Three perspectives on physical therapist managerial work. (Research Report).

As physical therapy practice evolves, the roles and skills of physical therapist managers who oversee clinical work must keep pace. Efforts to accurately describe physical therapy clinical practice are well under way, (1,2) but little has appeared in the physical therapy literature that describes the nature of physical therapist managerial work. The Guide to Physical Therapist Practice (2nd ed) (1) (Guide) provides a framework for organizing the work of the physical therapist who provides direct patient care, while providing only general descriptions of the work of physical therapy clinicians who take on managerial responsibilities. The Guide describes administration as a "skilled process of planning, directing, organizing, and managing human, technical, environmental, and financial resources effectively and efficiently" and includes "the management ... of resources for patient/client management and for organization operations." (1 (p49)) A Normative Model of Physical Therapist Education (2(pp128-129)) describes curricular content in administration and management, suggesting that new graduates should be able to delegate to and supervise or manage support personnel, as well as participate in planning, budgeting, billing, reimbursement, and marketing activities.

Fayol (3,4) introduced the concept of managerial work in 1916 by describing a manager as a person who plans, organizes, commands, coordinates, and controls. These descriptors continue to dominate the management vocabulary, appearing as chapters in management textbooks over the years. (5) Mintzberg (6,7) offered an alternative framework in the early 1970s that was supported by data. He used structured observations of 5 chief executive officers to describe 10 managerial roles, arranging these roles into 3 broad categories: interpersonal roles (figurehead, leader, and liaison), informational roles (monitor, disseminator, and spokesman), and decisional roles (entrepreneur, disturbance handler, resource allocator, and negotiator). Although Mintzberg's typology has been challenged over the years, (8,9) it has achieved classical status in the management literature. At about the same time that Mintzberg was defining managerial roles, Katz (10) also used structured observations to identify what administrators actually do. Katz advocated a skills-based approach, suggesting that all administrators use a minimum of 3 skills--technical, human, and conceptual--in executing their work.

Pavett and Lau (11) applied Mintzberg's (7) 10-role model to compare research and development (R&D) managers and non-R&D managers from both the federal government and the private sector. Based on interviews with public sector executives, they added an 11th role--technical expert--to those proposed by Mintzberg. Pavett and Lau created a 50-item questionnaire, with a minimum of 4 activities describing each of the 11 roles. They queried 438 managers on the importance of each item to their managerial job. Pavett and Lau found that, regardless of work setting, the most important roles were leader, resource allocator, and disseminator. They concluded that these roles did not appear to be career specific and could be applied to any manager in any industry.

In the absence of research describing the work of health care managers, Roemer (12) surveyed hospital-based middle managers in New England with a modified version of the Pavett and Lau (11) questionnaire. Roemer reported that some of these managers were rehabilitation directors, but did not identify the specific discipline. Roemer's work categories consisted of 11 roles (Mintzberg's 10 roles plus Pavet and Lau's technical expert role), 3 skills based on Katz's work, (10) and 2 skills derived from her own pilot study. The 5 skill categories were communications, operations, financial control, interpersonal relations, and strategic assessment. (12) Using a 75-item questionnaire, Roemer found that hospital-based middle managers (n = 196) rated the communication skill and the leader role as most important, while the roles of figurehead, technical expert, and spokesperson were rated as least important.

Even though no data are available that relate to physical therapist managerial work, I believe that interest in this area of study is growing. For example, in early 1999, the leadership, administration, and management preparation (LAMP) task force of the American Physical Therapy Association's (APTA's) Section on Administration reported its findings. (13) Using the Guide to Physical Therapist Practice (14) and A Normative Model of Physical Therapist Professional Education (2) as references, along with their own expertise, the task force produced a model of business problem solving that parallels that of clinical problem solving found in the Guide (examination, evaluation, diagnosis, prognosis, intervention, outcomes). Using this model, the LAMP task force developed a comprehensive list of potential content areas that could be addressed by faculty members and others in preparing physical therapists to become managers. At about the same time, another task force published a competency assessment of the physical therapy manager. (15) This group of physical therapy directors produced an assessment in the form of a checklist in response to various accreditation agencies that required competency assessment of all employees, including managers. Unfortunately, the validity of data obtained with these tools remains untested.

A thorough understanding of the scope of physical therapist managerial work, based on data, is necessary before identifying the competencies that should be addressed by professional (entry-level) education programs for physical therapists. In addition, health care administrators need to measure the competence of physical therapists in their managerial (versus clinical) duties as required by accreditation agencies. In order to target relevant physical therapist managerial competencies, agreed-upon work categories first should be identified. I began the process by measuring the relative importance that physical therapist managers and faculty members place on specified work categories.

I used Roemer's questionnaire (12) to compare the perceptions of faculty members in professional physical therapist education programs and 2 groups of physical therapist managers (hospital and private practice) on the importance of a set of managerial work categories (roles and skills) that had been tested in other industries. I also compared the responses among the 3 groups to determine whether the perceived importance placed on these work categories varied among work settings.

Methods

Three groups were simultaneously surveyed using the same instrument. (16-18) The first group consisted of faculty members in professional physical therapist education programs. One faculty member from each accredited professional physical therapist education program in the United States was solicited to participate in the study. I asked that this faculty member have taught management/administration content to professional physical therapist students. All accredited US programs as of October 1997 (n = 155) were included in this group. The second group consisted of 300 physical therapists who held management positions in US hospitals. This sample was randomly selected from the 672 members of APTA's Section on Administration who reported working in hospital settings in November 1997. The third group consisted of physical therapists who held management positions in private practices in the United States. A sample of 375 names was randomly selected from the total of 3,388 members listed in the 1997 membership directory of APTA's Private Practice Section.

The variables of interest were 11 work roles (disseminator, disturbance handler, entrepreneur, figurehead, leader, liaison, monitor, negotiator, resource allocator, spokesperson, technical expert) and 5 work skills (communication, financial control, interpersonal relations, operations, and strategic assessment) as defined by Roemer. (12) Each role or skill category was described by the same activities that Roemer used in her study. For example, an activity in the financial control category was "analyzing and using financial reports," and an activity in the negotiator category was "resolving disputes among subordinates or work groups." Faculty members were asked to rate how important they thought each activity was to physical therapist managers; both clinical manager groups were asked to rate the importance of each activity as part of their job.

Each activity (n = 75) was measured on a Likert-type scale of importance anchored by scores of 1 ("no importance") and 5 ("great importance"). Scores for activities within each of the 16 categories were averaged. The rank order of these averaged scores was explored by group in order to gain insights on how each group perceived the importance of each work category. Then, the same averaged scores became the dependent variables in a multivariate analysis designed to determine differences in perceived role or skill importance among the 3 groups.

Roemer (12) reported that activity assignments to each of the 16 categories demonstrated internal consistency (alpha coefficients > .60), but no data on reliability were reported. Construct validity was demonstrated in a study by Pavett and Lau, (11) who used factor analysis to determine which activities best described each work category. Therefore, all 75 activities of Roemer's survey were used in the present study with no edits.

The survey questionnaire with a cover letter was mailed to the selected sample from each group as described earlier. A follow-up reminder postcard was mailed 1 month later. Data were aggregated for each group by calculating the average importance scores of each of 16 work categories. Then, a multivariate analysis of variance (MANOVA) F value was calculated using the 16 work categories as the dependent variables and the 3 respondent groups as the grouping variable. A probability level of .05 was used to determine significance. Post hoc analyses were used to identify differences among groups when significant univariate F ratios were found. All data were analyzed using version 9.0 of SPSS. *

The total number of usable questionnaires returned was 343, for an overall response rate of 41.3%. The physical therapist education program faculty member response was 61 out of 155, representing a 39.4% within-group response rate. The private practice physical therapist manager response was 121 out of 375 (32.3% within-group response rate), whereas the hospital-based physical therapist manager response was 161 out of 300 (53.7% within-group response rate). These response rates were considered acceptable because they fell between 30% and 60%, which is considered realistic for mailed surveys. (19)

Of the 61 professional physical therapist education programs represented, 15 offered the bachelor's degree and 46 offered the master's degree. Fewer than half (42%) of the instructors of the administration/management content had a degree in management. The average number of years teaching this content was 6.2 years (range = 1-24).

The average practicing manager respondent had over 10 years of management experience, with private practice managers averaging 16 years (range = 1-47) and hospital-based managers averaging 11.5 years (range = 1-34). The vast majority of respondents reported receiving on-the-job management training (91% hospital, 86% private practice); continuing education (54% and 36%, respectively) and postprofessional education (31% and 21%, respectively) were the next most common ways to obtain management training. Over 90% of the private practice respondents were managers or owners of their clinics, spent 68% of their time in indirect patient care, and were predominantly male (60%). In the hospital setting, respondents reported varying degrees of managerial responsibility (ie, 31% were supervisors, 54% were middle managers, and 15% were administrators). Most (67%) of the hospital-based respondents were female.

Post hoc analyses to test the assumption of the internal consistency of each work category were conducted using the combined responses from all 3 groups. Results revealed that all 16 categories demonstrated acceptable internal consistency (alpha coefficients [greater than or equal to] .80 for 9 categories, [greater than or equal to] .70 for 13 categories, and [greater than or equal to] .64 for all categories). (19) The 3 categories with the lowest alpha coefficients were strategic assessment (.67), technical expert (.67), and communication (.64).

Results

The group means of importance scores by work category are depicted in Table 1. The averaged importance scores across all 3 groups are arranged in descending order from highest to lowest according to the unweighted mean score (sum of group means divided by 3). The results indicate that all 3 groups included communication, financial control, entrepreneur, resource allocator, and leader in their 5 most important categories, but not necessarily in the same rank order. For example, faculty members rated financial control as the most important category, whereas both clinical manager groups rated communication as the most important. When considering the least important categories, however, there was agreement among the groups concerning the bottom 2 categories. Without exception, figurehead was rated as least important, and technical expert was ranked second lowest.

The within-group rankings of work category importance appeared to be similar; however, the between-group comparisons of mean scores by category demonstrated noteworthy patterns (Figure). For example, private practice managers rated the importance of each category lower than either of the other groups, except for figurehead, whereas faculty members rated all work categories higher than the other groups, except for entrepreneur. To further explore these patterns, a MANOVA was done to test whether differences existed among groups on the mean scores of the 16 categories. The MANOVA was selected instead of 16 separate analyses of variance (ANOVAs) to reduce the probability of committing a type I error. (19)

[FIGURE OMITTED]

The MANOVA produced a Wilks lambda, which was converted to an estimated F statistic of 7.76 (P < .001), and a significant Box M test (P < .001), which indicated that the covariance matrices of the dependent variables were not equal across groups. (20) Because the omnibus F level was significant, univariate ANOVAs were conducted for each work category to determine where among the dependent variables the differences existed. (21) Significant univariate F scores were found in all categories except for figurehead (Tab. 2). Because the Levene test for equality of error variances was found to be significant (P < .001) in 13 of the 16 work categories, the Tamhane T2 multiple-comparison test was selected to determine the locations of differences among groups in each of the work categories. The Tamhane test can be applied in situations where variances or sample sizes are unequal. (20)

The results of the multiple-comparison tests for the 15 work categories that demonstrated significant univariate results are displayed in Table 3. The results suggest that (1) the 3 groups differed from each other on the level of importance for 4 work categories (leader, interpersonal relations, disturbance handler, and negotiator), (2) faculty members did not differ from hospital-based managers on level of importance, but did differ from private practice managers, for 7 work categories (communication, entrepreneur, operations, liaison, disseminator, spokesperson, and technical expert), and (3) the 2 groups of clinical managers did not differ on level of importance with each other, but did differ from faculty members, for 4 categories (financial control, resource allocator, strategic assessment, and monitor). Stated another way (Tab. 4), the 15 comparisons showed differences between faculty members and private practice managers in all 15 categories, 9 of 15 comparisons showed differences between hospital-based managers and private practice managers, and 8 of 15 comparisons showed differences between faculty members and hospital-based managers.

Discussion

Hierarchy of Work Categories

Communication was considered the most important work category when all 3 groups were considered together (Tab. 1) and was considered the most important category by both groups of clinical managers. Faculty members considered financial control to be most important, with communication a close second. Entrepreneur was considered the second most important category by hospital-based managers, the third most important by private practice managers, and fifth most important by faculty members. Resource allocator and leader were also among the 5 most important categories for all 3 groups. In Roemer's study, (12) hospital-based middle managers rated communication as most important, followed by leader, disturbance handler, entrepreneur, and operations. Pavett and Lau (11) identified the leader role as most important for both public and private sector managers. Their study did not include the 5 work skills of communication, financial control, operations, strategic assessment, and interpersonal relations. Technical expert and figurehead were considered the least important work categories by all 3 groups in the present study (Tab. 1). Roemer (12) reported similar findings, whereas Pavett and Lau (11) reported figurehead as the second least important category and negotiator as the least important category. These findings suggest the universality of some components of managerial work, such as communication and leader, across study populations, perhaps industries.

An unexpected pattern that emerged from the data in this study is that private practice managers consistently scored work categories as less important when compared with the other 2 groups (Figure). In addition, faculty members consistently scored the work categories as more important than both clinical manager groups, whereas hospital-based managers' scores were consistently between the other 2 groups' scores, with only one exception (entrepreneur). These patterns may be a function of the amount of time that the respondents actually engaged in managerial work in their respective work settings. In the context of their daily work, private practice managers may have to consider managerial work issues along with other clinical practice expectations that may compete for their attention because an average of 68% of their time was spent in direct patient care. Hospital-based managers, however, may encounter managerial issues more often than private practice managers but less often than faculty members think they do. The hospital-based managers in this study held positions at varying levels within their organizations (supervisor, middle manager, administrator), which may have confounded the results for this group. Preliminary analyses suggest that hospital-based managers at different levels in the organization perceive the importance of some work categories differently (DSS, unpublished data, 1999). Yet another explanation for these differences in patterns may be that the private practice group scores had much larger standard deviations in most categories when compared with the hospital-based manager scores, and especially when compared with faculty member scores. The heterogeneity of the private practice group may have resulted in lower scores. Overall, however, these results suggest that the work setting itself may be a factor in shaping the physical therapist manager's perception of work category importance.

Statistical Comparisons Among Work Settings

Upon further investigation of the 16 managerial work categories across the 3 physical therapist groups, differences were found in 15 of them (Tabs. 3 and 4). Interestingly, no difference in perceived importance of the figurehead role was found among the 3 groups (Tab. 2). All 3 groups rated the figurehead role as the least important of all 16 categories, suggesting that when compared with the other 15 roles or skills, the figurehead role is not perceived as a very important component of physical therapist managerial work.

No differences were found between hospital-based managers and private practice managers on 6 of the 15 categories that demonstrated differences among groups: communication, financial control, resource allocator, strategic assessment, monitor, and spokesperson (Tab. 4). These findings suggest that these 6 categories may be considered "universal" for physical therapist clinical managers in general, because their level of importance does not appear to be affected by type of clinical setting. The strongest case for universality may be made for communication, financial control, and resource allocator, because these work categories were ranked highly by all 3 groups.

Differences were found between hospital-based managers and private practice managers on the other 9 categories: entrepreneur, leader, operations, interpersonal relations, disturbance handler, negotiator, liaison, disseminator, and technical expert. Without exception, hospital-based managers rated the importance of these categories significantly higher than private practice managers. The results for the role of entrepreneur were the most surprising, because entrepreneurial activity is associated more with individuals who work in community-based private practices. Hospital-based managers' perceived high regard for entrepreneurialism may relate to the importance they place on intrapreneurialism (ie, being entrepreneurial within an organization). Regardless, these findings again suggest that type of clinical setting may influence the type of managerial roles or skills that are important to function in a particular setting. They also suggest that hospital-based managers may use these particular roles or skills more frequently or in a different way than private practice managers. Future studies should thoroughly differentiate type of setting in order to better understand the roles and skills of physical therapist managers within them. Factors such as age of the enterprise, the physical therapist manager's level within the organization (supervisor, middle manager, administrator), amount of time spent in direct patient care, and number of employees may affect the perceived importance of managerial roles and skills.

The viewpoint of faculty members as compared with both clinical manager groups is intriguing. Faculty members not only rated 11 of the 16 categories as 4 on a 5-point scale of importance, but they also rated all except one category higher than both clinical manager groups. Faculty members' ratings were higher than those of private practice managers on 15 of 16 work categories. Their ratings were higher than those of hospital-based managers on only 8 categories. These results suggest that the importance faculty members place on managerial roles and skills are more similar to those of hospital-based physical therapist managers than to those of private practice managers. These results may simply be related to the distribution of the data and warrant no further discussion. Other uncontrolled factors may be exerting their influence (eg, previous experience as a hospital-based manager or possession of a broader [systems] view of management that resulted in most work categories being considered very important). An equally likely interpretation of these results, however, is that faculty members responded to the importance that they themselves place on each work activity. They may have forgotten to follow the instruction to answer the survey in terms of what they thought was important to the clinical manager. In light of these multiple interpretations, further speculation about faculty members' scores in this study must remain guarded.

Conclusions

This study is the first to provide data on physical therapist managerial tasks and their perceived relative importance. Like previous studies on different managerial groups, the results confirm that communication is the most important work skill and the figurehead component is the least important role. In addition to communication, financial control, entrepreneur, resource allocator, and leader were considered the most important managerial work categories regardless of work setting. These findings suggest that these 5 categories are the ones that may be universal for physical therapist managers.

An unexpected finding was that work setting (hospital, private practice, higher education) appears to have an impact on the degree of importance placed on physical therapist managerial work categories. Marked differences were found among groups on 15 of 16 work categories, suggesting that future researchers should control for work setting when studying managerial work categories. Of note is the finding that the 2 clinical managerial groups agreed on the level of importance of 6 work categories: communication, financial control, resource allocator, strategic assessment, monitor, and spokesperson. Only the first 3 of these categories, however, appear in the top 5 noted for all work settings in this study. Researchers need to determine whether other work categories are "universal" to physical therapist managers regardless of work setting, because those used in this study may not represent an exhaustive set of choices.

The meaning of the agreement/disagreement findings between faculty members and both clinical manager groups remains unclear. In my opinion, however, the viewpoint of faculty members should not be disregarded in future research, because they are the ones who determine the managerial content taught in physical therapist education programs. Faculty members' decisions should be in line with what all potential clinical managers need to know regardless of setting.

This study contributes to the process of identifying a universal description of physical therapist managerial work. Researchers should take into account the potential influence of the practice setting, managerial hierarchy within an organization, and other factors on managerial work. By obtaining direct evidence from practicing managers on what physical therapist managers actually do rather than their perceptions, researchers will be better able to develop relevant measures of managerial competence. In addition, faculty members will be better able to develop curricula that include appropriate management content areas and will contribute to prepare physical therapists to be effective in their nonclinical, managerial roles.

Table 1.
Group Means by Work Category With Order Based on Unweighted Means
Across Groups

                       Importance Mean Scores, Standard Deviations,
                       and Ranks

                       Faculty Members          Hospital-Based
                       (n=61)                   Managers (n=161)

Work Category          X          SD     Rank   X      SD     Rank

Communication          4.50       0.51    2     4.44   0.59    1
Financial control      4.53       0.53    1     4.29   0.82    3
Entrepreneur           4.33       0.57    5     4.41   0.54    2
Resource allocator     4.40       0.49    3     4.10   0.69    5
Leader                 4.39       0.52    4     4.14   0.62    4
Operations             4.13       0.57   10     4.03   0.56    6
Strategic assessment   4.26       0.56    6     3.86   0.64    9
Interpersonal
  relations            4.19       0.55    8     3.92   0.57    7.5
Disturbance handler    4.16       0.52    9     3.81   0.72   11
Negotiator             4.24 (a)   0.55    7     3.75   0.71   12
Liaison                4.04       0.60   11     3.92   0.59    7.5
Monitor                3.98 (a)   0.66   12     3.64   0.66   13
Disseminator           3.94       0.58   13     3.83   0.56   10
Spokesperson           3.67       0.70   14     3.45   0.68   14
Technical expert       3.53       0.70   15     3.42   0.62   15
Figurehead             3.21       0.80   16     2.97   0.70   16

                        Importance Mean
                        Scores, Standard
                        Deviations, and
                        Ranks

                        Private Practice
                        Managers (n=121)         Unweighted

Work Category           X          SD     Rank   X      Rank

Communication           4.27       0.68    1     4.40    1
Financial control       4.15       0.73    2     4.32    2
Entrepreneur            4.07 (b)   0.72    3     4.27    3
Resource allocator      3.94       0.79    4     4.15    4
Leader                  3.86 (b)   0.74    5     4.13    5
Operations              3.79       0.72    6     3.98    6
Strategic assessment    3.71       0.75    7     3.94    7
Interpersonal
  relations             3.44       0.75    9.5   3.85    8
Disturbance handler     3.55       0.91    8     3.84    9
Negotiator              3.36 (b)   0.83   12     3.78   10
Liaison                 3.38 (b)   0.89   11     3.78   11
Monitor                 3.44 (b)   0.79    9.5   3.69   12
Disseminator            3.29       0.82   14     3.69   13
Spokesperson            3.35 (b)   0.78   13     3.49   14
Technical expert        3.18       0.81   15     3.38   15
Figurehead              2.99       0.86   16     3.06   16

(a) n = 60.

(b) n = 120.
Table 2.
Univariate Tests of Work Categories by Three Groups

                          Between   Within
Work Category             MS (a)    MS       F (b)   P (c)

Communication              1.60     0.36      3.87   *
Financial control          3.27     0.55      5.22   **
Entrepreneur               4.23     0.37     10.97   ***
Resource allocator         4.82     0.48      8.70   ***
Leader                     6.33     0.42     14.04   ***
Operations                 2.99     0.38      7.27   ***
Strategic assessment       6.71     0.44     14.04   ***
Interpersonal relations   14.05     0.40     33.72   ***
Disturbance handler        7.77     0.58     13.32   ***
Negotiator                15.75     0.54     29.39   ***
Liaison                   13.59     0.50     26.29   ***
Monitor                    5.78     0.50     11.56   ***
Disseminator              12.22     0.43     29.18   ***
Spokesperson               2.10     0.52      3.83   *
Technical expert           3.21     0.50      6.01   **
Figurehead                 1.78     0.58      2.32   NS

(a) MS=mean square.

(b) F statistic derived from one-way analysis of variance
(df=2,339).

(c) P < .05, ** P < .01, *** P < .001, NS=not significant.
Table 3.
Post Hoc Multiple Comparisons for Observed Means (a)

                           Importance Mean Scores

                                      Hospital-    Private
                           Faculty    Based        Practice
Work Category              Members    Managers     Managers

Communication              4.50       4.44         4.27
Financial control          4.53       4.29         4.15
Entrepreneur               4.33       4.41         4.07
Resource allocator         4.40       4.10         3.94
Leader                     4.39       4.14         3.86
Operations                 4.13       4.03         3.79
Strategic assessment       4.26       3.86         3.71
Interpersonal relations    4.19       3.92         3.44
Disturbance handler        4.16       3.81         3.55
Negotiator                 4.24       3.75         3.36
Liaison                    4.04       3.92         3.38
Monitor                    3.98       3.64         3.44
Disseminator               3.94       3.83         3.29
Spokesperson               3.67       3.45         3.35
Technical expert           3.53       3.42         3.18

(a) Tamhane T2 post hoc test, which compares mean differences, was
used because it adjusts for unequal variances and sample sizes.
Continuous lines represent means that did not differ (P > .05).
Means that did differ (P < .05) are separated by the break(s) in
the line. Overlapping lines mean that differences were found
between faculty members and private practice managers, but no
difference was found between hospital managers and either of the
other 2 groups.
Table 4.
Levels of Significance of Post Hoc Multiple Comparisons for
Observed Means

                           Multiple Comparisons (a)

Work Category              F/H    H/P    P/F

Communication                            *
Financial control          *             ***
Entrepreneur                      ***    **
Resource allocator         ***           ***
Leader                     **     **     ***
Operations                        *      **
Strategic assessment       ***           ***
Interpersonal relations    **     ***    ***
Disturbance handler        ***    *      ***
Negotiator                 ***    ***    ***
Liaison                           ***    ***
Monitor                    *             ***
Disseminator                      ***    ***
Spokesperson                             *
Technical expert                  *      **

(a) Tamhane T2 post hoc test, which compares mean differences, was
used because it adjusts for unequal variances and sample sizes.
F=faculty member, H=hospital-based manager, P=private
practice manager. * P < .05, ** P < .01, **** P < .001, blank=not
significant.

* SPSS Inc, 444 N Michigan Ave, Chicago, IL 60611.

References

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

(2) A Normative Model of Physical Therapist Professional Education: Version 97. Alexandria, Va: American Physical Therapy Association; 1997.

(3) Fayol H. Administration Industrielle et Generale. Paris, France: Dunod; 1916.

(4) Fayol H. General and Industrial Management. London, England: Pitman; 1949.

(5) Carroll SJ, Gillen DJ. Are the classical management functions useful in describing managerial work? Acad Manage Rev. 1987;12(1):38-51.

(6) Mintzberg H. Managerial work: analysis from observation. Manage Sci. 1971;18(2):B97-B110.

(7) Mintzberg H. The Nature of Managerial Work. New York, NY: Harper & Row; 1973.

(8) Hales CP. What do managers do? A critical review of the evidence. J Manage Stud. 1986;23(1):88-115.

(9) Stewart R. Twelve Managers. Suffolk, England: Ipswich Book Co; 1994.

(10) Katz RL. Skills of an effective administrator. Harv Bus Rev. 1974; 52(5):90-102.

(11) Pavett CM, Lau AW. A comparative analysis of research and development managerial jobs across two sectors. J Manage Stud. 1985; 22(1):69-82.

(12) Roemer L. Hospital middle managers' perceptions of their work and competence. Hosp Health Serv Adm. 1996;41:210-235.

(13) Kovacek P, Powers D, Iglarsh ZA, et al. Task force on leadership, administration, and management preparation (LAMP). The Resource. 1999;29(1):8-13.

(14) Guide to Physical Therapist Practice. Phys Ther. 1997;77:1163-1650.

(15) Brewer K. Competency assessment of the physical therapy manager. The Resource. 1999;29(4):10-13.

(16) Moseley RC. Defining the Work Scope of Physical Therapist Managers: The Educator's Viewpoint [unpublished professional paper]. Denton, Tex: Texas Woman's University; 1998.

(17) Scheib M. Defining the Work Scope of Physical Therapy Managers in Private Practice Settings: A Survey [unpublished professional paper]. Denton, Tex: Texas Woman's University; 1998.

(18) White P. Defining the Work Scope of Physical Therapist Managers: Perceptions of Clinical Managers in Hospital Settings [unpublished professional paper]. Denton, Tex: Texas Woman's University; 1998.

(19) Portney LG, Watkins MP. Foundations of Clinical Research: Applications to Practice. Upper Saddle River, NJ: Prentice Hall Health; 2000.

(20) SPSS Advanced Statistics 7.5. Chicago, Ill: SPSS Inc; 1997.

(21) Domholdt E. Physical Therapy Research: Principles and Applications. Philadelphia, Pa: WB Saunders Co; 2000.

This study was reviewed by the Institutional Review Board for the Protection of Human Subjects in Research of the Dallas campus of Texas Woman's University and received exempt status.

This article was submitted January 30, 2001, and was accepted July 11, 2001.

DS Schafer, PT, PhD, is Associate Dean, School of Physical Therapy, Texas Woman's University, 8194 Walnut Hill Ln, Dallas, TX 75231 (USA) ([e-mail]).