Individual DiJ/erences in a Model ot
Organizational Role Stress Robert H. Miles, Yale University
Role perceptions of conflict and ambiguity have received considerable attention by contemporary organizational researchers. Previous empirical studies have demonstrated the aversiveness of experienced role conflict and role ambiguity in terms of both indi~J~dual and organizational work outcomes. Now the focus of research is beginning to shift toward the identification of both organizational antecedents of ex. perienced role stress and individual sensitivity to stressful work conditions. ']he objective of the present investigation is to test the possible moderating effects of individual differences in a model of organiza. tional role stress. First, linkages between a potential objective stressor and experienced role stress, in the forms of role conflict and role ambiguity, are determined; then, these experienced stresses are related to vario~ls individual work outcomes; finally, the extent to which individual differences moderate both sets of these relationships is evaluated.
Theoretical Model The theoretical model of organizational role stress used in this study is presented in Figure l. It depicts linkages between (a) objective, or potential, and perceived role stress and, (b) between experienced stress and individual work outcomes. It also assumes that individual differences in self.perceptions moderate the linkage in the organizational, stress model. Potential Stressor---Perceived Role Stress The potential role stressor selected for use in this mode] is boundary relevance, or the extent to which members engage in boundary-spanning activities with persons outside their immediate organizational unit. Adams
Volume 4, Number 2
May, 1976
88
Journal o~ Business Research
F ~ r e 1: Theoretical Model THEORETICAL MODEL
¢ OBJECTIVE STRESSORS
WORK-RELATFD OUTCOMES
Boundaw-spanniag role. requimmen~,.s
i
I
Role Cor~lict
Job-induced tenst¢.n and anxiety
I
Degreeof experienced Rotegmblguity
Job satisfaction
II
It
"l
Interpersonal relaUons
INI),WIDUALDIFFERENCES IN SELF-PERCEPTIONS Supervisory ability Occupstional achievementmotivation Self-assurance
[1] has argued that persons occupyin~ organizational boundary rol~ are subject to conflicting demands from persons both inside ~and out~ide the organization. Persons occupying boundary roles are expeeted to experience higher levels of role conflict, especially of the inter~nder variety, than counterparts occupying "internal" organizational roles because they receive and initiate influence attempts. Furthermole, while experienced role conflict and boundary relevance are expected to be positively correlated, the strength of this relationship should be directly proportional to the amount of differentiation between the subunits or organizations spanned by boundary role per. sons [15]. For example, both Thompson ['23] and Corwin ['3] have argued that persons engaged in boundary.spanning activities must often maintain a delicate system of linkages across differential systems or subsystems and that this linking function is a major source oi strain and conflict in complex organ~ations. Some evidence for this re. ]afionship between stressor and perceived role stress has been reported by Kalm, et al. [13]. Kahn, et al. found a direct relationship between degree of boundary relevance and experiene~ role conflict. They also reported that freqmmey of interdepartmental contacts was only slightly less related to experienced conflict than the frequency of interorganizational contacts. The relationship between experienced role ambiguity and bound. ary relevance was not reported by Kalm and his associates, though both me,mutes were included in their Study.
Individual Di/]erertces and 013anizational Role Stress
89
Thus, there is some evidence suggesting that the extent of bound. ary-spanning role requirements is an objective organizational condition possessing high role conflict potential. In addition to testing this assumption, and investigating the possible moderating effects of individual differences, this study includes an exploratory study of boundary relevance as a possible source of experienced role ambiguity.
Role Perceptions--Work-Related Outcomes Considerable documentation exists to support the construct validity of the aversiveness of perceived role stress° Across a variety of samples and measures, role perceptions of both conflict and ambiguity have been lound to be unfaw~rably related to work outcomes of job-induced tension and anxiety, job satisfaction~ attitudes toward role senders, perceived effectiveness, job performance ratings, and voluntary turnover [4, 7, 8, 9, ]1, ]3, ]6, 18, 19, 22]. Thus, it is expected that both expel-fenced role conflict and experienced role ambiguity are unfavorably re]~ated to individual work outcomes of job-induced tension and anxiety, job satisfaction, and interpersonal relations. Individual Differences in Self-perceptions There is evidence suggesting that not all individuals perceive objective conditions in the same way or cope with experienced role stress wS~ththe same degree of effectiveness. Thus, individual differences in sensitivity to potentially stressful role conditions, and in coping ability u~nder experienced role stress, may substantially alter the later causal linkages specified in models of organizational role stress. Knowled~;e of these individual differences is a prerequisite for developing mechanisms for managing stress and for facilitating person-role fit in complex organizations. Kahn, et al []3] probed for answers to questions regarding the relationships between personality and role perceptions. They were concerned about personality differences in sensitivity to events in a person's objective environment and in his coping effectiveness under conditions of experienced stress. They found that differences in introversion, flexibility, need for clarity, and need for occupational achievement moderated the relationship between objective conflict and experSenced strain, but they did not systematically investigate the differences in these moderating e~fects across all three relationships shown in Figure ]L. Other researchers have reported individual difference moderators but they have tested only one linkage of the model in each study. Lyons [16] demonstrated that need for clarity moderated the relationship between experienced role clarity and voluntary turnover, but he did nw. report effects of experienced role conflict or the linkage between objective conditions and experienced stress. Johnson and Stinson [12] reported that both need for independence and need for
90
Journal o/Business Research
achievement moderate relationships between role perceptions and job satis:faetion, but thc.y did not report the moderator effects oZ" these i~:l. dividual differences on relationships between objective conditions al~ld experienced stress. Thus, there have Leen several studies which examined individual differences in one of the major linkages in the model depicted in Figure 1. As an extension of previous studies, the present study attempts to determine if selected individual differences (supervisory ability, need for occupational achievement, and self-assurance) distinguish between either sensitivity to objective condition~ or between coping effectiveness under conditions of experienced stress~ or both.
Method One hundred and forty.two persons, from a sample of 202 who participated in an earlier study [17, 19] attended a follow-up survey administration w|:~ich included variables used in the present study. The original subjects represented a voluntary participation rate of 96 percent of a sample of persons occupying five, distinct~ professional. level roles (GS-]I to GS.16) in nine governmental research and deve!opment organizations. All persons in the role of integrator, division manager, and group leader were invited to participate in the survey, with an equivalent-sized, random sample of nonsupervisory scientists and engineers. Thi~ type of sample was selected to ensure an adequate range on the objective or "potential" source of role conflict: the extent of boundary-spanning activities, ~;hich was presumed to be arrayed in decreasing order of intensity across the roles as listed above. In addition to the role requirement, role perception, and work-related outcome variables measured in the original survey, the follow-up survey included measures of individual differences in self-perceptions. The participation rate in the follow-up survey was 73 percent, and the relative distribution of the roles remained essentially the same. The researcher administered the survey to participants in conference rooms located near their places of work. The potential role stressor in this investigation is the degree of boundary relevance of the role occupied by a focal person. Boundary relevance was inferred from the extent of integration and boundary. spanning activities required of the focal person in the performance of his/her role. This measure was derived from a longer list oi job activities using a factor analytic procedure. The procedure, as well as construct validation evidence for this measure of boundary relevance, was provided in an earlier study []8]. As predicted this measure was found to be arrayed in ascending order of importance (role means ~ parentheses) across the five organi-
Individual DiHerences and Orga~Lizational Role Stress
91
zational roles sampled: from basic scientist (2.15), applied scientist (2.38), gloup leader (2.95), division manager (3.37), to integrator (3.45). Furtixermore, the overall difference in means acro!~s these roles was significant (F - - 21.28, p < 0.001) and the estimate of the statistical relation between organizational role and extent of integration and boundary-spanning activities in the one.way analysis of variance was strong enough to support the practical significance of these differences (t, 2 ~ 28 percent). Eleven items were included in the boundary relevance measure. Focal persons were asked to rate the extent to which these items were important in the performance of their assigned roles, with response alternatives ranging on a five-point scale from "definitely not part of my job" to "a most significant part of my job." Four representattive items were: (a) "Represent your directorate/laboratory to outsiders;" (b) "Travel at least 30 days per year as a represent/ttive of your organization;" (c) "Integrate or coordinate interdependent activities of others;" and (d) "Request performance obligations from persons outside your immediate unit. ''1 The role perceptions investigated in this study are role conflict, de:fined as the degree of incongruity or incompatibility of expectations associated with a role, and role ambiguity, defined as the lack of clarity of role expectations and the degree of uncertainty regarding the outcomes of role performance. These definitions are consistent with those of previous researchers []0, 13, 22]. These concepts were measured by scales developed by Rizzo, et al., [22] which were chosen because of their demonstrated factorial independence. The work.related outcomes investigated in this study are jobinduced tension and anxiety, job satisfaction and interpersonal relations. The job-induced tension and anxiety scales were designed to measure the existence of tensions expcrienced by focal persons on the job. A specific objective in selecting these scales was to avoid the construction of tension scales by rephrasing conflict and ambiguity items. This was a weakness of a previous study by Kahn, et al. [13] in which one measure of role arabiguity was, "Do you feel you are always as clear as you would like to be about what you have to do this job?" while a related tension item asked respondents to rate how ohen they felt bothered by "being unclear on just what the scope and responsibilities of your job are." The tension measures used in this study were those developed by Rizzo, et al. [22] to minimize the bias rchich could result from Kahn's item selection. Two sample items from the Rizzo, et al. scale are: "Problems associated with my job have kept me awake at night;" and "I often take my job home with me in the sense that I think about it when doing other things." The Bullock [2] job satisfaction scale was chosen for use in this
92
Journal of Business Research
study. In addition to having been used in previous studies of role stre,.~s, this scale has e:v.hibited published test-retest and split-half re!~abilities in excess of .90. Interpersonal relations or attitudes toward role senders [13] are composed of three attitudes of the focal person toward his important role senders, including h-'ust, respect, and liking for each sender. In the present study, the focal person was asked to list his important role senders up to a limit c,f 10 senders and to score his attitudes toward each. These scores were averaged for each sender and then combined into an unweighted role-set average. In order to reduce social desirabilisy, respondents were asked to destroy the list of role sender names a~ter they had co~mpleted the role-set mapping form. All objective role stressor, role perception, and work outcomes scales were scored using a five-point response mode ranging from 1 to 5. The Spearman-Brown reliability estimates for ~ese scales, exceeded .80 [17]. The Ghiselli Self-Report Inventory [5, 6] was used to measure self. perceptions of the participants. It is a forced-choice, adjective checklist consisting of 64, pairs of adjectives. The '~cale was chosen to minimize the effects cf transparency and nonindependenee threats. Both adjectives in each pair referred to traits approximately equal in ~ i a l desirability. For the first 32 pairs, the respondent chooses the one he believes most describes him, and for the last 32 he chooses the one he believes least describes him. Of the thirteen self-perception scales included in the Inventory, Ghiselli dscribes three as abilities, four as personality traits, and five as motives or need categories. The best predictor of managerial success in each of these three categories, as demonstrated in validation studies [5], was chosen for use in this study. They were the supervisory ability, self-assurance, and need for occupational achievement. According to Ghiselili, these three self-perceptions measure the following: (1) Supervisory Ability: " . . . the capacity to direct the work of others and to organize and integrate their activities so that the goal of the work can be attained [5:39]." (2) Need [or Occupational Achievement: the need to " . . . seek the responsibility and the prestige which is associated with high position [5 :~I~3]." (3) Self.Assurance: " . . . the extent to which the individual pereeives himself to be ~ffective in dealing with the problems that confront him [5:57]." The data were analyze, t by dichctomizing subjects on the basis of sdbperception sample r leans. CorIelation pairs for high and low individual.d~fference groups were tested using Fisher's one-tailed z-test
93
Individual DiHerences and Organizational Role Stress
for all directional predictions, In addition, both zero-order and partial correlations analyses were employed. Results
The correlations in Table ] demonstrate the independence of the Ghiselli Self-Report Inventory measures from the other variables. Selfperceptions are not correlated with the work outcome variables and are generally not related at a practical ]eve] to either boundary. spanning role requirements or role perceptions. This intercorre]ation pattern confirms the assumption that the design o~ the Ghise]li Self. Report Inventory is useful in avoiding confounding due to noninde. pendence of self-perceptions from either predictor or outcome variables, or to transparency which could cause faking in response to social desirability. Finally, while the three self-perceptions were intercorrelated, they were used separately in the following analyses rather than as a combined single index, because :their validities as predictors of managerial success were based on separate studies using different criteria [5]; and, there is suHicient variance unaccounted for by their over|aps to permit them to differentially moderate the important relationships in this study. ~Table 1: Intercorrelations Among Role Requirement, R01e Perception, Work-Related Outcome, ,,and Self-Perception Variables i
i
1.
2.
i
3.
4.
5.
6.
7.
8.
Mean
SD
2.75
0.85
2.79
0.80
2.86
0.96
2.19
0.69
3.83
0.66
3.78
0.47
27.07
6.20
36.37
8.77
26.10
5,39
Role Re~uirementp: Boundary-spannlng
Perceptions:
Role
Role c o n f l i c t
Role ambiguity
.36 a -.08
.11
-
N o ~ - r e l a t e d Outcomes: Job-lnduced tension
and anxiety
.02
.31 a
.16
-
Job s a t i s f a c t i o n
.26 b
-.28 a
-.39 a
-.26 b
Interpersonal r e l a t i o n s
.17 c
-.30 a
-.]5
-.I0
.06
.13
-.06
.15
.19 c
.17 c
.14
.47 a
Self:Perceptlgns: Supervisory ability
.11
.09
.09
-.02
.08
.04
.04
.42 a
-.08
-.16
.03
-.01
.35 a
Need for occupational
achievement Self-assurance
ap b
~ 0.001
p < 0.01
Cp Note:
I
.53 a
< 0.05
N - 142
III
m
I
P
94
Journal of Business Research
Potential Stressor---Perceived Role Stress As shown in Table 1, boundary-spanning role requirements were directly related to the level of perceived role conflict as predicted. This moderatel) s2rong correlation p~co~des some corustruct validity for the conflict potential of the obits:tire measure of boundary relevance. Boundary-spanning activities wer~ not related to the level of experienced role ambiguity, which is not surprising for two reasons. Boundary.spanning activities were selected for use in this study because of their conflict potential, as opposed to possible ambiguity. More importantly, "the present sample, which consists of various professional research roles, represents a resltriction in range in role ambiguity. The moderator effects of individual differences in self-perceptions are shown in 'Fable 2. The relationships between potential antecedents of stress attd the level of stress actually experienced by role incumbents appears to be relatively unresponsive to individual differences in self-perceptions. Only need for occupational achievement significantly moderates the relationship between boundary-spanning and experienced role conflict, and this moderator effect is opposite in direction from the prediction. Persons high in need for occupational achievement reported stronger direct relationships between these variables, indicating increased sensitivity to potentially conflicting, objective role requirements, as measured by the extent of boundarysparming activities. On the other hand, boundary-spanning activities were not correlated with perceived role ambiguity in the present sample and none of the self-perceptions appeared to significantly alte~ this relationship.
Role Perceptions---Work-Related Outcomes As shovm in Table 1, all of the relationships between role perceptions and work-related outcomes are in the predicted direction. Role conflict is positively related to job-inc~uced tension and anxiety and negatively related to both job satisfaction and interpersonal relations. Experienced role ambiguity is inversely related to job satisfaction and the role am. biguity correlations with tension and interpersonal relations, though not significant, are in the predicted direction. Thus, role perceptions of conflict and ambiguity appear to be adversely associated with work oL~tcomes, as could be expected from previous research on role stress.
The~e relationships between role perceptions and work-related outcomes (Figure 1) appear to be more sensitive to individual differences than the relationships between objective or potential sources of ~tress and expe~rienced role conflict and ambiguity. The rather substantial impact of individual differences in self-perceptions on these relationships may be seen in Table 3. Nine of the relationships
95
Individual Di//erences and Organizational Role Stress
Table 2 : Correlaiions, and Differences in Correlations, Between Boundary-spanning as a Potential Stressor and Perceived Role Stress for Dichotomous Svlf-Perception Groups
Correlations ..........
Supervisory Ability ~igh Low (n=,65) (n=77~
Need f o r O c c u p a ~ i o n a l Achievement High Low (n-68~ (n-76)
SelfAssurance High Low (n~64) .(nyT8)
Boundary-spanntn~ Role " RequlremenCswtCh: D~gree of e x p e r i e n c e d .role conflict
.40 a
.32 b
.51 b
.16 c
.35 b
.35 a
Degree of e x p e r i e n c e d role ambiguity a
b C
.07
-.19
-,,11
-.01
-.06
-.07
p < 0.001 ~, < 0 . 0 1
p < 0.05
No_9_~: D i f f e r e n c e s In c o r r e l a t l o n s f o r dichoLomous groups were e v a l u a t e d by t r a n s f o r m i n g the c o e f f i c t e n t s i n t o z - c o e f f i c i e n t s and p e r f o r m i n g the u s u a l o n e - t a i l e d t e s t f o r d i r e c t i o n a l h y p o t h e s e s .
between role perceptions and outcomes appear to be moderated by individual differences in selLpreceptions, and the moderator effects of all three selLperceptions appear to be operative and in the pre. dicted direction. Persons with low occupational achieveme~,t needs report significantly more job-induced tension and anxiety, and significantly less job satisfaction and less favorable interpersonal relations, in response to role conflict than persons with relatively high needs fo:- occupational achievement. These findings suggest that persons with high achievement needs may be better suited for roles which cause them to experience role conflict. This generalization for occupational achievement motivation does not appear, however, to apply to role ambiguity. In fact, using a two.tailed z.test, need for occupational achievement significantly moderates the ambiguity iiob satisfaction correlations in a direction opposite to that observed between role conflict and job satisfaction. This evidence suggests that the ef£vcts of self-perceptions may be selective. Persons with high needs for occupational achievement may be better able to cope with role conf|icts but may be more "~rustrated in their atgempt to satisfy these achievement needs under ambiguous role conditions. In contrast to these results for need for occl~pafiona] achievement, the Ghiselli measure of supervisory ability appears to discriminate mainly between the three work outcome correlates of experienced role ambiguity. As shown in Table 3, persons with low supervisory ability appear to cope less effectively with the experience of ambiguilly on the job. This demonstrates significantly less favorable correlations
96
Journal o/Business ~esearch H
iii
mill
i
ii
ii
Table 3: Differences in Correlation Between Bole Perceptions and Work.Related Outcomes for Dichotomo.s Sell-perception Groups
Correlated variables~
Supervisory Ability High Low (n=65) (n=77)
Need f o r Occupational Achievement High Low . (n=68) (n=74~
SelfAssurance High Low (n=64) (n=78)
Role C o n f l i c t with: Job-re]~ed
tension
.19
.39 a
.15
.44 a
.20
.41 a
Job sar s ~ a c t i o n
-.18
-.36 a
-.18
-.40 a
-.19
-.35 b
Interpersonal relations
-.23
-.36 a
-.06
-.45 a
-.19
-.39 a
.00
.27 c
Role A m b i g u ~ ~ : Joh-rela~ed tension Job s a t i s f a c t i o n Interpersonal relations a b c
-.26 c .08
.13
.19
.20
.12
-.48 a
-.50 a
-.30 b
-.50 a
-.32 b
-.20
-.13
-.18
-.27 c
-.07
p < 0.001 p < 0.01 p < 0.05
dA t w o - t a i l e d t e s t o f t h e moderator e f f e c t proaches e~.gnificance (.05 < p < .10).
o f need f o r o c c u p a t i o n a l
achievement ap-
Note: D i f ~ e r e n c e ~ i n c o r r e l a t i o n s f o r d i c h o t o m o u s groups were e v a l u a t e d by t r a n s f o r m i n g t h e c o e f f i c . ~ , e n t s ~nto z - c o e f f i c i e n t s and p e r f o r m i n g t h e u s u a l o n e - t a i l e d t e s t f o r d i r e c tional hypotheses.
between experienced role ambiguity and job outcomes of tension and anxiety, satisfaction, and interpersonal relations. The responses re. call stronger positive relationships between experienced role conflict and job-induced tension and anxiety. Also, the moderator effect was in the appropriate direction, though not significant, for the other two sets of conflict-outcome corre]ations. Finally, two of the role conflict-outcome relationships were altered in the predicted direction by self-assurance, suggesting that persons with low selLassurance may experience less favorable interpersonal relations and higher tension and anxiety in response to role conflict than persons with relatively high self-assurance. Collectively, the results in Table 3 suggest that the relationship between role perceptions of conflict and ambiguity with work-related outcomes are sensitive to differences in the self-perceptions of role incumbents, and that the moderath~g effects of the three self-perceptions used in this study are selective with respect to the nature of role stress actually experienced. Bom~dary Relevance---Work.Belated Outcomes It is also in. terestfng to look at the moderating effects of self-perceptions between
Individual Di#erences and Organizational Role Stress
97
potential or objective stressor and work-related outcomes. This can be accomplished in two ways: ignoring rnle perceptions of conflict and ambiguity, and controlling for role perceptions of conflict and ambiguity. The dirc,~t relationships between role requirement stressors and work.related outcomes, ignoring perceptions, are shown in the zero. order ~orre~ ~;,)ns in Table 4. These zero-order correlations demon. strate that ".,,undary.~tnning activity ~s positively related to both job ~atisfac~i,m and interpersonal re]atio~s. It appears that role requirements J~wolving intra, or inter.org~:~lizational linking requirements may provide favorable outcomes Io their incumbents, even though they may cause them to experience higher levels of role conflict or role ambiguity. This somewhat counter-intuitive finding is supported by previous research of Keller and Holland [14] and Orga,a [20, 21] which will be reported in the discussion section. Thus, there appear to be benefits of boundary relevance which may at least partially offset the role stress it arouses in individuals. Fur. Partial Correlations'Between Boundaryspanning as a Stressor and Work-Related Outcomes
T'able 4: Zero-Order i
i
Job-induced tension and
i
and i
i
i
i
Job satisfaction
•
i
,i
,m~:_
Interpersonal relatlons
__ a n x i e t y 02.a
.26 £
- ~1 0 D .03 c - . 08 °
.41 e .25 e
.17 g .32 e .16 g .31 e
.40 e
azero-order c o r r e l a t l o n between p o t e n t i a l s t r e s s o r spanning r o l e requirements) and w o r k - r e l a t e d outcome.
(boundary-
bFlrst-order partlal correlatlon, holdlng role eonfllct constant.
CFtrst-order p a r t i a l c o r r e l a t i o n , stant.
holding r o l e ambiguity con-
dsecond-order p a r t i a l c o r r e l a t i o n , and role ambiguity constant.
holding both r o l e c o n f l i c t
e
f
p < 0.001
p < 0.01
gp < 0 . 0 5
Note:
N = 142 III
II I
I
98
Journal o/ Business Research
thermore, the zero-order correlation pairs~ shown in the first row of each column in Table 5, indicate that these positive relationships be. tween boundary relevance and ,cork outcomes, ignoring role percep. tions, are not altered by individual differences in self-perceptions. The only exception is the significant moderator effect of supervisory ability on the interpersonal relations outcome. The partial correlations in Table 4, indicate the strength and direc. tion of relationships between boundary relevance and work outcomes, when controlling for role perceptions ot~ conflict and ambiguity. The pattern of correlations which emerges indicates that role conflict, not role ambiguity, appears to be an important intervening suppressor variable in the model. When the level of experienced role conflict is controlled, the relationships between boundary :relevance and work outcomes become even more favorable. On the other hand, a pattern of no change in zero-order and partial correlations emerges when experienced role ambiguity is controlled. Individual differences in these relationships, as shown in comparisons between partial correlations in Table 5, occur only for supervisory ability. In the case of this selLperception, the correlations between boundary relevance ~tnd both interpersonal relations and job satisfaction are significantly stronger for the high supervisory ability group, especially whell both role conflict and role ambiguity are controlled. Under these conditions, the partial correlations for job sati,,,faction and interpersonal relations are quite large. D~,cussion
The finding that boundary relevance is not a totally aversive state has been suggested by some previous research findings. While the present study found boundary relevance to be directly related to role perceptions of covflict, it also revealed positive relations between thi:~ antecedent condition and work outcomes, especia]ly when degree of experienced ro~e conflict was controlled. The w~rk of Organ [20, 21.] sugge~ that ~ e degree of visibility to role senders of the bound. ary rolo occupant's job behaviors may affect his/her role outcomes, with low ,risibility providing greater beilavi,~ral freedom and, con. seqaently, more job enjoyment and flexibility in dealing with po. tentially d:fficuh others. The boundary role may a]sc~ be viewed by iL~ occupants as being instrumental in enhancing career progress and status and in affording access to the gratificati~,n of power and mastery needs which are less accessible to persons confined to in. ternal organizat~ona| roles. In fact, the boundary role may be viewed as a reward--a l~sition of trust--for demonstrated performance effectiveness.
Individual Di//erences and Organizational Role Stress
99
Table 5: zero-Order and Pa'rtia] Correlations, and Differences 'in Correlations, Between Boundary-Spanning as a Potentia]t Stressor and Work-Related Outcomes for Dichotomous SelLperception Groups Selfperceptions . . . . .
Role requirements
Supervisory ability
Boundary-spannlng a b c d
Role requirements
Achievement motivation
Job-Induced t e n s i o n and anxiety High Low n=65 n=77
-.08 -.is -.08 -.z5
.07 -.06
Hlgh n=68
Low nffi74
Boundary-spanning a
.13
-.00
Job satisfaction
Interpersonal relations
.....
Hlgh n=65
Low n=77
.40 b .577 .45 b
.15 .30c :07 _ 6 " r ' - - _2"~
High n=78
Hlgh n=65
.33.c _si~ _:'3~_~.
.05a .18 .01
.52
"~"1'5"
Low nffi74
Hlgh nffi68
Low n=74
h
.04
.10
-.06
.28 d
.4S°.
,31~.
.22
.02
c
.11
-.06
.27~
.22 °
.12
d
.06
-.14
.48 b
.31 c
High n-64
Low n=78
.00 -.07 -.01 -.06
.05 --.09 .06 -.09
Role requirements Boundary-spanning a b c d
SelfAssurance
~ D t f f e r e n c e s in c o r r e l a t i o n a l
-.06
High n-64
.26 d .36 c .27 d .33 c
Low n=77
Low nffi78
.27~ .47~ .269 .46 n
27,d
.23 a
21 d
High nffi64
.34 c
Low n=78
.25 d .34! .25 a .32 c
p a i r s were e v a l u a t e d u s i n g F t s h e r ' s
.23
.34~
.12 .31 c .11 .30 c
two-tailed
z-.,~::, ~t. b
c d
p < 0.001 p < 0.01
p < 0.05
Note:
a = zero-order c o r r e l a t i o n s ; b = f i r s t - o r d e r p a r t i a l c o r r e l a t i o n , holding r-'ol-e c o n f l i c t constant; c ffi f i r s t - o r d e r p~rtial, holding role ambiguity constant; d = second-order p a r t i a l c o r r e l a t i o n , ho~,dtng b o t h r o l e c o n f l i c t a n d r o l e ambi-
guity c o n s t a n t .
It is also likely that £ature research will reveal that boundary relevance is multi-dimensional in nature. The construct may ininclude activities which vary considerably in their stress, especially conflict potential. These component actiivities may include representational, transactional, and integrative zole requirements, as well as buffering, mediative, technological gatekeeping, and information transfer and filtration activities. The result of a study by Keller and Holland [14], conducted in a single research and development or-
100
Journal o! Business Research
ganization, may illustrate this point. In contrast to the present results, Keller and Holland [14] did not find a significant correlation between boundary relevance and degree of experienced role conflict. However their measure of boundary relevance differed substantially from the one used in this study. It employed a four-item scale, borrowed from work on technological gatekeeping and technological transfer processes. Alternatively, :he present measure was based on specific job activities involving transactional, coordinative, and representational role responsibilities. The four items ,~ncluded in the Keller and Holland measure of boundary relevance were (a) "Indicate the total number of magazines, journals, and newspapers which you regularly read;" (b) "During the last month, what was the frequency with which you recommended ~specific ~information sources (e.g., journal article, knowledgeable colleague, report, magazine, etc.) to a colleague?" (c) "During the last month, indicate the frequency with which you have sought infor. mation/adviee from members of other organizations outside (name of organization); (d) "During the last month indicate the frequency with which members of organizations outside (name of organization) have sought information/advice from you." These items appear to tap only one aspect of boundary relevance: information transfer and filtration. This particular aspect of boundary relevance would not appear to place the boundary role person in a situation involving much conflict potential. Rather, it is the combination of representational, transactional and integrative requirements of the boundary role, as measured in the present study, which have been regarded as sources of conflict for its occupant [1, 13]. Future research on role stress should be directed to separating these dimensions of boundary relevance and relating them to role perceptions and outcomes. ~le absence of a direct relationship between boundary relevance and experienced role ambiiguity is probably an artifact of the setting selected for this study. Tile research and development environment, and the professional nature of all the roles sampled, may result in a restriction of range in tile ~legree of objective, if not experienced, role ambiguity. Future studiies in more conventional work organizations, and of more conventional roles, may be able to resolve this uncertainty. Finally, individual differences in self-perceptions appear to be better indicators of coping effectiveness under various types of experienced stress than otE sensitivity to potentially stressful, objective work conditions. It could be that other kinds of individual difference variables operate in the latter linkage between objective conditions and experienced role stress; e.g., Kahn et al. [13], found that persons with flexible as opposed to rigid personalities reacted more strongly,
l,dividual Dij]erencts and Organiz~ional Role Stress
101
i.e., experienced more conflict, under conditions of objecti'~,, role con. flict. Three situational variables which potentially could favorably moderate this linkaige are: (a) length of in.role se rvice especially, prior successful experiences with objective conditions; (b) whether role was assumed voluntarily; and (c) degree of social support from primary work group members. The present res~llts suggest that a conjunctive model may be required to improve the selection and placement of individuals into roles which may ext,ose them to high levels of role stress in the forms of conflict and ambiguity. The measures of self.perceptions included in this study exhibited differential moderating effects depending on whether the stress experienced was role conflict or role ambiguity. Need for occupationM achievement appeared to be an effective moderator in the case of the former type of experienced stress, while the supervisory abdity self.perception appeared to be most effective against role ambiguity. There was an unfavorable moderator effect of need for occupational achievement in one of the role ambiguity relationships with work outcomes. While the present results must be considered as tentative, they contribute to understanding individual differences in organizational role stres'~. Several su ~gestions for future research have been specified, among which are the need to: (a) replicate these findings in more conventional work olganizations and roles, (b) investigate dimensionality in the boundary relevance construct, (c) include additional individual difference and situational moderator variables, (d) in. corporate objective o Jtcomes, e.g., Fcrformance and health, in addition to the social-psychological ones included in the present study, and (e) replace assumptions of causality in the organizational role stress model with test., of causal inference. Together with the present accumulation of research evidence, the accomplishment of the suggested avenues of re.,earch should enhance our understanding and management of role stress in complex organizations.
Footnotes This research was completed while the author was Assistant Professor of Organizational Beht, vior on the Faculty of B~havioral Studies, Graduate School of Business, at the University of Alabama. 1A copy of the complete li~Lt of job activities, referred Io as ROLEREQS (R&D), may be obtained from the author.
102
Journal o/ Business Research
L Adanm, J. S. "The Stmetm'e and Dynamics of Organization Boundary Roles." In The Handbook of Industrial and Organizational Psychology, edited by M. D. Dunnette. Chicago: Rand-McNally, 1976. 2. Bullock, R. P. Social Facwrs Related to Job Satisfaction: A Technique ]or the Measurement o] Satisfaction. Research Report No. 70. Columbus, Ohio: Bureau of Business Research, The Ohio State University, 1952. 3. Corvdu, R. G. "Patterns of Organizational Conflict." Administrative Science Quarterly i4 (3,969) : 507-20. 4. Fon'~ N. M.; Walker, O. C., Jr.; and Churchill, G. A., Jr. "Expectation.Specific Measures of the Inter~nder Conflict and R~le Ambiguity Experienced by Industrial Salesmen." Journal of Business Research ~ (1975): 95-111. 5. Ghiselli, E. E. Explorations inManagerial Talent. Pacific Palisades, California: Goodyear Co., 1971. 6. Ghiselli, E. E. and Ba~thol, R. P. "Role perceptions of Successful and Unsuccessful Supervisors." Jour~nal of Applied Psychology 40 (1956). 7. Greene, C. N. and Organ. D. W. "An Evaluation of Causal Models Linking the Received Role with Job Satisfaction." Administrative S~:ience Quarterly 18 (1973): 95-103. 8. Gross, N.; Mason, W. S.; and McEachern, A. W. Explorations in Role Analysis: Studies of the School Superintendency Role. New York: Wiley, 1957. 9. Hamner, W. C. and Tosi, H. W. "Relationship of Role Conflict and Role Ambiguity to Job Involvement Measures." Journal of Applied Psychology 59 (1974): 497-99. 10. House, R. J. and RizT~, J. R. "Role Conflict and Ambiguity as Critical Variables in a Model of Organizational Behavior." Organizational Behavior and Human Perform. ance 7 (1972) : 467-505. 11. Johnson, T. W. and Graen, G. "Organizational Ass~lilation and Role Rejection." ~rganizational Behavior and Human Performance 10 (1973) : 72-87. 12~ Johnson, T. W. and Stinson, J. E. "Role Ambiguity, Role Conflict, and Satisfaction: Moderating Effe~:ts of Individua~ Differences." Journal o/ Applied Psychology 60 (1975) : 529-33. 13. Kahn, R.L.; Wolf, D. M.; Quinn, R. P.; Snoek, J. D.: and Rosenthal, R. A. Organizational Stress: Studies in Role Conflict and Ambiguity. New York: Wiley, 1964. 14. Keller, R. T. and Holland, W. E. "Boundary-spanning Roles in a Research and Development Organization: An Empirical Investigation." Academy of Management Journal 18 (1975): 38893. 15. Lawrence, P. R. and Lorsch, J. W. Organization and Environment: Managing Di]. ferentiation and Integration. Homewood, Illinois: Irwin, 1969. 16. Lyons, T. "Role Clarity, Need for Clarity, Satisfaction, Tension, and Withdrawal." Or.ganizutional Belwvior and Human Performance 6 (1971): 99-110. 17. Miles, R. H. Role Conflict and Ambiguity in Boundary and Internal Organizational Roles: .4 ~'ie[d Study ~]sing Rote.Set Analysis and a Panel Design. Unpublished doctoral dissertation. University of North Carolina, Chapel Hill, 1974. 18. MiJes, R. H. "Role Requirement~ as Sources of Organizational Stress." Journal of Ar,plie~ Psychology 61(i~76): 17~-179. 19. Miles, R. H. and Perreatdt, W. D., Jr. "Organizational Role Conflict: Its Antecedents and Consvquenees." Organizational Behavior and Human Performance, in press. 2~). Organ, D. W. "Some Variables Affecting Boundary Role Behavior." Sociometry 34 (1971) : 524-37. ~1. Organ, D. W. and Green, C. N. "The Boundary Relevance of the Project Manager's Job: Findings and Implications for R&D Management." R&DMauagement 3 (1972): 7-11. Y~,. Rizzo, J. R.; I-louse, R. J.; and Lirtzman, S. E. "Role Conflict and Ambiguity in Complex organizations." Administrative Science Quarterly 15 (1970): 150~3. 23. Thompson, J. D. "Organizational Management of Conflict. '~ Administrative Science Quarterly 4 (1960): 389-409.