Content uploaded by Carl P. Maertz
Author content
All content in this area was uploaded by Carl P. Maertz on Mar 01, 2016
Content may be subject to copyright.
UNCORRECTED PROOFS
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
The effects of perceived organizational
support and perceived supervisor support
on employee turnover
CARL P. MAERTZ JR
1
*
, RODGER W. GRIFFETH
2
,
NATHANAEL S. CAMPBELL
3
AND DAVID G. ALLEN
4
1
Department of Management, Saint Louis University, St. Louis, Missouri, U.S.A.
2
Departments of Psychology and Management Systems, Ohio University, Athens, Ohio, U.S.A.
3
Henderson State University, Arkadelphia, Arkansas, U.S.A.
4
Department of Management, University of Memphis, Memphis, Tennessee, U.S.A.
Summary This study examines mediated effects of perceived supervisor support (PSS) and perceived
organizational support (POS) on turnover cognitions, and their interactive effects on turnover
behavior in a sample of 225 social services workers. In this study, we address a seeming
contradiction in current findings regarding support and attachments to managers versus
attachments to the organization itself. The POS literature suggests fully mediated causal
paths to turnover from POS and PSS, through affective commitment. Whereas, the commit-
ment, LMX, and turnover literatures suggest alternative causal paths that imply broader effects
for POS and PSS on turnover. Contrary to earlier POS literature, findings showed that PSS had
independent effects on turnover cognitions not mediated through POS. Model tests also
indicated that POS had significant effects on turnover mediated through normative commit-
ment, as well as affective organizational commitment. Moreover, a new significant interactive
relationship was discovered such that low PSS strengthened the negative relationship between
POS and turnover, while high PSS weakened it. Implications for research and practice are
discussed. Copyright # 2007 John Wiley & Sons, Ltd.
Introduction
Voluntary employee turnover has been one of the most studied outcomes in organizational research.
This pervasive interest comes mainly from a recognition that voluntary turnover can be very costly, and
that understanding and managing it better can provide considerable benefits (e.g., Griffeth & Hom,
2001). In fact, retaining talent is now becoming more critical in a world where the organization’s
human capabilities are increasingly the key source of competitive advantage (e.g., Pfeffer, 2005). Also,
historical changes like the massive retirement of baby boomers (leading to a dearth of qualified workers
for some key jobs) and the erosion of societal norms favoring organizational loyalty promise difficult
challenges for turnover management (e.g., Cappelli, 2005; Ito & Brotheridge, 2005). Thus, better
Journal of Organizational Behavior
J. Organiz. Behav. (in press)
Published online in Wiley InterScience
(www.interscience.wiley.com) DOI: 10.1002/job.472
* Correspondence to: Carl P. Maertz, Jr, Department of Management, Saint Louis University, 3674 Lindell Boulevard, St. Louis,
MO 63108, U.S.A. E-mail: spiral3@charter.net
Copyright # 2007 John Wiley & Sons, Ltd.
Received 10 April 2006
Revised 21 February 2007
Accepted 14 April 2007
UNCORRECTED PROOFS
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
understanding turnover causes and how to control them will likely remain a primary concern into the
future.
Predictors of tur nover behavior have historically been the primary focus of the literature (Griffeth,
Hom, & Gaert ner, 2000), but focusing on predictors overlooks the multiple causal pathways proposed
in key turnover models (e.g., Lee & Mitchell, 1994; Maertz & Campion, 2004; Mobley, 1977; Mobley,
Griffeth, Hand, & Meglino, 1979; Price & Mueller, 1981). More theoretically fruitful streams of
turnover research have focused on the int ermediate causal linkages between established predictors and
turnover (e.g., Hom, Caranikas-Walker, Prussia, & Griffeth, 1992; Mobley, 1977). In this literature,
general causal linkages from predictors (e.g., job satisfaction, organizational commitment) to turnover
cognitions then to turnover behavior have been firmly established (Hom et al., 1992). That is, there is
little, if any, doubt that global attitudes influence turnover cognitions, which in turn, are the proximal
cause of turnover behavior. However, modeling the distinct causal mechanisms that drive relationships
between the multitude of predictors and turnover cognitions is a research direction that has emerged
more recently (Maertz & Griffeth, 2004). The first, general objective of this study is to contribute to this
growing stream of turnover research, which strives to better model the nomological network
surrounding voluntary turnover decisions.
March and Simon’s (1958) theory of organizational equilibrium suggests that individual decisions to
stay working in an organization are a function of the balance between the expected utilities of the
inducements offered by the organization and the contributions expected from the individual. Although
much of the subsequent research has focused on attitudinal (e.g., satisfaction) and tangible (e.g., pay)
inducements to stay, a growing body of work recognizes that relational inducements such as support
from the organization and from a supervisor can also play an important role in such decisi ons (e.g.,
Allen, Shore, & Griffeth, 2003). The second objective of this study is to expand understanding of how
perceived organizational support (POS) and perceived supervisor support (PSS) actually come to
influence turnover cognitions and behavior. In pursuing this objective we also provide needed
assessment of the potential practical utility of PSS and POS interventions for influencing turnover.
In the burgeoning POS/PSS research, findings suggest that POS and supervisors only influence
voluntary turnover decisions through their impact on affective organizational commitment. Yet,
research in commitment, mentoring, LMX, and turnover areas suggest that suppor t may impact
turnover decisions through its impact on normative obligations and through distinct supervisor-level
support effects. This seeming contradiction needs to be addressed because these additional mechanisms
of influencing turnover cognitions (i.e., obligations and supervisor-level effects), if supported, imply
new directions for researchers and practitioners that current POS/PSS models do not. For example, if
the full-mediation perspective is correct, this implies a simple causal model linking PSS to POS to
affective commitment to turnover cognitions, and finally, to turnover behavior, with little need for
further modeling. However, if the broader perspective is correct, researchers must look beyond
organizational attitudes to other mechanisms through which support influences turnover decisions.
Moreover, because researchers have not succeeded in explaining large percentages of variance in
turnover behavior (e.g., Peters & Sheridan, 1988), any potential causal paths to turnover that account
for variance beyond organizational attitudes are especially important to model (e.g., Mitchell, Holtom,
Lee, Sablynski, & Erez, 2001). Thus, the third objective of this study is to help resolve some seemingly
contradictory arguments and findings regarding the role of PSS and POS-driven obligations in the
causal paths to turnover.
With these objectives in mind, we test: (1) whether the effects of POS on turnover cognitions are
mediated through normative commitment, as well as established effects through affective commitment;
(2) whether PSS directly influences turnover cognitions, as well as established effects through POS;
and (3) whether PSS moderates the established negative POS–turnover relationship. Next, we briefly
review relevant background literature and then present arguments for each hypothesis in turn.
Copyright # 2007 John Wiley & Sons, Ltd. J. Organiz. Behav. (in press)
DOI: 10.1002/job
C. P. MAERTZ JR ET AL.
UNCORRECTED PROOFS
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
Literature Review and Hypotheses
Eisenberger, Huntington, Hutchinson, and Sowa (1986) suggested that employees form a global belief
concerning the extent to which the organization values their contributions and cares about their
well-being, called POS. POS has since become a pervasive construct in organizational research
(Rhoades & Eisenberger, 2002). Many recent studies have found significant empirical relationships
between POS and important outcomes (e.g., Allen et al., 2003; Eisenberger, Armeli, Rexwinkel, Lynch,
& Rhoades, 2001; Eisenberger, Stinglhamber, Vandenberghe, Sucharski, & Rhoades, 2002; Rhoades &
Eisenberger, 2002; Rhoades, Eisenberger, & Armeli, 2001; Settoon, Bennett, & Liden, 1996). In their
meta-analysis, Rhoades and Eisenberger (2002) found a mean corrected correlation of .51 between
POS and turnover intention, and a mean corrected correlation of .11 between POS and turnover
behavior.
Employees also develop general views concerning the degree to which supervisors value their
contributions and care about their well-being (Kottke & Sharafinski, 1988), called PSS. PSS has also
been significantly related to important outcomes (Eisenberger et al., 2002; Van Yperen & Hagedoorn,
2003). Payne and Huffman (2005) found that, among Army officers, mentoring support (68 per cent
from immediate supervisors) had a significant effect on turnover. Moreover, Eisenberger et al. (2002)
found that PSS was negatively correlated with turnover behavior (r ¼.11). Given the demonstrated
efficacy of POS and PSS as significant predictors, turnover researchers should attempt to understand the
relevant causal linkages and investigate potential higher order relationships involving PSS and POS.
Expanding POS linkages to turnover decisions
An emp loyee perceiving support from the organization should perceive greater inducements (March
and Simon, 1958) and these can potentially increase his/her instances of positive mood at work, which
could then cause positive emotional associations with the organization itself, thereby increasing
affective commitment (Eisenberger et al., 2001) and decreasing turnover intention (Mathieu & Zajac,
1990). In a key causal modeling study, Rhoades et al. (2001) confirmed this linkage empirically, finding
that the POS effect on turnover was fully mediated though affective commitment. Allen et al. (2003)
also found that commitment and job satisfaction fully mediated POS effects on turnover. Tekleab,
Takeuchi, and Taylor (2005) did not measure organizational commitment but found that all effects of
POS on turnover intentions were mediated through global job satisfaction. Thus, existing findings in the
POS literature indicate that POS effects on turnover constructs are fully mediated through global,
affect-based work attitudes.
However, based on the tenets of social exchange and reciprocity theory, receiving support from the
organization should also cause some experienced obligation to the organization (e.g., Meyer & Allen,
1991; Rhoades et al., 2001). Eisenberger et al. (2001) suggested that POS, through a norm of
reciprocity, can engender employee obligation to aid the organization. They maintain that this
mechanism (along with positive mood effects) helps explain the relationship between POS and
affective commitment. Although these obligations to stay with the organization, referred to as
normative commitment (Meyer & Allen, 1991), are narrower than the obligations to aid the
organization described by Eisenberger et al. (2001), most obligations to aid the organization certainly
subsume an obligation to remain. Maertz and Griffeth (2004) also theorized that such obligations are
distinct in their motivational potential from affective commitment; in that, employees can feel obliged
to stay without liking the organization curr ently, or like the organization without feeling any great
obligation. In any case, POS should inspire some obligation to remain, which leads to lower turnover
Copyright # 2007 John Wiley & Sons, Ltd. J. Organiz. Behav. (in press)
DOI: 10.1002/job
THE EFFECTS OF POS AND PSS
UNCORRECTED PROOFS
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
cognitions, apart from its effects on affective commitment (Meyer, Allen, & Smith, 1993; Meyer &
Allen, 1991). Moreover, empirical research has already supporte d that POS can lead to felt obligations
owed to the organization out of an adherence to a reciprocity norm (Tekleab et al., 2005). Despite
theoretical arguments and indirect empirical evidence, rese archers have not tested normative
commitment as a mediator of the POS–turnover cognitions relationship. Thus:
Hypothesis 1: POS effects on turnover cognitions will be mediated by normative commitment as
well as by affective commitment.
Expanding PSS linkages to turnover decisions
Most turnover research has focused primarily on global attitudes toward the organization or job as
antecedents (e.g., Mobley, 1977; Price & Mueller, 1981), rather than employee relationships with
managers and others within the organization (Mitchell et al., 2001). Nevertheless, anecdotal evidence
from practitioner circles and research evidence (e.g., Griffeth et al., 2000; Maertz, Stevens, &
Campion, 2003; Payne & Huffman, 2005) strongly indicate that immediate managers play a major role
in employee turnover decisions. In assessing their influence on turnover, a key question is: Do
immediate managers/supervisors embody the organization itself (Levinson, 1965), or do they forge
their own inde pendent attachments to employees (Reichers, 1985)? Here, we address this question with
respect to PSS and help resolve disagreement about how it influences turnover.
On one side, findings from the POS literature indicate that PSS effects on turnover are fully mediated
through POS (Eisenberger et al., 2002; Rhoades et al., 2001). Eisenberger et al. (2002) found that the
effects of PSS on turnover were fully mediated through POS. In a study of college graduates, Rhoades
et al. (2001) found that POS fully mediated the effects of PSS on affective commitment. Both of these
studies explain the full mediation of PSS through POS by assuming that the supervisor represents or
personifies the organization (Eisenberger et al., 2002). This causal path implies that support from a
supervisor only influences turnover by affecting the employee’s perceptions and attitudes regarding the
organization itself. In this view, the supervisor’s effect on turnover decisions is as a generat or of POS
and a developer of organizational commitment among his/her employees.
In contrast, theory and empir ical findings in the areas of leader–member exchange, commitment,
mentoring, and turnover suggest that supervis or support should have broader, independent effects on
turnover cognitions and behavior (e.g., Becker, 1992; Clugston, Howell & Dorfman, 2000; Maertz
et al., 2003; Maertz, Mosley, & Alford, 2002; Mitchell et al., 2001; Settoon, Bennett, & Liden, 1996;
Tekleab et al., 2005; Wayne, Shore, & Liden, 1997; Wayne, Shore, Bommer, & Tetrick, 2002).
Theorists have agreed that employees can distinguish relations with the immediate supervisor from
relations with the organizational entity or its leaders (Becker, 1992; Dirks & Ferrin, 2002; Reichers,
1985, 1986). When this distinction is made, employees can become independently attached to each
focus (Reichers, 1985). Maertz and Griffeth (2004) theorized that supervisor attachments have distinct
effects on turnover cognitions, apart from attachments and attitudes related to the organization itself.
Also, employees tend to direct their reciprocating actions toward the target from which benefits
originate (e.g., Hoffman & Morgeson, 1999). Thus, in response to PSS, any resulting obligation should
be primarily to the supervisor rather than to the organization.
Empirically, findings in these areas also indicate that supervisor-level constructs explain significant
incremental variance in outcomes beyond organization-level effects. For instance, Settoon et al. (1996)
found that LMX was significantly related to in-role performance and citizenship behavior, while POS
was not. Similarly, Wayne et al. (1997) found that POS and LMX were independently related to
different outcome variables. Constituent commitments have also demonstrated incremental effects on
Copyright # 2007 John Wiley & Sons, Ltd. J. Organiz. Behav. (in press)
DOI: 10.1002/job
C. P. MAERTZ JR ET AL.
UNCORRECTED PROOFS
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
outcomes beyond the effects of organizational commitment (Becker, 1992; Maertz et al., 2002).
Mitchell et al. (2001) found that relationships inside the organization (and out) explained variance in
turnover behavior beyond organizational commitment effects. From the mentoring literature, Payne
and Huffman (2005) found that the effects of mentoring support on turnover were only partially
mediated by affective commitment. They also found direct, non-mediated effects of mentor ing support
(largely from immediate supervisors) on turnover behavior. Similarly, Ito and Brotheridge (2005)
found that career support from supervisors had a significant direct path to turnover intentions after
controlling for affective commitment. While these findings involve LMX, commitment to the
supervisor, embeddedness, and mentoring rather than PSS per se, high levels of these former constructs
imply perceiving support from the manager in question (e.g., Rhoades & Eisenberger, 2002).
These theoretical arguments and findings all strongl y suggest that PSS should cause attachments
directly to the supervisor, which should influence turnover cognitions directly, in addition to its
established influence through increasing POS and affective commitment. Thus:
Hypothesis 2: PSS effects on turnover cognitions will be partially rather than fully mediated by POS.
Based on these two hypotheses and strong inference research (Platt, 1964), two alternative models
were constructed to be comparatively tested (See Figures 1 and 2). The first model represents the
existing full-mediation findings from the POS literature where effects of PSS are fully mediated by
POS, whos e effects on turnover cognitions are fully mediated by affective commitment. The second
hypothesized model adds a POS path to turnover cognitions mediated through normative commitment
and a direct path from PSS to turnover cognitions.
As our theoretical control in these models, we include perceived job alternatives in the model to
ensure that both primary factors, ease of movement (alternatives) and desirability of movement
(organizational commitment), are represented (March and Simon, 1958). We focus on turnover
cognitions as the dependent variable for this analysis for several reasons. First, turnover cognitions are
the more causally proximal result of the exogenous variables and mechanisms represented. Second,
with Hypotheses 1 and 2, we are focusing on competitively testing two sets of viable alternative causal
pathways rather than on confirming predictors of turnover behavior that have been establish ed
elsewhere (e.g., Rhoades & Eisenberger, 2002). Th ird, the dichotomous turnover behavior measure
would violate assumptions of the SEM causal modeling techniques employed (e.g., Allen et al., 2003)
PSS
POS
Affective
Commitment
Normative
Commitment
Turnover
Cognitions
Perceived Job
Alternatives
Figure 1. Model supported in the POS literature
Copyright # 2007 John Wiley & Sons, Ltd. J. Organiz. Behav. (in press)
DOI: 10.1002/job
THE EFFECTS OF POS AND PSS
UNCORRECTED PROOFS
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
and would provide lower power to detect effects (e.g., Peters & Sheridan, 1988). Thus, we believe it is
reasonable to use cognitions as the criterion in this test.
POS T PSS interaction
We do examine turnover behavior as the criterion in testing this hypothesis which examines an
interactive relationship for the first time. Presumably because previous POS research found that PSS
effects on turnover cognition are fully mediated by POS, there has been little interest in investigating
any higher order effects among PSS and POS. The failure to consider such interactions represents a gap
in our understanding, which inhibits our ability to correctly specify models or design interventions
involving both POS and PSS.
To address this, we first assume that most employees desire support from both sources. However, one
construct may affect the salience of the other in the minds of employees, separate from increasing or
reducing its average level. Specifically, more frequent day-to-day contact with employees provides the
supervisor with more opportunity than the organization itself to demonstrate support (or lack thereof) to
employees. Because of this, PSS levels become obvious to employees relatively quickly and may
fluctuate more than POS levels. Thus, employees experiencing low PSS are likely to focus more on the
organization as a source of needed support. This would make POS more salient for turnover decisions
under conditions of low PSS.
On the other hand, a ‘high PSS supervisor’ delivers important benefits on a regular basis (e.g.,
consideration, good assignments, flexible work schedules, feedback, recommendations, and
recognition). When supervisors are perceived to provide such support, they are likely to elicit
positive feelings and trust from employees (e.g., Dirks & Ferrin, 2002). This makes supervisors even
more able to position themselves in the forefront of their employees’ consciousness as the primary
provider (and potentially the inhibitor) of employee support. The supervisor’s providing regular
support may overshadow the organization as a source of support, making POS less salient to employee
decisions and thereby weakening the POS–turnover relationship.
Hypothesis 3: The negative relationship between POS and turnover will be stronger for those
experiencing low PSS and weaker for those experiencing high PSS.
PSS
POS
Affective
Commitment
Normative
Commitment
Turnover
Cognitions
Perceived Job
Alternatives
Figure 2. Model implied in the broader literature
Copyright # 2007 John Wiley & Sons, Ltd. J. Organiz. Behav. (in press)
DOI: 10.1002/job
C. P. MAERTZ JR ET AL.
UNCORRECTED PROOFS
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
Organizational Context
The data were collected in 1997 as part of a research proj ect to study employee relations in the
Department of Family and Children’s Services, in four Georgia counties surrounding a major
metropolitan area. Turnover data were collected 1 year later from three counties; one county did not
provide turnover data because our contact in that county left the organization. The study included a
wide variety of employees surrounding the organization’s primary mission of supporting children
and families (e.g., caseworkers, administrative support staff, HR, IT, and finance professionals,
managers). They were/are state employees, and many (particularly the caseworkers) are highly
involved in a very emotionally demanding work environment.
Methods
Sample and procedures
Participants were 375 social workers from four Counties of a state department of family and children’s
services located in the southeastern United States. Surveys were distributed to them in small groups to
be completed on a voluntary basis during work hours as part of an employee survey to improve
retention and working conditions. Since we asked respondents to identify themselves in order to match
their surveys with subsequent turnover data, the confidential nature of their responses was stressed on
the survey cover page and by the researcher collecting the surveys. Further, the surveys were returned
directly to the researcher. Complete surveys were received from 315 employees; however, one county
did not report subsequent turnover data. Thus, complete surveys with accompanying turnover behavior
data were received from 225 employees for a final response rate of 60 per cent. The average age of
respondents was 40.3 years (SD ¼ 10.6) and 82 per cent were female. Of our respondents, 64 per cent
were White, 32 per cent were African American, 2 per cent were Hispanic, and 2 per cent were from
other backgrounds.
Measures
Affective commitment
Employees’ level of affective organizational commitment was measured using a nine-item version of
the Organizational Commitment Questionnaire (Mowday, Steers, & Porter, 1979) suggested by
Bozeman and Perrewe (2001) to minimize conceptual overlap with withdrawal cognitions. A sample
item is ‘I feel, for me, this is the best of all possible organizations for which to work’.
Normative commitment
We measured normative commitment with three items from Meyer et al.’s (1993) normative
commitment scale. Items measured the employee’s felt obligation to stay with the current organization:
Copyright # 2007 John Wiley & Sons, Ltd. J. Organiz. Behav. (in press)
DOI: 10.1002/job
THE EFFECTS OF POS AND PSS
UNCORRECTED PROOFS
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
‘I would feel guilty if I left the organization now’, ‘I owe a great deal to my organization’, and ‘This
organization deserves my loyalty’.
Perceived alternative jobs
Perceived alternative jobs were measured as a key control (March and Simon, 1958) with a three-item
scale from Kim, Price, Mueller, and Watson (1996). The items assessed the employee’s perceived
ability to find another job. A sample item is, ‘It would be easy for me to find a job with another
employer that is as good as the one I now have’.
POS
POS was measured with 16-item short-form of the original 36 item SPOS developed and used by
Eisenberger and colleagues (1986). A sample item is, ‘The organization strongly considers my goals
and values’.
PSS
Similar to previous efforts to measure PSS (Eisenberger et al., 2002; Kottke & Sharafinski, 1988;
Rhoades et al., 2001), we adapted three items from the SPOS to assess employees’ perceptions that
their supervisor cares for their well-being and supports them at work. An example item is ‘My
supervisor is willing to listen to my job-related problems’.
Turnover cognitions
We used 10 items from Hom and Griffeth (1991) to measure this construct. Items were answered on a
five-point scale with either ‘no chance, 25 per cent chance, 50 per cent chance, 75 per cent chance, and
100 per cent chance’, or with ‘strongly disagree’ to ‘strongly agree’ as scale anchors. Sample item
stems for each anchor type include, ‘What are the chances that you will leave this organization in the
next 12 months?’ and ‘I am thinking about quitting my job’.
Turnover behavior
Turnover was measured 12 months after the survey collection. Twenty-six participants quit for a
turnover base rate of 12 per cent. Although somewhat low, this base rate is not unusual in turnover
research (e.g., Peters & Sheridan, 1988). ‘Leavers’ were coded as ‘1’ for our analyses and ‘stayers’
were coded as ‘0’.
Analyses
We ran the CFA measurement model with all exogenous causal variables and used SEM modeling to
test Hypotheses 1 and 2. We used structural equation modeling (SEM) with LISREL 8.30 (Joreskog &
Sorbom, 1993) to test the models and hypotheses. A covariance matrix was used as input for estimation
of models. We used scale scores as indicators of our measures rather than the individual items because
LISREL has some computing limitations with fitting models having more than 30 manifest indicators
(Bentler & Chou, 1987; Williams & Hazer, 1986). In our study, we had more than 60 manifest
indicators. Thus, following the procedure outlined by Williams and Hazer (1986), we calculated the
measurement loadings (square root of scale reliability) and error variances (one minus reliability times
the scale variance). We also controlled for tenure and age. For Hypothesis 3, because of the binary
nature of actual turnover behavior, we used logistic regression.
Copyright # 2007 John Wiley & Sons, Ltd. J. Organiz. Behav. (in press)
DOI: 10.1002/job
C. P. MAERTZ JR ET AL.
UNCORRECTED PROOFS
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
Results
CFA results
All estimated item loadings were significant ( p > .05). The model fit statistics were as follows:
x
2
¼ 1063.89, df ¼ 517, p ¼ .00; RMSEA ¼ .069; GFI ¼ .78; AGFI ¼ .75; NFI ¼ .80; NNFI ¼ .88;
CFI ¼ .89. Based on these findings and the fact that all scales had adequate reliabilities, we concluded
that there was sufficient evidence to proceed with the scales as proposed. Means, standard deviations,
scale internal consist ency reliabilities (alphas), and correlations among study variables are presented in
Table 1.
Hypothesis tests
The first model in Figure 3, hypothesizing that the POS effects on turnover cognitions will be fully
mediated by affective commitment and that POS would be fully mediated by POS, provided the
following fit to the data: x
2
¼ 295.86, df ¼ 21, p ¼ .00; RMSEA ¼ .24; GFI ¼ .75; AGFI ¼ .58;
NFI ¼ .58; CFI ¼ .59. We compared this mode l to the second model, where the POS effects on turnover
cognitions are mediated by normative commitment as well, and where PSS effects are only partially
mediated through POS. The results indicate that the partia lly mediated model in Figure 4, with paths
from POS to normative commitment and from PSS to turnover cognitions, fit the data significantly
better than the more restricted model (x
2
¼ 227.40, df ¼ 19, p ¼ .00; RMSEA ¼ .22; GFI ¼ .80;
AGFI ¼ .62; NFI ¼ .70; CFI ¼ .71). The x
2
difference between the partially and fully mediated models
was statistically significant (Dx
2
¼ 68.46, 2 df, p < .01). Further, the path coefficients for each of these
additional paths were significant in the expected direction (see Figures 3 and 4). Thus, Hypotheses 1
and 2 were supported in the sense that POS is directly related to normative as well as affective
commitment, and the effects of PSS on turnover cognitions are only partially mediated. However, the fit
of this model still failed to reach conventional standards for good fit.
Hypothesis 3 proposed that the strength of the negative relationship between POS and turnover
would be stronger for those low on PSS and weaker for those high on PSS. These logistic regression
results are presented in Table 2. The interaction coefficient was significant and the form of the
Table 1. Means, standard deviations, reliabilities, and correlations
Variable Mean SD 123456789
1. Age 40.29 10.56 —
2. Tenure 8.37 7.97 .51
—
3. PSS 3.48 .96 .11 .01 .89
4. POS 3.92 1.05 .15
.06 .62
.94
5. POS PSS 14.25 6.30 .15
.06 .87
.90
—
6. Affective commitment 4.45 1.17 .25
.19
.44
.64
.60
.91
7. Normative commitment 2.85 .90 .23
.24
.37
.50
.50
.63
.79
8. Perceived alternatives 2.97 .87 .24
.25
.09 .21
.17
.33
.41
.90
9. Withdrawal cognitions 2.57 1.08 .38
.34
.35
.46
.45
.58
.52
.47
.94
10. Turnover .12 .32 .13
.11 .04 .06 .03 .08 .01 .09 .11
Reliabilities (coefficient alphas) are reported on the diagonal.
n ¼ 225.
p < .05.
Copyright # 2007 John Wiley & Sons, Ltd. J. Organiz. Behav. (in press)
DOI: 10.1002/job
THE EFFECTS OF POS AND PSS
UNCORRECTED PROOFS
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
interaction was in the expected direction. The POS–turnover relationship was stronger for those with
low PSS and weaker for those with high PSS see (Figure 5).
Discussion
The purpose of this study was to clarify the relationships between two forms of perceived support and
voluntary turnover decisions. This stud y general ly confirms the place of POS and PSS in the
nomological network leading to turnover behavior. We contribute by expanding this network beyond
Note:
*
p < .05 (one-tailed).
PSS
POS
Affective
Commitment
Normative
Commitment
Turnover
Cognitions
Perceived Job
Alternatives
.68
*
.69
*
-.14
*
.61
*
-.34
*
-.16
*
.30
*
Age
Tenure
-.16
*
-.14
*
Figure 4. Partially mediated model implied in broader literature
PSS
POS
Affective
Commitment
Normative
Commitment
Turnover
Cognitions
Perceived Job
Alternatives
.68
*
.68
*
-.44
*
-.17
*
.29
*
Age
Tenure
-.17
*
-.13
*
Note:
*
p < .05 (one-tailed).
Figure 3. Fully mediated model from current POS literature
Copyright # 2007 John Wiley & Sons, Ltd. J. Organiz. Behav. (in press)
DOI: 10.1002/job
C. P. MAERTZ JR ET AL.
UNCORRECTED PROOFS
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
models previously supported in the POS literature. In particular, we added a path from POS to turnover
mediated through normative commitment, added a direct path from PSS to turnover cognitions, and
uncovered an int eraction of PSS and POS on turnover behavior. These findings have several
implications for researchers and practitioners.
Implications for predicting turnover behavior
Notably, our logist ic regression findings are the first to show that both POS and PSS are significantly
related to turnover behavior with antecedents reflecting both perceived desirability and ease of
movement (i.e., perceived alternatives) also included in the multivariate model. This finding with such
Table 2. Results of logistic regression on turnover behavior
Odds ratio Wald statistic SE
Organizational commitment .68 1.88 .29
Normative commitment 1.72 2.47 .34
Perceived alternative jobs 1.51 2.03 .29
POS .37 2.70
.61
PSS .37 2.86
.58
POS PSS 1.33 3.21
.16
Model x
2
8.73
2 log likelihood 152.364
Pseudo R
2
.054
n ¼ 225.
p < .05 (one-tailed).
0.00
0.10
0.20
0.30
0.40
0.50
0.60
0.70
0.80
0.90
1.00
12
34567
POS
Predicted Turnover
PSS =-1 st
d
PSS =+1 st
d
Figure 5. Interaction of PSS and POS on turnover behavior
Copyright # 2007 John Wiley & Sons, Ltd. J. Organiz. Behav. (in press)
DOI: 10.1002/job
THE EFFECTS OF POS AND PSS
UNCORRECTED PROOFS
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
antecedents controlled in the equation helps further establish the importance of POS and PSS as
turnover determinants (e.g., Mitchell et al., 2001). We also discovered that the POS–turnover
relationship was stronger under conditions of low PSS than under conditions of high PSS. When the
supervisor provides high support, POS becomes a less important predictor of turnover. POS becomes
significantly more important when support from the supervisor is relatively absent. Employees may
then seek out support from the organization itself, thereby making POS more relevant for turnover
decision-making. Theoretically, this interpretation suggests the possibility of a compensa tory model of
support where a certain amount of support is needed and sought by employees (to avoid quitting). This
minimum level of support may be supplied by the supervisor, the organization itself, or a combination.
Future researc h should directly investigate whether this compensatory relation for POS and PSS exists
with regard to turnover and other outcome s. The overall conclusion is that both constructs and their
interaction should be included in future predictive models, even with organizational commitment and
perceived job alternatives already included in the model.
Implications for turnover causal modeling research
Further, our findings indicate that the effects of POS on turnover are not always fully mediated through
affective commitment, as has been found in past POS studies (Eisenberger et al., 2002; Rhoades et al.,
2001). One main conclusion for research is that POS can influence turnover cognitions and behavior
through other mechanisms besides improving global affect-loaded work attitudes. Specifically, POS
can create obligations in the employee to reciprocate through remaining with the organization. These
obligations to stay may cause an employee to have fewer turnover cognitions and dismiss them more
quickly. Future research should investigate other potential causal mechanisms behind how
organizational support affects turnover tendencies. For example, current POS may give the employee
a feeling of confidence that support may also be forthcoming in the future. Such expectations of future
support may cause more favorable calculations of one’s ability to meet future career goals within
the organization, and thereby reduce turnover intentions (Maertz & Griffeth, 2004). Thus, future
researchers may want to investigate whether future-oriented, calculative motives of attachment also
mediate the effects of POS on turnover.
The effects of the supervisor
Our findings add to the growing body of research concluding that supervisor-related perceptions and
attitudes can influence outcomes separately from organization-related perceptions and attitudes (e.g.,
Becker, 1992; Ito & Brotheridge, 2005; Wayne et al., 1997), in contrast to the fully mediated effects for
PSS reported in the POS literature (i.e., Eisenberger et al., 2002; Rhoades et al., 2001). The implication
for turnover research is that future studies should always consider the potential for employees’
relationships with their supervisors to directly impact turnover decisions.
Given this, one wonders why the previous full-mediation findings occurred in the first place. Using
actual turnover behavior with a base rates of 13, 14, and 10 per cent may have hurt the power to detect
effects for PSS on turnover in Eisenberger et al. (2001) and Rhoades et al. (2001), compared to our
higher variance model criterion of turnover cognitions (which is also closer to PSS in the causal
sequence). In any case, significant evidence remains that employees may see supervisors acting as
organizational representatives who exert effects on turnover through helping to increase POS and
positive organizational attitudes.
However, it is clear from current and a larger body of findings that supervisors also form individual
relationships and related attachments with their employees that are distinct from employee attitudes
toward the organization as a whole. In other words, supervisors can act as agents of the organization or
Copyright # 2007 John Wiley & Sons, Ltd. J. Organiz. Behav. (in press)
DOI: 10.1002/job
C. P. MAERTZ JR ET AL.
UNCORRECTED PROOFS
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
independent actors or both when providing support and fostering attachments. Researchers must strive
to learn the personal and situational characteristics that determine whether supervisors are seen by
employees as independent objects of attachment versus agents of the organization causing positive
attitudes toward it.
Future research should also investigate other alternative causal paths to outcomes, not mediated
through organizational commitment and globa l job satisfaction. Following the pattern from the
commitment and LMX literature (i.e., coworker commitment or TMX), researchers should investigate
whether perceived coworker support adds incrementally to the prediction of turnover and other
outcomes. To comprehensively consider the effects of support, researchers should also investigate
support that comes from family or friends away from work (e.g., Adams, King, & King, 1996; Edwards
& Rothba rd, 2000) and how it impacts turnover decisions.
The current context
It was somewhat surprising that the established turnover predictors were not significantly related to
turnover behavior in the equation while POS and PSS were. One explanation may lie in the context.
Perhaps in this sample of social workers, commitment and perceived alternatives were less important
than POS and PSS. One important aspect of their jobs is to identify and provide support to those in need.
This sample of social workers may have been one where support was particularly needed and valued,
because they provide support to others using limited monetary resources. Thus, support was likely to be
salient in this context. Future studies should determine whether our findings generalize to other
contexts where support is less salient.
Implications for practice
POS and PSS generally suggest relatively straightforward interventions to managers (i.e., providing
more visible signs of organizational support and more visibly supportive supervisors), perhaps more so
than broad attitude constructs do. Increasing recognizable support is imminently practical for
supervisors and organizations to implement. Improving support would also be far less expensive and
complex than raising compensa tion or redesigning jobs to reduce turnover (e.g., McEvoy & Cascio,
1985). Despite this promising potential, if the full-mediation perspective from the POS literature were
true, it would be much more questionable whether increasing POS and PSS would really be a potent
method of retaining employees. In this view, perceived support is simply another antecedent of
organizational commitment, which may change with changes in many other perceptual antecedents
(see Mathieu & Zajac, 1990). Increasing POS and PSS may be no better at reducing turnover than any
other tempor ary, mood-boosting intervention. However, in this study, we rejected the fully mediated
model and supported additional causal pathways from support to turnover. Thus, the primary practical
implication of our findings is that increasing or bolstering POS and PSS is potentially quite valuable as
a part of a turnover management program.
For increasing POS to be successful, managers must strive to find out whether task-based or
socioemotional needs are best targeted through support (e.g., Rhoades et al., 2001) and tailor support
accordingly. Targeted efforts and policies to improve support in this way, and to publicize these efforts,
could play a part in reducing employee turnover. For example, providing adequate staff, increasing
other visible resources in the work environment (food, supply budgets), wellness facilities, and
providing child-care assistance/flextime could increase POS. Even communicating to employees
Copyright # 2007 John Wiley & Sons, Ltd. J. Organiz. Behav. (in press)
DOI: 10.1002/job
THE EFFECTS OF POS AND PSS
UNCORRECTED PROOFS
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
verbally and in writing that the organization wants to help them if possible may foster POS. The current
study increases our confidence that such interventions can help decrease turnover cognitions and
behavior.
Supervisors acting supportive and promoting themselves as such seems to be generally good for
retention. For instance, supervisors could regularly ask employees how they can help them do their job
better and show personal consideration. They could instill fair methods for making workgroup
decisions and even allow exceptions to rules in extreme circumstances to help employees. However,
increasing PSS could potentially have some mixed implications as well. Because of the relatively
frequent interactions between most employees and their supervisors, the supervisor becomes the
primary focus and a potential shaper of employees’ support perceptions. On the upside, it would seem
that a supportive supervisor can be a potential mitigating influence when POS is low. A supportive
supervisor may be able to cover for the shortcomings of organizational policies and top management
decisions that seem unsupportive, keeping turnover from increasing. On the downside, supervisors can
potentially manipulate how employees perceive support and its sources (Eisenberger et al., 2002). They
may portray themselves to employees as a supportive ‘savior’ fighting upstream against an
unsupportive organization. Supervisors can also take personal credit for true organizational-level
support (e.g., ‘good thing I got you that raise’), changing potential POS into PSS (Eisenberger et al.,
2002). Similar to the idea of conflicting commitments (Reichers, 1986), the interaction of POS and PSS
should be viewed partly as an agency problem that needs to be managed, rather than assuming that the
supervisor always represents the organization favorably in the minds of employees. Moreover,
attachment to the supervisor only means attachment to the organization as long as the employee sees
the supervisor as somewhat attached to the company (Maertz & Griffeth, 2004). Thus, distinctive
attachments to supervisors may even increase voluntary turnover when a loyalty-inspiring supervisor
leaves (Maertz et al., 2003). Still, a program of training and rewarding supervisors (who are personally
psychologically attached to the organization) for being supportive, and for presenting the organization
as supportive, is worth serious consideration as a method for reducing turnover.
Limitations and Conclusion
One limitation of this study was the low explained variance in turnover behavior. Nevertheless, this is a
common probl em in most turnover studies (e.g., Peters & Sheridan, 1988). Although the hypothesized
model was better than the fully mediated model, the overall model fit for both models was also poor by
typical standards; and therefore, our hypotheses were not supported at the highest level possible. This
poor fit is a shortcoming that we believe is primarily a function of two factors. First, there are literally
dozens of job, organization, and individual variables that influence turnover intentions that we were not
able to include in our model. Recall that our purpose here was not to comprehensively model turnover
causes, but instead to begin to examine the nature of relationships among POS, PSS, and withdrawal.
Still, although it would not generally be feasible to include all known predictors of turnover, future
research on these relationships would benefit from more completely specified models. Second, as noted
earlier, our context and results may indicate that certain industry or job types could represent boundary
conditions for traditional turnover models. Future research that directly assesses the effects of context
on turnover relationships would also be valuable.
One could potentially criticize our use of cognitions in our modeling analyses as promoting common
method bias, but we see this as a minor issue. The focus in our Hypotheses 1 and 2 was on competitively
testing alternative causal paths, not confirming previously established relationships with turnover
Copyright # 2007 John Wiley & Sons, Ltd. J. Organiz. Behav. (in press)
DOI: 10.1002/job
C. P. MAERTZ JR ET AL.
UNCORRECTED PROOFS
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
behavior. Moreover, turnover cognitions are the proximal cause of behavior in all major theoretical
models to date. This greatly mitigates concerns about common method bias, which do not apply at all to
our tes t of Hypothesis 3. Nevertheless, we could not truly assess causality with these data.
Despite these limitations, this study demonstrated that the effects of POS and PSS on turnover are
more expansive than previously thoug ht and that these constructs may have significant interactive
effects on turnover behavior. Hopefully, the current findings will encour age support researchers to
investigate this PSS POS interaction in the future and will prompt turnover researchers to include
both POS and PSS in future modeling efforts.
Author biographies
Carl P. Maertz, Jr is currently an Associate Professor in the John Cook School of Business at Saint
Louis University. He received his Ph.D. in Human Resources Management from Purdue University’s
Krannert Graduate School of Management and has published work in other outlets such as Academy of
Management Journal, Journal of Applied Psychology, Human Resource Management, and Journal of
Management. His current research interests include voluntary turnover, expatriate management, and
work–family conflict.
Rodger W. Griffeth, Ph.D. is the Byham Chair of Industrial/Organizational Psychology and Professor
of Management Systems at Ohio University. He received his Ph.D. in psychology from the University
of South Carolina and has published in other outlets such as Journal of Applied Psychology,
Psychological Bulletin, Personnel Psychology, Organizational Behavior and Human Decision Pro-
cesses, Academy of Management Journal, and Journal of Management. His current research interests
include investigating voluntary turnover processes and the propensity to turnover.
Nathanael S. Campbell is curr ently an Assistant Professor of Management in the Kelley College of
Business at Hardin-Simmons University. He received his Ph.D. in Human Resource Management and
Organizational Behavior from Mississippi State University’s College of Business and Industry. His
current research interests include mentoring, perceived support, and justice.
David G. Allen (Ph.D. Georgia State University) is an Associate Professor of Management in the
Fogelman College of Business and Economics at the University of Memphi s. His primary research
interests include the flow of people into and out of organizations, for example, turnover and retention,
recruitment communication, and Web-based recruitment. His research on these topics has been
published in Academy of Management Journal, Journal of Applied Psychology, Journal of Manage-
ment, Personnel Psychology, Organizational Research Methods, Human Relations, and other outlets.
References
Adams, G. A., King, L. A., & King, D. W. (1996). Relationships of job and family involvement, family social
support, and work-family conflict with job and life satisfaction. Journal of Applied Psychology, 81, 411–420.
Allen, D. G., Shore, L. M., & Griffeth, R. W. (2003). The role of perceived organizational support and supportive
human resource practices in the turnover process. Journal of Management, 29, 99–118.
Becker, T. E. (1992). Foci and bases of commitment: Are they distinctions worth making? Academy of
Management Journal, 35, 232–244.
Copyright # 2007 John Wiley & Sons, Ltd. J. Organiz. Behav. (in press)
DOI: 10.1002/job
THE EFFECTS OF POS AND PSS
UNCORRECTED PROOFS
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
Bentler, P. M., & Chou, C. P. (1987). Practical issues in structural modeling. Sociological Methods & Research,
16, 78–117.
Bozeman, D. P., & Perrewe, P. L. (2001). The effect of item content overlap on organizational commitment
questionnaire-turnover cognitions relationship. Journal of Applied Psychology, 86 , 161–173.
Cappelli, P. (2005). Will there really be a labor shortage? Human Resource Management , 44, 143–149.
Clugston, M., Howell, J. P., & Dorfman, P. W. (2000). Does cultural socialization predict multiple bases and foci of
commitment? Journal of Management, 26, 5–31.
Dirks, K. T., & Ferrin, D. L. (2002). Trust in leadership: Meta-analytic findings and implications for research and
practice. Journal of Applied Psychology, 87, 611–628.
Edwards, J. R., & Rothbard, N. P. (2000). Mechanisms linking work and family: Clarifying the relationship
between work and family constructs. Academy of Management Review, 25, 178–199.
Eisenberger, R., Armeli, S., Rexwinkel, B., Lynch, P. D., & Rhoades, L. (2001). Reciprocation of perceived
organizational support. Journal of Applied Psychology, 86, 42–51.
Eisenberger, R., Huntington, R., Hutchinson, S., & Sowa, D. (1986). Perceived organizational support. Journal of
Applied Psychology, 71, 500–507.
Eisenberger, R., Stinglhamber, R., Vandenberghe, C., Sucharski, I. L., & Rhoades, L. (2002). Perceived supervisor
support: Contributions to perceived organizational support and retention. Journal of Applied Psychology, 87,
565–573.
Griffeth, R. W., & Hom, P. W. (2001). Retaining valued employees. Thousand Oaks, CA: Sage Publications, Inc.
Griffeth, R. W., Hom, P. W., & Gaertner, S. (2000). A meta-analysis of antecedents and correlates of employee
turnover: Updated moderator tests, and research implications for the next millennium. Journal of Management,
26, 463–488.
Hoffman, D. A., & Morgeson, F. P. (1999). Safety-related behavior as a social exchange: The role of perceived
organizational support and leader-member exchange. Journal of Applied Psychology, 84, 286–296.
Hom, P. W., Caranikas-Walker, F., Prussia, G. E., & Griffeth, R. W. (1992). A meta-analytical structural equations
analysis of a model of employee turnover. Journal of Applied Psychology, 78, 890–909.
Hom, P. W., & Griffeth, R. W. (1991). Structural equations modeling test of a turnover theory: Cross-sectional and
longitudinal analysis. Journal of Applied Psychology, 76, 350–366.
Ito, J. K., & Brotheridge, C. M. (2005). Does supporting employee’s career adaptability lead to commitment,
turnover, or both? Human Resource Management, 44, 5–19.
Joreskog, K. G., & Sorbom, D. (1993). LISREL 8: User’s reference guide. Scientific Software International,
Chicago, IL.
Kim, S., Price, J. L., Mueller, C. W., & Watson, T. W. (1996). The determinants of career intent among physicians
at a U.S. air force hospital. Human Relations, 49, 947–975.
Kottke, J. L., & Sharafinski, C. E. (1988). Measuring perceived supervisory and organizational support.
Educational and Psychological Measurement, 48, 1075–1079.
Lee, T. W., & Mitchell, T. R. (1994). An alternative approach: The unfolding model of voluntary employee
turnover.
Academy of Management Review, 19, 51–89.
Levinson, H. (1965). Reciprocation: The relationship between man and organization. Administrative Science
Quarterly, 9, 370–390.
Maertz, C. P., & Campion, M. A. (2004). Profiles in quitting: Integrating content and process turnover theory.
Academy of Management Journal, 47, 566–582.
Maertz, C. P., & Griffeth, R. W. (2004). Eight motivational forces and voluntary turnover: A theoretical synthesis
with implications for research. Journal of Management, 30, 667–683.
Maertz, C. P., Mosley, D. C., & Alford, B. (2002). Does organizational commitment fully mediate constituent
commitment effects: A re-assessment and clarification. Journal of Applied Social Psychology, 32, 1300–1313.
Maertz, C. P., Stevens, M. J., & Campion, M. A. (2003). A turnover model for the Mexican maquiladoras. Journal
of Vocational Behavior, 63, 111–135.
March, J., & Simon, H. A. (1959). Organizations. New York: Wiley Press.
Mathieu, J. E., & Zajac, D. (1990). A review and meta-analysis of the antecedents, correlates, and consequences of
organizational commitment. Psychological Bulletin , 108, 171–194.
McEvoy, G. M., & Cascio, W. F. (1985). Strategies for reducing employee turnover: A meta-analysis. Journal of
Applied Psychology, 70, 342–353.
Meyer, J., & Allen, N. (1991). A three-component conceptualization of organizational commitment. Human
Resource Management Review, 1, 61–98.
Meyer, J., Allen, N., & Smith, C. (1993). Commitment to organizations and occupations: extension and test of a
three-component conceptualization. Journal of Applied Psychology, 78, 538–551.
Copyright # 2007 John Wiley & Sons, Ltd. J. Organiz. Behav. (in press)
DOI: 10.1002/job
C. P. MAERTZ JR ET AL.
UNCORRECTED PROOFS
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
Mitchell, T. R., Holtom, B. C., Lee, T. W., Sablynski, C. J., & Erez, M. (2001). Why people stay: Using job
embeddedness to predict voluntary turnover. Academy of Management Journal, 44, 1102–1121.
Mobley, W. H. (1977). Intermediate linkages in the relationship between job satisfaction and employee turnover.
Journal of Applied Psychology, 62, 237–240.
Mobley, W. H., Griffeth, R. W., Hand, H. H., & Meglino, B. (1979). Review and conceptual analysis of the
employee turnover process. Psychological Bulletin, 86, 493–522.
Mowday, R. T., Steers, R. M., & Porter, L. W. (1979). The measurement of organizational commitment. Journal of
Vocational Behavior, 14, 222–247.
Payne, S. C., & Huffman, A. H. (2005). A longitudinal examination of the influence of mentoring on organizational
commitment and turnover. Academy of Management Journal, 48, 158–168.
Peters, L., & Sheridan, J. (1988). Turnover research methodology: A critique of traditional designs and a suggested
survival model alternative. In G. R. Ferris & K. M. Rowland (Eds.), Theoretical and methodological issues in
human resource management (pp. 201–232). Greenwich, CT: JAI Press.
Pfeffer, J. (2005). Changing mental models: HR’s most important task. Human Resource Management, 44,
123–128.
Platt, J. R. (1964). Strong inference. Science, 146, 347–353.
Price, J. L., & Mueller, C. W. (1981). A causal model of turnover for nurses. Academy of Management Journal, 24,
543–565.
Reichers, A. (1985). A review and reconceptualization of organizational commitment. Academy of Management
Review, 10, 465–476.
Reichers, A. (1986). Conflict and organizational commitments. Journal of Applied Psychology, 71, 508–514.
Rhoades, L., & Eisenberger, R. (2002). Perceived organizational support: A review of the literature. Journal of
Applied Psychology, 87, 698–714.
Rhoades, L., Eisenberger, R., & Armeli, S. (2001). Affective commitment to the organization: The contribution of
perceived organizational support. Journal of Applied Psychology, 86, 825–836.
Settoon, R. P., Bennett, N., & Liden, R. C. (1996). Social exchange in organizations: Perceived organizational
support, leader-member exchange, and employee reciprocity. Journal of Applied Psychology, 81, 219–227.
Tekleab, A. G., Takeuchi, R., & Taylor, M. S. (2005). Extending the chain of relationships among organizational
justice, social exchange, and employee reactions: The role of contract violations. Academy of Management
Journal, 48, 146–157.
Van Yperen, N. W., & Hagedoorn, M. (2003). Do high job demands increase intrinsic motivation or fatigue or
both? The role of job control and job social support. Academy of Management Journal, 46, 339–348.
Wayne, S. J., Shore, L. M., Bommer, W. H., & Tetrick, L. E. (2002). The role of fair treatment and rewards in
perceptions of organizational support and leader-member exchange. Journal of Applied Psychology, 87,
590–598.
Wayne, S. J., Shore, L. M., & Liden, R. C. (1997). Perceived organizational support and leader-member exchange:
A social exchange perspective. Academy of Management Journal, 40, 82–111.
Williams, L. J., & Hazer, J. T. (1986). Antecedents and consequences of satisfaction and commitment in turnover
models: A reanalysis using latent variable structural equation methods. Journal of Applied Psychology, 71,
219–231.
Copyright # 2007 John Wiley & Sons, Ltd. J. Organiz. Behav. (in press)
DOI: 10.1002/job
THE EFFECTS OF POS AND PSS
