The relationship of role-related variables to job satisfaction and commitment to the organization in a restructured hospital environment. (Research Report).

Since the mid-1980s, hospital restructuring appears to have had a profound effect on the lives of health care practitioners, including physical therapists. As organizations changed, I believe clinicians experienced a great deal of anxiety as they faced potential and often real changes in their roles within the hospital. For some clinicians, including physical therapists, their worst fears were realized as their roles were altered and some positions were eliminated through restructuring and downsizing. (1-3) For other clinicians, the changes offered new opportunities for autonomy and collaboration, and even a renewed sense of professionalism. (1-3) Yet even though hospital environments have been undergoing major changes for over 15 years, we do not have a good understanding of the relationship among the factors that interact to shape the work experience of clinicians within this health care setting. (4-7)

If the work experience develops from a synthesis of employees' perceptions about the work they do, the organization they belong to, and the interpersonal relationships that bind these entities together, (8,9) then I believe many factors in today's restructured hospital environment have the potential to influence the perceptions of practitioners. Foremost among these factors, I believe the roles that clinicians assume, reflected in the work they perform, can be extremely influential. For example, some clinicians' roles have been changed in ways that require them to work in a more efficient and interdisciplinary manner. (5,10-12) For some clinicians, these role changes often symbolize a loss of traditional professional responsibilities, such as providing hands-on patient care. (11,13-15) For other clinicians, these changes represented an expansion of professional responsibilities into areas such as collaboration with other practitioners and coordination of services. (11,12,14,16,17) As roles changed, clinicians also experienced periods of stress during which role demands were too great, were too ambiguous, or conflicted with one another. (15,16,18,19) If these changes in roles influence how clinicians perceive their work, then role changes likely may influence 2 work-related outcomes: job satisfaction and commitment to the organization. Commitment to the organization is typically represented by the term "organizational commitment" in organizational literature. (8,20-22) These terms, which denote an employee's commitment to the employing organization, will be used interchangeably throughout this article.

A clinician's sense of professionalism also may shape that clinician's perception of the work experience. (13,23) Research carried out in a restructured hospital environment involving nurses and physical therapists suggests that the level of professionalism, in the form of commitment to a profession and a sense of responsibility for patients or clients, positively influenced how clinicians viewed their roles within the hospital and the manner in which they carried out their job responsibilities. (7,11,14) Professionalism even may mitigate the negative effects of the restructuring process on job satisfaction. (7,11,24)

I believe it is important to identify variables that influence job satisfaction and commitment to the organization because these 2 factors have been shown to influence an employee's work and to relate to other organizational outcomes, such as motivation, job performance, and turnover. (8,25-29) Collectively, these factors contribute to the efficiency of hospital function and the effectiveness of patient care. (7,13,30) My study was intended to add to our understanding of the nature of the work experience in restructured hospitals by examining the relationship between the changed role of physical therapists, stress, professionalism, and 2 work-related outcome variables: job satisfaction and commitment to the organization.

The Nature of Role Behaviors in Organizations

According to Katz and Kahn's Role Theory, (8) organizational structure can be viewed as a series of motivated, patterned behaviors associated with fulfilling organizational tasks. These behaviors form the organizational roles that link employees to their work groups in order to perform the assigned tasks. According to this argument, when role behaviors are performed in a predictable manner, the organization will operate effectively and efficiently. (8,13)

An employee discovers which role behaviors are appropriate through a cyclical process that primarily involves the work group to which he or she belongs. (8) A work group is generally defined as the group within which an employee functions and can include individuals from the same profession or from several professions, such as an interdisciplinary team. The work group sends the employee messages, in the form of cues, regarding the role behaviors they expect to see him or her exhibit. These role cues, I believe, can influence how the employee behaves and how he or she feels about his or her role, job, and employing organization. (8) If the employee perceives that the role behavior cues are congruent with his or her perceptions, experiences, and beliefs, I contend that the employee will conform to the work group's expectations and exhibit the behaviors they desire. (8) However, if the employee perceives that the cues are incongruent or coercive, he or she will resist meeting the work group's expectations. (8) In either case, the employee's response will either strengthen or alter the work group's subsequent role messages. (8)

The predictability in how role behaviors are performed is complicated further because each employee belongs to groups, other than the primary work group, that also send role messages that could influence the employee's behavior. (8) These groups include informal support systems and formal groups external to the organization, such as professional organizations. As employees attempt to meet the expectations of these varied groups, the role behaviors that emerge may be complex and may not always be directed toward the goals of the employing organization. (8)

Stress

The complexity of the behavioral demands of a particular role may lead to stress in the form of role overload, role conflict, or role ambiguity. (8) Role overload occurs when an employee perceives that too much is expected of him or her in the performance of the job. (31) Role conflict occurs when simultaneous and competing role expectations are received by the employee and complying with one set of expectations interferes with complying with the others. (8,32) Role ambiguity occurs when an employee perceives that there is a great deal of uncertainty about aspects of the role or membership in a work group. (8,32) Any of these forms of stress can contribute to organizational problems that have been shown to lead to a diminution in employee and organizational performance. (8,31,32)

Physical Therapist Role Behaviors and Hospital Restructuring

In previous research, (11,15) I identified 26 role behaviors of physical therapists working in restructured acute care hospitals. These role behaviors were organized into 3 broad categories: (1) behaviors exhibited during care production, or the hands-on delivery of care to the patient, (2) behaviors exhibited during care management, or the planning and coordination of care delivery, and (3) behaviors exhibited during the performance of administrative tasks. Care production and care management incorporate behaviors related to patient care, communication and collaboration with other practitioners, such as the members of an interdisciplinary team, in order to meet patients' needs, as well as the clinician's felt responsibility for his or her patients. (15-17,33) These behaviors have been described as reflecting professional responsibilities (33) and will be referred to in this article as "professional role behaviors." In contrast, the administrative tasks incorporated the behaviors directed at meeting corporate demands for timeliness, efficiency, and coordination of effort--tasks that some clinicians often perceive as regulating their work and detracting from their primary patient care focus. (7,15,19,33) These behaviors will be referred to as "organizational role behaviors" in this article.

The Role of Professionalism/Occupational Commitment

A practitioner's sense of professionalism appears to be an important part of clinical practice and may play a key role in clinical practice by mitigating the negative effects of a restructuring process on job satisfaction and commitment to the organization. (7,11,24) However, in organizational research, the construct of professionalism has been found to be ambiguous when used as a study variable. (34,35) To rectify this problem, occupational commitment has been used as a surrogate variable for professionalism because it is believed to reflect the essence of what the concept of professionalism represents (that is, the allegiance of a person to a profession and professional ideals) and is believed to be a more reliable construct. (34-38)

Occupational commitment has been defined as "one's attitude, including affect, belief and behavioral intentions, toward her/his occupation" (35(p311)) or "one's belief in and acceptance of the values of his chosen occupation or line of work, and a willingness to maintain membership in that occupation." (36(p535)) Although occupational commitment is generally considered to be multidimensional in nature, that is, there are several different conceptualizations of the construct, (34) in my study I focused on affective occupational commitment, which represents a strong emotional attachment to the individual's chosen occupation. (34,36,38)

The contemporary view of occupational commitment is that it is an antecedent of organizational commitment and a correlate of job satisfaction. (34,36,37) Thus, examining occupational commitment, as a moderating variable in the relationship between the role behaviors and the outcome variables, warrants study. That is, I believe that a person's level of occupational commitment will influence the relationship between his or her role behaviors and his or her job satisfaction and organizational commitment.

Job Satisfaction and Commitment to the Organization: Work-Related Outcome Variables

Many of the variables in this study have been found to relate to job satisfaction and commitment to the organization, (24,25,28,29,39) although the manner in which these variables actually influence these 2 work-related outcome variables has not been established. However, it has been argued that factors in the work environment that produce a positive work experience (eg, good feelings about the work or organizational role) will be more likely to produce positive outcomes such as greater job satisfaction and commitment to the organizationt. (9,25,39) In contrast, factors that produce a negative work experience (eg, stress) will likely produce negative outcomes or lower levels of job satisfaction and commitment to the organization. (9,25,39) It also has been posited that factors in the work environment that are closely related to an employee's actual work (eg, the creation of greater interaction and interdependence among practitioners) will have a direct and immediate influence on his or her perception of the work experience. Ultimately, these factors will influence the employee's satisfaction with the job and commitment to the organization more than factors that are more removed from the actual work (eg, the creation of a patient-focused care teams throughout the hospital). (5,7,9,11,25,34,40,41)

Job satisfaction is a complex construct and is often measured as a global attitude of an employee toward his or her work. That is, the employee is either satisfied or dissatisfied with the job. (6,28,39) Alternatively, many researchers, (24,28,39,42) believing that an employee's level of satisfaction varies with specific aspects of the job, have proposed that numerous elements (variables) underlie this construct. These elements have been classified into 5 distinct dimensions: satisfaction with work attributes (eg, the nature of the work, autonomy, responsibility), rewards (eg, pay, promotion, recognition), other people (eg, supervisors, coworkers), the organizational context (eg, policies, procedures, working conditions), and self or individual differences (eg, internal motivation, moral values). (24,26,28,39,42) I used this variable-specific classification scheme in an effort to ensure that all dimensions of job satisfaction were measured.

Organizational commitment as an empirical construct is generally regarded as a psychological state characterizing an employee's relationship with the organization that has implications for the employee's decision to remain or leave the organization. (29,34) Furthermore, this form of commitment reflects the employee's acceptance of the goals of the organization and willingness to engage in behaviors that are specified in the job description, as well as those that are considered to be beyond the job expectations. (34) As hypothesized, organizational commitment is believed to be multidimensional in nature, with affective organizational commitment representing a strong emotional attachment to the organization. (29,34) I chose affective organizational commitment for use in my study because it is the most widely accepted conceptualization of commitment to the organization. (25,34)

I tested 7 hypotheses to (1) assess the relationship among the variables discussed, (2) examine the influence of the role behaviors, role conflict, role overload, and role ambiguity on the outcome variables of job satisfaction and organizational commitment, and (3) ascertain whether the clinicians' occupational commitment, as a surrogate variable for professionalism, influenced the relationship between the role behaviors and the outcome variables. These hypotheses (depicted in Fig. 1) were:

Hypothesis 1: The professional role behaviors will be positively correlated with job satisfaction and organizational commitment.

Hypothesis 2: The professional role behaviors will make a positive contribution to the prediction of job satisfaction and organizational commitment.

Hypothesis 3: The organizational role behaviors will be negatively correlated with job satisfaction and organizational commitment.

Hypothesis 4: The organizational role behaviors will make a negative contribution to the prediction of job satisfaction and organizational commitment.

Hypothesis 5: The level of occupational commitment will interact with and positively influence the relationship between the role behaviors and both job satisfaction and organizational commitment.

Hypothesis 6: Role conflict, role overload, and role ambiguity will be negatively correlated with job satisfaction and organizational commitment.

Hypothesis 7: Role conflict, role overload, and role ambiguity will make a negative contribution to the prediction of job satisfaction and organizational commitment.

[FIGURE 1 OMITTED]

Methods

Identification of Restructured Hospitals

Hospital restructuring, defined as a major organizational change that alters the structure, reporting relationships, or operation of hospital departments and the delivery of patient care services, has occurred in many hospitals across the country and has taken many forms in those hospitals in which it has been implemented. (5,7,11) Therefore, to survey clinicians' perceptions of work attributes within restructured acute care hospitals, it first was necessary to identify those hospitals that had undergone restructuring. To do this, all physical therapy managers who were members of the American Physical Therapy Association's (APTA) Section on Administration or Section on Acute Care Hospital Practice were invited to participate in the study if their hospitals had undergone restructuring within the last 15 years and if the organizational changes had been implemented for at least 1 year prior to this study. A letter explaining the purpose of the study and containing a list of inclusion criteria for hospital and clinician participation and a form on which they could provide the names of clinicians working in their hospitals who might be willing to complete the questionnaire were sent to the managers.

Clinical managers representing 100 hospitals from 4 geographic regions of the United States (shown in Fig. 2) responded to the call for participation. All 100 hospitals were included in the study to maximize the pool of clinicians completing the questionnaire. Through a second mailing, the managers were asked to provide data on the nature of the hospital restructuring that occurred within their facilities and the organization of the physical therapy services at the time of the survey.

[FIGURE 2 OMITTED]

The data provided by the managers indicated that, on average, the hospitals had been engaged in restructuring for 4 years (SD=2.6, range=1-14), with physical therapy services included in the changes for 3.6 years (SD=2.3, range=1-13). Furthermore, 90% of the hospitals were still engaged in some form of restructuring. All of these changes resulted in the decentralization of services in 59% of the hospitals and in the reorganization of departments in 53% of the hospitals. In terms of the organization of the physical therapy services, 71% were organized at a hospital level and 20% were organized at a hospital system (ie, multi-hospital) level. Physical therapy services were managed as a rehabilitation department or physical therapy department in 95.7% of the hospitals, with physical therapists supervised by other physical therapists in 86.2% of the departments. The managers also indicated that hospital restructuring had a moderate to major effect on the delivery of physical therapy services in 72.8% of the hospitals.

Study Participants

To be included in the study, a clinician had to (1) be a licensed physical therapist, (2) occupy a position that primarily involved the delivery of direct patient care, and (3) have been employed in the hospital for at least 1 year. Therapists were not required to have been employed in the hospital prior to the initiation of restructuring. Thus, not all therapists experienced the same changes that may have occurred. I determined that at least 300 subjects should be surveyed for this study based on a ratio of 5 subjects for each questionnaire item and the fact that several different constructs would be measured. (43,44) The explanation of the nature of the study was kept general in the introductory letters and the questionnaire instructions (eg, the specific variables of interest were not disclosed) to preclude clinicians from forming preconceived perceptions about the study variables. (43)

The clinical managers identified a total of 360 clinicians. Therefore, to maintain an adequate pool of subjects, all identified clinicians were surveyed. Although this sample of convenience could introduce bias, I felt that it was important to maintain a large sample and as high a ratio as possible of subjects to questionnaire items to reduce the chance of Type I error. (43,44) A cover letter, a questionnaire, and a response envelope were mailed to each clinician identified by the clinical managers. In an effort to maximize the return rate, 2 follow-up questionnaires were mailed or faxed to those clinicians who did not respond. As indicated in the cover letter, agreement to participate in the study served as informed consent.

Two hundred seventy-three clinicians (75.8% of the sample) completed and returned the questionnaires, which provided, in my view, a large enough sample to achieve adequate statistical power (.80) given an estimated scale reliability of .70. (43) The return rate represented an average of 2.7 clinicians per hospital, and as indicated in Figure 2, the distribution of clinicians by geographic region was similar to the distribution of hospitals.

Research suggests that individual attributes may affect an employee's view of his or her job, organization, and occupation (36,38); therefore, salient demographic variables were collected for each clinician. These variables included: age, gender, ethnic background, professional experience (years of clinical experience), organizational tenure (years working in this organization), whether the clinician was working in an area of interest, and whether the clinician was a member of APTA.

The demographic data for the clinicians, shown in Table 1, indicate that the majority of the clinicians were Caucasian women, approximately 26 to 35 years of age ([bar]X=34.8, SD=8.6, range=23-63). Most had been physical therapists for 10 years or less ([bar]X=10.2, SD=8.8, range=1-41), had worked in the current hospital for less than 5 years, and were working in an area of interest. Fewer than half of the clinicians were members of APTA.

Instrument Development

The survey questionnaire was composed of 4 sections. Section 1, which included 25 items from the Professional Role Behaviors Survey (PROBES), (15) was used to ascertain which behaviors the clinicians felt were part of their role as physical therapists following hospital restructuring. Section 2 included items on job satisfaction, role conflict, role overload, and role ambiguity. This section was designed to reveal the clinicians' feelings about satisfaction with the job and role-related stress given the changes that had taken place. Section 3 included items from organizational commitment and occupational commitment scales intended to reveal the clinicians' feelings about their organization and their occupation or profession. Section 4 was designed to obtain demographic data for the clinicians.

The items selected for sections 2 and 3 were taken from scales that have been shown to have good reliability and validity as presented in Table 2. (28,29,31,32) Because a large number of variables were included in this study and using the entire scale for each variable could create problems of respondent fatigue and increase the number of non-responses, the total number of questionnaire items was reduced using guidelines set forth by Nunnally and Bernstein. (44) That is, items shown in prior research to have high Pearson r correlations (.60 or greater) with the underlying constructs were selected for use in the questionnaire. Thus, I contend that a sufficient level of construct validity for each variable could be maintained while limiting the number of items used for each variable. (43,44) To that end, a Cronbach alpha was calculated for each scale to measure how well the items fit with the underlying latent variable. (43) However, other forms of reliability of the modified scales were not examined. The Cronbach alpha was used as a measure of the how well the variables fit with the underlying latent variable; thus, it is appropriate to determine whether the test items still reflect the underlying construct. (43)

The response format for all items in sections I through 3 was modified to use a 5-point Likert-type scale (1=strongly disagree, 2=moderately disagree, 3=neutral, 4=moderately agree, 5=strongly agree). A consistent response format was used for all sections of the questionnaire to reduce respondent confusion, yet provide adequate variability for the various scales. (43)

Study Variables

Role behaviors. The clinicians' perceptions of the nature of physical therapy role behaviors in their current organizations were measured using a modified form of the PROBES. (15) For this study, the clinicians were asked to indicate their level of agreement with each role behavior statement. The correlations among the role behaviors were calculated, and the underlying dimensions of the role behaviors were determined using a principal component factor analysis.

A majority (53.6%) of the role behaviors were correlated with one another, although only 2 role behaviors ("Increase in delegation and supervision of others in providing physical therapy treatment" and "Continued emphasis on therapists providing patient care") had correlations above .50 (range=.12-.60). An exploratory factor analysis, accounting for 55.4% of the overall variance, suggested that a 6-factor solution would have the best potential for producing interpretable factors. This was verified through a Varimax rotation of the data matrix, which yielded 6 fairly distinct and interpretable factors with correlations of .40 or greater between the factors and their respective role behaviors. In addition, factor score correlations greater than .70 were found within individual factors, and small factor score correlations were found between factors. (43) The factors were named based on the nature of the role behaviors included in each. I determined that the first factor, "Integrate," represented behaviors involving the integration of therapists into health care teams. The second factor, "Interact," represented the interaction of therapists with others. The third factor, "Evaluate and Plan," included behaviors representing the evaluation and planning of patient programs, and the fourth factor, "Care," involved behaviors related to the provision of direct patient care. The fifth factor, "Educate," included the multiple forms of education in which the therapists participated. The final factor, "Organizational Responsibility," included behaviors associated with meeting organizational demands.

Because role behavior dimensions were to be used in hypotheses testing, further analysis of the factor loadings was necessary. Two changes were found that improved the distinction between the factors and the correlation of the role behaviors with the respective underlying construct of each factor. First, the role behavior "Increase in teaching patients, families, and other health care providers" was eliminated because it loaded almost equally on the factors "Integrate" and "Interact." Second, the internal consistency of the factor "Care" was improved (from [alpha]=.52 to [alpha]=.70) by the elimination of 2 role behaviors ("Increased pressure on physical therapists to assume formal responsibility of a case manager" and "Continued emphasis on professionalism").

The Cronbach alpha for the role behavior scale was found to be .72, indicating that the behaviors account for a large proportion of the variance of the underlying construct. (43) The factors with the role behaviors each represented and supporting statistics are presented in Table 3.

Stress variables. Stress was represented by the constructs of role overload, role conflict, and role ambiguity. Role overload, defined as having too much to do, was measured using the 3-item Work Overload Scale developed by Kim et al. (31) Role conflict, defined as the measure of the congruence of job requirements, was measured using the 3 items from Rizzo and colleagues' Role Conflict Survey (32) that were found to have a correlation of .60 or greater with the underlying construct. Role ambiguity is defined as the degree of certainty about duties, authority, time allocation, and relationships with others and the existence and clarity of guides, directives, and policies that allow individuals to predict sanctions as outcomes of behavior. (32) As with role conflict, the 3 items from Rizzo and colleagues' Role Ambiguity Survey (32) that were highly correlated with the underlying construct were used in this study. To provide consistency in the interpretation of these 3 variables, the scores for the role ambiguity items were reversed during data analysis. Therefore, high scores would reflect the perception of high levels of role overload, role conflict, and role ambiguity. Three separate composite scores were calculated by averaging the responses for the items for each variable.

Occupational commitment. Occupational commitment was measured using the 3 items that were highly correlated with this construct taken from the affective scale of Meyer and colleagues' Occupational Commitment Survey. (34) Responses were averaged for each clinician to yield a composite occupational commitment score. High scores reflected high levels of occupational commitment. According to Nunnally and Bernstein (44(pp14-30)) there are 2 schools of thought on what statistical measurements are appropriate for the analysis of data from Likert-type scales. The first school of thought asserts that a scale must demonstrate ostensive (visualizable) interval properties before a person can perform arithmetic operations on the data from it. The second school asserts that very few measures are ostensive and that a better criterion is the extent to which the scale fits a scaling model, such as having the response format reflecting a linear relationship and using anchoring to fix the points on the scale. The analysis of the data from the Likert-type scales used in the current study represent the view expressed by the second school of thought; thus, the data have been subjected to arithmetic operations. (44) All of statistical operations used in this study have commonly been utilized with the scales I used, as demonstrated in the references listed.

Outcome variables: job satisfaction and organizational commitment. Items from 2 survey instruments were used to obtain an overall measure of job satisfaction that included items for each dimension. First, satisfaction with the dimensions of work attributed, rewards, other people, and organizational context was measured using 12 items (3 per dimension) selected from Spector's Job Satisfaction Survey. (28) Second, satisfaction with the dimension of individual differences was measured using 3 items from Hackman and Oldham's Job Diagnostic Survey. (45)

Commitment to the organization, referred to as organizational commitment, was measured using the affective scale from Meyer and colleagues' Organizational Commitment Survey. (34) As with other variables, 3 items shown to have a high correlation (.60 or higher) with the underlying construct were selected from this scale.

Composite scores for job satisfaction and organizational commitment were obtained by averaging the clinician's responses to the items for each scale. High scores represented high levels of job satisfaction or organizational commitment.

Data Analysis

Descriptive statistics for all study variables were calculated to characterize the variables. The correlations (Pearson r) among the variables were calculated to determine the relationships among constructs, that is, to test hypotheses 1, 3, and 6. Stepwise, hierarchical regression analyses were used to determine the influence of the independent variables on the prediction of job satisfaction and organizational commitment (hypotheses 2, 4, and 7). For these analyses, the demographic and stress variables were entered into the regression analyses in the first 2 steps to control for their effect on the outcome variables. The role behavior dimensions were entered in the third step and occupational commitment was entered in the fourth step to test for their effect on each outcome variable. The criteria used for the stepping method were based on the probability of the F value, with P [less than or equal to] .05 used for entering or retaining any variable in a particular step and with P [greater than or equal to] .1 used for variable removal. (47) The contributions of each variable to the prediction of the outcome variables, as measured by their beta weights, were calculated.

General linear modeling (GLM) (47) was used to test hypothesis 5, the effect of the interaction between the role behavior dimensions and occupational commitment on the outcome variables. This procedure was used because it allowed me to analyze both the main effects and the interactive effects of independent variables on the prediction of dependent variables. (47) In using GLM, only independent variables that have a main effect on a dependent variable are included in the analysis. (44,46)

Results

The descriptive statistics and reliability coefficients for the study variables are shown in Table 4. The data indicate that, on average, the clinicians were experiencing moderately low levels of stress and had a strong sense of commitment to their occupation. Furthermore, they were moderately satisfied with their jobs, but were fairly neutral in their commitment to their organizations following hospital restructuring. The items included in all scales, except for role ambiguity, were found to correlate well with one another and the underlying construct.

Several correlations were found between the demographic variables and the study variables (Tab. 5). Interestingly, only 1 role behavior dimension ("Educate") was correlated with the demographic variables. It was positively correlated with being an APTA member, but negatively correlated with professional experience and with age. In addition, job satisfaction and organizational commitment were positively correlated with 2 demographic variables (organizational tenure and working in an area of interest). Role overload was positively correlated with 2 demographic variables that are reflective of the clinicians' work status (professional experience and organizational tenure), but it was negatively correlated with working in an area of interest.

Correlations among the study variables are shown in Table 6. In relation to the hypotheses, 3 of the professional role behavior dimensions were positively correlated with job satisfaction, whereas only 2 dimensions were positively correlated with organizational commitment. Thus, hypothesis 1 was partially supported for job satisfaction, yet minimally supported for organizational commitment. The organizational role behavior dimension, "Organizational Responsibility," was negatively correlated with job satisfaction and organizational commitment, providing support for hypothesis 3. In addition, all 3 of the stress variables were negatively correlated with both outcome variables, providing support for hypothesis 6. Finally, 3 of the professional role behavior dimensions--"Integrate," "Interact," and "Educate"--were found to have negative correlations with the stress variables, whereas one professional role behavior dimension, "Care," and the organizational role behavior dimension "Organizational Responsibility" were found to have positive correlations with the stress variables.

The results of the final steps of the regression analyses used to test hypotheses 2, 4, and 7 (ie, the influence of the study variables on the prediction of the outcome variables) are depicted in Tables 7 and 8. For job satisfaction (Tab. 7), the regression analysis indicated that 54.4% of the total variance of job satisfaction was explained by the independent variables. The 3 stress variables accounted for the majority of the job satisfaction variance (43.9%), and all 3 stress variables made negative contributions to its prediction. Therefore, hypothesis 7 was supported for job satisfaction. Four of the 5 professional role behavior dimensions made a small (6.2%) positive contribution to the prediction of job satisfaction, providing partial support for hypothesis 2. However, the data did not support hypothesis 4 (ie, the influence of the organizational role behavior dimension on job satisfaction). Finally, occupational commitment made a positive contribution to the prediction of job satisfaction.

The regression analysis for organizational commitment (Tab. 8) indicated that 33.9% of the total variance for this variable was explained by the independent variables. Two demographic variables, organizational tenure and professional experience, made a positive contribution to the prediction of this outcome variable. Again, the stress variables accounted for the majority of the variance (19.1%), but only role ambiguity and role conflict made negative contributions to the prediction of organizational commitment. Thus, hypothesis 7 was partially supported for organizational commitment. Two of the professional role behavior dimensions "Interact" and "Educate" made a small (4.6%) positive contribution to the prediction of organizational commitment. These findings offer only minimal support for hypothesis 2. No support was found for the influence of the organizational role behavior dimension on organizational commitment (hypothesis 4). Occupational commitment made a positive contribution to the prediction of organizational commitment.

The results of the test of the effect of an interaction between the role behavior dimensions and occupational commitment on job satisfaction or organizational commitment showed no significant interactions. Thus, hypothesis 5 was not supported for either outcome variable. The results of the hypothesis testing are summarized on Table 9.

Discussion

Demographics of Participating Hospitals and Clinicians

I believe 2 aspects of the clinical environment are important to an understanding of the organizational context of this study. First, because 90% of the hospitals in this sample were still engaged in the process of restructuring, with some managers indicating that they had undergone several major changes, my data support the assertion that hospital restructuring is an ongoing process (1,18); thus, changes continue to occur as clinicians join and leave the respective organizations. Second, despite the fact that a majority of the hospitals had decentralized services and reorganized departments, the clinicians' identification with the profession of physical therapy remained strong in terms of organization and management. For example, although most physical therapy services were managed as rehabilitation departments, the majority of these departments were organized at the hospital level and the majority of the clinicians were directly supervised by other physical therapists.

The relationships between the demographic variables and the other study variables shown in Table 5 reveal 4 interesting points. First, my data indicate that younger clinicians, those with less professional experience, and those who are members of APTA had a strong sense of a continued emphasis on education. This finding, I contend, could be due to the fact that clinicians with these characteristics are generally greater participants in educational activities, view education as essential to fulfilling their professional roles, or are influenced by the emphasis placed on continued professional education as part of their early professional socialization. Second, clinicians with more professional experience and longer organizational tenure experienced a greater sense of role overload than those with less experience. This finding could be because more experienced clinicians compared their current workloads with those they had prior to the implementation of changes. Less experienced clinicians, however, were more likely to have been socialized into the today's hospital environment with its demands for greater productivity. Third, the correlation between working in an area of interest and job satisfaction agrees with previous research findings, in that working in an area of interest contributes to an employee's psychological comfort with his or her job, enhances his or her sense of competence, and produces a more positive work experience. (25,34,41) Finally, the correlation between organizational tenure and organizational commitment is consistent with the relationships described by Meyer and Allen (29) and Hackett et al (48) in that an individual with longer organizational tenure tends to have a stronger affective attachment to an organization because this attachment forms well after the period of organizational socialization.

The Nature of Role Behavior Change

The role behaviors identified in my study (Tab. 3) indicate that this sample of clinicians agreed with the views about the nature of the changes in role behaviors held by physical therapy managers. (11,15) This agreement is important from 2 perspectives. First, it challenges the belief that individuals at higher organizational levels (ie, managers) may not be able to accurately reflect the perception of organizational life experienced by their subordinates. (26,28) Second, although some researchers have shown that there has been considerable variability in the changes across hospitals and over time, the role of physical therapists has changed in similar ways across these institutions and periods. (5,11,15,18) That is, the clinicians' work appears to remain focused on the professional responsibilities that reflect what the Guide to Physical Therapist Practice (49(pp42-49)) refers to as the primary elements in patient/client management and related professional roles. Nonetheless, the pressure to be more responsive to organizational demands also was reflected in the changes in organizational responsibilities. (1,10,11,15,18)

The Relationship Among Study Variables

The results of the hypothesis testing (shown in Tab. 9) provide support for the existence of several relationships discovered in prior research and add to our understanding of the factors that influence job satisfaction and commitment to the organization. (7,12-14,24,50) Four of the 5 professional role behavior dimensions ("Interact," "Evaluate and Plan," "Integrate," and "Educate") contributed to the prediction of job satisfaction, providing partial support for hypothesis 2. The positive influence of the role behavior dimensions that represent interactions with other people on job satisfaction agrees with findings from nursing research. (7,12,14,24) These findings may support the view that these role behaviors reflect accommodations made by practitioners to facilitate one another's role performance and professional accountability for patient care. As such, they may mediate the effects of organizational change on job satisfaction. (13,24,50)

There was minimal support for hypotheses 1 and 2, in relation to organizational commitment. However, the 2 professional role behavior dimensions that contributed to the prediction of this outcome variable ("Interact" and "Educate") had not undergone substantial change. The finding that some stable aspects of the job may contribute to an employee's organizational commitment also has been demonstrated in prior research. (25,29) In my study, the continued emphasis on interaction among clinicians may reflect factors such as group cohesiveness and group-leader relations that help employees feel "psychologically comfortable" with their work. Likewise, the continued participation in professional and educational activities outside of patient care may reflect factors that enrich the work experience and enhance the employee's sense of competence. (29)

The negative correlations found between the "Organizational Responsibility" dimension and both outcome variables (hypotheses 3 and 4) are also consistent with the results of prior research. Organizational research (25,51) and research in nursing (7,24) have shown that changes in the context in which the work takes place can contribute to decreased job satisfaction and an employee's disengagement from the organization. In my study, the role behaviors included in the "Organizational Responsibility" dimension represented changes in the contextual factors in the work environment, such as the requirements for weekend and holiday work or work in multiple organizational settings and the ability to plan and control work. If clinicians view fulfilling these role behaviors as an encroachment on professional autonomy, then the role behaviors may diminish the clinicians' job satisfaction and commitment to the organization. (13,24)

As demonstrated in prior research, (36,41,44) I found that occupational commitment had an effect on the prediction of both outcome variables (Tabs. 7 and 8). However, the moderating influence of occupational commitment on the relationship between the role behaviors and the outcome variables proposed in the conceptual model and tested as hypothesis 5 was not confirmed. Perhaps the absence of significant interactions was due to the fact that these independent variables had small main effects on the outcome variables. (46)

As predicted in hypothesis 6, role conflict, role overload, and role ambiguity were negatively related to both outcome variables. This finding is consistent with other research (25,28,29) using global measures of job satisfaction and supports the view that employees who perceive their roles as having higher levels of stress experience lower levels of job satisfaction. The results agree with previous research (25,29) that showed that employees who experience stress in the form of role ambiguity and role conflict tend to have less psychological attachment to their employing organization. My study, however, does not help clarify the question of whether stress affects commitment to the organization directly or whether the effect is mediated by other variables. (25,29)

Study Limitations

The findings from my study are generally consistent with those of previous research that examined these organizational and occupational variables, (28,29,36) although my study is unique in that specific role behaviors that are part of the physical therapist's role in today's hospital environment were examined. There are limitations to my research that must be acknowledged.

Although the correlations found were small (none greater than .60), finding relationships with this magnitude are not uncommon in organizational behavior research, especially in studies including the variables of job satisfaction, organizational commitment, and occupational commitment. (28,29,42) In addition, many other variables, both within and outside of the work environment, can affect job satisfaction and commitment to the organization. (27-29) Together, these 2 factors may have contributed to the rather small amount of variance for both outcome variables explained by the study variables. (27-29) Perhaps including other salient study variables would have enhanced the amount of variance found for job satisfaction and commitment to the organization. However, studying the effect of new variables, such as role behaviors, along with the numerous other variables that might influence these 2 variables would have created a survey instrument that was exceedingly long. In doing so, completing the instrument would have become an onerous task for respondents, which would have affected the return rate. (43,44) Researchers must strike a reasonable balance between questionnaire length and psychometric soundness because problems with insufficient response rates can affect the credibility of studies as much as the properties of the tools used. (43) To control length, I chose to include the stress variables only, because they consistently have been shown to have an important effect on both outcome variables. (28,29,39)

To ensure psychometric soundness, I selected measurement scales that previously were reported to be psychometrically sound and I selected items from those scales that were reported to have strong associations with the underlying constructs. (52) This selection process, however, may have reduced the reliability (Cronbach alpha) of several of the scales to a point that the size of the relationships found among the variables studied could have been overstated or understated, thus raising concerns about the credibility of the findings.

Several factors could have contributed to sample bias in this study and limited the generalizability of my findings. Asking hospital managers to identify clinicians who could participate in the study and using a sample of convenience to maintain a large number of potential respondents may have created selection bias. This strategy, in my view, was necessary to compensate for the limited number of participating hospitals, the large number of survey items, typical questionnaire return rates, and necessary statistical power. Additionally, because only survivors of the restructuring process participated in this study, the perceptions of those who are no longer with these hospitals were not included. This is a common problem with research on organizational change and has been noted in several meta-analyses on the effect of restructuring on employees. (4,24)

Finally, I used a cross-sectional study design, which can only provide a snapshot of perceptions of the practitioners studied at a particular point in time. This method may limit the credibility of the study findings because it created a situation in which not all of the respondents experienced the same changes and not all respondents worked in the hospitals throughout the restructuring process. At face value, these limitations might seem substantial unless the context of the change process itself and the nature of the data from the participants are taken into account. Because restructuring is a process and not an event, either the changes were occurring over a long period of time or a series of separate changes were taking place within the hospitals. Given this context, it would have been impossible to find a group of therapists who had experienced the same changes and had worked in the environment throughout the process. To compensate for this situation, the therapists were asked to reflect on their perceptions of their roles and personal experiences with the changes that had taken place in their respective environments. Therefore, the question posed (implicitly) was: How have the changes that you experienced affected your feelings about your work, organization, and profession during your tenure at the hospital? Future research using a longitudinal design could examine the effect over a longer period and would allow for the testing of causal hypotheses.

Conclusions

The roles of practitioners are changing due to pressures from both within and outside of today's hospitals. In this study, it was shown that stress is a major factor as organizations continue to change. In addition, although a causal relationship between the evolving role behaviors and job satisfaction and commitment to the organization cannot be inferred, the role behaviors appear to be related to these important organizational outcomes, and thus may contribute to the efficiency and effectiveness of hospital operation. The professional role behaviors reflect the responsibilities expressed in the Guide to Physical Therapist Practice (49(pp42-49)) as well as many of the characteristics that draw individuals to the field of physical therapy. Together with occupational commitment, or professionalism, they appear to be factors worth nurturing at all levels of professional development.

Table 1.
Demographic Data for the Clinicians Participating in the Study

                                 [bar]X   SD     Median   Range

Age (y)                            34.8    8.6   32       23-63
Professional experience (y)        10.2    8.8    7        1-41
Organizational tenure (y)           6.6    5.9    5        1-30

                                        n%

Gender
  Female                            228   83.5
  Male                               45   16.5

Age (y)
 [less than or equal to] 25          22    8.1
  26-30                              87   32.1
  31-35                              73   26.9
  36-40                              28   10.3
  41-45                              19    7.0
  46-50                              25    9.2
  >50                                17    6.3

Working in an area of interest
  Yes                               260   95.2
  No                                 13    4.8

Member of APTA (a)
  Yes                               126   46.2
  No                                147   53.8

Race/ethnic background
  African American                    2    0.7
  Asian American                      9    3.3
  Caucasian                         247   91.8
  Hispanic                            2    0.7
  Native American                     2    0.7
  Other                               7    2.6

(a) APTA=American Physical Therapy Association.

Table 2.
Internal Consistency and Test-Retest Reliability of Scales Used
in the Study

                                                        Subjects
                                          Test-Retest   Used in
                               Cronbach   Reliability   Reliability
Scale                          Alpha      (a)           Study

Job Satisfaction Survey (28)   .91        .71           43 employees
Organizational Commitment
  Survey (40) (Affective
  Scale)                       .85        .94           NA (b)
Occupational Commitment
  Survey (34) (Affective
  Scale)                       .82        .71           NA
Role conflict (32)             .82        .71           148 employees
Role ambiguity (32)            .78        .65           148 employees
Role overload (31)             .73        .65           NA

(a) Statistic used to measure test-retest reliability not identified.

(b) NA=sample size not available from reference source.

Table 3.
Factor Analysis of the Clinicians' Role Behaviors

                                       Factors

                                                              Evaluate
Role Behaviors                         Integrate   Interact   and Plan

Increased integration of physical
  therapists into multidisciplinary
  teams                                   .736
Increased communication/
  collaboration with other health
  care professionals                      .696
Increased use of critical pathways/
  care paths to guide care                .678
Increased role as a consultant,
  specialist, or advanced clinician       .635
Continue to interact with and assist
  other physical therapists with
  patient care                                       .726
Continue to interact socially with
  other physical therapists in the
  facility                                           .625
Continue to attend staff meetings                    .462
Teach groups in the community                        .447
Increased documentation of the
  results of patient care                                       .761
Increased focus on the functional
  needs of patients                                             .672
Increased focus on efficiency/
  productivity in the performance of
  activities                                                    .606
Increased patient evaluation and
  program planning                                              .580
Maintain focus on providing direct
  patient care
Increased delegation and supervision
  of physical therapy treatment
Participate in the clinical
  education of students
Participate in self-education
Be involved in professional
  activities outside of work
Decreased ability to plan and
  control how the work will be done
Increased work on weekends/holidays
  on a rotating schedule
Increased work in more than one
  clinical area of the hospital or
  more than one site
Spend time with patients providing
  care other than physical
  therapy (eg, nursing duties)
Increase in administrative
  activities

% variance explained                    11.283      9.862      9.682

Eigenvalues                              2.482      2.170      2.130
Multiple correlation ([R.sup.2])          .841       .722       .872

                                       Factors

                                                         Organizational
Role Behaviors                         Care    Educate   Responsibility

Increased integration of physical
  therapists into multidisciplinary
  teams
Increased communication/
  collaboration with other health
  care professionals
Increased use of critical pathways/
  care paths to guide care
Increased role as a consultant,
  specialist, or advanced clinician
Continue to interact with and assist
  other physical therapists with
  patient care
Continue to interact socially with
  other physical therapists in the
  facility
Continue to attend staff meetings
Teach groups in the community
Increased documentation of the
  results of patient care
Increased focus on the functional
  needs of patients
Increased focus on efficiency/
  productivity in the performance of
  activities
Increased patient evaluation and
  program planning
Maintain focus on providing direct
  patient care                          .804
Increased delegation and supervision
  of physical therapy treatment         .754
Participate in the clinical
  education of students                          .749
Participate in self-education                    .715
Be involved in professional
  activities outside of work                     .568
Decreased ability to plan and
  control how the work will be done                           .605
Increased work on weekends/holidays
  on a rotating schedule                                      .558
Increased work in more than one
  clinical area of the hospital or
  more than one site                                          .536
Spend time with patients providing
  care other than physical
  therapy (eg, nursing duties)                                .500
Increase in administrative
  activities                                                  .473

% variance explained                   8.690    8.423        7.476

Eigenvalues                            1.912    1.853        1.645
Multiple correlation ([R.sup.2])        .792     .883         .942

Table 4.
Descriptive Statistics for the Study Variables

Study Variable              [bar]X    SD    Median   [alpha]

Role behaviors scale        N/A (a)   N/A   N/A          .72
Stress
  Role overload             3.3       0.9   2.3          .78
  Role conflict             2.5       0.9   2.3          .71
  Role ambiguity            2.2       0.7   2            .59

Occupational commitment     4.5       0.6   4.7          .75

Job satisfaction            3.8       0.5   3.7          .79

Organizational commitment   3.3       0.9   3.3          .77

(a) N/A=not applicable.

Table 5.
Correlations Among the Demographic Variables and the Study Variables

                                                            Organi-
                                             Professional   zational
                                  Age        Experience     Tenure

Professional role behaviors
  Integrate
  Interact
  Evaluate and plan
  Care
  Educate                         -.13 (b)   -.16 (b)

Organizational role behaviors
  Organizational responsibility

Stress
  Role overload                               .13 (b)       .12 (b)
  Role conflict
  Role ambiguity

Occupational commitment

Job satisfaction

Organizational commitment                                   .18 (c)

                                  Working in
                                  Area of Interest   APTA (a)

Professional role behaviors
  Integrate
  Interact
  Evaluate and plan
  Care
  Educate                                            .14 (b)

Organizational role behaviors
  Organizational responsibility

Stress
  Role overload                   -.14 (b)
  Role conflict                   -.15 (b)
  Role ambiguity

Occupational commitment

Job satisfaction                   .15 (b)

Organizational commitment          .15 (b)

(a) APTA=American Physical Therapy Association.

(b) Correlation is significant at the .05 level (2-tailed t test).

(c) Correlation is significant at the .01 level (2-tailed t test).

Table 6.
Correlations Among the Study Variables

                                  Role       Role       Role
                                  Overload   Conflict   Ambiguity

Professional role behaviors
  Integrate                                  -.18 (a)   -.14 (b)
  Interact                        -.22 (a)   -.23 (a)   -.17 (a)
  Evaluate and plan
  Care                             .23 (a)    .14 (b)
  Educate                         -.21 (a)              -.13 (b)

Organizational role behaviors
  Organizational responsibility    .33 (a)    .29 (a)    .22 (a)

Stress
  Role overload
  Role conflict
  Role ambiguity

Occupational commitment

                                                 Job        Organi-
                                  Occupational   Satis-     zational
                                  Commitment     faction    Commitment

Professional role behaviors
  Integrate                                       .21 (a)
  Interact                         .17 (a)        .29 (a)    .28 (a)
  Evaluate and plan
  Care
  Educate                          .20 (a)        .23 (a)    .19 (a)

Organizational role behaviors
  Organizational responsibility                  -.26 (a)   -.17 (a)

Stress
  Role overload                   -.20 (a)       -.45 (a)   -.29 (a)
  Role conflict                   -.21 (a)       -.60 (a)   -.41 (a)
  Role ambiguity                  -.22 (a)       -.54 (a)   -.43 (a)

Occupational commitment                           .37 (a)    .40 (a)

(a) Correlation is significant at the .01 level (2-tailed t test).

(b) Correlation is significant at the .05 level (2-tailed t test).

Table 7.
Regression Analysis: Prediction of Job Satisfaction from the
Independent Variables

                                                           Change
                                Model Summary              Statistics

                                Standardized
                                [beta]         Adjusted    [R.sup.2]
Variable (Constant)             Coefficient    [R.sup.2]   Change

Age                             -.009          .018        .018
Professional experience          .043
Organization tenure              .060
Working in area of interest      .023
APTA (a)                        -.051

Role ambiguity                  -.147 (b)      .457        .439
Role conflict                   -.364 (b)
Role overload                   -.161 (b)

Integrate                        .112 (c)      .519        .062
Interact                         .126 (b)
Evaluate and plan                .091 (c)
Care                             .085
Educate                          .124 (b)
Organizational responsibility   -.053

Occupational commitment          .172 (b)      .544        .025

                                Change
                                 Sta-
                                 tis-
                                 tics

                                F        Significant
Variable (Constant)             Change   F Change      df

Age                              1.926   .091           5,253
Professional experience
Organization tenure
Working in area of interest
APTA (a)

Role ambiguity                  27.657   .000           8,253
Role conflict
Role overload

Integrate                       20.488   .000          14,253
Interact
Evaluate and plan
Care
Educate
Organizational responsibility

Occupational commitment         21.093   .000          15,253

(a) APTA=American Physical Therapy Association.

(b) Coefficient is significant at the .01 level (2-tailed t test).

(c) Coefficient is significant at the .05 level (2-tailed t test).

Table 8.
Regression Analysis: Prediction of Organizational Commitment From
the Independent Variables

                                                           Change
                                Model Summary              Statistics

                                Standardized
                                [beta]         Adjusted    [R.sup.2]
Variable (Constant)             Coefficient    [R.sup.2]   Change

Age                              .047          .049        .049
Professional experience         -.224 (b)
Organization tenure              .293 (c)
Working in area of interest      .049
APTA (a)                        -.005

Role ambiguity                  -.217 (c)      .240        .191
Role conflict                   -.110 (b)
Role overload                   -.034

Integrate                        .005          .286        .046
Interact                         .173 (c)
Evaluate and plan                .038
Care                             .009
Educate                          .108 (b)
Organizational responsibility   -.071

Occupational commitment          .250 (c)      .339        .053

                                Change
                                 Sta-
                                 tis-
                                 tics

                                F        Significant
Variable (Constant)             Change   F Change      df

Age                              3.621   .004           5,253
Professional experience
Organization tenure
Working in area of interest
APTA (a)

Role ambiguity                  10.984   .000           8,253
Role conflict
Role overload

Integrate                        8.235   .000          14,253
Interact
Evaluate and plan
Care
Educate
Organizational responsibility

Occupational commitment          5.325   .000          15,253

(a) APTA=American Physical Therapy Association.

(b) Coefficient is significant at the .05 level (2-tailed t test).

(c) Coefficient is significant at the .01 level (2-tailed t test).

Table 9.
Results of Hypothesis Testing

                                                    Organizational
Hypothesis                    Job Satisfaction      Commitment

1. The professional role      Partially supported   Minimally supported
   behaviors will be            (3 of 5)              (2 of 5)
   positively correlated
   with job satisfaction
   and organizational
   commitment.
2. The professional role      Partially supported   Minimally supported
   behaviors will make a        (4 of 5)              (2 of 5)
   positive contribution to
   the prediction of job
   satisfaction and
   organizational
   commitment.
3. The organizational role    Supported (1 of 1)    Supported (1 of 1)
   behaviors will be
   negatively correlated
   with job satisfaction
   and organizational
   commitment.
4. The organizational role    Not supported         Not supported
   behaviors will make a
   negative contribution to
   the prediction of job
   satisfaction and
   organizational
   commitment.
5. The level of               Not supported         Not supported
   occupational commitment
   will interact with and
   positively influence the
   relationship between the
   role behaviors and both
   job satisfaction and
   organizational
   commitment.
6. Role conflict, role        Supported (3 of 3)    Supported (3 of 3)
   overload, and role
   ambiguity will be
   negatively correlated
   with job satisfaction
   and organizational
   commitment.
7. Role conflict, role        Supported (3 of 3)    Partially supported
   overload, and role                                 (2 of 3)
   ambiguity will make a
   negative contribution to
   the prediction of job
   satisfaction and
   organizational
   commitment.

References

(1) Arndt M, Bigelow B. Reengineering: deja vu all over again. Health Care Management Review. 1998;23(3):58-66.

(2) D'Aunno T, Alexander JA, Laughlin C. Business as usual? Changes in health care's workforce and organization of work. Hospital and Health Service Administration. 1996;41(1):3-18.

(3) Leatt P, Baker GR, Halverson PK, Aird C. Downsizing, reengineering, and restructuring: long-term implications for healthcare organizations. Frontiers in Health Services Management. 1997;13(4):3-37, 52-54.

(4) Johnson RA. Antecedents and outcomes of corporate refocusing. Journal of Management. 1996;22:439-483.

(5) Effken JA, Stetler CB. Impact of organizational redesign. Journal of Nursing Administration. 1997;27(7/8):23-32.

(6) Kinneman MT, Hitchings KS, Bryan MA, et al. A pragmatic approach to measuring and evaluating hospital restructuring efforts. Journal of Nursing Administration. 1997;27(7/8):33-41.

(7) Davidson H, Folcarelle PH, Crawford S, et al. The effects of health care reforms on job satisfaction and voluntary turnover among hospital-based nurses. Med Care. 1997;35:634-645.

(8) Katz D, Kahn R. The Social Psychology of Organizations. 2nd ed. New York, NY: John Wiley & Sons Inc; 1978: chap 7.

(9) Fishbein M, Ajzen I. Belief, Attitude, Intention and Behavior: An Introduction to Theory and Research. Reading, Mass: Addison-Wesley; 1975.

(10) Lopopolo RB. The effect of hospital restructuring on the roles of physical therapists in acute care. Phys Ther. 1997;77:918-932.

(11) Lopopolo RB. Hospital restructuring and the changing nature of the physical therapist's role. Phys Ther. 1999;79:171-185.

(12) Shindul-Rothschild J, Duffy M. The impact of restructuring and work design on nursing practice and patient care. Best Practices and Benchmarking in Healthcare. 1996;1:271-282.

(13) Georgopoulos BS. Distinguishing organizational features of hospitals. In: Weiland GF, ed. Improving Health Care Management. Ann Arbor, Mich: Health Administration Press; 1981: chap 1.

(14) Lepine S, Ahola-Sidaway J. An exploratory study of trends in hospital nursing and their impact on professional development initiatives. Paper presented at: 3rd International Interdisciplinary Qualitative Health Research Conference; October 30-November 1, 1996; Bournemouth, England.

(15) Lopopolo RB. Development of the Professional Role Behaviors Survey (PROBES). Phys Ther. 2001;81:1317-1327.

(16) Schwiekhart SB, Smith-Daniels V. Reengineering the work of caregivers: role redefinition, team structures, and organizational redesign. Hospitals and Health Services Administration. 1996;41(1):19-35.

(17) Wynn KE. Embracing change: hospital restructuring revisited. PT Magazine. 1997;5(1):38-50.

(18) Sochalski J, Aiken LH, Fagin CM. Hospital restructuring in the United States, Canada, and Western Europe. Med Care. 1997;35: OS13-OS25.

(19) Armstrong-Stassen M, Cameron SJ, Horsburgh ME. The impact of organizational downsizing on the job satisfaction of nurses. Canadian Journal of Nursing Administration. November-December 1996:8-32.

(20) Porras JI. Hoffer SJ. Common behavioral changes in successful organization development efforts. Journal of Applied Behavioral Science. 1986;22:477-494.

(21) Robertson PJ, Roberts DR, Porras JI. Dynamics of planned organizational change: assessing empirical support for a theoretical model. Academy of Management Journal. 1993;36:619-634.

(22) Lopopolo RB. The relationship between practitioner role change and outcomes in the restructured hospital environment. Dissertation Abstracts. July 29, 2000.

(23) Strauss AL. Cited by: Vollmer HM, Mills DL. Professionalization. Englewood Cliffs, NJ: Prentice-Hall; 1966.

(24) Blegen MA. Nurses' job satisfaction: a meta-analysis of related variables. Nursing Research. 1993;42(1):36-41.

(25) Mathieu JE, Zajac D. A review and meta-analysis of the antecedents, correlates, and consequences of organizational commitment. Psychol Bull. 1990;108:171-194.

(26) Locke EA. The nature and causes of job satisfaction. In: Dunnette, MD, ed. Handbook of Industrial and Organizational Psychology. Chicago, Ill: Rand-McNally College Publishing Co; 1976: chap 30.

(27) Price JL, Muller CW. Absenteeism and Turnover Among Hospital Employees. Greenwich, Conn: JAI Press; 1986.

(28) Spector PE. Job Satisfaction: Application, Assessment, Causes, and Consequences. Thousand Oaks, Calif: Sage Publications; 1997.

(29) Meyer JP, Allen NJ. Commitment in the Workplace: Theory, Research, and Application. Thousand Oaks, Calif: Sage Publications; 1997.

(30) Chassin MR, Galvin RW, the National Roundtable on Health Care Quality. The urgent need to improve health care quality. JAMA. 1998;280:100-105.

(31) Kim S, Price J, Mueller C, Watson T. The determinants of career intent among physicians at a US Air Force hospital. Human Relations. 1996;49:947-976.

(32) Rizzo JR, House RJ, Lirtzman SI. Role conflict and ambiguity in complex organizations. Administrative Science Quarterly. 1970;15: 150-163.

(33) Lathrop JP, Seufert RJ, MacDonald RJ, Martin S. The patient-focused hospital: a patient care concept. Journal of the Society for Health Systems. 1991;3(2):33-50.

(34) Meyer JP, Allen NJ, Smith CA. Commitment to organizations and occupations: extension and test of a three-component conceptualization. Journal of Applied Psychology. 1993;78:538-551.

(35) Blau G, Paul A, St John N. On developing a general index of work commitment. J Vocat Behav. 1993;42:298-314.

(36) Vandenberg RJ, Scarpello V. A longitudinal assessment of the determinant relationship between employee commitments to the occupation and the organization. Journal of Organizational Behavior. 1994;15:535-547.

(37) Blau G. Testing the generalizability of a career commitment measure and its impact on employee turnover. J Vocat Behav. 1989;35: 88-103.

(38) Irving PG, Coleman DF, Cooper CL. Further assessment of a three component model of occupational commitment: generalizability and differences across occupations. Journal of Applied Psychology. 1997;82: 444-452.

(39) Agho AO, Mueller CW, Price JL. Determinants of employee job satisfaction: an empirical test of a causal model. Human Relations. 1993;46:1007-1027.

(40) Allen NJ, Meyer JP. Affective, continuance, and normative organizational commitment: an examination of construct validity. J Vocat Behav. 1996;49:252-276.

(41) McDermott K, Spence-Laschinger H, Shamian J. Work empowerment and organizational commitment. Nursing Management. 1996;27: 44-47.

(42) Cook JD, Hepworth SJ, Wall TD, Warr PB. The Experience of Work: A Compendium and Review of 249 Measures and Their Use. New York, NY: Academic Press; 1981.

(43) DeVellis RF. Scale Development: Theory and Application. Thousand Oaks, Calif: Sage Publications; 1991.

(44) Nunnally JC, Bernstein IH. Psychometric Theory. 3rd ed. New York, NY: McGraw-Hill Inc; 1994.

(45) Hackman JR, Oldham GR. Development of the Job Diagnostic Survey. Journal of Applied Psychology. 1975;60:159-170.

(46) Hair JF, Anderson RE, Tatham RL, Black WC. Multivariate Data Analysis. 3rd ed. New York, NY: Macmillan Publishing Co; 1992.

(47) SPSS Base 8.0 Applications Guide. Chicago, Ill: SPSS Inc; 1998.

(48) Hackett RD, Bycio P, Hausdorf PA. Further assessment of Meyer and Allen's (1991) three-component model of organizational commitment. Journal of Applied Psychology. 1994;79:15-23.

(49) Guide to Physical Therapist Practice. 2nd ed. Phys Ther. 2001;81: 9-744.

(50) Bethel S, Ridder J. Evaluating nursing practice: satisfaction at what cost? Nursing Management. September 1994:441-448.

(51) Wallace JE. Professional and organizational commitment: compatible or incompatible? J Vocat Behav. 1993;42:333-349.

(52) Robinson JP, Shaver PR, Wrightsman LS. Measures of Personality and Social Psychological Attitudes. New York, NY: Academic Press; 1991.

RB Lopopolo, PT, PhD, MBA, is Associate Professor, Department of Physical Therapy, Arcadia University, 450 S Easton Rd, Glenside, PA 19038 (USA) ([e-mail]).

The Temple University Institutional Review Board approved this study. The rights of human subjects were protected.

This article was submitted January 25, 2002, and was accepted April 9, 2002.