See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/11284309

Perceived Supervisor Support: Contributions to Perceived Organizational

Support and Employee Retention

Article  in  Journal of Applied Psychology · July 2002

DOI: 10.1037//0021-9010.87.3.565 · Source: PubMed

CITATIONS READS

1,563 40,691

5 authors, including:

Florence Stinglhamber Christian Vandenberghe

Université Catholique de Louvain - UCLouvain HEC Montréal - École des Hautes Études commerciales
99 PUBLICATIONS   6,253 CITATIONS    194 PUBLICATIONS   9,100 CITATIONS   

SEE PROFILE SEE PROFILE

Ivan Laars Sucharski

Ivan Sucharski Consulting
8 PUBLICATIONS   1,795 CITATIONS   

SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Performer, Innover par le Bien-Etre (PIB) View project

The dark sides of modern office designs: Towards de-humanization View project

All content following this page was uploaded by Florence Stinglhamber on 17 November 2015.

The user has requested enhancement of the downloaded file.

Journal of Applied Psychology Copyright 2002 by the American Psychological Association, Inc.
2002, Vol. 87, No. 3, 565–573 0021-9010/02/$5.00 DOI: 10.1037//0021-9010.87.3.565

Perceived Supervisor Support: Contributions to Perceived Organizational

Support and Employee Retention

Robert Eisenberger Florence Stinglhamber and Christian Vandenberghe

University of Delaware Catholic University of Louvain

Ivan L. Sucharski and Linda Rhoades

University of Delaware

Three studies investigated the relationships among employees’ perception of supervisor support (PSS),
perceived organizational support (POS), and employee turnover. Study 1 found, with 314 employees
drawn from a variety of organizations, that PSS was positively related to temporal change in POS,
suggesting that PSS leads to POS. Study 2 established, with 300 retail sales employees, that the PSS–POS
relationship increased with perceived supervisor status in the organization. Study 3 found, with 493 retail
sales employees, evidence consistent with the view that POS completely mediated a negative relationship
between PSS and employee turnover. These studies suggest that supervisors, to the extent that they are
identified with the organization, contribute to POS and, ultimately, to job retention.

Organizational support theory (Eisenberger, Cummings, Armeli, & pervisor’s favorable or unfavorable orientation toward them as
Lynch, 1997; Eisenberger, Huntington, Hutchison, & Sowa, 1986; indicative of the organization’s support (Eisenberger et al., 1986;
Rhoades & Eisenberger, in press; Shore & Shore, 1995) supposes that Levinson, 1965). Additionally, employees understand that super-
to meet socioemotional needs and to determine the organization’s visors’ evaluations of subordinates are often conveyed to upper
readiness to reward increased work effort, employees develop global management and influence upper management’s views, further
beliefs concerning the extent to which the organization values their contributing to employees’ association of supervisor support with
contributions and cares about their well-being (perceived organiza- POS. Although over a dozen studies have reported positive rela-
tional support, or POS). Accordingly, employees showed a consistent tionships of POS with PSS (e.g., Hutchison, 1977a, 1997b; Kottke
pattern of agreement with various statements concerning the extent to & Sharafinski, 1988; Malatesta, 1995; Rhoades, Eisenberger, &
which the organization appreciated their contributions and would treat Armeli, 2001; Yoon, Han, & Seo, 1996; Yoon & Lim, 1999; Yoon
them favorably or unfavorably in differing circumstances (Eisen- & Thye, 2000) and related measures (e.g., Allen, 1995; Hutchison,
berger, Fasolo & Davis-LaMastro, 1990; Eisenberger et al., 1986; Valentino, & Kirkner, 1998), little attention has been given to
Shore & Tetrick, 1991; Shore & Wayne, 1993). Employees evidently assessing the direction of causality between POS and PSS, the
believe that the organization has a general positive or negative orien- mechanisms responsible for this association, or the behavioral
tation toward them that encompasses both recognition of their con- consequences of the POS–PSS relationship.
tributions and concern for their welfare.
Just as employees form global perceptions concerning their
Temporal Relationships Between PSS and POS
valuation by the organization, they develop general views concern-
ing the degree to which supervisors value their contributions and On the basis of organizational support theory, findings of a
care about their well-being (perceived supervisor support, or PSS; positive relationship between PSS and POS have usually been
Kottke & Sharafinski, 1988). Because supervisors act as agents of interpreted to indicate that PSS leads to POS (e.g., Hutchison,
the organization, who have responsibility for directing and evalu- 1997a; Malatesta, 1995; Rhoades et al., 2001; Yoon et al., 1996;
ating subordinates’ performance, employees would view their su- Yoon & Lim, 1999). Yoon and Thye (2000) suggested that cau-
sality might also occur in the reverse direction, with POS increas-
ing PSS: employees’ perception that the organization values their
contribution and cares about their well-being might lead them to
Robert Eisenberger, Ivan L. Sucharski, and Linda Rhoades, Department believe that supervisors, as agents of the organization, are favor-
of Psychology, University of Delaware; Florence Stinglhamber and Chris- ably inclined toward them. A relationship between the initial value
tian Vandenberghe, Department of Psychology, Catholic University of
of one variable and changes in a second variable over time pro-
Louvain, Louvain-la-Neuve, Belgium.
Linda Rhoades is now at The Working Data Group, Wayne,
vides stronger causal evidence than is afforded by the simulta-
Pennsylvania. neous measurement of the two variables (Finkel, 1995). We there-
Correspondence concerning this article should be addressed to Robert fore used a panel design to examine the relationship of PSS to
Eisenberger, Department of Psychology, University of Delaware, Newark, temporal change in POS and the relationship of POS to temporal
Delaware 19716. E-mail: [email protected] change in PSS.

565
566 EISENBERGER ET AL.

Hypothesis 1: PSS is positively related to temporal change in POS. mediate a negative PSS–turnover relationship. However, Malatesta
did not test the mediating effect of POS on the possible association
Supervisor’s Perceived Organizational Status as a between PSS and turnover.
Moderator of the Relationship Between PSS and POS Hypothesis 3: POS mediates a negative relationship between PSS and
voluntary employee turnover.
Because supervisors act as organizational agents in their treat-
ment of subordinates, PSS should contribute to POS (Levinson,
1965). The strength of this relationship would depend on the Study 1: Temporal Relationship Between PSS and POS
degree to which employees identify the supervisor with the orga- Organizational support theory assumes that PSS leads to POS.
nization. Supervisors who appear to be highly valued and well Because prior studies assessed PSS and POS simultaneously, ev-
treated by the organization would be highly identified with the idence is unavailable concerning causal direction. We therefore
organization’s basic character and would therefore strongly influ- measured PSS and POS at two points in time 3 months apart. We
ence POS. Of course, an employee may attribute a supervisor’s used structural equation modeling to assess Hypothesis 1, holding
high perceived status to the organization’s misperception of the that initial PSS would be related to a temporal change in POS. We
supervisor’s character. But, on average, supervisors who appear to also examined the relationship between POS and temporal change
be highly regarded by the organization would be assumed by in PSS.
workers to strongly embody the organization’s character.
Employees’ perception of the status accorded their supervisor Method
by the organization, and therefore the employees’ belief that su-
pervisor support also represents organizational support, would Sample and Procedure
increase with employees’ perceptions concerning (a) the organi-
The final sample consisted of 314 alumni of a Belgium university who
zation’s positive valuation of the supervisor’s contributions and its
graduated between 1997 and 1998. To obtain a diverse sample of job types
concern about the supervisor’s well-being, (b) the supervisor’s and organizations, names and addresses of a random sample of 578 alumni
influence in important organizational decisions, and (c) the auton- were obtained from university records. We sent prospective participants a
omy and authority accorded the supervisor in his or her job questionnaire packet containing a cover letter, the survey, and a postage-
responsibilities. The supervisor’s informal organizational status, as paid return envelope. The cover letter explained to the alumni that the
conveyed by these features of the organization’s favorable treat- purpose of the study was to examine the opinions of the University’s
ment of supervisors, should moderate the relationship between graduates about their work environment, informed them that they would
PSS and POS. receive the questionnaire a second time 3 months later, stressed the
importance of responding to both questionnaires, and provided assurances
Hypothesis 2: The relationship between PSS and POS is positively of confidentiality. Two weeks later, we mailed follow-up letters to non-
related to the supervisor’s perceived status within the organization. compliant individuals stressing the value of the survey and the importance
of their participation. Three months after we received a completed ques-
tionnaire from any participant, we sent that participant the second version
POS as a Mediator of the Relationship Between PSS and of the questionnaire. If the follow-up questionnaire was not returned
Employee Turnover within 2 weeks, we sent a letter encouraging participants to complete and
return the questionnaire.
Insufficient consideration has been given as to why PSS and Fifty-four percent of the initial sample returned questionnaires at Time 1
POS have both been found to be related to employee withdrawal and then again at Time 2. This final sample of employees had an average
behaviors. According to organizational support theory, PSS should age of 30.8 years (SD ⫽ 4.1) and had been employed by their organization
decrease voluntary employee turnover by increasing POS. The an average of 4.1 years (SD ⫽ 3.4) before receipt of the first questionnaire.
POS resulting from PSS would strengthen employees’ felt obliga- Seventy-one percent of participants were male. Overall, 20% of the re-
tion to help the organization reach its goals and increase affective spondents were working in heavy and chemical industry, 13% in public
administration, 11% in computer science, 11% in banking and insurance,
organizational commitment, with a resultant reduction in turnover
9% in consulting, 8% in research, 7% in law, 7% in transportation and
and other withdrawal behaviors (Eisenberger, Armeli, Rexwinkel, communication, 5% in construction, 3% in media and advertising, 2% in
Lynch, & Rhoades, 2001; Eisenberger et al., 1986; Rhoades et al., public health, 2% in environmental agencies, and 2% in commerce. A total
2001; Shore & Shore, 1995). of 35% of the respondents worked in large organizations (more than 1,000
Malatesta (1995) maintained that based on the reciprocity norm, employees), 35% in midsize organizations (100 –1,000 employees), and
PSS should increase obligations to the supervisor and to the 30% in small organizations (less than 100 employees).
organization. Malatesta found evidence for both relationships: PSS
increased extra-role performance beneficial to supervisors, and Measures
PSS increased POS, which, in turn, led to greater extra-role per-
formance beneficial to the organization. Concerning voluntary Because participants had graduated from a French-speaking university,
employee turnover, Malatesta (p. 23) suggested that when PSS was all survey measures were first translated from English into French and then
independently back-translated into English by a second translator, follow-
low, employees would believe that they could deal with the un-
ing the procedure recommended by Brislin (1980). The few discrepancies
pleasant situation by switching to a new supervisor or, we would between the original English version and the back-translated version re-
add, minimizing contact with the supervisor while continuing to sulted in adjustment in the French translation based on direct discussion
carry out usual job responsibilities. However, by reducing POS, between the translators.
low PSS would have general unfavorable implications for the Control variable. Because we used employees drawn from a variety of
employees’ future. Thus, in Malatesta’s view, POS would entirely organizations, we controlled statistically for organizational size. We also
 PERCEIVED SUPERVISOR SUPPORT 567

controlled for employee tenure. This information was obtained from the the error covariances of identical terms administered at both
first questionnaire. Time 1 and Time 2 (Finkel, 1995, p. 61).
Perceived organizational support. Studies surveying many occupa- Evidence consistent with an effect of PSS on POS would be
tions and organizations provided evidence for the high internal consistency provided by a statistically significant path between Time 1 PSS
and unidimensionality of the survey generally used to assess POS (Survey
and Time 2 POS. Because this path controlled for Time 1 POS,
of Perceived Organizational Support, or SPOS; Eisenberger et al., 1986,
such an effect would be interpretable as a relationship between
1990; Shore & Tetrick, 1991; Shore & Wayne, 1993). The scale has been
found to be related to, but distinguishable from, measures of similar beliefs PSS and temporal change in POS (Finkel, 1995, p. 27). We also
and attitudes (e.g., Aquino & Griffeth, 1999; Cropanzano, Howes, examined the relationship between Time 1 POS and Time 2 PSS,
Grandey, & Toth, 1997; Eisenberger et al., 1990; Randall, Cropanzano, which, controlling for Time 1 PSS, would provide evidence con-
Bormann, & Birjulin, 1999; Rhoades et al., 2001; Settoon, Bennett, & cerning the influence of POS on PSS.
Liden, 1996; Shore & Tetrick, 1991; Vandenberghe & Peiro, 1999; Wayne, Figure 1 gives the estimated path model with standardized
Shore, & Liden, 1997). To assess employees’ perception that the organi- regression coefficients. For ease of presentation, we show the
zation valued their contribution and cared about their well-being, we structural model rather than the full measurement model, and we
selected three high-loading items from the SPOS (Items 1, 4, and 9; describe the effects of the two control variables (tenure and orga-
Eisenberger et al., 1986) with factor loadings, respectively, of .71, .74, and
nizational size) in the text rather than the figure. Tenure was not
.83. For this and the PSS measure, respondents rated their agreement with
significantly related either to Time 2 PSS (␤ ⫽ .05) or Time 2 POS
each statement using a 5-point Likert-type scale (1 ⫽ strongly disagree;
5 ⫽ strongly agree). (␤ ⫽ .01). Similarly, organizational size was not significantly
Perceived supervisor support. To assess employees’ perception that related to Time 2 PSS (␤ ⫽ .01) or Time 2 POS (␤ ⫽ ⫺.04). As
their supervisor valued their contribution and cared about their well-being, predicted, controlling for employees’ tenure and organizational
we adapted the SPOS in the same manner as Kottke and Sharafinski size, PSS was positively related to the temporal change in POS. In
(1988), Hutchison (1997a, 1997b), Rhoades et al. (2001), and others, contrast, POS was not associated with the temporal change in PSS.
replacing the word organization with the term supervisor. We adapted The overall model showed adequate fit to the data: root-mean-
three items from the SPOS (Items 10, 27, and 35; Eisenberger et al., 1986) square error of approximation (RMSEA) ⫽ .04; comparative fit
on the basis of their high loadings (respectively, .72, .76, and .80). index (CFI) ⫽ .98; goodness-of-fit index (GFI) ⫽ .96; adjusted
goodness-of-fit index (AGFI) ⫽ .93; ␹2(63) ⫽ 91.8, p ⬍ .01. In
Results and Discussion sum, consistent with Hypothesis 1 that PSS leads to POS, PSS was
found to be positively related to the temporal change in POS. In
Descriptive Statistics and Correlations contrast, the relationship between initial POS and the temporal
change in PSS was not statistically significant.
Means, standard deviations, internal reliabilities, and intercor-
relations among variables at Times 1 and 2 are displayed in
Table 1. As predicted, PSS at Time 1 was positively related to POS Study 2: Moderating Effect of Perceived Supervisor
at Time 2. Status on the PSS–POS Relationship

Temporal Relationships Between PSS and POS Those supervisors believed to be most favored by the organiza-
tion would be perceived by employees as most strongly embody-
To assess the temporal relationships between PSS and POS, we ing the organization’s basic character. Thus, support received from
estimated a two-wave panel model with PSS and POS measured at supervisors who appear highly favored by the organization should
both Time 1 and Time 2 (see Figure 1). We used the latent variable have the greatest impact on POS. Study 2 examined the moderat-
structural equation modeling approach of LISREL 8 (Jöreskog & ing effect of the supervisor’s perceived status in the organization
Sörbom, 1993). Paths connected the antecedent variables (Time 1 on the PSS–POS relationship. The supervisor’s informal organi-
PSS, Time 1 POS, tenure, and organizational size) with Time 2 zational status, as perceived by employees, was assessed by em-
PSS and Time 2 POS. As recommended by Finkel (1995, p. 29), ployees’ judgments of (a) the organization’s positive valuation of
Time 2 latent variable variances were allowed to covary. Addi- the supervisor and care about the supervisor’s welfare, (b) the
tionally, we allowed for autocorrelated error variances by freeing supervisor’s influence over important organizational decisions,

Table 1
Study 1: Descriptive Statistics and Intercorrelations

Variable M SD 1 2 3 4 5 6

1. Tenure 4.08 3.34 — .05 ⫺.13* ⫺.10 ⫺.12* ⫺.10

2. Organizational size — — — .10 .09 ⫺.11 ⫺.10
3. Perceived supervisor support at Time 1 3.37 .87 (.81) .74*** .47*** .47***
4. Perceived supervisor support at Time 2 3.34 .92 (.82) .37*** .55***
5. Perceived organizational support at Time 1 2.85 .77 (.74) .65***
6. Perceived organizational support at Time 2 2.85 .74 (.75)

Note. N ⫽ 314. Internal reliabilities (coefficient alphas) are given in parentheses on the diagonal.
* p ⬍ .05. *** p ⬍ .001.
568 EISENBERGER ET AL.

Of the final sample, 27% were hourly paid sales-support employees,

45% were hourly paid salespeople, 21% were salaried sales-support em-
ployees, and 7% were salaried salespeople. These employees had an
average age of 34.3 years (SD ⫽ 11.9) and had been employed by their
organization an average of 4.6 years (SD ⫽ 4.2) before receipt of the
questionnaire. Twenty-seven percent of respondents were women.

Measures
Tenure. We controlled statistically for employee tenure in the organi-
zation. The number of months that each employee had worked for the
organization before the questionnaire administration was obtained from
company records.
Perceived organizational support. We selected eight high-loading
items (Items 4, 8, 9, 13, 20, 22, 23, and 25; loadings from .66 to .84) from
the SPOS to assess POS (Eisenberger et al., 1986). For this and the
following measures, respondents indicated their agreement with each item
Figure 1. Structural equation model of the relationship between per-
using a 7-point Likert-type scale (1 ⫽ strongly disagree, 7 ⫽ strongly
ceived supervisor support (PSS) and perceived organizational support
agree).
(POS) over time. (Study 1). *** p ⬍ .001.
Perceived supervisor support. We measured PSS with the same 8
items used to assess POS, as modified by replacing the word organization
with supervisor.
and (c) the authority and autonomy allotted supervisors to carry Supervisor’s organizational status. We developed a 12-item scale,
out their job responsibilities. given in Table 2, to assess employees’ perceptions of their supervisor’s
organizational status. Four items examined the organization’s valuation of
the supervisor and care about his or her well-being. Four items examined
Method the supervisor’s contributions to important organizational decisions. And
Sample and Procedure the final four items considered the supervisor’s authority and autonomy in
carrying out job responsibilities.
The final sample consisted of 300 employees who worked for a chain of
large discount electronics and appliance stores located in the northeastern
Results and Discussion
United States. These employees comprised 98% of the 313 employees who
received questionnaires and returned them completed during their regularly Dimensionality of the Supervisor’s Perceived
scheduled working hours in the organization’s cafeterias. To encourage
candidness, we gave employees verbal and written assurances that their
Organizational Status
individual responses would be kept confidential and that only group data We carried out a principal-components analysis to determine
would be reported to the organization. Surveys were distributed and col-
whether the 12 supervisor’s organizational status items comprised
lected by the researchers in sealed envelopes. We attribute the extraordi-
a unitary dimension or multiple dimensions. The results suggested
narily high return rate in this and the following study to the survey’s
administration during paid working hours and to the favorable long-term a single factor, as indicated by a strong loading on the first factor,
professional relationship between the investigators and the employees. accounting for 47.4% of the total variance, a related break in the
Employees trusted that their individual identities would be kept confiden- scree plot, and eigenvalues far less than 1.0 for the remaining
tial, that their views would be accurately represented, and that they would presumptive factors. The resultant single-factor scale had an ac-
have access to the same information given to management. ceptable internal consistency (Cronbach’s alpha of .89). As shown

Table 2
Study 2: Factor Analysis of Supervisor’s Perceived Organizational Status Items

Item no. Statement Factor loading Item categorya

1. The organization holds my supervisor in high regard. .81 VAL

2. The organization gives my supervisor the chance to make important decisions. .80 INF
3. The organization values my supervisor’s contributions. .79 VAL
4. The organization gives my supervisor the authority to try new things. .78 AUT
5. The organization supports decisions made by my supervisor. .76 AUT
6. My supervisor participates in decisions that affect the entire organization. .72 INF
7. My supervisor influences decisions made by upper management. .69 INF
8. The organization allows my supervisor to run things the way he wants. .64 AUT
9. The organization consults my supervisor when deciding on new policies and procedures. .61 INF
10. The organization gives my supervisor the freedom to determine how to treat me. .57 AUT
11. If my supervisor decided to quit, the organization would try to persuade him to stay. .54 VAL
12. Even if my supervisor did well, the organization would fail to notice. .45 VAL
a
VAL designates the organization’s high valuation and positive regard for supervisor. INF designates supervisor’s influence in important organizational
decisions. AUT designates supervisor’s authority and autonomy in carrying out job responsibilities.
 PERCEIVED SUPERVISOR SUPPORT 569

in Table 2, principal-components loadings for the items ranged The strength of the relationship between PSS and POS depended
from .45 to .86, with a mean loading of .68. The factor loadings of on employees’ perception concerning their supervisors’ organiza-
the three types of items (organization’s positive valuation of su- tional status. PSS was positively related to POS at both high and
pervisor; supervisor’s influence over important organizational de- low organizational status, but the effect was stronger when the
cisions; and authority and autonomy afforded supervisors) over- supervisor was highly favored by the organization. These findings
lapped in magnitude. Because each item loaded highly on the suggest that supervisors with high perceived organizational status
single factor, all items were used to compute the total perceived are taken by subordinates to more completely embody the organi-
organizational status score. zation’s basic character, leading to a stronger relationship between
PSS and POS.
Descriptive Statistics and Correlations
Study 3: POS as a Mediator of the PSS–Employee
Means, standard deviations, internal reliabilities, and intercor-
relations among variables are displayed in Table 3. Of greatest Turnover Relationship
interest, PSS was positively related to POS. Both PSS and POS have been found to be related to employee
withdrawal behaviors such as absenteeism and turnover. Accord-
Moderating Effect of Supervisor’s Perceived ing to organizational support theory, POS should mediate a neg-
Organizational Status ative relationship between PSS and employee turnover (Hypothe-
sis 3). Therefore, using sales employees, we examined the
We used standard hierarchical regression analysis to assess the relationships among PSS, POS, and turnover.
moderating effect of the supervisor’s perceived organizational
status on the PSS–POS relationship (see Table 4). To reduce the
potential collinearity between the interaction term and its compo- Method
nent variables, we followed Aiken and West’s (1991) recommen- The final sample consisted of 493 employees who worked for the same
dation to center the independent variables (PSS and supervisor’s organization investigated in Study 2. A total of 532 employees were
perceived organizational status) involved in the presumptive inter- administered the questionnaire under the same conditions as in the second
action. Tenure was entered in the first step as a predictor of POS; study; 306 of these employees were those surveyed in Study 2 plus an
PSS and supervisor’s organizational status were added in the additional 226 employees who received the same PSS and POS scales
second step; and the product term (i.e., PSS ⫻ Supervisor’s Or- given the employees in Study 2 but who did not receive the scale items
used in Study 2 to assess the supervisor’s organizational status. Five
ganizational Status) was entered in the third step to assess the
hundred twenty-one employees (98%) returned completed questionnaires.
interaction between these two variables. As shown in Table 4,
We used the entire available sample to maximize statistical power for
controlling for tenure, PSS and supervisor’s perceived organiza- assessing the relationships among PSS, POS, and turnover. Because we
tional status were reliably related to POS. Of more direct interest, were interested in voluntary employee turnover, we excluded from the
the interaction between PSS and supervisor’s organizational status analysis 28 of the respondents who were laid off during the 6 months after
was statistically significant. the survey.
To examine this interaction in more detail, regression lines Twenty-nine percent of the final sample were hourly paid sales–support
representing the relationship between PSS and POS were plotted, employees, 44% were hourly paid salespeople, 20% were salaried sales–
as shown in Figure 2, at high and low levels of supervisor’s support employees, and 7% were salaried salespeople. The average tenure
perceived organizational status (i.e., 1 SD above and below the of these employees was 60 months (SD ⫽ 49 months). Seventy-two percent
were men. We obtained voluntary employee turnover data from organiza-
mean; cf. Aiken & West, 1991; Cohen & Cohen, 1983). Simple
tional records 6 months after the survey administration.
effects tests (Aiken & West, 1991) indicated reliable positive
relationships between PSS and POS at both one standard deviation
above, t(295) ⫽ 7.56, p ⬍ .001, and one standard deviation below, Results and Discussion
t(295) ⫽ 6.66, p ⬍ .001, the mean supervisor’s organizational
Descriptive Statistics and Correlations
status score. Consistent with Hypothesis 2, the slope of the rela-
tionship between PSS and POS was greater when supervisors had Means, standard deviations, internal reliabilities, and intercor-
high perceived organizational status than when they had low relations among variables are displayed in Table 5. Thirteen per-
status, t(295) ⫽ 2.76, p ⬍ .01. cent of the employees voluntarily left the organization during the 6

Table 3
Study 2: Descriptive Statistics and Intercorrelations

Variable M SD 1 2 3 4

1. Tenure 55.50 50.09 — .04 .18** .21***

2. Perceived supervisor support 4.41 1.21 (.88) .47*** .60***
3. Supervisor’s perceived organizational status 3.90 1.15 (.89) .69***
4. Perceived organizational support 3.54 1.33 (.88)

Note. N ⫽ 300. Internal reliabilities (coefficient alphas) are given in parentheses on the diagonal.
** p ⬍ .01. *** p ⬍ .001.
570 EISENBERGER ET AL.

Table 4
Study 2: Hierarchical Regression Analysis for Perceived Organizational Support

Regression steps B SE B ␤ ⌬R2

Step 1 .04**
Tenure .01 .00 .21***
Step 2 .54***
Tenure .00 .00 .10**
Perceived supervisor support .39 .05 .36***
Supervisor’s organizational status .58 .05 .50***
Step 3 .01**
Tenure .00 .00 .09*
Perceived supervisor support (PSS) .44 .05 .40***
Supervisor’s organizational status .57 .05 .49***
PSS ⫻ Supervisor’s Organizational Status .09 .03 .11**

Note. Final model: F(4, 295) ⫽ 106.55, p ⬍ .001; total R2 ⫽ .59.

* p ⬍ .05. ** p ⬍ .01. *** p ⬍ .001.

months after the questionnaire administration. As predicted, PSS between PSS and employee turnover. Employees appear to infer
was positively related to POS and negatively related to turnover. POS from PSS based on their perception of their supervisors’
This pattern of correlations satisfies Kenny, Kashy, and Bolger’s status in the organization, leading to reduced turnover. With due
(1998, p. 260) first two conditions required to provide evidence regard to the limitations of correlational research for assessing
consistent with mediation, stating that the exogenous variable causality, the findings are consistent with PSS’s role in POS’s
(PSS) should be related to both the mediator (POS) and the development, as maintained by organizational support theory.
outcome variable (turnover). Consistent with the view that PSS leads to POS, we found that
PSS was positively related to the temporal change in POS. The
Mediating Role of POS results supplement prior findings involving the simultaneous as-
sessment of PSS and POS that left the association’s direction
Because turnover was a binary variable, we used hierarchical indeterminate. Such a finding of a positive cross-lagged relation-
logistic regression to provide evidence consistent with a mediating ship would occur if the interval between the assessments of PSS at
role of POS in the PSS–turnover relationship. As shown in Ta- Time 1 and POS at Time 2 were long enough for PSS to influence
ble 6, we entered tenure in the first step as a control variable, PSS POS but not so long that the effect would have substantially
in the second step, and, finally, POS in the third step. Because we decayed (Maruyama, 1998, pp. 115–117).
used logistic regression, statistical significance was assessed with
the Wald statistic, which approximates a Z2 distribution. The third
mediation requirement of Kenny et al. (1998), that POS should be
associated with turnover while controlling for PSS, was met (see
Step 3 in Table 6). Fulfilment of the fourth and final mediation
requirement, that the relationship between PSS and turnover be
reduced when POS is entered into the model, is suggested by the
nonsignificant relationship between PSS and turnover in the final
regression step. More precisely, using the Kenny et al. formula (p.
260), we found that the drop in the relationship of PSS with POS
from Step 2 to Step 3 was statistically significant, b ⫽ .32, SE ⫽
.08, Z ⫽ 3.94, p ⬍ .001. The presumptive mediation effect was
complete, with only a negligible relationship between PSS and
turnover remaining after controlling for POS.
The findings were consistent with POS’s mediation of a nega-
tive relationship between PSS and voluntary employee turnover.
This result follows from organizational support theory, holding
that PSS leads to POS, which in turn, reduces turnover by strength-
ening felt obligation toward the organization and affective orga-
nizational commitment.

General Discussion Figure 2. The relationship between perceived supervisor support (PSS)
and perceived organizational support (POS) as a function of the supervi-
The present findings provide a greater understanding of the sor’s perceived organizational status (Study 2). High status is indicated by
relationship between PSS and POS, including evidence concerning the top, bold line, and low status is indicated by the bottom line. High and
the causal direction, the mechanism underlying employees’ gen- low supervisor’s perceived organizational statuses are, respectively, 1 SD
eralization of PSS to POS, and POS’s role in the association above and 1 SD below the mean.
 PERCEIVED SUPERVISOR SUPPORT 571

Table 5
Study 3: Descriptive Statistics and Intercorrelations

Variable M SD 1 2 3 4

1. Tenure 60.22 48.83 — .00 .21*** ⫺.21***

2. Perceived supervisor support 4.34 1.24 (.90) .58*** ⫺.11*
3. Perceived organizational support 3.38 1.31 (.87) ⫺.24***
4. Turnover .13 .34 —

Note. N ⫽ 493. Internal reliabilities (coefficient alphas) are given in parentheses on the diagonal. For turnover,
stayers were coded as 0 and voluntary leavers as 1.
* p ⬍ .05. *** p ⬍ .001.

Although the relationship between PSS and temporal change in Study 2 found that the relationship between PSS and POS was
POS provides stronger evidence of causality than does simulta- greater for employees who perceived their supervisors to have high
neous measurement of PSS and POS, the observed association informal status within the organization. Organizations generally
could still be due to variables other than those for which we have a high regard for members who are reputed to embody
controlled (Finkel, 1995). The finding of a positive relationship favored characteristics. Therefore, perceived high standing of a
between PSS and change in POS is nevertheless consistent with supervisor within the organization would generally be seen by
organizational support theory, according to which supervisors, as employees as indicating the supervisor’s exemplification of the
representatives of the organization, contribute to POS. Our use of organization’s character. Perceived high supervisor status was
individual scale items as indicators of PSS and POS at both times found to involve beliefs concerning the organization’s positive
had the benefit of taking into account autocorrelated error vari- valuation of the supervisor’s contributions and its concern about
ances between successive measurements of the same latent vari- the supervisor’s well-being; the supervisor’s influence in impor-
ables over time (Maruyama, 1998, pp. 111–112). Future research tant organizational decisions; and the autonomy and authority
might include more distinctive indicators of PSS and of POS to accorded the supervisor in his or her job responsibilities.
better assess measurement error (Farkas & Tetrick, 1989). From the viewpoint of organizational support theory, the indi-
The failure to find a cross-lagged relationship between POS and vidual employee’s attribution of informal status to the supervisor,
temporal change in PSS is less definitive. We cannot assume that whether accurately reflecting the beliefs of others in the organiza-
the duration required for PSS to influence POS is the same as that tion or not, should moderate the relationship between PSS and
required for POS to influence PSS (Maruyama, 1998, p. 116). One POS. Employee perceptions of supervisor status would be based
might argue that the relationship should be bidirectional. To the on personal observation of upper management’s treatment of su-
extent that employees believe that supervisors are influenced by pervisors as well as the communicated views of upper manage-
the organization’s views, support for employees by the organiza- ment, supervisors, and fellow employees. Future research might
tion should increase PSS (cf. Yoon & Thye, 2000). The present examine the relative contributions of these sources of belief con-
research examined changes of POS and PSS over a 3-month cerning informal supervisor status.
interval, which may have been too short to observe a relationship PSS may also contribute to perceived supervisor status. Being
between POS and change in PSS. Future research might assess the viewed favorably by a supervisor who plays an important role in
possibility of such a longer-term relationship by examining PSS the organization may enhance fulfillment of socioemotional needs
and POS several times, using the same 3-month interval between and increase expectations of future favorable treatment from the
successive measurements as in our study. POS could then be organization. Therefore, employees who believe that their super-
related to changes in PSS over 3 months, 6 months, or longer. visor values their contributions may be motivated to view the

Table 6
Study 3: Hierarchical Regression Analysis of Turnover on Perceived Supervisor Support and
Perceived Organizational Support

Regression steps B SE B Wald

Step 1
Tenure ⫺.02 .00 20.18***
Step 2
Tenure ⫺.02 .00 20.69***
Perceived supervisor support ⫺.27 .11 6.59*
Step 3
Tenure ⫺.02 .01 16.31***
Perceived supervisor support .00 .12 0.01
Perceived organizational support ⫺.53 .13 16.18***

Note. Wald coefficient is Z2.

* p ⬍ .05. *** p ⬍ .001.
572 EISENBERGER ET AL.

supervisor as having an important organizational role. However, it References

should be noted that the relationship between PSS and perceived
Aiken, L. S., & West, S. G. (1991). Multiple regression: Testing and
supervisor status was moderate, indicating that some employees
interpreting interactions. Thousand Oaks, CA: Sage.
were reporting high supervisor support even when supervisors Allen, M. W. (1995). Communication concepts related to perceived orga-
were not perceived to have high status. Perceived supervisor nizational support. Western Journal of Communication, 59, 326 –346.
status, whatever its sources, enhanced the relationship between Aquino, K., & Griffeth, R. W. (1999). An exploration of the antecedents
PSS and POS. Support from a supervisor who is perceived to and consequences of perceived organizational support: A longitudinal
strongly embody the organizational ethos is more likely to be taken study. Unpublished manuscript, University of Delaware, Newark.
as organizational support than is support from a supervisor whom, Brislin, R. W. (1980). Translation and content analysis of oral and written
material. In H. C. Triandis & J. W. Berry (Eds.), Handbook of cross-
the employee believes, less well represents the organization.
cultural psychology (pp. 398 – 444). Boston: Allyn & Bacon.
Study 3 found evidence consistent with POS’s mediation of a Cohen, J., & Cohen, P. (1983). Applied multiple regression/correlation
negative PSS–voluntary turnover relationship. Employees who analysis for the behavioral sciences (2nd ed.). Hillsdale, NJ: Erlbaum.
believed that the supervisor valued their contributions and cared Cropanzano, R., Howes, J. C., Grandey, A. A., & Toth, P. (1997). The
about their well-being showed increased POS, which in turn relationship of organizational politics and support to work behaviors,
was related to decreased turnover. This finding is consistent attitudes, and stress. Journal of Organizational Behavior, 22, 159 –180.
with organizational support theory, which holds that beneficial Eisenberger, R., Armeli, S., Rexwinkel, B., Lynch, P. D., & Rhoades, L.
(2001). Reciprocation of perceived organizational support. Journal of
treatment received from supervisors should increase POS, lead-
Applied Psychology, 86, 42–51.
ing to felt obligation to aid the organization and to affective Eisenberger, R., Cummings, J., Armeli, S., & Lynch, P. (1997). Perceived
organizational commitment, both of which should reduce turn- organizational support, discretionary treatment, and job satisfaction.
over (Eisenberger et al., 2001; Rhoades et al., 2001). The Journal of Applied Psychology, 82, 812– 820.
presumptive mediation effect was complete in that PSS did not Eisenberger, R., Fasolo, P., & Davis-LaMastro, V. (1990). Perceived
have a statistically significant relationship with turnover be- organizational support and employee diligence, commitment, and inno-
yond the mediational role played by POS. Employees with low vation. Journal of Applied Psychology, 75, 51–59.
Eisenberger, R., Huntington, R., Hutchison, S., & Sowa, D. (1986). Per-
POS may believe that their prospective success in the organi-
ceived organizational support. Journal of Applied Psychology, 71, 500 –
zation is greatly limited. Thus, a reduction of POS, resulting 507.
from low PSS or other sources, may increase employees’ like- Farkas, A. J., & Tetrick, L. E. (1989). A three-wave longitudinal analysis
lihood of quitting the organization. of the causal ordering of satisfaction and commitment on turnover
In all three studies, PSS was higher on average than POS. decisions. Journal of Applied Psychology, 74, 855– 868.
Because supervisors have greater daily contact with most employ- Finkel, S. E. (1995). Causal analysis with panel data. Thousand Oaks, CA:
ees than do upper level managers, they may be able to more readily Sage.
convey positive valuations and caring. To foster personal loyalty, Hutchison, S. (1997a). A path model of perceived organizational support.
Journal of Social Behavior and Personality, 12, 159 –174.
many supervisors may exaggerate their positive valuation of their
Hutchison, S. (1997b). Perceived organizational support: Further evidence
subordinates and their own role in obtaining benefits for subordi- of construct validity. Educational and Psychological Measurement, 57,
nates, resulting in greater PSS than POS. Future research might 1025–1034.
examine how supervisor self-presentational behaviors influence Hutchison, S., Valentino, K. E., & Kirkner, S. L. (1998) What works for
PSS and POS. the gander does not work as well for the goose: The effects of leader
Study 1 of the temporal relationships between PSS and POS behavior. Journal of Applied Social Psychology, 28, 171–182.
involved Belgian university graduates, whereas Studies 2 and 3, Jöreskog, K. G., & Sörbom, D. (1993). LISREL 8: Structural equation
modeling with the SIMPLIS command language. Chicago: Scientific
concerning the moderating effect of perceived supervisor organi-
Software International.
zational status on the PSS–POS relationship and the mediating Kenny, D. A., Kashy, D. A., & Bolger, N. (1998). Data analysis in social
influence of POS on the PSS–turnover association, involved U.S. psychology. In D. T. Gilbert, S. Fiske, & G. Lindzey (Eds.), The
retail sales employees. The confirmation of organizational support handbook of social psychology (pp. 233–265). New York: McGraw-Hill.
theory’s predictions with samples of differing nationality and job Kottke, J. L., & Sharafinski, C. E. (1988). Measuring perceived supervi-
type adds to the generality of the findings. Future studies might sory and organizational support. Educational and Psychological Mea-
assess organization-level and national-level differences in PSS– surement, 48, 1075–1079.
Levinson, H. (1965). Reciprocation: The relationship between man and
POS relationships. For example, because small organizations usu-
organization. Administrative Science Quarterly, 9, 370 –390.
ally have fewer levels of hierarchy than large organizations, em- Malatesta, R. M. (1995). Understanding the dynamics of organizational
ployees in small organizations might generally identify their and supervisory commitment using a social exchange framework. Un-
supervisors more with the organization’s basic character than those published doctoral dissertation, Wayne State University, Detroit, MI.
in large organizations, resulting in stronger PSS–POS relation- Maruyama, G. M. (1998). Basics of structural equation modeling. Thou-
ships. Also, in some collectivistic nations, employees may prefer sand Oaks, CA: Sage.
to view themselves as part of a highly integrated hierarchy of Randall, M. L., Cropanzano, R., Bormann, C. A., & Birjulin, A. (1999).
Organizational politics and organizational support as predictors of work
organizational authority (termed vertical collectivism by Triandis
attitudes, job performance, and organizational citizenship behavior.
& Gelfand, 1998). Such identification could increase employees’ Journal of Organizational Behavior, 20, 159 –174.
perception that supervisors embody the organization’s character, Rhoades, L., & Eisenberger, R. (in press). Perceived organizational sup-
resulting in a greater PSS–POS relationship than in individualistic port: A review of the literature. Journal of Applied Psychology.
cultures. Rhoades, L., Eisenberger, R., & Armeli, S. (2001). Affective commitment
 PERCEIVED SUPERVISOR SUPPORT 573

to the organization: The contribution of perceived organizational sup- Vandenberghe, C., & Peiro, J. M. (1999). Organizational and individual
port. Journal of Applied Psychology, 86, 825– 836. values: Their main and combined effects on work attitudes and percep-
Settoon, R. P., Bennett, N., & Liden, R. C. (1996). Social exchange in tions. European Journal of Work and Organizational Psychology, 4,
organizations: Perceived organizational support, leader-member ex- 569 –581.
change, and employee reciprocity. Journal of Applied Psychology, 81, Wayne, S. J., Shore, L. M., & Liden, R. C. (1997). Perceived organiza-
219 –227. tional support and leader-member exchange: A social exchange perspec-
Shore, L. M., & Shore, T. H. (1995). Perceived organizational support and tive. Academy of Management Journal, 40, 82–111.
organizational justice. In R. S. Cropanzano & K. M. Kacmar (Eds.), Yoon, J., Han, N-C., & Seo, Y-J. (1996). Sense of control among hospital
Organizational politics, justice, and support: Managing the social cli- employees: An assessment of choice process, empowerment, and buff-
mate of the workplace (pp. 149 –164). Westport, CT: Quorum. ering hypotheses. Journal of Applied Social Psychology, 26, 686 –716.
Shore, L. M., & Tetrick, L. E. (1991). A construct validity study of the Yoon, J., & Lim, J. C. (1999). Organizational support in the workplace:
survey of perceived organizational support. Journal of Applied Psychol- The case of Korean hospital employees. Human Relations, 52, 923–945.
ogy, 76, 637– 643. Yoon, J., & Thye, S. (2000). Supervisor support in the work place:
Shore, L. M., & Wayne, S. J. (1993). Commitment and employee behavior: Legitimacy and positive affectivity. Journal of Social Psychology, 140,
Comparison of affective commitment and continuance commitment with 295–316.
perceived organizational support. Journal of Applied Psychology, 78,
774 –780.
Triandis, H. C., & Gelfand, M. J. (1998). Converging measurement of Received February 27, 2001
horizontal and vertical individualism and collectivism. Journal of Per- Revision received October 18, 2001
sonality and Social Psychology, 74, 118 –128. Accepted October 24, 2001 䡲

View publication stats