Journal of Vocational Behavior 8, 247-258 (1976)
Organizational Tenure, Work Environment Perceptions, and Employee Mental Health
JAMES F. GAVIN Colorado State University
JEFFREY H. GREENHAUS Stevens Institute of Technology
The mediating effect of organizational time investments on the relationship between work environment perceptions and mental health was explored in two work settings. Participants were 257 managerial-level employees in a line organization and 214 in a staff setting. Results support the hypothesis of greater responsivity to perceived organizational stimuli for longer tenured employees, but only in the line setting. The mixed findings in the staff organization suggest that other characteristics of the environment (e.g., systems function) and of its members (e.g., organizational versus occupational identification) need to be considered. Implications for the social responsibilities of organizations to their members are offered in conclusion.
The work roles which individuals occupy in society have demonstra- ble consequences for their physical and psychological well-being (Special Task Force, 1973). A significant amount of research has highlighted the importance of such factors as ambiguity, conflict, work overload, participation, and utilization of abilities in accounting for the physical and mental strains of industrial workers (Caplan, Cobb, French, Van Harrison, & Pinneau, 1975; French, 1974; French & Caplan, 1970; Kahn, 1974). In addition, the individuals level in the organization (Dunn & Cobb, 1962; Kasl & French,
An earlier version of this article was presented at the meeting of the Rocky Mountain Psychology Association in Salt Lake City, Utah, May 1975.
The authors are indebted to the following individuals for their contributions to the completion of this manuscript and the research on which it is based: Dr. Abraham K. Korman, Dr. Saul B. Sells, Mr. David L. Toole, Dr. Lee B. Murdy, and Ms. Wendy Axelrod.
Requests for reprints should be sent to James F. Gavin, Department of Psychol- ogy, Colorado State University, Fort Collins, CO 80523.
Copyright @ 1976 by Academic Press, Inc. All rights of reproduction in any form reserved.
1962; Kornhauser, 1965) size of the work force (Kornhauser, 1965; Porter & Lawler, 1965), and shift work (French, 1963; Mott, Mann, McLaughlin, & Warwick, 1965) manifest significant relationships with various health out- comes. Past research has also indicated that the relationships between stressful environmental factors and mental health tend to be mediated by other considerations in the environment (e.g., social support), as well as by certain characteristics of the individual (e.g., biographical or personality factors).
One variable which merits attention for both theoretical and practical reasons is organizational tenure. Although a number of studies have examined the implications of tenure for various job attitudes (Gibson & Klein, 1970; Hall & Mansfield, 1975; Herzberg, Mausner, Peterson, & Capwell, 1957; Hulin & Smith, 1965) the process by which length of service influences individual responses has received relatively little attention.
One possible conception, deriving from a model presented by Gavin and Howe (1975), is that tenure affects the relationship between the individuals perceptions of the work environment and his or her responses to it. While the possibility that tenure may also covary with either perceptions or responses is not to be ignored, what is of concern here is how tenure might serve to mediate the perceptual-response linkage. The significance of an employees perceptions can be seen in Likerts (1961) remark that the way in which a person reacts to any situation is always a function . . . of his perception of it. It is how he sees things that counts, not objective reality (p. 102). Empirical research also supports this position (Gavin & Howe, 1975; Lawler, Hall, & Oldham, 1974; Likert, 1961).
As one illustration of the proposed mediating effect, employees differ- ing in seniority may perceive similar characteristics in a newly impiemented participative management program; however, their responses may vary according to the extent of their experiences with new management programs. As a further example, an employees affective response to a perceived threat of labor layoffs will probably depend on how much time the individual has invested in the organization. A new member may perceive a high likelihood of being laid off, but will not feel unduly upset about it, whereas someone with 20 years with the company may be considerably disturbed.
In proposing such an effect of tenure, one must also consider the manner in which tenure might influence the perceptual-response relationship. One suggestion is that the effect of tenure is to sensitize the employee to perceived organizational influences so that the longer the person is in the organization, the more responsive he or she becomes (Gibson & Klein, 1970; Katz & Kahn, 1966, pp. 119-121; Schein, 1968). In another perspective, one might argue that the longer an individual is employed by an organization, the more he or she builds a private world to protect himself/herself from the variety of organizational influences. With the current state of knowledge,
TENURE,PERCEPTIONS ANDMENTALHEALTH 249
hypothesizing one of these effects over the other would be mere speculation, while asserting in a more general sense that the relationships between perceptions and the employees affective and behavioral responses may vary as a function of time investments seems more justifiable.
These considerations in the context of studies on employee mental health gave rise to the present investigation. An earlier report (Gavin, 1975a) indicated that employees who perceived the environment as having well- defined organizational structures, minimal interference in work processes, equitable reward systems, and an atmosphere of trust and consideration tended to have more favorable scores on mental health indices. As an extension of this research, the present study set out to examine the proposed mediating effects of tenure on the relationship between six orthogonal dimensions of work environment perceptions and eight interrelated criteria of mental health. In so doing, it was hoped that some of the implications of increased service would be clarified.
Data were gathered in two subsystems of a 35,000 employee domestic airline as part of an organizational study. One subsystem was a line organization (L) or, more specifically, a regional center for the airline, while the other was a staff division (S) serving the airline in its management information and data processing needs. Both employed approximately 3000 members.
Research was conducted on the managerial segments of these organiza- tions. The 257 management members in L tended to be slightly older (1, = 40; Xs = 35), more tenured (XL = 12; Xs = 6), and somewhat less well-educated than the 214 managers in S. These differences appear to be consistent with the professional tone of the staff organization vis-ri-vis the production orientation of the line division.
Three sets of measures were relevant in this investigation: work environ- ment perceptions, mental health criteria as described by the worker, and mental health criteria as assessed by the individuals supervisor. Since these have been adequately described elsewhere, they will only be summarized here.
Work Environment Perceptions (WEP). A 106-item questionnaire was used to assess six orthogonal factors derived from a principal-axis analysis and varimax rotation. Items were developed from current organizational literature
250 GAVIN AND GREENHAUS
and from the models proposed by Sells (1968) and Litwin and Stringer (1968). Similar forms of this instrument had been pretested and used on over 4000 employees. All items were answered on a five-point Likert scale ranging from strongly agree to strongly disagree. Additional information regarding this instrument can be obtained in Gavin and Howe (1975) and Gavin and Hodapp (Note 1).
The labels for the six factors and their reliability estimates (coefficient alpha) were as follows: I. Clarity and Efficiency of Organizational Structure (.89), II. Hindrance (.83), III. Rewards (.90), IV. Esprit (.84), V. Managerial Trust and Consideration (.91), and VI Challenge and Risk (.67). Correlations between factors ranged from -.15 to 20, since an approximate scoring method was used to obtain the factor scores. Also, factors were reflected so that high scores indicated more favorable perceptions, e.g., high Hindrance meant a low degree of administrative interference.
Mental Health (Ml!). The measurement of mental health is extremely problematic owing to the inherent issue of values (Smith, 1961) and difficul- ties in operationalizing its component constructs for work environments (Gavin, Note 2). As Smith (1961) notes, there is no single set of mental health criteria, so at best our measures only suggest what is essentially defined on an individual and subjective basis as the good life.
Based on a comprehensive review of literature on worker mental health (Gavin, Note 2) eleven categories of positive mental health (Jahoda, 1958) variables were noted. The categories of interest in the present study were: (a) growth, development, and self-actualization (French & Kahn, 1962; Jahoda, 1958; Maslow, 1943); (b) environmental mastery, including measures of performance and job strain (French & Kahn, 1962; Indik, Seashore, & Slesinger, 1964; Zander & Quinn, 1962); (c) interpersonal relations (French & Kahn, 1962; Jahoda, 1958; Zander & Quinn, 1962); and (d) satisfaction (French & Kahn, 1962; Gurin, Veroff, & Feld, 1960; Zander & Quinn, 1962). Assessment of these proposed areas of mental health was achieved through instruments administered to the workers, as well as to their supervisors (cf. Gavin, 1975a, for more complete details).
The self-report measures and the mental health categories from which the measures were derived (letters corresponding to above categories are listed after measures) are reported below. Also, the number in parentheses refers to the reliability estimates (coefficient alphas based on combined samples) for scales containing more than one item: 1. Self-Development, a (.65), 2. Job Pressure, b (.71), 3, Interpersonal Relations, c (.57), and 4. Job Satisfaction, d. The supervisory-based measures, their corresponding mental health cate- gories and reliabilities were: 5. Strain Symptoms, b (.83), 6. Interpersonal Relations, c, 7. Planning and Decision making, b (.74), and 8. Performance, b (.89). As in the scoring of WEP factors, MH scaIes were reflected so that high scores indicated a more positive state.
TENURE,PERCEPTIONSAND MENTALHEALTH 251
Questionnaires containing the WEP factors and self-report MH scales were administered under controlled testing conditions on company premises and were returned to an independent, university-based research group, ensur- ing confidentiality. Supervisory-based MH scales were available for approxi- mately half of each sample within three months after survey administrations. An analysis of differences between participants with and without supervisory evaluations indicated some notable discrepancies. In the line organization, participants with ratings tended to be at a higher job level as well as having more favorable perceptions on the Challenge/Risk factor. Participants who were rated in the staff setting showed more extreme scores on WEP factors II-VI, and on the self-rated MH criteria of Self-Development, Job Pressure, and Job Satisfaction. In this light, caution should be exercised in generalizing from the correlational data involving supervisory assessments.
A median split on tenure was used in defining long and short-term employees; the median in L was 142.5 months and in S, 44.5 months. While the medians were quite discrepant, it seemed more appropriate to use relative definitions of high-and low-tenure in each sample than to have numerically equivalent ones (cf. Festinger, 1954, for basis of rationale).
Tests of significance between means for the high- and low-tenure subgroups within each organization indicated that average scores on WFP factors and MH scales were generally comparable. Only two significant differences were found in the 28 tests (i.e., 14 in each setting). High-tenured employees in S perceived more challenge and encouragement of risk taking (Factor VI) than did low-tenured members. In L, long-tenured employees experienced more job pressure than short-tenured employees.
Tables 1 and 2 report the correlations and correlation-comparisons (using Fishers r to z transformation) for low- and high tenure groups in the two organizations. It should be noted that the less conservative .lO level was used in tests of significance between correlations since sample sizes in the tenure subgroups (particularly when supervisory-based criteris were involved) were such that considerable differences between correlational values would have been necessary to achieve significance at the traditional .05 level. While in a global sense L and S had a similar number of significant differences between zero-order correlation pairs (i.e., 10 out of 48, or 21%), the variables involved and the direction of differences varied across settings. In L, for example, all 10 of the differences were such that the longer service employees manifested more positive covariation between measures, while in S this was so in only half the cases.
252 GAVIN AND GREENHAUS
l 2 0