12: 554 –593. Feij, J., Whitely, W., Peiro, J., & Taris, T. 1995. The development of career-enhancing strategies and content innovation: A longitudinal study of new workers. Journal of Vocational Behavior, 46: 231–256. Feldman, D. C. 1976. A contingency theory of socialization. Administrative Science Quarterly, 21: 433– 452. Feldman, D. C. 1981. The multiple socialization of organization members. Academy of Management Review, 6: 309 –318. Ferris, G. R., & Judge, T. A. 1991. Personnel/human resources management: A political influence perspective. Journal of Management, 17: 447– 488. Fichman, M., & Levinthal, D. A. 1991. Honeymoons and the liability of adolescence: A new perspective on duration dependence in social and organizational

542

Academy of Management Journal

June

relationships. Academy of Management Review, 16: 442– 468. Fishbein, M., & Ajzen, I. 1975. Belief, attitude, intention, and behavior: An introduction to theory and research. Reading, MA: Addison-Wesley. Fisher, C. D. 1986. Organizational socialization: An integrative review. In G. R. Ferris & K. M. Rowland (Eds.), Research in personnel and human resource management, vol. 4: 101–145. Greenwich, CT: Jai Press. Gabarro, J. J. 1987. The development of working relationships. In J. W. Lorsch (Ed.), Handbook of organizational behavior: 172–189. Englewood Cliffs, NJ: Prentice-Hall. Graen, G. B. 1976. Role-making process within complex organizations. In M. Dunnette (Ed.), Handbook of industrial and organizational psychology: 1201– 1245. Chicago: McNally. Graen, G. B., & Uhl-Bien, M. 1995. Relationship-based approach to leadership: Development of leadermember exchange (LMX) theory of leadership over 25 years: Applying a multilevel multi-domain perspective. Leadership Quarterly, 6: 219 –247. Green, S. G., Anderson, S. E., & Shivers, S. L. 1996. Demographic and organizational influences on leader-member exchange and related work attitudes. Organizational Behavior and Human Decision Processes, 66: 203–214. Griffeth, R. W., Hom, P. W., & Gaertner, S. 2000. A metaanalysis of antecedents and correlates of employee turnover: Update, moderator tests, and research implications for the next millennium. Journal of Management, 26: 463– 488. Gundry, L. K., & Rousseau, D. M. 1994. Critical incidents in communicating culture to newcomers: The meaning is the message. Human Relations, 47: 1063– 1087. Hackman, J. R., & Oldham, G. R. 1980. Work redesign. Reading, MA: Addison-Wesley. Hanisch, K. A., & Hulin, C. L. 1991. General attitudes and organizational withdrawal: An evaluation of causal model. Journal of Vocational Behavior, 39, 110 – 128. Harris, M. M., & Schaubroeck, J. 1988. A meta-analysis of self-supervisor, self-peer, and peer-supervisor ratings. Personnel Psychology, 41: 43– 62. Harrison, D. A., Newman, D. A., & Roth, P. L. 2006. How important are job attitudes? Meta-analytic comparisons of integrative behavioral outcomes and time sequences. Academy of Management Journal, 49: 305–325. Hu, L., & Bentler, P. M. 1999. Cutoff criteria for fit indexes in covariance structure analysis: Conventional criteria versus new alternatives. Structural Equation Modeling, 6: 1–55. Jablin, F. M. 2001. Organizational entry, assimilation,

and disengagement/exit. In F. M. Jablin & L. L. Putnam (Eds.), The new handbook of organizational communication: Advances in theory, research, and methods: 732– 818. Thousand Oaks, CA: Sage. Judge, T. A., Thoresen, C. J., Bono, J. E., & Patton, G. K. 2001. The job satisfaction-job performance relationship: A qualitative and quantitative review. Psychological Bulletin, 127: 376 – 407. Kammeyer-Mueller, J. D., & Wanberg, C. R. 2003. Unwrapping the organizational entry process: Disentangling multiple antecedents and their pathways to adjustment. Journal of Applied Psychology, 88: 779 –794. Kammeyer-Mueller, J. D., Wanberg C. R., Glomb T. M., & Ahlburg D. 2005. The role of temporal shifts in turnover processes: It’s about time. Journal of Applied Psychology, 90: 644 – 658. Katz, R. 1980. Time and work: Toward an integrative perspective. In B. M. Staw & L. L. Cummings (Eds.), Research in organizational behavior, vol. 2: 81– 127. Greenwich, CT: JAI. Kozlowski, S. W., Gully S. M., Salas, E., & Cannon-Bowers, A. 1996. Team leadership and development: Theory, principles, and guidelines for training leaders and teams. Advances in Interdisciplinary Studies of Work Teams, 3: 253–291. Lance, C. E., Vandenberg, R. J., & Self, R. M. 2000. Latent growth models of individual change: The case of newcomer adjustment. Organizational Behavior and Human Decision Processes, 83: 107–140. Lee, D. M. & Allen, T. J. 1982. Integrating new technical staff: Implications for acquiring new technology. Management Science, 28: 1405–1420. Louis, M. R. 1990. Acculturation in the workplace: Newcomers as lay ethnographers. In B. Schneider (Ed.), Organizational climate and culture: 85–129. San Francisco: Jossey-Bass. Louis, M. R., Posner, B. Z., & Powell, G. N. 1983. The availability and helpfulness of socialization practices. Personnel Psychology, 36: 857– 866. Major, D. A., Kozlowski, S. W., Chao, G. T., & Gardner, P. D. 1995. A longitudinal investigation of newcomer expectations, early socialization outcomes and moderating effects of role development factors. Journal of Applied Psychology, 80: 418 – 431. McPherson, M., Smith-Lovin, L., & Cook, J. M. 2001. Birds of a feather: Homophily in social networks. In K. S. Cook & J. Hagan (Eds.), Annual review of sociology, vol. 27: 415– 444. Palo Alto: Annual Reviews. Moore, M. L. 1974. Superior, self, and subordinate differences in perception of managerial learning times. Personnel Psychology, 27: 297–305. Morrison, E. W. 1993. Longitudinal study of the effects of information seeking on newcomer socialization. Journal of Applied Psychology, 78: 173–183.

2009

Jokisaari and Nurmi

543

Morrison, E. W. 2002. Newcomers’ relationships: The role of social network ties during socialization. Academy of Management Journal, 45: 1149 –1160. Muthen, L. K., & Muthen, B. O. 2005. MPlus user’s guide ´ ´ (3rd ed.). Los Angeles: Muthen & Muthen. ´ ´ Ostroff, C., & Kozlowski S. W. 1992. Organizational socialization as a learning process: The role of information acquisition. Personnel Psychology, 45: 849 – 874. Reichers, A. E. 1987. An interactionist perspective on newcomer socialization rates. Academy of Management Review, 12: 278 –287. Rhoades, L. A., & Eisenberger, R. 2002. Perceived organizational support: A review of the literature. Journal of Applied Psychology, 87: 698 –714. Robinson, S. L., Kraatz, M. S., & Rousseau, D. M. 1994. Changing obligations and the psychological contract: A longitudinal study. Academy of Management Journal, 37: 137–152. Rogosa, D. 1995. Myths and methods: “Myths about longitudinal research” plus supplemental questions. In J. M. Gottman (Ed.), The analysis of change: 3– 65. Mahwah, NJ: Erlbaum. Saks, A. M., & Ashforth, B. E. 1997. Organizational socialization: Making sense of the past and present as prologue for the future. Journal of Vocational Behavior, 51: 234 –279. Schafer, J. L., & Graham, J. W. 2002. Missing data: Our view of the state of the art. Psychological Methods, 7: 147–177. Schein, E. H. 1978. Career dynamics: Matching individuals and organizational needs. Reading, MA: Addison-Wesley. Seidel, M.-C. L., Polzer, J. T., & Stewart, K. J. 2000. Friends in high places: The effects of social networks in discrimination in salary negotiations. Administrative Science Quarterly, 45: 1–24. Sparrowe, R. T., & Liden, R. C. 1997. Process and structure in leader member exchange. Academy of Management Review, 22: 522–552. Thoresen, C. J., Kaplan, S. A., Barsky, A. P., de Chermont, K., & Warren, C. R. 2003. The affective underpinnings of job perceptions and attitudes: A meta-analytical review and integration. Psychological Bulletin, 129: 914 –945. Trevor, C. O., Gerhart, B., & Boudreau, J. W. 1997. Voluntary turnover and job performance: Curvilinearity and the moderating influences of salary growth and promotions. Journal of Applied Psychology, 82: 44 – 61. Tubre, T. C., & Collins, J. M. 2000. Jackson and Schuler (1985) revisited: A meta-analysis of the relationships between role ambiguity, role conflict, and job performance. Journal of Management, 26: 155–169. Wanous, J. P. 1992. Organizational entry: Recruitment,

selection, orientation, and socialization of newcomers (2nd ed.). Reading, MA: Addison-Wesley. Watson, D., Clark, L., & Tellegen, A. 1988. Development and validation of brief measures of positive and negative affect: The Panas scales. Journal of Personality and Social Psychology, 54: 1063–1070. Zaheer, S., Albert, S., & Zaheer, A. 1999. Time scales and organizational theory. Academy of Management Review, 24: 725–741.

APPENDIX A Items in Scales
Supervisor Support (Feij et al., 1995). (1) To what extent does your supervisor provide helpful advice on how to perform your job tasks? (2) To what extent does your supervisor give feedback about your job performance? (3) To what extent does your supervisor provide task assignments which improve skills and knowledge? (1, “seldom/never,” to 5, “very often”). Role Clarity (Dallner et al., 2000). (1) Have clear goals and objectives been defined for your job? (2) Do you know what your responsibilities are? (3) Do you know exactly what is expected of you at work? (1, “seldom/never,” to 5, “very often”). Work Mastery (Dallner et al., 2000). (1) Are you content with your ability to solve problems in your work tasks? (2) Are you content with the quality of work you do? (3) Are you content with the amount of work that you get done? (1, “seldom/never,” to 5, “very often”). Job Satisfaction (Cammann et al., 1983). (1) All in all, I’m satisfied with my job. (2) In general, I don’t like my job. (reverse-coded) (3) In general, I like working here. (1, “strongly disagree,” to 5, “strongly agree”).

APPENDIX B Additional Analyses
We also examined whether two individual attributes, proactive behavior and negative affect, and job characteristics (i.e. job demands) were related to change in supervisor support and attitudinal socialization outcomes. Among the individual-level variables, newcomer proactivity in particular has gained currency in the field (for a review, see Saks and Ashforth [1997]). In this view, newcomers are purposive agents in organizational socialization: they try actively to direct their transition into organizations. Proactive behaviors have been operationalized as, for example, information seeking, feedback seeking, and relationship building (e.g., Ashford & Black, 1996). Our measures included feedback seeking as an indicator of proactive behavior at time 1 (after Ashford & Black, 1996). Furthermore, earlier research suggests that people high in negative affectivity, which refers to aversive mood states such as contempt and nervousness, are less satisfied with their jobs and work environments (for

544

Academy of Management Journal

June

a review, see Thoresen et al. [2003]). We measured negative affect (Watson, Clark, & Tellegen, 1988) at time 1. Subsequently, we performed latent growth modeling (LGM) analyses related to perceived supervisor support and socialization outcomes, as reported above, with the addition that the feedback seeking and negative affect variables at time 1 were in the models as a predictor of change in perceived supervisor support and socialization outcomes. The results showed that neither feedback seeking nor negative affect predicted change in supervisor support or socialization outcomes over time (all p’s .05). Finally, we examined whether change in job content was related to change in supervisor support and socialization outcomes. The job characteristics model (Hackman & Oldham, 1980) suggests that job demands, which refer to task requirements and workload, affect work attitudes. We measured job demands (e.g., “Are your work tasks too difficult for you?” [Dallner et al. 2000]) over times 1– 4 and subsequently modeled change in job demands over time by using latent growth modeling. The results showed that job demands increased, on average, between time 1 and time 4, albeit this increase was only marginally significant ( 0.06, t 1.88). Furthermore, we ran multivariate latent growth models to examine the relations between job demands, supervisor support, and attitudinal socialization outcomes. In the multivariate LGM models, the growth factors for job demands and perceived supervisor support and the focal socialization outcome variable were intercorrelated. The models showed adequate fit to the data, excepting the multivariate LGM model related to supervisor support, job satis-

faction, and job demands, which failed to converge to a proper solution. Consequently, in this model we allowed the time-specific factors between job satisfaction and job demands to correlate. Thereafter the model showed adequate fit with the data. The results showed that change in job demands was not related to change in perceived supervisor support and socialization outcomes (all p’s .05). The results showed further that, after we included the job demands variable in the models, the relations between perceived supervisor support and socialization outcomes were analogous to those reported above.

Markku Jokisaari (mjokisaari@gmail.com) is a researcher at the Finnish Institute of Occupational Health. He received his Ph.D. in psychology from the University of Jyvaskyla. His research interests include newcomer ¨ ¨ adjustment, social networks, work transitions, and working relationships. Jari-Erik Nurmi (jari-erik.nurmi@psyka.jyu.fi) is a professor of psychology at the University of Jyvaskyla, Fin¨ ¨ land, and the director of the Finnish Center of Excellence in Learning and Motivation Research. He received his Ph.D. in psychology from the University of Helsinki. His research interests include topics such as motivation and coping at school, parenting, adolescent socialization, and modeling of developmental processes.