. u . €1.33}?! {.14). Influx??? ‘ . ‘1.th \i . \ {A 0 00(0): bk. 0!. ..\I.1:.000§.|01.0(.00 :0.0.0U«..0.i~00d.u\..5i- '0 ... .011050 A MULTIPLE CORRELATION AND * A ‘IRECRESSION ANALYSIS COMPARING ' FACULTY ACTIVITY SURVEY WITH - ;- OTHER MET HODOLOGIES FOR , ,.; 71,5;‘s:':::s~:iii:;;2:; ' ALLOCATIIIC COSTS T0 COURSES 1‘1} 5335;}; _7 DIssertatIonfortheDegreeofPh I} ‘ MICHIGAN STATE UNIVERSITY WILLIAM CLARK momma - “ 1973 tiff; ...... ".1 warvw‘ ‘ " ' :I , , TI g H“ x, , ,, WEEEEEEEEEEEEEEEEEMQ/EEEII . Michigan in ate 4: 3 1293 ,00812 University This is to certify that the ‘ - 1.3 thesis entitled A MULTIPLE CORRELATION AND REGRESSION ANALYSIS COMPARING FACULTY ACTIVITY SURVEY WITH OTHER METHODOLOGIES FOR ALLOCATING COSTS TO COURSES presented by William Clark Crothers has been accepted towards fulfillment of the requirements for m—degree in DeEartment of Admin— istration and Higher Education fl/f? flmm j Major profesi/ { Date y/2 3/73 93:4 JUN llfiQQ 0 7 Wink?) 0 “2; I 99‘? W E f.§@31‘%31399 a4 w ABSTRACT A MULTIPLE CORRELATION AND REGRESSION ANALYSIS COMPARING FACULTY ACTIVITY SURVEY WITH OTHER METHODOLOGIES FOR ALLOCATING COSTS TO COURSES ‘ BY William Clark Crothers Efficient resource projection and allocation requires a thorough analysis of the faculty activities and their related costs. Effective utilization of faculty is a major responsibility of college and uni- versity administrators. Even in light of this, many institutions have little idea of how the faculty allocate their time or how this distribution effects costs. Few institutions have devised a meaningful method of gather- ing this information, analyzing the data, or costing out these activities. Only now is the concept of faculty activity analysis emerging with general interest and acceptance. To manage the cost of instruction requires the identification of the major variables of costs. Yet research on costing methodology in higher education has lacked a thorough examination assessing what results when different costing methodologies are employed. To William Clark Crothers generate a cost figure is not sufficient when it is realized that the costing.methodology will to a degree determine that cost figure. A need for research which inquires into the nature of the costing methodology used for allocating costs to the individual units chosen has been reflected in the literature. Methodology of the Study The population from which the data were drawn for this study were the faculty and course offerings of one college at Michigan State University. This college offers both undergraduate and doctoral level programs, and its five departments are diversified in both the nature of their activities and course offerings. The entire faculty were surveyed and a usable return rate of 86.3 per cent was realized. All of the data, except salaries, were collected for the fall term of 1972 through a faculty activity survey instrument designed by the National Center for Higher Education Management Systems. The objectives of this dissertation were divided into two parts. The first was to examine the inter- relationships of selected instructional workload factors through determining the distribution of faculty time among several activities and assigning costs to these activities based on the salaries of the faculty. The profile analysis was develOped by department and by rank. William Clark Crothers The second objective was to compare four costing methodologies used in allocating costs to courses. The bases of allocating costs were: (1) the total course time, (2) formal contact hours, (3) student credit hours, and (4) course credit hours. This comparison was not only based on the costs per course, but it also considered the differences in selected variables of costs and the respective costing methodology. The five variables of costs which were examined were: (1) faculty rank, (2) course level, (3) class size, (4) number of courses and/or sections taught by the faculty member, and (5) the method of instruction. The comparison of costing method- ologies as related to these variables centered on answer- ing four research questions which addressed the compara- bility of these methodologies. Findings of the Study The findings must be interpreted in light of the limitations of the study. The costs were developed from only the instructional portion of the faculty's time and from one college at one university. l. A large portion of faculty time (45%) was devoted to noninstructional activities. The distribution of time spread over activities was consistent with the mission of the departments and repre— sentative of other similar studies. William Clark Crothers 2. It appeared that the total workweek was relatively fixed as instructional workload factors increased or decreased. The total time devoted to a par- ticular course correlated highly with formal contact hours, but not with other factors such as class size, course credit hours, student credit hours, or level of instruction. 3. As the level of instruction increased, the size of the classes tended to get smaller, while the total time devoted to the class changed little. 4. As faculty rank increased there was a slight decrease in total course time, formal contact hours, and the number of sections taught. 5. A very high correlation was realized in comparing costs allocated on total course time and costs allocated on faculty reported formal contact hours. Costs develOped on total course time correlated higher than any other method with each of the other costs except those developed on credits. 6. Costs developed by level of instruction were highly similar when based on total course time and formal contact hours. 7. Although previous research suggested that graduate instruction was four times the cost of lower In, 10. William Clark Crothers undergraduate instruction when determined on a per student credit hour, this research, while supporting that conclusion for that one methodology, found lower undergraduate instruction more costly under the other methodologies. However, these other methodologies were costing on the basis of a unit related to the course, whereas the student credit hour is related to the student and is really a subset of course credit hours. This finding supports the contention that the costing methodology will to some extent determine the cost. Class size, the number of sections taught by the faculty member, and method of instruction were found to be significant variables of course costs. The dependent variable used in the multiple regression analysis which explained the greatest amount of variance in costs was the method for which costs were developed on total course time as reported in the faculty activity survey. A costing methodology should not be developed on the presumption that a relationship exists between time and some other basis of allocation. A MULTIPLE CORRELATION AND REGRESSION ANALYSIS COMPARING FACULTY ACTIVITY SURVEY WITH OTHER METHODOLOGIES FOR ALLOCATING COSTS TO COURSES BY William Clark Crothers A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Administration and Higher Education 1973 Copyright by WILL IAM CLARK CROTHERS 1973 DEDICATION This dissertation is dedicated to my parents whose expectations have always been worth fulfilling and to my wife whose love, patience, encouragement, and confidence have been the foundation of my efforts. ii ACKNOWLEDGMENTS The writer wishes to express his sincere appre- ciation to the chairman of his doctoral committee, Dr. walter F. Johnson. His friendship, encouragement, advice, and sensitive criticism have been a major motivator for completing the program and dissertation. Dr. Richard L. Featherstone has also provided gracious support, friendship, and guidance throughout the program and dissertation. Both of these men have had a profound influence on this writer. A word of appreciation is due Dr. Thomas M. Free- man for suggesting this research problem and for his helpful criticism of the manuscript. Dr. Gardner M. Jones provided important assistance in developing the program and helpful suggestions on this dissertation, for which the writer is grateful. Dr. Ted W. Ward pro- vided constructive suggestions as the research was being formulated. The enjoyable learning experiences through his classes and informal contacts will long be remembered. Grateful appreciation is also extended to Dr. Paul L. Dressel for the opportunity of employment iii in the Office of Institutional Research and to all the members of the staff, especially Mr. Lynn H. Peltier for his helpful advice. It has been a unique pleasure for me to have attended Michigan State University and to have had the privilege of working with distinguished scholars and professionals. However, none of this would have been possible without the loving assistance of my wife, Rilla, who spent many hours typing, in addition to raising our family and maintaining employment. iv TABLE OF CONTENTS Chapter Page I. RATIONALE FOR THE STUDY . . ._ . . . . . l The Problem and Need for the Study . . . l Purposes of the Study. . . . . . . . 5 Research Objectives . . . . . . . . 6 Definition of Terms . . . . . . . . 7 Overview of the Dissertation . . . . . 10 II. REVIEW OF THE LITERATURE. . . . . . . . l4 Search.Approach . . . . . . . . . l4 Societal Pressures. . . . . . . 15 Purposes of Faculty Load Studies . . . . 18 Faculty WCrkload Studies . . . . . . 19 Teaching Load Studies. . . . . . . 19 Faculty Activity Analysis . . . . . 21 Cautions and Observations . . . . . 32 Costing Methodology . . . . . . . . 36 Early History . . . . . . . . . 39 Recent Studies . . . . . . . 42 Selected Variables and Relationships. . 51 Alternative Management Tools . . . . 56 Advantages and Disadvantages of Cost AnalYSj-S O O O O O O O O O O 60 Comparison of Dissertation and Literature . 62 III. RESEARCH DESIGN. . . . . . . . . . . 77 Summary of the Rationale for the Study . . 77 Purposes of the Study. . . . . . . . 78 Research Objectives . . . . . . . . 79 Parameters of the Data . . . . . . . 80 Data Collection. . . . . . . . . . 82 Assumptions of the Study. . . . . . . 86 86 Limitations and Scope of the Study . . . Chapter IV. V. Instrument Reliability and Validity Profile Development . . . . Dependent Variables . . . . Costing Methodology . . . . First Level of Allocation . Second Level of Allocation Independent Variables. . Research Questions and Statistical Approaches . . . . . . Correlation Coefficient . . One-way Analysis of Variance Multiple Regression Analysis Kendall Coefficient of Concordance Summary . . . . . . . . ANALYSIS OF RESULTS . . . . . Objectives of the Research . . Activity Profile . . . . . Time and Cost Distribution . Student, Faculty, and Course Relationships. . . . . Cost Relationships. . Costing Methodologies and Variables Cost . . . . . . . . Summary . . . . . . . . THE PROBLEM, FINDINGS, CONCLUSIONS, MENDATIONS FOR FUTURE RESEARCH The Problem . . . . . . . Research Objectives . . . Parameters of the Data . . Findings . . . . . . . . Conclusions . . . . Recommendations for Future Research vi 102 103 104 109 109 114 114 115 115 124 128 133 161 166 166 167 168 169 171 174 Page APPENDICES Appendix A. Unit Costs for Method of Instruction by Course Costs, Course Credit Hour, and Student Credit Hour . . . . . . . . . . . 177 B. Analysis of Variance for Method of Instruction . . . . . . . . . . . 180 C. Scheffé Contrasts on Method of Instruction. . 183 D. Multiple Regression Results for Addition and Deletion Analysis with a Minimum of .05 Level of Significance (Considering Four Variables) . . . . . . . . . . . 185 E. Multiple Regression Results for Addition and Deletion Analysis with a Minimum of .05 Level of Significance Without Graduate Assistant Costs (Considering Eight Variables) . . . . . . . . . . . 189 F. Multiple Regression Results for Addition and Deletion Analysis with a Minimum of .05 Level of Significance (Considering Eight Variables) . . . . . . . . . . 193 G. Faculty Activity and Outcome Survey Instru— ment . . . . . . . . . . . . . 198 SELECTED BIBLIOGRAPHY . . . . . . . . . . . 203 LIST OF TABLES 1. Returned questionnaires . . . ,. . . . 2. Test for significant differences in total hours between 1970 green form and 1972 NCHEMS form 0 O O O O O O O O O 3. Cost calculation . . . . . . . . . 4. Time and cost distribution to activities by rank for total college . . . . . . 5. Time and cost distribution to activities by rank for department 1. . . . . . . 6. Time and cost distribution to activities by rank for department 2. . . . . . . 7. 'Time and cost distribution to activities by rank for department 3. . . . . . . 8. Time and cost distribution to activities by rank for department 4. . . . . . . 9. Time and cost distribution to activities by rank for department 5. . . . . . . 10. Simple correlations of instructional work- load characteristics . . . . . . . 11. Course cost patterns by level of instruction for four costing methodologies (in dollars) . . . . . . . . . . . 12. Course costs by level converted to unit cost per course credit hour, student credit hour, contact hour, and total course time . . . . . . . . . . . . viii Page 85 9O 98 116 117 118 119 120 121 125 130 131 Table 13. Selected correlations among variables and costing methodologies . . . . . . . . 14. Correlations among bases of allocating costs and course costs develOped on four methodologies . . . . . . . . . . 15. Analysis of variance F statistic for method of instruction . . . . . . . . . . 16. Multiple regression results for costs allocated on total course time reported (four variables). . . . . . . . . . 17. Multiple regression results for costs allo- cated on formal contact hours (four variables). . . . . . . . . . . . 18. Multiple regression results for costs allo— cated on student credit hours (four variables). . . . . . . . . . . . 19. Multiple regression results for costs allo- cated on course credit hours (four VariableS) o o o o o o o o o o o o 20. Summary of multiple correlation coefficients . 21. Multiple regression results for costs allo- cated on total course time reported (eight variables) . . . . . . . . . 22. Multiple regression results for costs allo— cated on formal contact hours (eight variables). . . . . . . . . . . . 23. Multiple regression results for costs allo— cated on student credit hours (eight variableS) o o o o o o o o o o o o 24. Multiple regression results for costs allo— cated on course credit hours (eight variableS) o o o o o o o o o O O 0 25. Comparison of multiple correlation coefficients including and excluding graduate assistants cos-ts o o o o o o o o o o o O 0 ix Page 134 138 141 144 145 146 147 150 152 154 156 157 159 AAA-“IA ‘- Table Page 26. Course cost by method of instruction (in dollars). . . . . . . . . . . . 177 27. COStS by method Of instruction, course costs allocated by three methods and converted to per course credit hour . . . . . . 178 28. Costs by method of instruction, course costs allocated on total course time and con- verted to per student credit hour cost. . 179 29. Analysis of variance for method of instruction (including graduate assistant costs) . . . . . . . . . 180 30. Analysis of variance for method of instruc- tion (excluding graduate assistant cost) . 181 31. Analysis of variance for method of instruc- tion (per credit cost including graduate assistant cost) . . . . . . . . . 182 32. Scheffé simple contrasts (six methods — course costs) . . . . . . . . . . 183 33. Scheffé complex contrasts (three methods — course and credit costs) . . . . . . 184 '34. Multiple regression results for costs allo- cated on total course time reported (considering four variables) . . . . . 185 35. Multiple regression results for costs allo- cated on contact hours (considering four variables) . . . . . . . . . . . 186 36. Multiple regression results for costs allo- cated on student credit hours (consider- ing four variables) . . . . . . . . 187 37. Multiple regression results for costs allo- cated on course credit hours (consider- ing four variables) . . . . . . . . 188 38. Multiple regression results for costs allo- cated on total course time without graduate assistant costs . . . . . . 189 Table 39. 40. 41. 42. 43. 44. 45. E 46. Multiple regression results for costs a110- cated on formal contact hours without graduate assistant costs . . . . . Multiple regression results for costs allo— cated on student credit hours without graduate assistant costs . . . . . Multiple regression results for costs allo- cated on course credit hours without graduate assistant costs . . . . . Multiple regression results for costs allo— cated on total course time reported (considering eight variables) Addition and Deletion at .05 . . . . . . Multiple regression results for costs allo- cated on formal contact hours (considering eight variables) Addition and Deletion at .05 o o o n o o n o o o o a Multiple regression results for costs allo- cated on student credit hours (considering eight variables) Addition and Deletion at .05 o o o c o o o o o c o a Multiple regression results for costs a110- cated on course credit hours (considering eight variables) Addition at .05 . . Multiple regression results for costs allo— cated on course credit hours (considering eight variables) Deletion at .05 . . xi Page 190 191 192 193 194 195 196 197 CHAPTER I RATIONALE FOR THE STUDY Institutions of higher education are facing some of the most significant issues they have had to face with regard to their own survival and the nature of their existence. The literature is filled with concepts of accountability, collective bargaining, tenure, resource allocation, program budgeting, and evaluation. Effective resource allocation and projection requires a thorough understanding of faculty activities and the related costs. To manage the cost of instruction requires the identification of the major variables of cost. Yet research on costing methodology is so limited that major questions concerning the use of cost studies cannot be answered. However, administrative decisions continue to be made, often without a full understanding of the cost- ing methodology employed. The Problem and Need for the Study The institutional resource which will realize the greatest impact from the application of the current concepts of accountability, resource allocation, program budgeting, and evaluation will be the faculty. Yet many institutions have little idea of how the faculty allocate their time or how this distribution effects costs. "One has only to raise the question as to what professors in a given department actually do, to learn that in most departments and most universities only the professor can provide ananswer."l Few institutions have devised a meaningful method of gathering this information, analyzing the data, or costing out these activities. Research to this end must be conducted if higher education is to meet the issues it is facing in a manner that is constructive and educationally sound. Although faculty workload has been discussed in the literature for some time, recently it has taken on a new emphasis because it is becoming an integral part of management systems. Cost studies which focus on faculty utilization are being conducted due to the new financial squeeze facing higher education. This move— ment toward financial accountability is accelerating without a good understanding of costing methodology or of the activities of the faculty. Yet the faculty represents the most significant factor in the educational process and it often accounts for the allocation of 60 to 80 per cent of the budget. There is a great need for research in costing methodology. In a 1969 paper in which he examines the value and validity of cost studies, Alfred D. Cavanaugh, from the Office of Institutional Research at the Uni- versity of California, Berkeley, stated that, Any study of cost analysis procedures is severely handicapped by the scantiness of the literature. The work of cost analysis is largely done by staff personnel associated with colleges and universities, with state budget offices and with coordinating councils. The work of most analysis is tied down to administrative demands for information and to the exigencies of state budget procedures; the pressure of time precludes significant publication in method- ological problems although it is sorely needed.2 The need for information related to costs and faculty has been recognized by the federal and state government as well as the National Center for Higher Education Management Systems (NCHEMS). For example, the 1972 Federal Education Bill has authorized a com- mission to study per unit cost in higher education. The demand for this kind of information is already being felt in Michigan.. The 1972 Michigan Senate Bill number 1141 of the legislature has called for an "academic staff performance audit." The bill states that, "The academic staff performance audit shall include measures of experience, training, salary and other com- pensation, rank, and productivity in terms of instruction and other duties of all academic staff."3 It has been pointed out by NCHEMS that while cost studies have been conducted since the early 1920‘s, each study approaches the costing in methodologically different ways.4 Furthermore, the literature is very limited in research on the differences in methodologies which is one purpose of this study. This need for better research in cost analysis was confirmed in a recent publication by the Carnegie I Commission on Higher Education. In 1971 Howard R. Bowen and Gordon K. Douglass authored a Carnegie Commission report entitled Efficiency in Liberal Education in which they present a comprehensive cost study of liberal arts colleges. The methodology which they employed was based on a faculty workload study. "If professional labor is the primary input into higher education's processes of production, then learning how faculty members spend their professional time ought to help us calculate the costs of alternative instructional modes."5 A factual profile of workweek activities is only now beginning to emerge from faculty activity analyses at a few universities. The use of faculty activity surveys in calculat— ing costs was one of the reasons the National Center for Higher Education Management Systems developed an instru— ment for gathering this information. Data provided through a survey using this instrument will allow a comparison of several variables of cost given different methodologies. Purposes of the Study This study was intended to fill, in part, the need as already expressed for research in faculty activity analysis and costing methodology. The need has been stated at the federal, state, and local levels. The purposes of this study were the following: 1. DevelOp a faculty time allocation profile, by department and rank, of the major activities in which the faculty participate; assign costs to these activities based on the allocation of time as determined by a faculty activity survey; and examine the interrelationships between several workload factors. 2. Compare four costing methodologies as a basis for examining the relative importance of some major variables in determining instructional costs and for considering their interrelation- ships. This was a descriptive study, as opposed to an experimental or evaluative study. Evaluative research is a comparison with a standard which has been developed from a value position.6 Whereas the purpose of experi- mental research is to test hypotheses, descriptive research seeks to examine relationships of basic information in order to identify probable cause and effect relationships or to characterize the nature or status of a phenomena. The faculty activities of one college at Michigan State University were examined in this study through the develOpment of a faculty activity profile. This study sought to examine the relationships between different costing methodologies as they were applied to the instructional costs of the faculty activity profile. The relationships of several major variables of costs with the different costing methodologies were examined. This study not only generated costs based on different methodologies, but it also explored what type of dif- ferences, depending on the methodology employed, resulted in several independent variables. Research Objectives The objectives of this dissertation were divided into two parts. The first was to determine the allocation of faculty time among several activities and to assign a cost to these activities based on the salaries of the faculty. The profile analysis was developed by depart— ment and by rank for one college at Michigan State Uni- versity during the 1972 fall term. The interrelation- ships of selected instructional workload factors were examined. .W77 “I“, I The second objective was to compare the four costing methodologies used in allocating costs to courses. The scheduled teaching section of the faculty activity profile served as the source of information for this part of the research. This comparison was not only based on the costs generated per course, but also considered the differences in selected variables of costs and the reSpective costing methodology. The comparison was made by answering the following research questions: 1. Is there a relationship between each independent variable and section cost for each costing methodology? 2. What is the relative importance of the independent variables for each costing methodology? 3. Under which costing methodology do the variables explain the greatest amount of the variance in the costs? 4. To what extent is there agreement in the rank ordering of independent variables across the costing methodologies? Definition of Terms For the purposes of this study the following definitions were assigned to the terms used: Faculty Activity and Outcome Survey.--is a questionnaire designed to measure faculty time allo- cations for different functions. For the purposes of this study the outcomes section of the questionnaire was not used. Activities.--include those functions normally considered part of the responsibility of faculty. They are the tasks actually performed. These include: teach- ing activities, unscheduled teaching, academic program advising, course and curriculum research and development, research, scholarship, creative work, public service activities, internal service. Each of these activities are defined in Chapter III under "Profile Development." Workload.--is defined as the "full professional life," that is, the full range of faculty activities as defined under activities above.7 It is clear that the workload is much greater than simply the assigned teach- ing or committee work. It is the total professional involvement of the faculty member as included in the profile analysis. WOrkweek.--is the total number of hours worked on an average per week. Formal contact hours.-—are the scheduled contact or class hours per week. Other contact hours.-—are the unscheduled contact hours with students that are related to the particular course . Total course time.—-is the total number of hours devoted to the course. Frequency.——is the term used to indicate the number of times the teacher has taught the course. Method of instruction.——indicates the means and extent of communication used between the teacher and the student. Course.——is a unit of the curriculum covered in a quarter or ten—week period.8 Course section.-—applies to one group of students in a course which is given concurrently to more than one instructional group. Sectioning of classes may be due to class size or for convenience of scheduling or for both 9 reasons. Course leve1.-—suggests the general level of maturity which the course demands of the students. Three levels are used in this study: lower undergraduate, upper undergraduate, and graduate.lo 10 Course credit hours.-—are the number of units granted to a student toward a degree for taking one course. Student credit hours.—-are the total number of credits for which students are registered in a particular course. A student credit hour would be a unit of measure- ment which represents one student taking a course for one credit. Class enrollment times course credit hours gives the student credit hours for one class. Fixed credit sections.—-are organized classes or sections meeting for a specific number of hours per week and assigned fixed credit values. Variable credit sections.-—are classes taught by independent study or for different credits depending upon the nature and extent of the study. Cost.-—is the measure in dollars of instructional resources (faculty salaries) used in the process of pro- ducing outputs during the term.ll Overview of the Dissertation In Chapter II, the pertinent literature is reviewed. This includes a review of both the costing literature in higher education and the methods used in assessing faculty activities. The design of the study 11 is presented in Chapter III which includes the description of the data, variables, and procedures of analysis. In Chapter IV the results of the analysis are reported. The summary, findings, conclusions, and recommendations are included in Chapter V. NOTES--CHAPTER I 1Paul L. Dressel, F. Craig Johnson, and Philip M. Marcus, The Confidence Crisis (San Francisco: Jossey- Bass, 1970), p. 186. 2Alfred D. Cavanaugh, "A Preliminary Evaluation of Cost Studies in Higher Education" (Berkeley: Office of Institutional Research, University of California, October, 1969), p. 1. 3State of Michigan, 76th Legislature Regular Session of 1972, Michigan Senate Bill 1141 (1972), Sec. 16. 4Gordon Ziemer, Michael Young, and James Topping, Cost Finding Principles and Procedures (Boulder, Colorado: National Center for Higher Education Management Systems at WICHE, November, 1971), p. 6. 5Howard R. Bowen and Gordon K. Douglass, Effi— ciency in Liberal Education, Carnegie Commission on Higher Education (Hightstown, N.J.: MCGraw—Hill, 1971), p. 23. 6Malcolm Provous, Discrepancy Evaluation (Berkeley, Calif.: McCutchan, 1971), p. 373. 7R. J. Henle, "To Devise and Test Simplified Ade- quate Systems of Measuring and Reporting Financial, Man— power, Facilities, Research, and Other Activities in Colleges and Universities, A Final Report," National Science Foundation and National Institute of Health, June, 1965. 8Floyd W. Reeves, Nelson B. Henry, and John Dale Russell, Class Size and University Costs (Chicago: Uni- versity of Chicago Press, 1933), p. 8. 12 13 9rbid., p. 9. 10James I. Doi, "The Analysis of Class Size, Teaching Load and Instructional Salary Costs," in College Self Study: Lectures on Institutional Research, e . y Richard G. Axt and Hall T. Sprague (BouIder, Colorado: Western Interstate Commission for Higher Education, 1960), p. 191. llZiemer, op. cit., p. 286. CHAPTER II REVIEW OF THE LITERATURE Search Approach The literature reviewed for this dissertation was approached from several different directions. A systematic perusal of leading journals, publications, and dissertation abstracts was conducted to search out the pertinent liter- ature. In addition, the private library of the Office of Institutional Research at Michigan State University, which consists of many of the unpublished reports and analyses conducted within the universities across the country, was used to identify the related research. Of course references cited by the writers in the field were obtained and evaluated for their inclusion in this review. Finally, the services of two computer search corporations were employed to identify related research. Although an extensive bibliography could be assembled, the bibliog- raphy included in this dissertation consists only of those references actually cited. This review of the literature will include both the kinds of relationships found in previous research and the techniques employed in faculty activity analysis 14 15 and costing methodologies. The ideas and observations of past and present scholars in the field will be con— sidered. At the conclusion of this chapter a comparison will be made between the literature and the purpose of this dissertation. Societal Pressures Societal issues are having a significant effect on institutions of higher education and therefore, the faculty. The economic pressures resulting from inflation, unemployment, and increasing interest rates are being felt by the tax payer and the educational institutions. Decreasing population growth, changes in public policy-- both federal and state, and signs of the general public's questioning of extensive higher education all indicate problems ahead for colleges and universities.1 Since higher education is a societal institution, it is not surprising that such anti-establishment sentiment should be transferred to the colleges and universities. The citizenry are concerned with high taxes, inflation, and their cost of living. Higher education must now compete for its funds with other higher priority functions with which the government is concerned. Added to these societal issues are student disenchantment, faculty involvement in national issues, the onset of collective bargaining, and the seeming lack of control by the 16 administration which have all culminated in a "Confidence Crisis" in higher education.2 There appears to be a growing public sentiment that higher education is a right, not a privilege. The parents of many young people feel that since it has been their tax money which has supported the state university, their sons and daughters have the right to try. The community college's open-door concept has been a response. Societal accountability is not restricted to public higher education for the private schools have had to adjust in light of the changes in the public institutions. The relationship between the community and higher education is of primary importance given the concept of accounta- bility. Briner emphasized this point when he stated, "Accountability in education must be the result of rational understanding and communication between the public and educators about the discharge of reSponsi- bility for determining educational purpose, defining function, judging results, and taking corrective actions to improve learning."3 The implications of these societal issues for the administration and the faculty are significant. The administration will have to be prepared to discuss the utilization of the faculty, their activities, the related costs, and the rationale for that cost. Perhaps one of the more important administrative functions will be to 17 assist and protect the faculty so that the pressures will not reduce their effectiveness. The early history of higher education reflects the vulnerability of the faculty. Falvey observes that "The first medieval universities were owned and operated by the students, who hired the faculty, chose the towns in which the universities were set up, formulated the rules by which the schools were governed, and dealt directly with the municipalities when difficulties arose."5 Haskins indicates that the students at Bologna organized the university to protect themselves from usury by the townspeople. Once they were successful with the townspeople, they turned their attention toward the professors, requiring that they live up to a set of regulations which guaranteed the students a fair return for the payments they made.6 Doi reminds us that the origin of the pocket in the hood of the academic attire was to allow the students to pay for the master's ser- vices without the embarrassment of direct hand-to-hand transfer of money.7 Rudolph documents the exploitation of faculty in American colleges and universities. Not only were work- loads often increased and salaries cut, but one insti— tution closed down for a year so the professors could go out and raise money. 18 In light of the societal pressures which higher education will face in the next few years, it is impera- tive that the most efficient utilization of faculty resources be achieved. "The really essential task for the university does not lie in convincing the public by some sleight of hand ratio that its instruction is 'efficient' or cheap but in conveying the full scope of university activity-~its research, its public service, and its instruction——to a variety of new publics."9 Faculty activity analysis will assist in conveying to the public the nature and extent of the many activities in which the faculty are justifiably involved. Purposes of Faculty Load Studies Doi suggests several purposes of faculty load studies. The studies may be used: 1) for assessing the general efficiency and economy of institutional programs . . . 2) to assist in development of objective criteria for determining instructional loads and staffing needs. . . . 3) to stimulate experimentation with instructional tech- niques and various class sizes. . . . 4) to provide information for the planning of future expansion and changes in instructional programs. . . . and 5) to provide the kinds of information necessary for making the wisest allocation of instructional funds.lo Two purposes of measuring the faculty load are seen by Bolton, "1) to acquire adequate faculty, and 2) to divide responsibilities among faculty members."ll Young sees such studies as being useful for providing information for decisions related to 1) increases in 19 staff, 2) additional course offerings, 3) salary increases, 4) realignment of teaching duties, and 5) adjusting inequity in faculty load.12 Hicks believes that the faculty will gain through a re—examination of their own priorities. He states that, Perhaps the greatest value which can be gained from faculty workload studies comes when we, in our institutions, take the time to analyze and define what each of our faculty is doing. We will find, I think, that our good professors are always really "overworked." It is their nature to be so. The duty of the administrator is to protect the time of these professors so that it may be used to the fullest extent for what they can do best.13 Faculty Workload Studies Studies of faculty workload generally fall into two categories, 1) those that are restricted to the teaching or instructional load, narrowly defined, and 2) those that deal with the total service load or all the professional activities.14 Teaching Load Studies The issue of faculty workload has been in the literature for decades and it is still a current concern. The American Association of University Professors has recently set maximum workload limits of twelve hours per week undergraduate instruction or nine hours of graduate instruction. They went on to say that this maximum workload "presumes no unusual additional expec- tations in terms of research, administration, counseling, 20 or other institutional responsibilities."15 Although this standard has been stated in teaching hours, other units of measurement are often used. Young reported on a faculty load profile used at Capital University and built upon six criteria which included: (1) student credit hours, (2) semester credit hours, (3) class- contact hours, (4) number of class preparations, (5) hours of upper division work, and (6) committee, research, and administrative assignments as determined by the administration. Ratings of light, average, and heavy were assigned to each criteria and the individual was then compared to his colleagues in his department.16 A 1968 survey of 206 colleges and universities, conducted by Bolin and McMurrain, found that the average full—time instructional workload was 12.76 credit hours, a little less than that of the students. The range, however, extended from a low of seven hours to a high of twenty. The standard deviation was 4.08 credit hours. The average instructional workload declined as the size of the institution increased. At the small institutions the average was 13.73 hours; at the'medium institutions, 12.10 hours; and at the large universities the average credit hours per week was 10.81.17 Hobbs found that although two—thirds of the colleges in her sample reported a maximum teaching load of 12 hours per 21 semester, the median average load was 10 hours and the mean was 10.1 hours.18 Williams reported that contact hours decreased as rank increased. In a study primarily based upon an analysis of faculty workload in several midwestern uni— versities, he found that the average contact hours per full-time equivalent instructor was 13.19; assistant professor, 12.02; associate professor, 12.12; and pro- fessor, 10.84.19 Through a questionnaire sent to 500 teachers and administrators in 200 colleges and universities across the country, of which 228 were returned, Knowles and White found that the actual clock hours per credit hour ranged from 2.9 to 5.5. "There appeared to be no sig— nificant relationship between the semester hour credits carried by a course and the amount of time and effort required of an instructor assigned to teach it." This led the researchers to conclude that, "In short, each instructor's assignment needs to be individually evalu— ated in clock hours per week if it is desired to measure his load in a reasonably accurate and meaningful way."20 Faculty Activity Analysis One of the major weaknesses of teaching load studies is that it ignores the many other activities of the faculty member. The faculty activity analysis 22 involves the total professional life of the faculty member including his instruction, research, service, and administration. One of the first articles calling for the adoption of a faculty activity survey which requested a time distribution for all activities was written by Heilman in 1925. He argued forcefully for the total activity survey and against surveying only the teaching load, which he considered “about as reliable a method of determining the teacher's total timeload as is the method of getting a man's height from the measurement of his head."21 Prior to this study, in 1919, a time study of instruction was made by Koos to see if the method of instruction correlated with the time required for effec— tive teaching. He found that the lecture method required the most time followed by discussion, recitation, and laboratory.22 In a recent book on the evaluation of faculty, Miller points out the lack of understanding by the com- munity outside of higher education of the workload for the average professor. The overall academic load of the average professor is seriously misunderstood by those outside the academic community. The hours spent in the class— room need to be considered on a two—to—one basis: two hours of preparation and evaluation for every hour in the Classroom. A twelve—hour classroom teaching load, considered normal for most under- graduate teaching assignments, becomes thirty-six hours of teaching and teaching—related activities. Add a conservative estimate of five hours a week for 23 advising and counseling, five hours for committee and departmental activities, four hours for cor- respondence and other academic housekeeping details, and the total becomes fifty hours per week—-and the professor has not even begun to do the reading, studying, and research that are essential to keep him on the growing edge of his field. This being the case, and with every field of knowledge literally bursting at the seams, it is not difficult to make the case for professional leaves and sabbaticals for all academic personnel.23 One of the outstanding scholars in faculty activity analysis is John E. Stecklein. He reported that the Uni— versity of Minnesota had conducted three comprehensive faculty workload studies since 1940. The 1941—42 study asked the faculty to allocate their time among the many activities included in the total professional life of a faculty member. They found that a fourth of the returns reflected forty hours or less and a fourth reflected seventy hours or more. The average was between fifty- five and sixty hours. Just as the 1941—42 study was requested by President Coffey, the 1950-51 study was initiated by President Morrill. "He believed a comprehensive study of faculty activities would be helpful in appraising programs, in evaluating or equalizing faculty work loads, and in documenting needs to the Board of Regents and the legislature."24 In this second study it was decided to use a diary maintained by a sample of the faculty to validate the time estimates on the questionnaire. "Analysis of these reports showed very close agreement 24 with the estimates which the faculty members had.made on the survey forms and the results seemed to substantiate the ability of individuals to estimate the allocation of time accurately."25 The third study was done in 1959-60. It requested the data as percentages of total. After con- verting the previous studies to percentages the three studies seldom differed by more than 5 per cent. "Although additional comparative studies have not yet been com- pleted, the initial impression is that, even with the variability expected due to changes in personnel and programs, the data gathered in the two just described provide essentially the same kind of results."26 During the fall term 1970 the Office of Insti- tutional Research surveyed the faculty at Michigan State University to obtain information on faculty time dis- tribution. A return was realized of 2,012 or 90.5 per cent of the regular faculty and 1,232 or 49 per cent of the temporary faculty. The analysis indicated an average work week by rank of 59.1 hours for professors, 58.2 hours for associate professors, 57.8 hours for assistant pro- fessors, and 54.2 hours for instructors. Peltier noted that, These values appear to be typical of other uni- versities which have surveyed workloads. The Uni- versity of Minnesota in the fall of 1969 reported an average of 57 hours per week for all ranks. The University of Wisconsin reported 54 hours per 25 week and the institutions in California reported a range of 50 - 71 hours per week, depending on the type of institution and rank of faculty."27 This study involved the total professional life of the faculty including teaching, research, and service. The distribution on a percentage basis for all ranks shows 43.8 direct instruction; 4.4, curriculum develop- ment; 11.0, additional instruction; 20.4, research; 2.8, public service; 11.7, administration; and 5.9, other. Faculty activity analysis is increasingly being accepted as a methodology for assessing the activities of faculty in medical colleges. Some of the leading studies include a 1964 study of thirty-nine dental schools, a 1969 study of seven medical centers, and a 1971 HEW study of eleven medical centers receiving financial distress grants. All of these studies used data collected through a faculty activity survey.28 Sexson conducted a study on faculty load by employing a letter of inquiry, questionnaire, diary, and personal interviews. He found that none of the seventy-five colleges and universities surveyed had a method or formula for computing load on positively identified load factors. Fifty-two per cent had a policy of assigning the same number of contact hours to all instructors regardless of discipline. 26 A faculty activity survey was conducted at a college of approximately 300 faculty over the four quarters in the academic year. A random sample of the faculty also kept a diary which was used to verify the activity survey. A personal interview with each faculty member was held after the diaries were completed. Five load factors were identified which included: (1) class- room function, (2) advising students, (3) committee assignments, (4) official correspondence, and (5) special assignments. Research was considered a special assign- ment and was granted upon presentation of a request to an internal committee. A faculty workload was developed based on the average time for each function as it related to factors of contact hours, class size, number of advisees, and number of committees.29 Sullivan compared the bias associated with faculty self-reporting of time and effort through a diary and a questionnaire. As an initial step he sur- veyed the forty largest universities in 1971 to deter- mine how other universities obtain information on faculty activity distribution. It was found that a mail questionnaire was the most common method of collecting faculty time information.30 The results of the poll led to the decision to use a faculty activity survey at Florida State Uni- versity. Sullivan notes that, 27 The reaction to the questionnaire itself, the faculty concepts of the reasons for it, their apprehensions about the uses or the misuses of the data, and their View of the process as a threat impinging upon academic freedom were revealed as real forces at work among the faculty. These factors were seen to be significant and unavoidable forces impinging upon the reporting process.31 This study led to the comparison with a diary maintained by forty-eight faculty members. Unfortunately only three categories of instructional activities were examined. It was found that the time reported on the questionnaire exceeded the diary in all three categories. However, there were no significant differences between ranks. That is, rank made no difference in the bias in reporting. There were differences between disciplines with the greatest bias in Arts and Letters. Sullivan points out that the diary approach is quite costly to carry out and that with SubStantial savings, and with somewhat less precision in results, one might consider using a questionnaire approach. He goes on to say, "The researcher believes that for most I applications it is neither necessary nor desirable to incur the costs associated with obtaining precise estimates of employee time/effort expenditure."32 The Pennsylvania State University was required by the 1972-73 Appropriation Act to submit "an analysis of the average weekly workload of each full-time faculty member" along with a series of reports summarizing course credit, student credit hours, enrollment, degrees 28 awarded, course and section size, and salary costs. A form entitled Analysis of Faculty Activity was completed by each full-time regular faculty member and reviewed by the department heads, deans, and campus directors, to insure completeness and accuracy. The form asked each faculty member to report his average weekly workload in instruction, research, public service, and other activi- ties for summer and fall terms, 1972 and estimated activity for winter and spring terms, 1973.33 The composite data for fall term 1972 show a total average hours per week of 50.6 with the average hours and percentage of total for instruction as 27.5 hours and 54.3 per cent, research as 13.6 hours and 26.9 per cent, public service as 3.1 hours and 6.1 per cent, and other as 6.4 hours and 12.7 per cent. In addition to submitting the individual forms and summary reports, interviews were conducted for a sample of the faculty randomly selected from the survey forms. A written summary of these interviews was also submitted to the State.34 The state of Tennessee conducted a faculty time distribution for the fall term of 1972 for all of the public institutions of higher education. Bogue, Stovall, and Norman found that there were no major variations in the average hours per week for the four faculty ranks and that the pattern of time distribution was consistent 29 with general role expectations. The top three ranks spent an average of seven hours per week beyond the 48—hour week in professional but unassigned activities in public and institutional service and research. Instructors spent an additional five hours. It was also found that there was a heavy emphasis on instruction and the faculty were less involved in research and ser- vice than in other states.35 Eckert reported on a survey of Minnesota insti— tutions. In the spring of 1968 a questionnaire was sent, based on a 20 per cent random stratified sample, to the faculty members in Minnesota's 43 recognized private and public colleges. Of the 1,678 faculty members surveyed, 82 per cent completed and returned the questionnaire. In the state colleges and private liberal arts colleges approximately three—fourths of the faculty time was devoted to teaching and other instructional tasks, whereas the university faculty devoted slightly more than half of their time to instructional tasks. The other major difference was in research. Forty-five per cent of the faculty in liberal arts colleges said they gave no time to research, whereas the comparative figure for the university was 21 per cent. The median percentage of time spent by faculty on various professional functions is: teaching activities, 64.9; counseling, 9.6; services to student groups, 0; research and scholarly writing, 10.4; 30 committee and administrative duties, 11.0; and off-campus services, 4.1. Eckert also found that, "Compared with correlate figures for 1956, the four—year college faculty is spending less time now (by 10 per cent) in teaching and somewhat more time (6 per cent) in research."36 This finding of Eckert is partially supported by a study cited by Light entitled Careers of Ph.D.'s. He states that: From 1940 to 1963, the proportion of time spent on teaching decreased overall from 66 per cent to 50 per cent. Research remained constant at 25 per cent, but administrative duties increased from 8 per cent to 20 per cent. . . . These percentages indicate a reduction in hours spent teaching over three decades, but a change in the total work week is not evident.37 Light also cites a study by Parsons and Platt in which they compared the actual use of time and the faculty ideals about time. At every level, faculty desired more time to teach graduate students and to do research, with less time involvement in administration.38 One of the significant contributions in the last decade was a report Sponsored by the National Science Foundation and the National Institute of Health. This study established criteria for testing and developing the activity categories. The criteria included: (1) consistency with a sound philosophy of university activity, (2) realistic categories, (3) differentiation from the purely personal, (4) consistency with other data systems, (5) fundamental or irreducible categories, and (6) practicality. 31 The term "full professional life" originated with this report. In approaching the problem of understanding, interpreting and measuring the activities of academic/academic—professional personnel, it is important that whatever measures or systems of measurement are adopted be in accord with, if not based on, a sound philosophical understanding of the academic or academic—professional man. These two types (or perhaps this one type) fall within a roughly recognizable group of people in society called "professional" people. Professional persons play a peculiar kind of role in society. They have a kind of pervasive and general obligation to society which is, in a sense, a public trust. In playing this role in society the professional person carries on a wide range of activities. His total activities can be divided into two groups, (1) those which are related to, or in function of, his professional life, and (2) those which are not and which are personal to him, or at least relate to some aspect of his life not part of his pro- fessional role. Thus, a physician may have a determinate private practice, do charitable work in a clinic or asylum, sit on local A.M.A. com— mittees, work in public health education programs, and advise national bodies. All these activities relate to his professional role and these activities, plus all like them, constitute the full professional life of the doctor. They can rather easily be dis— tinguished from the purely personal phases of his life. His picnicking with his family, his fishing, his interest in rare stamps or rare steaks—-all such matters are personal to him, not professional. It is important, therefore, that both the pro- fessional person himself and those dealing with him, either as an active participant in an oper- ational setting or as an abstraction in a theoreti— cal theme, should think of his professional activi— ties as organically interrelated, as a totality of activity. This totality we here refer to as his "full or total professional life."39 The most recent significant impetus given to faculty activity analysis is from the National Center for Higher Education Management Systems (NCHEMS). The 32 purposes of the faculty activity project at NCHEMS are to develop techniques for collecting data and to stan- dardize procedures for analyzing faculty activity.40 Thus far the center has published an overview of faculty activity analysis and a preliminary unedited draft of a procedures manual has been assembled.41 The pro- cedures manual will be published in the fall of 1973 followed by an analysis manual. The questionnaire used in this dissertation was developed and pilot tested by this organization. Cautions and Observations Although faculty activity analysis has a rela- tively long history and is becoming increasingly useful, not everyone supports the concept. "Some feel that the teaching profession is lowered in dignity and prestige when the service loads of its membership are subject to such an evaluation."42 Even though Toombs supports the concept of faculty activity analysis,43 he points out that higher education is classified as an industry con- sidered to be "labor intensive." This means that the personnel represent a large part of the operation. Although most service industries are in this classifi- cation, colleges, hospitals, and to some extent govern- ment, with their large professional staffs, are often . . . 44 distinguished within the class1f1cat1on. He further states, 33 It is not only labor intensive but also "quality intensive." That is to say, the way in which edu- cation is carried out has qualities that must be preserved. How it is done, what happens between input and output, is the heart of the matter. Earl Cheit and others have used the analogy of the sym- phony orchestra, also a labor intensive and quality intensive organization. The orchestra is limited in how many engagements it can play per week before its repertoire deteriorates and the performance declines. It cannot play faster or louder. The number of members cannot be increased to 200 to yield more output. In short, the quality of the process itself, not just the outcome, is a part of "production."45 Dodds suggests that there is a certain incom- patibility between "economy and efficiency" and the 46 inspired teacher. Although Bailey recognizes the importance of efficiency, he believes, as most authori- ties do, that it can be over emphasized. Our supreme function is not to improve managerial efficiency in education. We cannot countenance obvious waste, and we have obligations to the public to see that money is not used frivolously. But our supreme obligation is to remind ourselves and our public and private benefactors that a partially unquantifiable and inherently untidy system of higher education must routinely make legitimate demands upon the treasuries of the purse in order to nourish the treasuries of the mind and spirit. For freedom is the condition of nobility, and knowledge is the condition of freedom.47 Etzioni and others have demonstrated the pro- . 48 fessional status of college teach1ng and because of the very nature of a professional Organization, change, if it is to result, must be produced by the faculty. Simonds argues that all of the effort toward increasing faculty productivity has failed to include one essential component—-the faculty member. It is his position 34 that faculty members need to be given incentives for finding new ways for increasing.their productivity. In industry several approaches have been used to enlist the minds of the employees. "Among them are l) a guarantee that no one will be laid off as a result, and 2) provision for reward to individuals for the improve- ment." This reward is usually a percentage of the first year's savings resulting from the change or a profit Sharing such as a bonus to all the workers based on a ratio of labor costs and total sales. Executives often are rewarded with bonuses, salary increases, and pro- motions in return for their efforts toward increased productivity.49 It is not enough to develop better methods of educating students, the faculty need to be convinced that it is in their best interest to make use of the new methods. Ruml believes that the faculty lack the necessary information to cause them to respond posi- tively. If information about teaching loads, course offer- ings and enrollments is available to administrative officers, it is not likely to be distributed routinely to the faculty. Lacking this basic information, it is small wonder that the individual teacher does not see the possibilities of improving his economic status by means of an institutional program utiliz- ing total faculty resources more efficiently. The lack of productivity increase in higher edu- cation has been well documented by O'Neill. Historically, most industries in the American economy have increased 35 steadily their productivity per unit of input. However, higher education has not done so well. Between 1930 and 1967, instructional inputs and credit hours appear to have increased more or less proportionately. If growth in credit hours is a reliable indicator of growth in real instructional output, then there is the strong possibility that there has been no productivity change in the pro- duction of higher education over the time period—— and this despite the very rapid growth rates in higher education.51 A change has been called for by several writers including Reeves, Russell, and Dressel in the plan for administering the faculty workload whereby each faculty member would normally be given a teaching assignment to occupy his entire working time, but he would receive release time from teaching while involved in research and administration. The amount of release time would be a matter of individual adjustment. Under this plan the distribution of faculty time would be fixed administra- tively.52 Millett has made some more general recommendations based on the financial pressures higher education is facing. These include: (1) preparation for a slower rate of growth, (2) more attention to cost reduction through increased faculty instructional loads, and (3) more self—consciousness about their academic plan- ning and about the management of their resources.53 Faculty activity or workload studies have been used for assessing productivity, effectiveness, and 36 efficiency. The limitations discussed in this chapter need to be kept clearly in mind as this method of data collection is utilized. The faculty activity survey can be a meaningful instrument for planning and for costing. Misused or misunderstood, it can be destruc- tive; but if wisely used it can be constructive. John Dale Russell states: A vital issue is at stake. If unwise formulas, standards, or other measures are devised and employed, great damage can be done to higher edu- cation. At the same time, unnecessary and ill- founded opposition to the use of any measure, no matter what its form, is a barrier to good govern- ment and the pursuit of excellence in public ser- vices. Good will alone won't solve the problem. What is necessary is thorough analysis and clear thinking as a basis for consensus.54 Costing Methodology The term "cost" has several different meanings.55 McConnell states that the economist's notion of costs is built upon the fact that resources are scarce and they have alternative uses. Thus cost in an economic sense implies missed opportunities or foregone alternatives.56 The American Institute of Certified Public Accountants define cost as the "amount, measured in money, or cash expended or other property transferred, capital stock issued, services performed, or a liability incurred, in consideration of goods or services received or to be . 5 . . . . received.“ 7 Bus1ness managers, in addition to 37 economist and accountants, have made use of costs. Moore and Jaedicke state, Costs cess, value to management in controlling and planning business operations. Costs can be defined in various ways, and cost data can be rearranged or adapted to serve different purposes.58 In general financial accounting costs are are important in the income measurement pro- but they are also important because of their reported in the aggregate; but in unit cost studies in higher education, as in cost accounting in business, the costs are broken down on a unit basis. Horngren states that, A unit cost is calculated by dividing a total cost by some related base. A unit cost is a useful communicative device because it often expresses costs as they are best understood. unit costs should be expressed in terms most meaningful to the people who are responsible for incurring the costs.5 The National Center for Higher Education Man- agement Systems defines cost as, of institutional resources used in the process of pro— viding institutional outputs during a given time 60 period." It is apparent that there is no established definition which extends across all institutions which make use of the term. It is also clear that costs must be understood in relationship to the purposes for which they are to be used.61 Although the American Institute of Certified Public Accountants will be publishing a document in the next few months entitled Audits on Educational . . . Generally, "The measure in dollars 38 Institutions, they have published very little directly related to educational institutions.62 They have, how- ever, made use of College and University Business Admin- istration, a manual revised in 1968, which attempts to bring standards to the accounting and financial operation of higher education. This manual represents a consensus among the college and university business officers. It defines principles and procedures for fund accounting in higher education.63 It has been pointed out by several writers, including Cavanaugh, Jones, and Withey, that the fund accounting procedures of higher education do not provide an adequate way of identifying the resources used in different activities.64 Withey states that, There has been growing concern in recent years that reporting of expenditures on a natural or objective classification by nonprofit organizations does not present a meaningful statement of stewardship. Any substantial organization which reports to the public only in terms of salaries, rent, telephone, travel, supplies, and so on, is not informing its readers on the real activities of the organization. It should be obligatory to inform the public of its expenditures in terms of program, project and management functions. Collier points out that now institutions are being asked to report in terms of a broader concept of accounta— bility than simply their fiduciary responsibility. The public wants to know what was accomplished with the dollars received, not simply, were the dollars spent on the activities for which they were given.66 39 Early History Doi traces the concepts of "course offerings," "class size," "teaching loads,“ and "instructional costs" to the ancient teachers. The professors of the medieval universities were quite concerned about class size, since their pay usually was determined by the number of stu— dents they served.67 Sherer noted that unit instructional costs were being developed in this country as early as 1894.68 John Dale Russell, in The Finance of Higher Edu— cation, traces the historical development of attempts at standardizing income and expense accounts. The first major contribution was in 1910 by the Carnegie foundation. In its annual report, they made a series of recommen- dations for improving financial reporting. The second land mark was in 1917 and became known as the Christen- sen report. It was a committee report on uniform classification of expenditures of the Business Officers Association of the Middle Western Universities. The third major contribution was the work of Arnett in 1922 which called for a separate fund for endowments and plant. The fourth land mark mentioned by Russell was the 1935 National Committee on Standard Reports. Only the 1935 study dealt with unit costs.69 One of the early attempts in research on unit costs in higher education was conducted by the Educational 40 Finance Inquiry Commission and published in 1925. The commission was organized as a result of a meeting of the Department of Superintendents of the National Edu- cation Association and it was funded by several foun- dations. The unit of study was mainly the student clock hour defined as one hour of instruction for one student. The conclusions and findings of this commission with respect to unit cost studies were the following: 1. That there is a need of systematizing the accounting method of higher education as a basis for intelligent and effective economy in internal administration. 2. That there is a need of differentiating between instruction and non-instructional service, such as research and extension. 3. That unit costs of instruction point to ways and means for promoting economy. 4. That the unit costs of the same kinds of work in different institutions tend toward similar levels. 5. That the unit costs in curricula and in depart- ments with small enrollments will tend to be high. 6. That the purpose of unit costs is to make more effective the work of higher education. 7. That unit costs have decreased during the recent period of increasing enrollments. 8. That demands for better service will, in the next few years, tend to increase costs as the equipment catches up with the enrollment. 9. That the unit-cost figures are needed to permit the claims of higher education to be presented to the public in terms of service. 10. That there is a need of standardizing the financial accounting of the institutions of higher education in order to make possible satisfactory cost comparisons and judgments based on such com- parisons looking to wise public policy with reference to the support of higher education. 11. That such work should be undertaken and developed by a representative agency such as the American Council on Education.7 Ten years following this commission another major publication was released. It was an attempt at 41 standardizing costing methodology by the National Com— mittee on Standard Reports for Institutions of Higher Education in 1935. This report was later reprinted by the American Council on Education for use with College and University Business Administration. Some of the purposes of unit cost studies and the methodology util— ized were discussed in the report. If properly conducted, cost studies should be of value in the internal administration of edu- cational institutions. The determination of costs may well be considered one of the first steps in a complete analysis of the administrative and financial practices within an institution. Variations in costs between departments of instruction, schools and col- leges, curriculums, and levels of student achieve- .ment, or variations in costs for the institution as a whole over a period of years, should lead at once to a further examination of enrollment, size of classes, number of faculty members, teaching loads, salary schedule of faculty members, curricular offerings, and efficiency of use of the facilities of the educational plant. Unit-cost studies, furthermore, may be of value in the determination of the rates of student fees, in the preparation of the budget, in educational surveys, in the accreditation of educational institutions, and in the determination of desirable reorganization within an institution or within systems of higher education.71 Since 1935 some individuals in higher education have come to realize the important uses that industry has made of cost accounting. Much of the increased pro- ductivity can undoubtably be linked to the costing methodologies developed. Higher education has in the past found it easier to engage in fund raising than to examine its operations to discover places where savings could be achieved. 42 However, it is interesting, but not unexpected, to note that higher education has historically turned its attention to finance and cost studies when the country was experiencing economic depression. Of the six studies mentioned in this section, four of them, 1894, 1910, 1922, and 1935 were during depression years.72 Only the studies of 1917 and 1925 were during economically good years; and of these, World War I would have cut enrollments making 1917 a financially difficult year. Russell notes that not much was accomplished in the 1940's. The Financial Advisory Service of the American Council on Education was discontinued and by midcentury standardization appeared to be a "distant, though desirable goal."73 Recent Studies Of the cost studies conducted in the last twenty years, one study which continues to be cited was a cooperative venture of ten universities which is some— times referred to as the "council of ten study." The California and Western Conference Cost and Statistical Study for the Year 1954-55, under the direction of William T. Middlebrook of the University of Minnesota, concluded among other things that (1) cost per student is affected by the composition of the student body, instructional level, curriculum, and so on: the "student mix," (2) other factors are of greater 43 importance than teaching salaries in relation to cost per student, and (3) although methods of instruction definitely affect cost, it is in terms of their influence on class size, teaching load, and other cost factors.74 Regrettably, this cost study and others75 have disseminated the idea that the product of higher edu- cation is the environment produced.76 The environment is not the end but only a means to an end. It is the educated student that is the product of higher education. Doi classified cost studies into two categories, "a) those that are limited only to salaries paid persons who actually taught a class or classes during a given term or year, and b) those that attempt to take into account not only salaries for teaching but also expen— ditures for instructional supplies, faculty benefits, secretarial and clerical services for faculty members, and for other items directly related to the function of instruction."77 Of the two classifications, Tydall and Barnes argue that allocation of salary costs only is preferred for the inclusion of indirect costs may make the results less useful.78 Hull and McWhirter conducted a cost study in 1962-63 at Indiana University using a faculty activity survey to determine the cost per student by class level. To allocate salary to each course taught by the faculty member, a ratio of time spent on each course to the 44 total work week was made and multiplied by the salary paid. The salary was then allocated to each student in the class by class level to produce a cost per student.79 A cost study which used a faculty time allocation survey was also reported by Hubbard. The three basic steps in allocating cost to courses were: (1) determine time allocation, (2) convert time reported to a per- centage basis, and (3) multiply the percentage of time for each course by the appropriate individual's salary.80 He noted that there was some apprehension in allowing faculty to allocate their time among the functions. "The reporting of suSpiciously high per- centages of time to 'filling out questionnaires' by a couple of faculty members suggests that the concerns were not unfounded." It appeared that faculty became increasingly accurate in their judgments of the per- centage distribution as they made the judgments over several terms.81 In one of the better methodology allocation studies, Bogue compared the effects of allocating instructional salaries based on course credit and based on faculty time. He found that there is a ten- dency toward higher student credit hour costs at the doctoral level when faculty effort is the basis of allocation as opposed to the course credit value. This 45 cost study conducted at Memphis State University in 1970 also found that there is a definite stairstep increase in unit costs when allocation to instructional level is based on course number rather than student classifi- cation.82 This study resulted in the development of a more complete program of academic management at Memphis State University, one which, in Bogue's words, "will a) permit the identification of faculty resource potential, b) make clear the pattern of faculty workload assign— ments, and c) provide bases for qualitative assessment of both personnel and program performance."83 Anderson studied the relative influence of selected factors upon instructional unit cost of higher education. The factors included in the study were: individual institutions, institutional types, faculty rank, class size, level of instruction, type of instruction, and subject field. Only one costing methodology was employed which reduced the data to the mean unit instructional costs which were then tested for significant differences. Anderson found that, Statistically significant differences existed between the subdivisions of each of the selected factors. University-type institutions incurred costs 1.6 times greater than those of the state colleges and municipal institutions which were comparable. Class size costs were inversely proportional to class enrollments. Level of instruction differences approximated a 1:2:4 ratio for lower, upper, and graduate divisions, respec— tively, although the non-state universities had 46 lower ratios at the graduate level. Lecture- discussion type of instruction, by far the most popular, was also least costly with a unit cost .7 of laboratory—type costs and .2 that of "other" modes of instruction. Subject field costs varied to the extent that the one extreme was less than half the other. The specialized subject fields of agriculture, art, engineering, and law comprised a costly group while the more universal offerings of social studies, business, science, and mathematics were least expensive. A median group included the humanities, home economics, and education. Individual institutional and faculty rank dif- ferences were less strong and were overshadowed by more powerful factors in several of the analyses. Only the state universities showed regular cost differences among the several ranks and institutional type differences tended to outweigh individual institutional differences. Siegel reported that during the fall term of 1964 and 1966 the University of Oregon was the subject of a pilot unit cost study. Built upon a "course enrol— lment matrix," costs were developed on a course basis using only faculty salaries. This direct salary cost is allocated to each course based on university reports of the faculty members' distribution of their teaching time. These costs were then allocated by student level. In this study Siegel found, among other things, that, (l) the average cost per course taken rises with the level of student, (2) there is great variability in average cost per course, (3) there is a lack of flexi— bility of resources in a university, and (4) rigidities limit the reallocation of existing faculty and necessi- tate a more optimum allocation of new staff members.85 47 Witmer conducted a unit cost study of the state universities in Wisconsin in 1966—67. After comparing the cost per credit between the state universities he concluded that since the large universities had a lower cost per credit than the smaller universities, the economy of scale was operating.86 However, at least one of those small universities in the Wisconsin sys— tem, Stout, has a highly specialized curriculum. Witmer failed to note that the academic mission of an insti- tution may be just as important as size in cost relation- ships between institutions. The Oklahoma State System of Higher Education developed a teaching load and cost study based on the number of semester hours of classes taught. However, they did recognize that this measure may not adequately reflect the actual teaching load since the methodology did not consider actual time or contact hours. The analysis was aggregated on the three academic levels-- lower division undergraduate, upper division under- graduate, and graduate--so comparison between insti— tutions could be made by level of instruction. Instruc— tional salary costs were examined with the suggestion that salary costs could be reduced by (l) filling vacan- cies with individuals of lower academic rank, (2) limit- ing courses, which forces students into fewer courses 48 thus raising class size, and (3) absorbing additional students without adding faculty or courses.87 Adkins examined the volume and cost of instructional services in the colleges of Virginia. In this cost study costs are allocated on a student credit—hour basis with- out reference to student contact hour or clock hours. Only faculty time devoted to teaching is included, other functions have been excluded based on the institution's allocation of their faculty members' time. Allocating student credit hour costs by student level and by course level revealed that "the two methods result in substantial differences for some institutions while the differences at others are slight."88 Scheerer developed a formula to produce a cost per equivalent student credit hour. The formula takes into consideration the level of instruction by assigning weights which represent the collective judgment of the deans as to the relative demand on faculty time. Addi- tional direct and indirect costs are included to give a complete cost figure. Unfortunately, the formula was not built upon research data and it ignores other variables such as contact hours, class size, faculty rank, and so forth.89 Peltier and Ingall examined degree costing for the Michigan Council of State College Presidents. They suggested that one alternative to degree costing is to 49 track individual degree recipients through their course work and accumulate the costs by course, the total of which would be the degree cost. The basic requirement is differential cost data for each course based upon student credit hours (or some other unit) if equal expenditures for all levels are to be avoided. This presents several problems, including the actual tracking of students through the courses. After conducting a pilot study, they came to the conclusion that "the pattern of cost per degree is influenced by the choice of programs, as well as the design of the alternative procedures."90 Austin conducted a degree costing study in the College of Education at Michigan State University which included five component cost categories: (1) instruc— tional costs, (2) faculty support costs, (3) research costs, (4) space costs, and (5) administrative costs. The dollar costs for the B.A., M.A., and Ph.D. were cal— culated and then they were used as the dependent variables in a multivariate regression analysis to test the hypothe- sis that the following factors would explain the dif- ferences in degree program cost: (1) class size, (2) level of study, (3) curriculum, (4) number of College of Education student credit hours in the degree program, and (5) ratio of graduate to total student 50 credit hours in the degree program. Each of the factors were found to be statistically significant at the 5 per cent level.91 The lead article in the first publication of a new research journal in higher education involves the use of a faculty activity survey and cost studies. Yet this research does not compare methodologies, but simply adopts an approach for degree costing. In this article Blackburn and Trowbridge reported on their research of faculty workload and cost analysis with the unit of study consisting of a Ph.D. degree. Through a faculty activity survey, time was allocated to the production of a Ph.D. The authors note in conclusion that there is no evidence of the faculty falling short with regard to accountability. "They are hard at work, a number of hours exceeded by no other occupation."92 Stuart examined three representative types of formulas used in state-wide budgeting to determine which procedures might be relevant and appropriate for improving the internal budgeting process of academic departments. Using fifteen departments at Michigan State University and data derived from the 1962-65 records of the Office of Institutional Research, he found that, The variation in budget projections derived from different combinations of objective procedures and various combinations of normative data indicate that the degree of equity achieved depends upon the 51 way in which workload is defined and measured. For some departments it makes little difference which procedure is employed, but for certain depart- ments the budget allocation resulting from the appli— cation of objective procedures varies considerably, depending upon whether objective measures of depart— mental input or output are the primary determinants of staffing requirements. The results of applying state—wide budget pro— cedures to the internal budgeting of academic depart- ments seems to indicate 1) that university budgeting can be moved in the direction of greater objectivity and equity through the selective application of various types of formula or cost analysis techniques; 2) that a more rational approach to effective resource allocation and utilization is possible with objective budget procedures than with traditional budget approaches; 3) that given adequate study and careful testing, objective procedures can be developed to the point where they are both sensitive to depart- mental differences and flexible enough to accom— modate the inevitable changes that occur from year to year in any dynamic institution; and 4) that by making explicit the various relationships that enter into resource utilization, more effective management of resources and better evaluation of management effort are possible.93 0 o Selected Variables and Relationships Whether there are severe external pressures or not, higher education should be interested in maximizing the use of its resources. Unfortunately the organizational structure and patterns in higher education do not lend themselves to cost cutting and retrenchment. Decision by consensus is a tradition in the management style of college administrators and faculty. New special interest groups are emerging such as minorities, women, and the bargaining units for faculty, all of which are insisting upon a voice in decision-making. The input from the 52 academic side of higher education, no matter how slow and indecisive, must be maintained if the quality of higher education is to be maintained. Balderston believes that, We are beginning to develop a new breed of analyti- cally trained persons who can operate with some grace ‘ at the crossing points between the academic and the I administrative sides of our institutions. There is a considerable way still to go, both in developing the techniques of cost analysis and in finding ways to weave into the pattern of decision the systematic judgements of educational effectiveness that are needed from the teachers and scholars in each discipline and profession.9 One of the outstanding contributions to the literature on cost analysis in higher education was written by John Dale Russell and James I. Doi and pub- lished in a series of twelve articles in College and University Business between September, 1955, and August, 95 1956. They recommend nine kinds of data for examining the efficient use of faculty. (1) Extent of different courses taught; (2) Semester hours of classes taught; (3) Student credit hours produced; (4) Percentage of credit hours taught in small classes; (5 v Credit hours taught in unneeded duplicate sections; 53 (6) Unnecessary repetition of courses during the regular academic year; (7) Average (weighted) size of classes; (8) Average student credit hours produced per full- time-equivalent faculty member; (9) Instructional salary cost per student credit hour produced.96 They believe this data will provide the necessary infor- mation to examine the cost factors related to proper utilization of faculty. The use of faculty manpower is the most relevant factor in measuring economy. Kettler suggests several factors which offset instructional costs. Some of these factors are: (l) the scope and type of the educational program, (2) the dis- tribution of faculty time, (3) the level of instruction, (4) faculty rank, (5) class size, (6) teaching load, and (7) the general level of faculty salaries. It is Ket- tler's opinion that, "these factors affecting costs are so inter-related that each.must be considered separately and in relation to all other factors if an accurate cost analysis is to be made."97 Several units have been used in unit cost studies. Perhaps the most common unit has been the credit hour. Just how representative this unit is of academic output has been questioned.98 McGrath states that, “One f ' " "" 54 accepted index of instructional costs is obtained by dividing total expenditures for full-time faculty salaries by the aggragate number of student credit hours of instruction given."99 Toombs challenges the use of student credit hours, or other similar factors for two reasons. First, it suggests that faculty only teach and it ignores other areas such as research and public service. Secondly, it confuses instruction and learning, since activity outside the classroom may contribute to 100 Other units that have been used are the contact hour,101 the degree,102 the student,103 the 104 the matriculated student,105 and the program.106 learning. class, In addition to the units already discussed, other factors are often considered as variables of costs. Evans and Hicks state that, "Any discussion of the major factors affecting instructional costs would necessarily include faculty teaching loads, faculty nonteaching functions, class size, salary analysis, and other "107 instructional department expense analyses. Tydall and Barnes argue that the most important single determi- nant of instruCtional cost is salaries of the teachers.108 Reeves found that class size was not a major factor, for as class size increased from nine or less up to thirty, there were only moderate increases in time required for the course. Beyond this point there was no increase.109 However, McGrath found that institutions 55 with relatively small class size had a relatively large teaching load.llo Method of instruction is a major cost factor to Ruml and Morrison. They state, "The selection of the most effective and efficient methods of instruction is a matter of overwhelming importance to the liberal college since, at any level of tuition income, faculty salaries are sharply affected by how the curriculum is administered."111 They recognize that class size may determine the method of instruction. Eckert reviewed the pertinent research concern- ing the relationship between class size and instructional quality and concluded that, in general, large group instruction is as effective as small group instruction as reflected in student achievement.112 An increase in class size would decrease the cost per student and this does not necessarily mean a reduction in instructional quality. Ikenberry states that, Quality can be sacrificed on the altar of instructional economy if (1) faculty salaries fail to receive adequate increments; (2) faculty members are requested to teach an increasing number of classes and also expected to perform creditably many other functions such as student advising, research, committee work, and public service; and (3) large group instructional methods are used irrespective of educational objectives and subject- matter requirements and merely because a large group of students is enrolled and available for instruction. Instructional quality, on the other hand, can be increased and costs of instruction controlled when curriculum reorganization reduces the number so that (l) institutional objectives can of courses, 56 become useful guides in the selection of educational experiences; (2) educational experiences can be better integrated, both during a single point in time and over the four years of college; and (3) the majority of courses enroll a sufficient number of students to allow a choice of instructional method when all factors, incIuding costs, are considered.113 Dressel suggests that, if course proliferation is stopped as a result of a decrease in educational costs, 114 then the effect may even be an improvement in education. It appears that course proliferation and class size are 115 two areas where significant savings are possible. Saupe reported on a study of the relationships between selected variables and the amount of total time devoted to instruction as assigned by the department chairmen. He found correlations between time and the .28- student credit I following variables to be: credits, hours, .40; class hours, .32; student class hours, .46, and enrollment, .33. In addition he found faculty rank had a negative relationship with: (1) percentage of (2) class hours, (3) enrol- class hours.116 time assigned to instruction, lment, and (4) student credit and Alternative Management Tools Berdahl distinguishes between cost analyses and formulas by stating, "cost analyses are attempts to . . by dividing measure past actual costs per unit . total institutional expenditures into various cost cate- gories; whereas formulas are attempts to estimate future fiscal needs on the basis of certain assumptions about 57 enrollments, faculty/student ratios, average teaching salary, ratios of instructional expenses to other insti- tutional outlays, etc."ll7 Formula budgeting has been one method used in statewide allocations to higher education. Rourke and Brooks note that while cost analyses are extremely useful for internal management, they are used less frequently than formulas for external budgeting at the statewide level.ll8 James Miller, who has made the major contri- bution to the research in this field, states that, . . . the greatest single limitation of formulas and cost analysis procedures is that they cannot make policy. . . . The intervening value judgements—- the policy decisions——must still be made by responsible individuals: higher education officials, state fiscal officials, governors, and legislators.1 It has been generally agreed, that formulas used in the analysis of budgets are useful if: "(1) their purpose is clearly understood; (2) they are based on adequate, dependable data concerning institutional activities and the context in which institutions grow and change: they are sensitive to differences, and are subject to change when change is required; and (3) they are applied with good judgment and are acceptable to parties concerned with their use."120 One of the alternatives to unit cost studies for resource allocation is internal pricing. Breneman, in a report describing the proceedings of the "Conference on Internal Pricing" held at Berkeley in July, 1971, stated that, ". . . it should be clear that internal "121 pricing is not synonomous with cost accounting. The theory of internal pricing is built upon the economic principles of the interaction of the market place. Each department "earns" a budget based upon the value of its output (measured by a schedule of internal prices), and uses this budget to purchase resources, internally priced at opportunity cost. Each department has an incentive to minimize costs for any level of output, since it seeks to maximize output from available resources in order to increase its budget.1 2 This theory is sound and supported by a signifi— cant base of research in economics. However, there are several problems in applying it to higher education including the inability to define and measure outputs and the inflexibility due to tenure. It may have some value in simulations after some solutions are found to some very real problems. Another alternative which has received much dis- cussion in the literature in recent years is (PPBS), planning-programming—budgeting systems. Judy believes it can "substantially contribute to the efficient allo— cation of resources in higher education."123 Weathersby and Balderston state that, The key conceptual components of a PPB System are: systematic long—range planning (5—15 years) which clearly articulates objectives and carefully examines the costs and benefits of alternative courses of action which meet these global objectives; a selection process for deciding on a specific course of action (1—5 years) in objectives ( ro- gramming); translating these decisions into immediate 59 (0-1 years), specific financial, manpower, and policy plans (budgeting); and recognizing a multi- year planning Horizon and incorporating to the fullest extent possible the total long—term costs and benefits attributable to each course of action.124 The application of PPBS at the University of California was not totally successful. The political considerations took over and the emphasis became one of input—control rather than output. However, . . . it contributed to, but was not necessarily the controlling influence on, the policy decisions that were made."125 Weathersby and Balderston stated that, to their knowledge, "a total, comprehensive implementation of PPBS has not been achieved in any college or university in the United States."126 Related to PPBS is the work being done in mathe- matical model building. Koenig developed one of the more comprehensive and often cited management systems for higher education. He states that, The objective of the development is to structure a mathematical model of an educational institution that will provide the "logic" of information pro- cessing programs to aid university administrators in the overall allocation of resources. . . . The model itself, then, consists of sets of equations which describe the relationship of resources to production, and, based on these, the associated unit costs of production. It is therefore a mathematical description of the way the university utilizes its resources in production. The resources of the university are described, broadly, as per- sonnel, space, and equipment. The products are identified as developed manpower, research, and public or technical services. 60 Basic to the systems approach to management is the concept of cost—effectiveness, relating inputs and outputs.128 This has also been referred to as the value— added. To determine what an institution does for its students requires knowing their condition as they enter and leave. The value-added is assessed and a cost per unit of value—added determined.129 All of the alternatives discussed in this section are built upon some costing methodology. The way costs are defined and allocated varies but each approach requires some methodology. Unfortunately, with all of the literature in costing, there is little research in comparing the effects of different methodologies. Advantages and Disadvantages of Cost Analysis The advantages of cost analysis in higher edu- cation has been cited by many writers. It has been suggested that cost analysis could be used for evalu- ating efficiency and economy, comparing purposes inter- nally and externally, studying alternatives, reporting to legislators and others, justifying the fees charged, budgeting, program planning, assisting in policy modifi- cation or formulation, evaluating different methods of instruction, answering questions on costing methodology, introducing logic rather than snap judgment into decision making, and serving as an indicator where thought is 130 needed. 61 Although generally supportive of the need for analysis, the weaknesses, pitfalls, and disadvantages of cost analysis have been expressed by several writers. Cooper says, for example, Preoccupation with detail and accountability takes the vitality out of administration, and creative dreams are drowned out, leaving only a shell without life, direction, or purpose. . . . And some in administration have found a certain kind of safety in their retreat from reSponsibility, a retreat behind the curtains of electronic gadgetry and technical detail.l3l Johnson points out that, "The 'human' element of management is still a prime ingredient." Systems are necessary for complex organizations, but " . . . they are not in and of themselves a substitute for effective leadership."132 Although he sees the need for unit cost studies, Logan Wilson emphasizes caution in how cost data are used. The use of averages as norms, for example, carries with it the virtue of standardization, but it also may lead to the vice of leveling down to mediocrity. . . . When applied indiscriminately, without regard to institutional differences in role and scope and heedless of quality, they can nullify the meaning of the one adjective in the phrase "higher edu— cation."133 The disadvantages mentioned by the writers in the field are varied. Some writers mention that cost studies lack accuracy, may imply cost is the most impor- tant aspect, are opinions not facts, are crude instru- ments, are partial measurements, are quantitative rather than qualitative in nature, often are not pointed toward 62 specific objectives and thus waste time, may lead to excessive zeal at reducing costs, and are vulnerable to misuse and misinterpretation.134 John Dale Russell succinctly stated his position with regard to the use of cost analysis in higher edu— cation and it best represents the consensus today. It is the considered judgment of the writer that the calculation of unit—expenditure data has a distinct place in the administration of higher institutions. The data must, of course, be treated somewhat cautiously; but this is true of almost every kind of factual data used in adminis— tering a college or university. Unit costs cannot be considered a substitute for administra— tive intelligence. As the financial reports of institutions take on a larger degree of uniformity than they have displayed in the past, an increased use of unit—expenditure data may be expected. The result may well be an improved efficiency in the Operation of American colleges and universities.l35 Comparison of Dissertation and Literature A perusal of the cost analysis literature in higher education reveals several phenomena including the following: 1. Authorities recognize the need to examine cost- ing methodologies, but they are quick to mention the limitations. 2. There are several cost studies, each with its own methodology; but a serious lack of compara— tive analysis of methodologies is evident. 63 3. There are several variables considered indi- vidually to be significant cost factors, but little comparison is being done with different costing methodologies. 4. Faculty activity surveys have been in the literature for a long time, but they are just now emerging in general acceptance and appli— cation. What is needed at this juncture in the development of cost analysis is a study to compare the several costing methodologies. This dissertation has the major objective of comparing four costing methodologies used in allocat— ing costs to courses in conjunction with examining the importance of several variables of cost under each of these methodologies. Therefore, this research has some elements in common with previous research. The four costing methodologies used in this study have been pre- viously employed and reported in the literature. This study will use a common set of data to compare the four costing methodologies and the relative importance of several variables of cost. NOTES--CHAPTER II lLeonard C. Romney, Faculty Activity Analysis: Overview and Major Issues (Boulder, Coloradg: National Center for Higher Educatibn Management Systems at WICHE, 1971), pp. 4-9; David D. Henry, "Accountability: To Whom, For What, By What Means?" Educational Record, Fall, 1972, pp. 287-88. 2John Minter and Ben Lawrence, eds., Management Information Systems: Their Development and Use in Higher Education (Boulder, Colorado: Western Interstate Com- m1551on for Higher Education, 1969), p. vii. 3Conrad Briner, "Administrators and Accounta- bility," Theory Into Practice, October, 1969, p. 206. 4Herman C. Bumpus, "Efficiency in the University," School and Society, I, No. 19 (May 8, 1915), 665; John W. Hicks, "Faculty Workload——An Overview," Faculty Work Load: A Conference Report, ed. by Kevin Bunnel (Wash— ington, D.C.: AmeriEEn Council on Education, 1960), pp. 10—11; Romney, 9p. 915., p. 9. 5Francis E. Falvey, Student Participation in College Administration (New York: Bureau of Publications, Teachers College, Columbia University, 1952), p. 34. 6Homer Charles Haskins, The Rise of Universities (Ithaca, N.Y.: Cornell University Press, 1957), p. 9. 7James I. Doi, "The Analysis of Class Size, Teaching Load and Instructional Salary Costs," in College Self—Study: Lectures on Institutional Research, e . by Richard G. Axt and Hall T. Sprague (Boulder, Colorado: Western Interstate Commission for Higher Education, 1960), p. 183. 64 65 8Frederick Rudolph, The American College and University (New York: Vintage Books, 1962), pp. 193-98. 9William Toombs, Productivity and the Academy: 'The Current Condition, Center for the Study of Higher Education, Report 16 (University Park, Penn.: The Pen- nsylvania State University, April, 1972), p. viii. 10James I. Doi, "The Use of Faculty Load Data Within An Institution," in Faculty WOrk Load: A Con- ference Report, ed. by Kevin Bunnell (washington, D.C.: AmeriCan Council on Education, 1960), pp. 40—42. 11Dale L. Bolton, "Measuring Faculty Load," Improving College and University Teaching, Summer, 1965, p. 157. - 12William L. Young, "Six Criteria Form a Composite Profile Chart of Faculty Load," College and University Business, XXXVI, No. 4 (April, 1964), 60. l3Hicks, 9p, cit., pp. 10-11. 4James I. Doi, "The Analysis of Class Size, Teaching Load and Instructional Salary Costs," op. cit., p. 185. —_ 15American Association of University Professors, "Statement on Faculty WOrkload," AAUP Bulletin, March, 1970, p. 31. l6Young, op, cit., pp. 59-60. 17John G. Bolin and Tom McMurrain, Student- Faculty Ratios in Higher Education (Athens: Institute of Higher Education, University of Georgia, 1969), p. 10. 18Mary T. Hobbs, "Teaching Loads in Selected Liberal Arts Colleges," Liberal Education, December, 1966, pp. 419-21. 19Robert L. Williams, The Faculty WOrk Load-- Alternate Methods of Evaluation, Report No. 19 (Denver: Education Commission of the States, December, 1970). 66 20Asa S. Knowles and William C. White, "Edu- cational Research and Statistics: A New Approach to the Evaluation of Faculty Loads," School and Society, XLIX (May 27, 1939), 684. ZlJ. D. Heilman, "Methods of Reporting the Col- lege Teacher's Load and Administrative Efficiency," Educational Administration and Supervision, XI, No. 3 (March, 1925), 175. 22Leonard V. Koos, ”Adjustment of the Teaching Load in a University," U.S. Bureau of Education Bulletin, No. 15 (1919), cited by Edwin B. Stevens and Edward C. Elliott, Unit Costs of Higher Education (New York: The MacMillan Company, 1925), pp. 113—15. 23Richard 1. Miller, Evaluating Faculty Per— formance (San Francisco: Jossey—Bass, 1972), p. 10. 24John E. Stecklein, "Research on the Faculty," in College Self—Study: Lectures on Institutional Research, ed. by Richard G. Axt and Hall T. Sprague (Boulder, Colorado: Western Interstate Commission for Higher Education, 1960), p. 79. ZSIbid. 26Ibid., p. 80. 27Lynn H. Peltier, "Faculty Activity Analysis" (East Lansing: Office of Institutional Research, Michigan State University, July, 1971), p. 5. 28Cost Study of Dental Education, The American Association of Dental Schools, 1963-64, n.p.; T. J. Campbell, Program Cost Allocation in Seven Medical Centers (Washington, D.C.: Association of American Medical Colleges, 1969); Harris Committee Report, Robert Harris, chairman (Washington, D.C.: Department of Health, Education, and Welfare, Office of the Assistant Secretary Comptroller, December, 1971). 29Jack E. Sexson, "A Method for Computing Faculty Load," Improving College and University Teach— ing, Autumn, 1967, pp. 219-22. 67 30Patrick H. Sullivan, "A Study on Bias in Faculty Reports of Time and Effort Expenditure," paper presented at the 1973 Convention of the Association for Institutional Research, Vancouver, Canada, May, 1973. 3lIbid., p. 3. 321bid., p. 5. 33The Pennsylvania State University, "Report of Enrollment, Student Credit Hours, Course Credits, and Faculty Activity, Academic Year 1972-73" (University Park: Office of the President, The Pennsylvania State University, March 30, 1973), p. 36. 34Ibid., p. 40. 35E. Grady Bogue, Thomas F. Stovall, and Brenda Norman, Faculty Time Distribution and Evaluation on Per— formance, Fall, 1972, A Report Prepared for the Tennessee General Assembly by the Tennessee Higher Education Com- mission, April, 1973, pp. 17—35. 36Ruth E. Eckert and Howard Y. Williams, College Faculty View Themselves and Their Jobs (Minneapolis: College of Education, University of Minnesota, 1972), pp. 19-20. 37D. W. Light, L. R. Marsden, and T. C. Carl, The Impact of the Academic Revolution on Faculty Careers (Washington, D.C.: American Association for Higher Edu— cation, February, 1973), p. 37. 38Ibid., p. 38. 39R. J. Henle, coordinator, "To Devise and Test Simplified Adequate Systems of Measuring and Reporting Financial, Manpower, Facilities, Research, and Other Activities in Colleges and Universities, A Final Report,“ National Science Foundation and National Institutes of Health, June, 1965, chapter 3, p. 1. 4ORomney, op. cit., p. v. 68 41Charles W. Manning, Faculty Activity Analysis: Procedures Manual (Boulder, Colorado: NationaI Center for Higher Education Management Systems at WICHE, April 12, 1973), mimeographed. 42Knowles and White, gp. cit., p. 682. 43Toombs, op. cit., p. 95. 44Ibid., p. 26. 45Ibid., p. 29. 46Harold W. Dodds, The Academic President: Educator or Caretaker (New York: McGraw—Hill, 1962), p. 69. 47Stephen K. Bailey, "Combating the Efficiency Cultists," Change, V, No. 5 (June, 1973), 9. 48Amitai Etzioni, Modern Organizations (Englewood Cliffs, N.J.: Prentice-Hall, 1964), p. 81; T. Leggatt, "Teaching as a Profession," in Professions and Profes- sionalization, ed. by A. J. Jackson TLondon: Cambridge University Press, 1970), pp. 255-60. 49Rollin H. Simonds, "To Increase Man-Hour Output in Higher Education," The Educational Record, XLIX, No. 4 (October, 1958), 338. 50Beardsley Ruml and Donald H. Morrison, Memo to a College Trustee (New York: McGraw-Hill, 1959), p. 60. 51June O'Neill, Resource Use in Higher Education: Trends in Output and Inputs, 1930 to 1967 (Berkeley: Carnegie Commission on Higher Education, 1971), p. 49. 52Floyd W. Reeves, Nelson B. Henry, Frederick J. Kelly, Arthur J. Klein, and John Dale Russell, The Uni- versity Faculty (Chicago: The University of Chicago Press, 1933), p. 277; Floyd W. Reeves, Nelson B. Henry, and John Dale Russell, Class Size and University Costs (Chicago: University of ChiCago Press, 1933), p. I49; Paul L. Dres— sel, private discussions on this research (East Lansing: Michigan State University, July, 1973). . _____,____.__-._...-_ _- _.. .... _ ,a_. - ___77 ._i .. 69 53John D. Millett, Financing Current Operations of American Higher Education (washington, D.C.: Manage— ment Division, Academy for Educational Development, December, 1972), p. 26. 54Western Interstate Commission for Higher Edu- cation, Yardsticks and Formulas in University Budgeting (Boulder, Colorado: Western Interstate Commission for Higher Education, February, 1959), p. 15. 55Paul A. Samuelson, Economics (New York: McGraw- Hill, 1970), p. 441. 56Campbell R. McConnell, Economics: Principles, Problems and Policies (New York: McGraw—Hill, 1972), p. 439. 57Paul Grady, Inventory of Generally Accepted Accounting Principles for Business Enterprises (New York: American Institute of Certified Public Accountants, 1965), p. 228. 58Carl L. Moore and Robert K. Jaedicke, Managerial Accounting (Cincinnati, Ohio: South—Western, 1967), p. 286. 59 Charles T. Horngren, Cost Accounting: A Man— agerial Emphasis (Englewood Cliffs, N.J.: Prentice Hall, 1962 , pp. 25-26. 60Gordon Zeimer, Michael Young, and James Topping, Cost Finding Principles and Procedures (Boulder, Colorado: National Center for Higher Education Management Systems, November, 1971), p. 286. 61Robert N. Anthony, "What Should 'Cost' Mean?" Harvard Business Review, XLIX (May-June, 1970), 121. 62Howard A. Withey, "Financial Reporting for Non— profit Organizations," The Journal of Accountancy, December, 1967, pp. 40—53. 63American Council on Education, College and University Business Administration, Revised (Washington, D.C.: American Council on Education, 1968). 70 64Alfred D. Cavanaugh, "A Preliminary Evaluation of Cost Studies in Higher Education" (Berkeley: Office of Institutional Research, University of California, October, 1969), p. 17; Gardner Jones and wayne Cunningham, "Cost Analysis of Instruction by Closed Circuit Tele— vision" (East Lansing: Michigan State University, December 9, 1964), p. l (mimeographed); Howard A. Withey, op. oip., p. 48. 65Withey, op. cit., p. 48. 66Douglas J. Collier, Higher Education Finance Manual: An Overview (Boulder, Colorado: National Center for Higher Education Management Systems at WICHE, December, 1972), pp. 4—5. 67Doi, "The Analysis of Class Size, Teaching Load, and Instructional Salary Costs," op. cit., p. 183. 68Harvey Sherer, "U.I.C.'s Are Dangerous," College and University Business, XVIII, No. 2 (February, 1955), 29-31. 69John Dale Russell, The Finance of Higher Edu— cation (Chicago: The University of Chicago Press, 1954), pp. 39—41. 70Edwin B. Stevens and Edward C. Elliott, Unit Costs of Higher Education, The Educational Finance Inquiry Vol. 13 YNew York: The MacMillan Company, 1925), p. 129. 71National Committee on Standard Reports for Institutions of Higher Education, Financial Reports for Colleges and Universities, Lloyd Morey, chairman (Chi- cago: University of Chicago Press, 1935), p. 177. 72Business Trends and Progress, 1967 Edition, American Credit Indemnity Company of New York (Toledo, Ohio: Century Press, 1968). 73Russell, op. cit., pp. 43—44. 74California and Western Conference Cost and Sta— tistical Study for the Year 1954-55 (Berkeley: University of California, n.d.), PP. 30—31. 71 75John M. Evans and John W. Hicks, An Approach to Higher Educational Cost Analysis, Studies in Higher Edu- cation, XCI (Lafayette, Indiana: Purdue University, 1961), p. 26. 76California and Western Conference Cost and Sta- tistical Study for the Year 1954-55, op. cit., pp. 94-95. 77Doi, "The Analysis of Class Size, Teaching Load and Instructional Salary Costs," op. cit., p. 186. 78Gordon D. Tyndall and Grant A. Barnes, "Unit Costs of Instruction in Higher Education," in Basis for Decision, ed. by L. J. Lins (Madison, Wisconsin: Dembar Educational Research Service, 1963), p. 115. 79L. E. Hull and D. A. McWhirter, Unit Cost Analysis Procedure, Indiana University (Bloomington, Indiana: Bureau of Institutional Research, Indiana Uni- versity and Indiana University Foundation, May, 1964). 80Robert E. Hubbard, "An Approach to Instructional Cost Analysis," in Basis for Decision, ed. by L. J. Lins (Madison, Wisconsin: Dembar Educational Research Ser- vice, 1963). 811bid., p. 110. 82E. G. Bogue, "An Inquiry into the Relationship Between Instructional Cost Patterns and Assumptions Influencing Analysis of Basic Data in Unit Cost Studies," Paper presented at the 11th annual Forum of the Associ- ation for Institutional Research, 1971. 83Ibid., p. 93. 84Waldo Keith Anderson, "Factors Associated with Instructional Costs in Kansas Public Higher Education, 1958-59" (unpublished Ph.D. dissertation, University of Minnesota, 1963), Dissertation Abstracts, Vol. 2507, p. 3908. 85Barry N. Siegel, "Costing Students in Higher Education--A Case Study, A Progress Report" (Eugene, Oregon: University of Oregon, August, 1967), ERIC Document, ED 014143. 72 86David R. Witmer, Unit Cost Studies (Madison: Wisconsin Board of Regents of State Universities, 1967), p. 32, ERIC Document, ED 013492. 87John J. Coffelt, Faculty Teaching Loads and Student Credit Hour Costs: Oklahoma State System of Higher Education, 1962-63 Academic Year (Oklahoma City: Oklahoma State Regents for Higher Education, n.d.), pp. 3-13. 88Chase M. Adkins, Jr., The Volume and Cost of Instructional Services at Virginia Coileges (Richmond: Virginia State Council of Higher Education, February, 1969), p. 34. 89Anne E. Scheerer, "A Formula for Determining Instructional Costs," College Management, November, 1972, pp. 25—29. 90Lynn H. Peltier and David L. Ingall, "Costing Degree Programs," A pilot study for the Michigan Council of State College Presidents (Lansing, Michigan, May, 1971), p. 12. 91Philip Edward Austin, "Resource Allocation in Higher Education: A Study of University Costs" (unpub- lished Ph.D. dissertation, Michigan State University, 1969). 92Robert T. Blackburn and Keith W. Trowbridge, "Faculty Accountability and Faculty Workload: A Pre- liminary Cost Analysis of Their Relationships as Revealed by Ph.D. Productivity," Research in Higher Education, I, No. 1 (1973), 11. 93Douglas Allen Stuart, "The Application of Formula and Cost Analysis Procedures to the Budgeting of Academic Departments" (unpublished Ph.D. dissertation, Michigan State University, 1966), pp. iii, 105-06. 94E. F. Balderston, Cost Analysis in Higher Edu- cation, Ford Foundation Program for Research in Uni— ver81ty Administration, P-33 (Berkeley: University of California, July, 1972), p. 34. 73 95John Dale Russell and James I. Doi, "Analysis of Institutional Expenditures" (a series of twelve articles on finance) College and University Business, XIX, No. 3 (September, 1955), 19-21; No. 4 (October, 1955), 27—29; No. 5 (November, 1955), 44-47; No. 6 (December, 1955), 39-41; XX, No. 1 (January, 1956), 41-45; No. 2 (February, 1956), 47-52; No. 3 (March, 1956), 41—43; No. 4 (April, 1956), 35-37; No. 5 (May, 1956), 47-48; No. 6 (June, 1956), 48-51; XXI, No. 1 (July, 1956), 43-46; No. 2 (August, 1956), 42—47. 96Ibid. (July, 1956), 43-44. 97Raymond W. Kettler, "The Analysis of Class Size, Teaching Load, and Instructional Costs," in College Self Study: Lectures on Institutional Research, ed. by Richard G. Axt and Hall T. Sprague (Boulder, Colorado: western Interstate Commission for Higher Education, 1960), pp. 203-04. 98Toombs, o . cit., pp. l8-19; M. F. Lorimer, "How Much Is a Credit fiafir?" Journal of Higher Edooation, XXLIII (June, 1962), 302-06; James M. Heffernan, "The Credibility of the Credit Hour: The History, Use, and Shortcomings of the Credit System," Journal of Higher Education, XLIV (January, 1973), 61, 72. 99Earl J. McGrath, Memo to a College Faculty Member (New York: Teachers College, Columbia University, 1961), p. 28. 100William Toombs, Productivity: Burden of Success, ERIC/Higher Education Research Report No. 2 (Washington, D.C.: American Association for Higher Education, May, 1973), pp. 17-18. lOlBolton, op, cit., pp. 157—58. lOZPeltier and Ingall, 3p. oip.; Philip Edward Austin, Resource Allocation in Higher Education: A §§udy of University Costs, op. cit.; Blackburn and Trowbridge, op. oip., pp. 17126——~ 103Hull and McWhirter, op. cit. 104Evans and Hicks, op. cit., p. 10. 74 105Merlin George Duncan, "Institutional Unit Costs in Selected Central American Universities" (unpub— lished Ph.D. dissertation, Michigan State University, 1964), p. 97. 106Ralph Nelson Calkins, "The Unit Costs of Pro- grams in Higher Education" (unpublished Ph.D. disser- tation, Columbia University, 1963), Dissertation Abstracts, Vol. 24, p. 3578. 107Evans and Hicks, op, cit., p. 10. 108Tyndall and Barnes, op. cit., p. 115. 109Reeves, et a1., The University Faculty: 22- cit., pp. 276-76. lloMcGrath, op. cit., p. 30. lllRuml and Morrison, op. cit., pp. 19-22. 112Ruth E. Eckert, "College and University—— Programs," in Encyclopedia of Educational Research, ed. by Chester W. Harris—(New York: The MacMillan Company, 1960), p. 279. ll3Stanley Ikenberry, "Instructional Cost and Quality," College and University, XXLVII, No. 3 (Spring, 1962), 50. 114Paul L. Dressel, "Measurement and Evaluation of Instructional Objectives," The 17th Yearbook, National Council on Measurements Used in Education (Ames, Iowa: National Council on Measurements Used in Education, 1960), p. 5. 115Dexter M. Keezer, ed., Financipngigher Edu- gotion 1960—70 (New York: McGraw-Hill, 1959), p. 6. 116Joe L. Saupe, "Relationships Among Various Measures of Instructional Load" (East Lansing: Office of Institutional Research, Michigan State University, April 20, 1961), pp. 4-5. 75 117Robert O. Berdahl, Statewide Coordination of Higher Education (washington, D.C.: AmeriCan Council on Education, l97I), pp. 122-23. ll8Frank Rourke and Glen Brooks, Managerial Revolution in Higher Education (Baltimore: Johns Hopkins Press, 1966), p. 81. lngames L. Miller, Jr., State Budgeting for Higher Education: The Use of Formulas and(Cost AnaIysis, Michigan Governmental Study No. 45 (Ann Arbor: Institute of Public Administration, University of Michigan, 1964), pp. 155-56. 120Yardsticks and Formulas in University Budgeting, Op. Cit.’ p0 41. — 121David W. Breneman, Internal Pricing Within the University, A Conference Report, Ford Foundation Program for Research in University Administration, P-24 (Berkeley: University of California, December, 1971), p. 13. 122Ibid., p. 29. 123Richard W. Judy, A Research Progress Report on: Systems Analysis for Efficient Resource Allocation in Higher Education (Toronto, Canada: Institute for the Quantitative Analysis of Social and Economic Policy, University of Toronto, January, 1970), p. l. 124George B. Weathersby and Frederick E. Balder- ston, PPBS in Higher Education Planning and Management, Ford Foundation Program for Research in University Administration, P-31.(Berkeley: University of California, May, 1972), pp. 5-6. lZSIbid., pp. 47—48. 126Ibid., p. 2. 127Herman E. Koenig, Martin G. Keeney, and R. Zemach, A Systems Model for Management, Planning, and Resource Allocation in Institutions of Higher Education (East Lansing: Division of Engineering Research, Michigan State University, September 30, 1968), pp. i, 7. 76 128Philip H. Coombs and Jacques Hollak, Managing Educational Costs (London: Oxford University Press, 1972), p. 82. 129Harold L. Hodgkinson, "How Can We Measure the 'Value Added' to Students by a College Education?" The Chronicle of Higher Education, November 13, 1972; JBHH KelIer, rrHigher Education Objectives: Measures of Per- formance and Effectiveness," Management Information Sys- tems (Boulder, Colorado: The Natibnal Center for Higher Education Management at WICHE, October, 1969), pp. 79-84. 130Seymour E. Harris, Higher Education: Resources and Finance (New York: McGraw-Hill, 1962), p. 503; Thomas Mason Freeman, "A Multiple Correlation Analysis of the Supplies and Services General Fund Budgets for Selected Academic Departments at Michigan State Uni- versity: 1964—65 and 1965—66" (unpublished Ph.D. disser— tation, Michigan State University, 1967), pp. 27-29; Thad L. Hungate, Finance in Educational Management of Colleges and Universities (New York: Teachers College, Columbia University, 1954), p. 100; Balderston, op. 933., pp. 2—3; Witmer, op. oi:., p. 3. 131Lloyd G. Cooper, "Decision Ability, Not Accountability," Journal of Higher Education, XLIII, No. 8 (November, I972), 656-59. 132Walter F. Johnson, "Some Comments on Efficient College Management," review of Efficient College Manage- ment, by William W. Jellema, in Educational Administrative Quarterly, Spring, 1973. 133Logan Wilson, "Analyzing and Evaluation Costs in Higher Education," The Educational Record, XLII, No. 2 (April, 1961), 102. 134L. E. Hull, "Pitfalls in the Use of Unit-Cost Studies," Journal of Higher Education, XXLII (October, 1961), 373—76; M. M. Chambers, Financing Higher Edu- cation (Washington, D.C.: The Center for Applied Research in Education, 1963), pp. 85—86; Toombs, Productivity: Burden of Success, op. cit., pp. 20—21; Evans and Hicks, op. oip., p. 7. —— ___ l35Russell, op. cit., p. 155. .___ __.__ __.;._..... __- . . . ._ _- .___._... ,___-,,-t .— CHAPTER III RESEARCH DESIGN Summary of the Rationale for the Study Effective resource projection and allocation requires a thorough analysis of the faculty activities and their related costs. Wise use of faculty is a major reSponsibility of college and university administrators. Evans and Hicks remind us that, "Faculty teaching load is unquestionably the most important factor influencing instructional costs and therefore should be subjected to careful periodic analysis."1 The control of instruc— tional costs requires the identification of the major variables of costs and their interrelationships. Yet, costing methodologies have lacked a thorough examination to assess just what results when different methodologies are employed. To generate a cost figure is not sufficient when it is realized that the costing methodology will, to a great degree, determine that cost figure. What is needed is research which will inquire into the nature of the costing methodology used for allocation of costs to the individual units chosen. 77 78 Purposes of the Study The purposes of this study were the following: 1. Develop a faculty time allocation profile by department and rank of the major activities in which the faculty participate; assign costs to these activities based on the allocation of time as determined by a faculty activity survey; and examine the interrelationship between several workload factors. 2. Compare four costing methodologies as a basis for examining the relative importance of some major variables in determining instructional costs and for considering their interrelationships. This was a descriptive study which considered the major faculty activities of one college at Michigan State University. The intention of this study was to consider different costing methodologies and the importance of five major variables in determining costs. Most cost studies concentrate on arriving at a final figure, but few try to determine the reasons for a particular figure appearing as it is reported. This study not only gen- erated costs based on different methodologies; but it also explored what type of differences, depending on the methodology employed, resulted in several independent variables. 79 Research Objectives The objective of this dissertation were divided into two parts. The first was to examine the inter- relationships of selected instructional workload factors through determining the distribution of faculty time among several activities based on the salaries of the faculty. The profile analysis was developed by depart- ment and by rank. The second objective was to compare the four costing methodologies used in allocating costs to courses. This comparison was not only based on the costs generated per course, but also considered the differences in selected variables of costs and the respective costing methodology. The comparison was made by answering the following research questions: 1. Is there a relationship between each independent variable and section cost for each costing methodology? 2. What is the relative importance of the indepen- dent variables for each costing methodology? 3. Under which costing methodology do the variables explain the greatest amount of the variance in the costs? 4. To what extent is there agreement in the rank ordering of independent variables across the costing methodologies? 80 Parameters of the Data The population from which the data were drawn for this study were the faculty and course offerings of one college at Michigan State University. This college offers the Bachelor of Arts, Master of Arts, and Doctor of Philosophy degrees. Composed of five departments, it is diversified in both the nature of its departments and course offerings. It ranges from nonlaboratory departments to more scientific laboratory departments, and from the lower undergraduate courses to the advanced graduate programs. Some faculty and departments have gained international recognition for the contributions they are making to their field. There were fifty-one full-time faculty members who taught during the 1972 fall term. They were all included in the faculty activity survey which was the method for collecting the research data. The part-time faculty were not included since the focus of this study was on the regular full-time faculty. All of the full- time faculty were within the ranks of instructor, assistant professor, associate professor, and professor, except one. There was one full-time lecturer who was considered an instructor for the purposes of this study. Although graduate assistants were excluded from the profile analysis, they were included as a cost factor in the costing methodology portion of the dissertation. 81 The salary for the graduate assistant was allocated to the course section based upon what the faculty member reported as the graduate assistant's total hours spent on the course. Although some of the data can be verified with the records of the Office of Institutional Research, much of the data is a reflection of the faculty members' perception. Therefore, the data used for analysis reflects the faculty members' view of their workload with regard to activities and the distribution of the several variables. Costs for this study were limited to salary expenditures; thus supplies, equipment, and capital out- lay were excluded. The human resource is the major cost element of higher education and therefore has the greatest influence on the financial operation. Dressel and associates emphasized this point in The Confidence Crisis when they said, "Since salaries account usually for seventy per cent or more of departmental expenditures and represent that aspect of expenditures which is most immediately subject to modification, emphasis on person- nel time and salary expenditures is the key to manage- ment."2 A major thrust of this study was to examine the costing of faculty resources, and therefore salaries are our primary cost consideration. 82 Data Collection A faculty activity survey of each faculty member in the college, which is the most predominant method for gathering data,3 was the means for collecting all but the salary data. The survey was conducted during the 1972 fall term. Over a period of several months the Office of Institutional Research staff and the advisory council for the college had been considering approaches to research within the college which would provide meaning— ful data for a cost study. The college faculty, repre- senting five departments, received their forms during November 1—10 and were asked to return the completed forms to their respective chairmen by November 16. The survey forms were distributed by the chairmen of each department. The questionnaire which was used is entitled Faculty Activity and Outcome Survey4 and was designed by the National Center for Higher Education Management Systems. Financial data from the Office of Institutional Research were used in conjunction with the data provided through the questionnaire. Although it is estimated to require one hour to complete the question- naire, a very high return was expected since this was a faculty initiated study and they were very interested in examining the allocation of their time and the associated costs. The questionnaire permits a wide 83 range of activities to be included, thus facilitating a comprehensive look at the total faculty activity as it relates to time. Although Campbell and Stanley in Experimental and Quasi-Experimental Designs for Research5 present a series of factors which can jeopardize the validity of various designs, for this research two appeared to be of real concern. Since some of the data were collected by a survey, it was possible that a self—selection factor could result. For this reason an attempt was made to contact those faculty not responding to assess whether they were uniquely different from those responding. Of the three faculty not responding, two were contacted. One reported no teaching responsibilities with his full- time responsibility related to a funded research project. The other faculty member taught only one course and spent most of his time advising a student organization. There- fore, two of the three not responding were not typical teaching faculty. The reactive effect of faculty being asked to provide this kind of information could result in dis- torted data. However, since this was a study which was initiated by the college with faculty participation, it was likely that the faculty would conscientiously com- plete the questionnaire. 84 One of the classical studies which illustrates the reactive effect was conducted at the Western Electric Company's Hawthorne WOrks in Chicago from 1927 to 1932. These studies have come to be known as the Hawthorne studies. Among other things, they found that, "The employees being tested were reacting to changes in light intensity in the way in which they assumed that they were expected to react."6 The very fact that they were aware of the experiments affected their performance. Since the data for this study came mainly from the faculty, it is possible that the reported hours could be distorted. As was noted by Hubbard,7 Sullivan,8 and others, some faculty may react negatively to a request for activity data and therefore not report accurately their time allocation. Although some express full con- fidence in the faculty reporting their time, many acknowledge that the possibility for error remains. Therefore, in this study, those returned questionnaires which exceeded a plus or minus three standard deviations from the mean of their respective departments were excluded from the analysis. In a normally distributed pOpulation 99.73 per cent will be within three standard deviations from the mean. All of the other factors considered by Campbell and Stanley were either not relevant to this study or have been considered and ruled out as plausible extraneous variables. 85 The questionnaires were sent to all the faculty by the department chairman and returned to him. Of the fifty—one full-time faculty during fall term 1972, forty- eight returned the questionnaire which is a return rate of 94.1 per cent. Of the forty-eight returned question- naires, three were not sufficiently complete to make them usable. One questionnaire exceeded the three standard deviations from the mean and was excluded. The total reported hours per week on this questionnaire were 126. Three standard deviations for this department totaled 53 hours and the mean was 69.8 hours. Thus, the range was 16.4 hours to 123.2 hours. The nearest high to the one which exceeded three standard deviations did not exceed two standard deviations. Therefore, the total usable returned questionnaires were forty-four, which was 86.3 per cent of the total full-time faculty during the 1972 fall term. Table 1 summarizes the return rates. TABLE l.--Returned questionnaires 1972 Fall Term Usable Returned Department Full-Time Faculty Questionnaires Percentage Dept. 0 6 6 100.0 Dept. l 13 10 76.9 Dept. 3 15 13 86.7 Dept. 5 10 8 80.0 Dept. 7 7 7 100.0 Total 51 44 86.3 1% I 86 Assumptions of the Study The assumptions of this study are the following: The faculty member can work only 100 per cent, i.e., this study considers the total hours worked regardless of what should be expected as a normal work week. Cost may be assigned to the total hours worked, i.e., the salary is for the total professional services of the faculty member. There is a linear relationship between independent and dependent variables with the exception of one independent variable-—method of instruction. Limitations and Scope of the Study The limitations and scope of this study are the following: 1. This study was limited to the faculty and course offerings of one college at Michigan State Uni— versity. Since human resources are the major cost factor in higher education, the costs for this study were limited to faculty salaries thus excluding supplies, equipment, and overhead. The question of quality and value of activities cannot be evaluated solely on the basis of this study. 2,71,- 1. 87 4. Generalizations to other colleges concerning faculty allocation of time or the cost of par- ticular activities cannot be made on the basis of this study. These are highly institutionalized considerations. The distribution of time and cost patterns need to be evaluated as they relate to the mission and objectives of the department and the institution. For example, a department or institution which stresses research as opposed to undergraduate instruction will certainly have different cost patterns and distributions of time among activities. However, the procedure of analysis is applicable for use by other colleges and universities. Instrument Reliability and Validity As was pointed out in the review of the literature, there is no agreement as to reliability and Validity of faculty activity surveys. However, some of the most current research is beginning to support such procedures. The University of Michigan has conducted a university—wide faculty activity survey for several years. They conducted an experiment using the Faculty Activity and Outcome Survey to assess the reliability of the instrument. A comparison was made between a sample of the faculty who completed the NCHEMS form and the faculty who completed the Michigan form which they have used for 88 several years. Only two activity categories, credit instruction and professional development, were signifi- cantly different. The other four categories, noncredit instruction, research, service, and administrative activities, were not significantly different. It is meaningful that when the questionnaires were compared at the departmental level for the three departments which participated, there were no significant differences; the two differences only resulted in the aggregate. One possible explanation of these differences may be due to reporting in percentages instead of average hours. Since the Michigan form allows report- ing in percentages, the University's Office of Insti- tutional Research has examined this issue. In the most recent comparison, made in the Uni- versity's largest unit, the initial indications are that those individuals who report in percentages tend to report more time in credit instruction than This means that the those who report in hours. possibility exists that the pilot study differences in credit instruction may be due to the unit of measurement used to report time rather than in the questionnaire used. In their conclusion, the observation is made that, Because of the consistency of the pilot results with other analyses in the area of differences caused by the use of hours or percentages as the one could conclude that the reporting measure, differences caused by the survey forms themselves are actually much.smaller than those reported. . . . When it is remembered that virtually the same information concerning courses, sections, contact hours, etc. is reported in the category on both questionnaires, this conclusion seems more warranted. Therefore, to the extent that one is 89 willing to attribute the credit instructional dif- ferences on the two forms to differences caused by the units of measurement, the two questionnaires are comparable and measure the same phenomena.lo Other institutions which have conducted faculty activity surveys such as Ohio State and the University of Missouri have found that the data can be a reliable indicator of faculty distribution of time.11 The environment in which the survey is conducted and how the faculty are approached are critical factors in the The survey can be useful for success of the survey. If changes monitoring activities but not for control. in activities are desired, they will only be brought about by adjusting the reward systems and policies of the individual college. Although Michigan State University does not rou- tinely conduct a faculty activity survey as most of the other "big ten" universities do, a university-wide survey was conducted in the fall of 1970.12 This survey, which became known as the "green form" survey, showed an average of 57.6 hours per week per faculty member. Since the basic data on the college used as the popu- lation for this dissertation were available from the the means were compared with the 1972 1970 survey, A t—test was used to test for significant dif- survey. ferences in the mean total hours reported per faculty member between the two surveys for each department and __— — -' _—o—-—‘—_ 90 for the total college. No statistically significant differences exist. Table 2 below shows the mean number of total hours reported by department for each survey and the t score. TABLE 2.--Test for significant differences in total hours between 1970 green form and 1972 NCHEMS form 1970 Mean 1972 Mean Degrees of Department Green Form NCHEMS Form Freedom t Score Dept. 0 57.8 61.9 12 - .87 Dept. 1 69.9 59.4 23 1.13 Dept. 3 68.6 69.9 26 - .19 Dept. 5 72.9 67.4 16 .83 Dept. 7 63.4 54.6 12 1.52 Total College 67.5 63.5 97 1.26 No significant differences The question of validity was approached by Lorents in Minnesota. A sample of faculty were asked to estimate, before the term and again after the term, how many hours per week they would spend in eleven cate— gories. During the term a "work-study experiment" was conducted in which the faculty carried a "beep" device at all times. When the device beeped they wrote down their activity at that time. Only two categories were significantly different in the pre-term estimates and the post-term estimates. The point is that the survey approach, which is much less 91 costly, provided close to the same results as a relatively expensive work study experiment. The debate over the issue of reliability and validity will likely continue, but from a practical administrative point of View, the loss in accuracy may be offset by the advantages of 1555 cost and time required in procuring the data. clear that a survey, "However, it does seem if properly conducted, can result in data that are reliable and reasonably valid and that provide institutional managers with at the very least some rough answers rather than undocumented guesses." 14 Profile Development One of the two purposes of this dissertation was to develop a faculty profile and for each department. of activities for each rank The activities and their defi- nitions come directly from the survey instrument and include the following: Activity A. Teaching Activities A.l Scheduled teaching . . . A.2 Unscheduled teaching . . Academic program advising . . . . . . . Course and curriculum research and development Definitions All activities related to courses given in the current term. Teaching not associated with the specific courses listed in A.1. - Giving advice to students concerning course scheduling and academic programs. Developing and preparing for future courses. 92 B. Research, Scholarship and Creative WOrk Activity Institutes and Research Centers . . . . . . . . .Activity that is carried on for an institute or center. (This may include B.2 activity which is directly linked to a separate unit of adminis- _ . tration) SpelelC projects . . . .WOrk related to a specific project such as writing research proposals, books, articles; giving recitals; and performing your profes- sional skill. General scholarship and professional development.Wbrk activities related to keeping current in a pro- fessional field. C. Public Service Activities General professional service/advice directed outside the institution .Activities that would not be considered C.2 and meant to benefit the community outside the institution, e.g., con- sulting, advising, and lectures for the public. Extension service (not instructional) . . . . .Activity directed outside the institution where fiscal con- trol is shared by the insti- tution and government agen- cies. D. Internal Service Activities Student—oriented service.Activities related to stu— dent non-academic activi- ties, e.g., preparing recom- mendations, social inter- action, sponsoring student organizations and personal, career, and financial coun— seling. Service reports and records . . . . . . . . .Fulfilling institutional information requests, e.g., writing and answering memo- randa, preparing minutes, and completing questionnaires. HM“- .. *«.‘.._ .—,r Mu -_ _.___ . _ 93 E. Administrative and Committee Activities E.l Administrative duties . .Administrative functions which include helping during registration, gathering data, preparing budgets, and per- forming the duties of a department chairman. E.2 Committee participation .Committee involvement which might include committees for planning, budgeting, admis- sions, candidate selection, and faculty senate.15 Each questionnaire was examined by a representa- tive of the Provost office for the purpose of noting which activities were not within the purposes of the university. Although many kinds of public service activities are encouraged by the university, some are up to the individual faculty member and may not be directly related to the disciplines of the college or to the purposes of the university. Examples might be political activities or church activities. Only three questionnaires included such activity which amounted to one hour each, two for church activities and one for political activities. Among the above activities there were four dis- tributions made by rank and department which are: (1) average number of hours per week, (2) percentage of the average hours per week, (3) average faculty salary, and (4) total salaries. This profile provides an approach for examining the utilization of faculty and the associated costs. 94 Since the second purpose of this dissertation was to compare costing.methodologies which deal with the scheduled teaching section of this profile, several cor- relations were made between many of the instructional load factors. This instructional load correlational study provides some information for administrative decision—making. Dependent Variables The other purpose of this dissertation was to compare four costing methodologies as applied to the scheduled teaching section of the profile analysis. The dependent variables for this objective of the study were the course section costs generated by the four costing methodologies. The four costing methodologies are based on the course section's percentage of the faculty member's total number of: (1) hours of total course time, (2) formal contact hours, (3) student credit hours, and (4) course credit hours. Each course section received four separate cost figures, one for each costing methodology. In this study the four methodologies for generating costs were the dependent variables. 95 CostingVMethodology The unit of study was the individual course. The focus of this research was on the methods used in allocat- ing the costs to each of these course sections. Often cost studies concentrate on generating a cost per unit and the unit is frequently a contact hour, a student credit hour, or a course credit hour. Each of these units were considered in one of the costing methodologies used for allocating costs to the course sections in this study. The method of allocating the costs to the indi- vidual courses will to some extent determine what the cost figure will be when it is calculated for one of the smaller units such as per student credit hour, per course credit hour, or per contact hour. Although all costs including secretarial assistance, supplies and services, and overhead could be prorated to each course, for the purposes of this study only faculty and graduate assistants' salaries were included in the course section costs. The primary focus of this dissertation was on faculty workload and costing of the instructional activities; therefore, salaries were the primary cost consideration. The salaries were taken from the Michigan State University Salary Budget for 1972-73.16 Included in the usable returned questionnaires were 124 course sections. For each course the faculty 96 member reported the total number of hours he spent on the course, including preparation, administration, formal contact hours, and other contact hours. He also reported the enrollment and the number of credits. By multiplying the enrollment times the credits, the stu- dent credit hours were derived. First Level of Allocation Two levels were developed for allocating the faculty member's salary to each course section for each of the four costing methodologies. The first level was used for all four of the costing methodologies. This first level of allocation was developed by multiplying the faculty member's salary times a ratio of total scheduled teaching activities divided by the total work- week. For example, if a faculty member spent half of his time in scheduled teaching, one-fourth in public service, and one—fourth in research, then his salary would be multiplied by 50 per cent. The scheduled teaching portion of his salary is the amount of salary which was then allocated by the four costing methodolo— gies. Second Level of Allocation There were four methodologies for allocating salary under the second level of allocation. Each of the four methodologies is described below. 97 1. The first methodology was to allocate the scheduled teaching salary based on the course section's percentage of the faculty member's total hours devoted to scheduled teaching referred to as total course time. 2. The second methodology for allocating costs was to assign the scheduled teaching salary to the course sections by calculating the percentage between the faculty member's total formal contact hours and the formal contact hours for the given course section. 3. The third methodology was to assign the scheduled teaching salary based on the course section's percentage of the faculty member's total student credit hours produced. 4. The fourth.methodology was to allocate scheduled teaching salary based on the percentage of the total credits taught by the faculty member and the credits for the given course section. Table 3 provides an example of how the costs were calculated for each of the four costing methodolo- gies and assigned to the particular course section. Some issues of interest related to this study involve the value of a faculty survey as a methodology for costing. The value of such a survey for costing 98 could be challenged if the relationships between variables and costing methodology differ little. If on the other hand there are significant differences, then perhaps the faculty activity survey can provide the answers to the question of why there are differences. TABLE 3.-—Cost calculation Professor John Doe Term Salary $8,000 Total hours per Scheduled teaching hours 15 25% week 60 Scheduled teaching costs $2,000 Statistics . . Total Course/Section Credits SCH Contact Hours Eoui§E_Time A 3 60 2 6 B 3 32. 4 9 6 9O 6 15 Costs Course/Section Method 4 Method 3 Method 2 Method 1 A $1,000 $1,333 $ 667 $ 800 B 1,000 667 1,333 1,200 $2,000 $2,000 $2,000 $2,000 Independent Variables The independent variables were factors considered to have a relationship to costs. Two of the variables, faculty rank and course level, were on the ordinal scale. Class size and number of sections taught were on a ratio scale. Method of instruction was on a nominal scale and was treated separately. After reading the literature, 99 several variables were added to the initial list and included in a correlation analysis from which a selection was made for additional analysis. All of these variables were on a ratio scale except graduate assistants, which was on an ordinal scale. The initial list of variables was the following: 1. Faculty rank, with four levels: instructor, assistant professor, associate professor, and professor. 2. Course/section level, with three levels classified as: lower undergraduate, upper undergraduate, and graduate. 3. Class size, which is the enrollment in the course section. 4. Number of courses and/or sections taught by the faculty member during the term. 5. Methods of instruction, which are defined in the questionnaire as: Lecture Formal presentation-primarily one-way communication Laboratory Instructing, preparing, and supervising student investi- gations Recitation/Discussion Two-way communication of course materials Seminar Students carry the major responsibility for prepar- ation 100 Independent Study Students work independently with only minimal faculty direction Tutorial Students work one-to—one with the instructor Programmed Instruction Course contents presented 1 through programmed materials The other variables which were included in a correlational analysis were: (1) the faculty member's years of experience, (2) the number of fixed credit course sections taught by the faculty member, (3) his total workweek in hours, (4) the number of times he has taught the course, (5) the course credit hours, (6) the student—credit hours per course section, (7) the formal contact hours per course section, (8) the other contact hours per course section, (9) the preparation and administrative time per course section, (10) the total teaching time per course section, and (11) whether or not a graduate assistant was assigned to the course section. All but one of the variables, method of instruction, were at least on an ordinal scale, which indicated a rank ordering of categories within the variable. Method of instruction was on a nominal scale which is simply the assignment of numbers to a qualita- tive category and it therefore had none of the properties of the other scales.l8 Hence it was treated by a separate analysis. 101 Research Questions and Statistical Approaches Although this is a descriptive study in which a profile of faculty activity was developed and costs assigned to these activities, it was also the intention of this study to examine the influence of the four costing methodologies on the importance of five variables in determining cost. As part of the profile analysis, the instructional portion of the profile was examined by a correlational analysis since it related to the second purpose of this dissertation, namely, instructional cost- ing methodology. Comparing the costing methodologies and selected variables of cost using the instructional costs of the profile analysis was accomplished by answering the four research questions. The statistical approach follows each question as listed below. 1. Is there a relationship between each independent variable and cost for each of the costing methodologies? Approach Simple bivariate correlations and One-way Analysis of Variance 2. What is the relative importance of the variables within each costing methodology? 102 Approach A multiple regression analysis--The relative importance was determined by the contribution each variable made to the total R2 as developed by the stepwise addition and deletion methods.‘ 3. Under which costing methodology do the variables explain the greatest amount of the variance in the costs? Approach Coefficient of multiple determination—R2 4. To what extent is there agreement in the rank ordering of independent variables across the costing methodologies? Approach Kendall Coefficient of Concordance Each of the statistical approaches mentioned as the method for answering the research questions is dis- cussed below. The basic statistical measures are dis- cussed here to facilitate reporting the results in sta- tistics. Correlation Coefficient The simple bivariate correlation coefficient (r) is a single number which measures the extent to which two things are related, i.e., the extent variations in one variable change with variations in the other. 103 Correlations may vary from a +1.00, which is a perfect positive correlation, through zero, which means complete independence, to a -l.00, which is a perfect negative correlation.19 A general verbal description of correlation coefficients by Guilford provides a guide for interpre- tation. Less than .20 Slight; almost negligible relationship .20 -- .40 Low correlation; definite but small relationship .40 -- .70 Moderate correlation; substantial relationship .70 —- .90 High correlation; marked relationship .90 --l.00 Very high correlation; very dependable relationshipzo One-Way Analysis of Variance The basic purpose of the one-way analysis of variance was to determine if the means for each method of instruction for each costing methodology varied signifi— cantly from the total population mean. Since method of instruction is on a nominal scale it was not apprOpriate to include it as a variable in a regression analysis. Although this was a population study and it was not therefore necessary to deal with inferential statistics, in this case an F test was used at a .05 level of sig- nificance as a guide for assessing the meaningfulness of the analysis of variance. 104 A significant F means that there is a relation- ship between the dependent and independent variables and it is not a result of chance. The F test does not tell where the differences are between the groups.21 For this purpose the Scheffé method of multiple contrasts was employed. "According to Scheffé's theorem, it is possible to study observed data and search out meaningful contrasts and confidence intervals to help identify possible reasons for the rejection of a tested hypothe- sis."22 By the Scheffé method each mean for each method of instruction was compared with each other. With this method it was possible to investigate each contrast while keeping the probability of making a type I error (the denial of a true hypothesis) equal to .05.23 Fifteen contrasts were made based on the cost per course. These course costs were converted to cost per course credit for the costing methodologies developed on total course time, contact hours, and student credit hours. Course costs developed on course credits could not be converted to costs per course credit since the mean for each method would be the same. The Scheffé method was then used on the cost per course credit for each of the three costing methodologies. Moltiple Regression Analysis Multiple regression analysis was applied with cost as the dependent variable and the four independent 105 variables composed of faculty rank, course level, class size, and the number of sections taught by the faculty member. A multiple regression coefficient (R) indicates the extent of relationship between a dependent variable and two or more independent variables. "In the many areas of research in which controlled eXperiments are not practicable, multiple regression analyses are exten- sively used in attempts to disentangle and measure the 24 effects of different X variables on some response Y." The multiple regression model which will be used is: Yi = a + lel + b2X2 + b3X3 + b4X4 + e where: Yi = (l...4) Cost based on the four methodologies X1 = Faculty rank X2 = Course level X3 = Class size X4 = Number of sections taught e = the residual, assumed to be distributed independently of the X's with zero mean and . 2 variance 0 a = a constant 106 b = the expected change in Y when X decreases assuming all other X's remain unchanged b = the same as bl respectively. 2’ 3’ Since it was not necessary to deal with infer- ential statistics because the entire pOpulation was included in the survey, the correlations of each variable on each of the costing methodologies were compared. The objective was not only to discover the relationship of the variables with cost, but to also rank the variables in order of their importance. Because high correlations between the X variables can upset the cal- culations, the methods employed were both the stepwise addition and the stepwise deletion. Correlation and regression analysis is especially suitable for studies where no single variable has been established as a totally apprOpriate evaluation criterion. As Freeman points out, "Correlation and regression analy- sis, while used extensively in industry as a tool of budget evaluation, has not been effectively utilized in higher education budget analysis or at least it has not been reported in the literature."25 Stepwise Addition.--After a regression is com- puted, the utility of a variable may be questioned and its omission proposed. The most thorough approach is 1 increases or 107 to work out the regression of Y on every variable singly. The procedure for the stepwise addition process is to sequentially build the equation. The variable giving the greatest reduction in the sum of squares of Y is selected. This is called X1. Then the bivariate regressions in which Xl appear are worked out. The variate which gives the greatest additional reduction in the sum of squares after fitting X1 is selected. This is called X2. All trivariate regressions that include both X1 and X2 are computed, and the variate that makes the greatest additional contribution to them is selected. This process is continued for all independent variables.26 Stepwise Deletion.--The stepwise deletion method is basically the reverse process of stepwise addition. This process may not necessarily select the same inde- pendent variables nor will it necessarily account for the same amount of variation in the dependent variable. The regression of Y on all variables is worked out. The variable that gives the least additional reduction in sum of squares is then dropped, and so on. Coefficient of Multiple Determination.--The R2 is the coefficient of multiple determination of variance (the square of the multiple correlation coefficient). This measure indicates the proportion of variance in the depen- dent variables combined with the regression weights used. 108 For example, an R2 of .8759 indicates that 87.59 per cent of the variance in variable Y is explained by the variation in the combined X's. The methodology which yields the highest R2 is the best fit relationship between the independent variables. That is, it provides the best weighting of the variables for predictive ability.27 Beta weights were calculated for each variable. The beta weights are the constants in the regression equation and they provide a means of measuring the change in the dependent variable for a given change in an inde- pendent variable assuming the other independent variables are held constant. The coefficient of multiple determi- nation (R2) is the sum of the beta weights times their reSpective simple r.28 Shortcomings of regression analysis are recognized.. In multiple regression analysis the value of any regression coefficient depends on the other variables included in the regression. Difficulty may arise because one can never be sure that there are not other X variables related to cost. Even if the regression coefficients are clearly meaning— fully significant, it is not uncommon to find that the fraction of the variance of Y attributable to the regression is much less than 50 per cent. This indicates that in much of the research, most of the variation in Y is due to variables not included in the regression. To 109 further complicate the problem, one must recognize that the relationships may not be linear. Even with all of the above mentioned problems, multiple regression analysis is among the best available statistical methodologies to meet the objectives of this aSpect of the study. Kendall Coefficient of Concordancezg The Kendall Coefficient of Concordance was used to assess the degree to which there was agreement in the rank ordering of variables across the several costing methodologies. A high agreement indicates that no one methodology gives a meaningfully different rank ordering of variables. The degree of agreement among the four costing methodologies, as to their rank ordering of the variables, was reflected by the degree of variance among the four independent variables. The coefficient of con- cordance is a function of that degree of variance. A high value of the Coefficient of Concordance may be interpreted as meaning that the costing methodologies apply essentially the same standard in ranking the four variables. Summary In this chapter the design of the study was pre- sented. A summary of the rationale for the study, the purposes, and research objectives were discussed as were 110 the parameters of the data and the methods of collecting the data. The assumptions, limitations, and scope of the study were listed and a discussion on the instrument reliability and validity was presented. The dependent and independent variables were defined and discussed. Also included was a discussion on the profile develop- ment, the costing methodology, and the statistical approach. NOTES--CHAPTER III 1John M. Evans and John W. Hicks, An Approach to Higher Education Cost Analysis, "Studies in Higher Edu- cation," XCI (Lafayette, Indiana: Purdue University, 1961), p. 11. 2Paul L. Dressel, F. Craig Johnson, and Philip M. Marcus, The Confidence Crisis (San Francisco: Jossey- Bass, 1970), pp. 2-5. 3Leonard C. Romney, Faculty Activity Analysis: Overview and Major Issues (Boulder, Colorado: National Center for Higher Education Management Systems at WICHE, 1971), p. 85; Patrick H. Sullivan, "A Study on Bias in Faculty Reports of Time and Effort Expenditure" (paper presented at the 1973 Convention of the Association for Institutional Research, Vancouver, Canada, May, 1973). 4Faculty Activity and Outcome Survey (Boulder, Colorado: National Center for Higher Education Management Systems at WICHE, 1972). 5Donald T. Campbell and Julian C. Stanley, Experimental and Quasi—EXperimental Designs for Research (Chicago: Rand McNally, 1963), pp. 5, 20. 6Amitai Etzioni, ed., Readings on Modern Organi- zations (Englewood Cliffs, N.J.: Prentice-Hall, 1969), p. 100. 7Robert E. Hubbard, "An Approach to Instructional Cost Analysis," Basis for Decision, ed. by L. J. Lins (Madison, Wisc.: Dembar Educational Research Service, 1963), p. 110. 8Sullivan, op, cit., p. 3. 111 112 9 . . . . UniverSity of Michigan, "Report on the Pilot Test of the NCHEMS Faculty Activity and Outcome Survey at the University of Michigan" (Ann Arbor: Office of Institutional Research, University of Michigan, January 29, 1973), P. 13. l01bid., pp. 15—16. 11Charles W. Manning, Faculty Activity Analysis: Procedures Manual (Boulder, Colorado: National Center for Higher Education Management Systems at WICHE, April 12, 1973), p. 34 (mimeographed). 12Lynn H. Peltier, Faculty Activity Analysis, Fall 1970 (East Lansing: Office of Institutional Research, Michigan State University, July, 1971), p. 14. 3Manning, op, cit. 14Ibid., p. 36. 15Faculty Activity and Outcome Survey, op, cit. 16Michigan State University, Salary Budget 1972-73 (East Lansing: Michigan State University, 1972). 17Faculty Activity and Outcome Survey, op, cit., 18William A. Mehrens and Irvin J. Lehmann, Measure- ment and Evaluation in Education and Psychology (New York: Holt, Rinehart and Winston, 1973), p. 79. ng. P. Guilford, Fundamental Statistics in Psy- chology and Education (New York: McGraw-Hill, I956), pp. 135-36. 20Ibid., p. 145. ZlIbid., p. 263. 22Leonard A. Marascuilo, Statistical Methods for Behavioral Science Research (New York: McGraw-Hill, 1971), pp. 358-59. 113 23ibid., p. 360. 24George W. Snedecor and William G. Cochran, Statistical Methods (Ames, Iowa: The Iowa State Uni- versity Press, 1967), p. 393. 25Thomas Mason Freeman, "A Multiple Correlation Analysis of the Supplies and Services General Fund Budgets for Selected Academic Departments at Michigan State Uni- versity: 1964-65 and 1965-66" (unpublished Ph.D. disser- tation, Michigan State University, 1967), p. 134. 26Snedecor, op. cit., 413. 27Guilford, 9p. cit., pp. 397-99. 28Ibid., pp. 395-96. 29Sidney Siegel, Nonparametric Statistics for the Behavioral Sciences (New York: McGraw-Hill, 1956i, pp. 229-39. CHAPTER IV ANALYSIS OF RESULTS Objectives of the Research This study was intended to fill, in part, the need for research on costing methodology and faculty activity analysis. The basic objectives of this descrip- tive research consisted of two parts. The first was to examine the allocation of faculty time and salary costs among several activities and to consider the interrelation- ships of several instructional workload factors. The second objective was to compare four costing methodologies used to allocate costs to courses while considering the importance of selected variables of cost. The analysis for the first objective was approached in two ways. A profile of faculty activity by department and rank was developed and the relationships of several workload factors were examined by employing the Pearson product moment correlation. The analysis for the second objective centered on answering four research questions which address the problem of comparing the four costing methodologies. 114 115 Activity Profile Time and Cost Distribution Tables 4 - 9 show the distribution of faculty time by rank for the total college and for each depart- ment. The distribution of the forty-four faculty members among the four ranks for the total college is shown on Table 4. The distribution was five instructors, thirteen assistant professors, nine associate professors, and seventeen professors with an average reported total hours per week for all ranks in the college of 63.5 hours. As discussed in Chapter III, there were no statistically significant differences in the college or by department from a similar survey done in 1970. The greatest dif— ferences between ranks for the college as reflected in this survey were between professors at 66.4 hours and instructors at 59.6 hours, a difference of 6.8 hours. Scheduled teaching accounted for 41.2 per cent of the time; and including other instructional related functions, the total teaching activities amounted to 55.3 per cent of the average week. Research consumed 21.2 per cent of the time, slightly more than half of the scheduled teaching time. Public service functions were only 3.5 per cent, but internal service and administrative and committee work together amounted to 20 per cent of the average week. 1.1.6 NMIawqumvmm No I Nmmm OI OOH m I mo I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I'm/«FOP. mm.~mmwm mw.mmo o.mH p.n . . . . . . . . . . . . . . . meHH>HHo<. mmHeszoo o7... m>HH¢mHmHszo< aceoe .m Hm.mmHmH um.onm w.m v.m . . . . . . . . . . . . . . . . oneHeHeo< moH>mmm Hmmm m.o mp.mva Ho.~v 0.0 0.0 . . . . onHseHemzH mze om<3oH .omeommHo moH>oaxmoH>mmm aaonmmmeoai ammm omeszmonezmopem H o um.quw mm.HoH m.m m.m . . . . . . . . . . . . . . . . . . . . . . mmHeH>Heo¢ moH>mmm oHampi naeoH .o “HOmwN mmOO NIG H00 0 O I C C O O . I O O O O O I O I O C O O U O O O O mogmmm ZOHmEme NIIU mk.omHm us.me m.m m.m . . . oneoHHHm H may moneao omeommHo moH>o<\mmoH>mmm aHeo¢ smoz m>He<1zo oza iHmmmauozom .momammmm ameoe .a »~.omovm ww.oom w.oH 0.6 . . . . . . . . . . .Hzmzioum>mo. 3Heo< oszomo .oza .:om¢mmmm znaaonmao o2< mmmooo v.< oo.H¢oo om.mom m.m v.m . . . . . . . . . . . . . . . . oszH>o< zamoomi onmoHH>HHU< niece mo< ezmo mmm mo< mewoo muse: mxzam 44¢ .11: venz ~n.ooeo9Hso.Honp .oa.asH 3.06 99.9Hmum 31.:nom so..aaH e.go no.on~:m Hm.aH~o can..oH ~.no nn.n~¢oH No.3oon 99¢.aeH o.om nape» Hm.~nm~« mo.cmeH oum.oH a.HH -:.Hmsn ~o.oH: ¢o«.~ v.3 Ns.:aco Ho.~mo Ho~.HH e.» oo.n¢k ¢~.ocH -a.: «.w m a<~o~ so.ms~m n~.man om~.: ~.n mo.mHe~ cI.¢- 53o.» n.~ a-.-n: Hn.onn soo.s a.» ~o.~mm om.««« o~o.n o.« «.u Ho.o-~H ~o.a:~ Hns.HH o.H ~¢.HnsH on.~mH Hm~.n e.~ H~.~na~ on.omH :H~.n n.~ Hm.moH oH.~n use.H o. H.m ca.oaso ~:.n~m ooa.o =.o :o.~m:~ em.-~ H~S.I o.~ ~H.moom ~o.mon cm~.o o.m Ho.~nnH om.oo~ mH~.H n.: o Japan en.ouo: °~.H- sm~.: o.~ oo.onn «3.5» H30. 3. s~.o~mH so.HHH oos.~ o.H no.:o: oo.~m sHm.~ m.H n.o 9:.oos a~.m: nah. m. mH.oa mo.aH noH. H. 4o.nmo Hn.nH H:s.H H.H oo.Hn u~.o ooH. H. ~.o s~.esn: so.sm~ o~s.: H.~ ma.mHo~ I:.:- H's.» n.~ o..mom~ Hw.cmH :-.I a.» no.0». on.~o« .nm.a ~.~ H.o sm.o:a: Ha.ew~ .om.¢ a.» nn.~HaH a~.oHH ~ms.~ ~.H Ho.o:H~ oa.moH Hwa.n m.~ om.~c~ om.o: «In.H o. o 4< aqpoh moan 4¢ «cu m>< pzuo aua mom osa Hzmo «we «on osa wzmo «ma «cu uxa pzmo «ma mbmoo muse: mumoo manor MHmoo manor memoo mane: aommuuoaa commuuoma uHfluom ou Gowusnflnumflo umoo tam wEHBII. «0 magma 117 99.90999 99.0099 0.099 9.99 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 99909 09.9099 99.999 9.9 0.9 . . . . . . . . . . . . mmHeHsHeo¢ mmeeszoo 929 m>99¢999HzH=o< 99909 9 09.0999 99.999 9.9 9.9 . . . . . . . . . . . . . . . . . . . . . . . . oneHH§HmHZHZQ< #Iw 2.9.992.” 99.99m 0.09 9.9 . . . . . . . . . . . . . . . . . . . mmHHH>HPU< mogmmw .H<7.mm9.zH .2909 o vm.mHHH mm.OQH v.m ~.N . . . . . . . . . . . . . . . . . . . . . . momoomm 02¢ mfimommm mUH>mmm m.o 99.999 99.99 990 9.0 H ” zomeoemem2m 9:9 999309.999099Ho mom>p¢mmeommm 9mmm omezmHmo-9299999 9.9 90.999 99.909 9.9 9.9 . . . . . . . . . mmHHH>HHo< moH>mmm oHamsm 99909 o 00.0 OOIO OIO OIO I I I I I I I I I I I I I I I I I I I I I I I I I mogmmw ZOHmzmem NIU 90.999 99.99 9.9 9.0 . onHDHHemzH 9:9 9999990 omeummHo 909>a<\mmoH>mmm 99209mwmmomm 9999290 H.o 99.9999 99.999 9.99 9.9 H H H . . . . mmH9H>Heo< 9903 m>Hemo 9H8< OZEU§H \H9¢9¢w 99909 no 9993 999 >9H>Heo< 99909 mu9 9290 999 90999>9 99900 9990: 99299 99< ouz 99.99999 99.9999 999.999 9.99 99.9999 99.9999 999.999 9.99 99.99999 99.9999 999.999 9.99 99.9 99.9 999.9 9.9 99999 99.9999 99.999 999.99 9.9 99.999 99.999 999.99 9.9 99.999 99.999 999.9 9.9 99.9 99.9 999.9 9.9 9 99999 99.999 99.999 999.9 9.9 99.999 99.999 999.9 9.9 99.999 99.999 999.9 9.9 99.9 99.9 999.9 9.9 9.9 99.9999 99.999 999.9 9.9 99.999 99.999 999.9 9.9 99.9 99.9 999.9 9.9 99.9 99.9 999.9 9.9 9.9 99.9999 99.999 999.99 9.9 99.999 99.999 999.99 9.9 99.999 99.999 999.9 9.9 99.9 99.9 999.9 9.9 9 99999 99.999 99.999 999.9 9.9 99.999 99.999 999.9 9.9 99.999 99.99 999.9 9.9 99.9 99.9 999.9 9.9 9.9 99.99 99.99 999. 9. 99.99 99.99 999.9 9.9 99.9 99.9 999. 9. 99.9 99.9 999.9 9.9 9.9 99.999 99.999 999.9 9.9 99.999 99.999 999.9 9.9 99.999 99.999 999.9 9.9 99.9 99.9 999.9 9.9 9.9 99.999 99.999 999.9 9.9 99.99 99.99 999.9 9.9 99.99 99.99 999. 9. 99.9 99.9 999.9 9.9 9 99999 99.9 99.9 999.9 9.9 99.9 99.9 999.9 9.9 99.9 99.9 999.9 9.9 99.9 99.9 999.9 9.9 9.9 99.999 99.999 999.9 9.9 99.99 99.99 999.9 9.9 99.99 99.99 999. 9. 99.9 99.9 999.9 9.9 9.9 99.9999 99.999 999.99 9.9 99.999 99.999 999.9 9.9 99.9999 99.999 999.99 9.99 99.9 99.9 999.9 9.9 9 99999 99.999 99.999 999.9 9.9 99.999 99.999 999.9 9.9 99.9999 99.999 999.9 9.9 99.9 99.9 999.9 9.9 9.9 99.999 99.999 999.9 9.9 99.999 99.999 999.9 9.9 99.9999 99.999 999.9 9.9 99.9 99.9 999.9 9.9 9.9 99.9 99.9 999.9 9.9 99.9 99.9 999.9 9.9 99.9 99.9 999.9 9.9 99.9 99.9 999.9 9.9 9.9 99.9999 99.9999 999.99 9.99 99.9999 99.9999 999.99 9.9 99.9999 99.9999 999.99 9.99 99.9 99.9 999.9 9.9 9 9999» 99.999 99.999 999.9 9.9 99.9 99.9 999.9 9.9 99.999 99.999 999.9 9.9 99.9 99.9 999.9 9.9 9.9 99.999 99.999 999.9 9.9 99.9 99.9 999.9 9.9 99.999 99.999 999.9 9.9 99.9 99.9 999.9 9.9 9.9 99.9999 99.999 999.99 9.9 99.9 99.9 999.9 9.9 99.999 99.999 999.9 9.9 99.9 99.9 999.9 9.9 9.9 99.9999 99.9999 999.99 9.99 99.9999 99.9999 999.99 9.99 99.9999 99.9999 999.99 9.99 99.9 99.9 999.9 9.9 9.9 99999 9999 99999 99 999 99999 9999 99999 99 999 99999 9999 99999 99 999 99999 9999 99999 99 999 «99 999 9299 «we «99 999 9299 999 «99 999 9299 «mi «99 999 9299 999 99999 9999: 99999 9999: 99999 9999: 99999 9999: «99999999 999999999 999999999 999999999 92999H99< 9999999929 9uz Huz muz ouz 9 9292999999 H ucmsuummmt 90m xdmu an wwwufl>wpom 09 coqusnwuumfip umoo 0:9 mEHBII.m mqmda W -__9.-.. ud—u-I—L _ ll8 mMImmowv mwumovv OIOOH Voom I I I I I I I I I I I I I I a I o I I I I I I I I I I I I I I I I J<§ 99.9999 99.999 9.99 9.9 H . . . . . . . . . . . . 90999>9900 009992200 929 m>99<99092929< 9<909 99.9999 999.999 9.9 9.9 . H H H H H H H . . . . . . . . . . . . . . . . 20029909995 099,992.80 00.VCHM omoOHm 5.0 09.“. o n o o o o o o u o o o o o o o o o o 9 WMHHDQ m>HH§HmHZHZQ¢ 99.9999 99.909 9.9 9.9 . . . . . . . . . . . . . . . . 99999>990< 009>mmm 0<299929 90909 Neohme ObommH N‘Im N'N I I I I I I I I I I I I I I I I I I I I I I 8m8mm Qz¢ mmemm mUH>mmm 90.999 99.99 9.9 9.9 . . 20990999929 029 099209 99909999 909>9<\009>mmm 992090909099 9999200 99.0999 90.999 9.9 9.9 . . . . . . . . . . . . . . . . . . . . . . . .009>mmm 90920990-9209090 09.9099 99.099 9.9 9.9 . . . . . . . . . . . . . . . . . . . . 9099H>990< 909>2mm 099099 99909 00.0 00.0 0.0 0.0 . . . . . . . . . . . . . . . . . . . . . . . . . . . 909>909 onmzmem 09.9099 99.099 9.9 9.9 . 20990999929 9:9 0999990 90900999 909>9<\9009>909 042090909090 0990290 99.9999 99.999 9.99 9.9 . . . . . . . 00999>9909 2903 m>999090 92¢ .99299990009 .20999909 90909 99.9999 99.999 9.0 9.9 . . . . . . . . . . . .92029099>09 9<2099909009 92¢ 99299900209 0990200 ONIhmON NNIIwON N00 KtIm H I I I I I I I I I I I I I I I I I I I I I I I I I I thomm UHmHummw 00.0 00.0 0.0 0.0 . . . . . . . . . . . . . . . . . . . . .9909200 20990900 92< 099999929 bMIOhth ¢OIthN OIOm wImm I I I I I I I I I I I I I I I I I I I I I I 3HIHH>H8< OZEUEH J99 920 20990909 2009099990 920 099900 00.9999 99.999 0.9 9.9 . . . . . . . . . . . . . . . . . . . . . . 02999>9< 2920099 09209909 OMIOMNN MOIMNN . .HIm OIm I I I H H I I I I I I I I I I I I I I I I I I I IOZEUE aggmg 99.90009 99.0909 9.99 9.99 . . . . . . . . . . . . . . . . . . . . .02920909 900990209 wMHmm<4¢w AHH>HHU¢ .2909. 00§m>< 929.0 90.. 9099959. 99900 9900: 00:43 09uz 99.99999 99.9999 999.999 9.99 99.9999 99.9999 999.999 9.99 99.99999 99.9999 999.999 9.99 99.9999 99.9999 999.999 9.99 99.9999 99.999 999.99 9.9 99.9 99.9 999.9 9.9 99.9999 99.999 999.99 9.99 99.99 99.99 999. 9. 99.999 99.999 999.9 9.9 99.9 99.9 999.9 9.9 99.9999 99.999 999.9 9.9 99.99 99.99 999. 9. 99.9999 99.999 999.99 9.9 99.9 99.9 999.9 9.9 99.9999 99.999 999.9 9.9 99.9 99.9 999.9 9.9 99.9999 99.999 999.9 9.9 99.999 99.999 999.9 9.9 99.9999 99.999 999.99 9.9 99.999 99.999 999.9 9.9 99.999 99.999 999.9 9.9 99.9 99.9 999.9 9.9 99.999 99.999 999.9 9.9 99.999 99.999 999.9 9.9 99.99 99.99 999. 9. 99.9 99.9 999.9 9.9 99.999 99.999 999.9 9.9 99.99 99.99 999. 9. 99.999 99.999 999.9 9.9 99.999 99.999 999.9 9.9 99.999 99.999 999.9 9.9 99.999 99.99 999.9 9.9 99.9999 99.999 999.9 9.9 99.999 99.999 999.9 9.9 99.9999 99.999 999.99 9.9 99.9 99.9 999.9 9.9 99.9 99.9 099.9 9.9 99.9 99.9 999.9 9.9 99.9 99.9 999.9 9.9 99.9 99.9 999.9 9.9 99.9999 99.999 999.9 9.9 99.999 99.999 999.9 9.9 99.9999 99.999 999.99 9.9 99.9 99.9 999.9 9.9 99.9999 99.9999 999.99 9.99 99.999 99.999 999.9 9.9 99.9999 99.999 999.99 9.9 99.999 99.999 999.99 9.9 99.9999 99.999 999.9 9.9 99.999 99.999 999.9 9.9 99.9999 99.999 999.9 9.9 99.999 99.999 999.9 9.9 99.9999 99.999 999.99 9.9 99.999 99.999 999.9 9.9 99.999 99.999 999.9 9.9 99.999 99.999 999.9 9.9 99.9 99.9 999.9 9.9 99.9 99.9 999.9 9.9 99.9 99.9 999.9 9.9 99.9 99.9 999.9 9.9 99.9999 99.9999 999.99 9.99 99.9999 99.9999 999.99 9.99 99.9999 99.9999 999.99 9.99 99.9999 99.9999 999.99 9.99 99.999 99.999 999.9 9.9 99.999 99.999 999.99 9.9 99.999 99.99 999.9 9.9 99.9 99.9 999.9 9.9 99.999 99.999 999.9 9.9 99.9 99.9 999.9 9.9 99.999 99.999 999.9 9.9 99.9 99.9 999.9 9.9 99.9999 99.999 999.9 9.9 99.9 99.9 999.9 9.9 99.9999 99.999 999.9 9.9 99.99 99.99 999. 9. 99.9999 99.9999 999.99 9.99 99.9999 99.9999 999.99 9.99 99.9999 99.9999 999.99 9.99 99.9999 99.9999 999.99 9.99 99909 9099 99909 90 099 99909 9099 99909 90 099 99909 9099 99909 90 099 99909 9099 99909 90 099 909 999 9290 «09 909 999 9200 909 909 999 9200 999 909 099 9200 900 99900 99002 99900 99002 99900 9900: 99990 9900: 909999099 909999099 099900999 909999099 929999999 0090099929 muz Huz vuz Nuz HN . I <¢9000 00 cowusnfiuumflw 0900 000 meqeul.9 mqm9909 999992200 929 9>999999929299 99909 9 99.9999 99.999 9.9 9.9 . . . . . . . . . . . . . . . . . . . . . . . . . 2099999099999 999992200 9.9 99.9999. 99.999 9.9 9H9 . . . . . . . . . . . . . . . . . . . . . . . ._. . 999999 9>999999929299 9.9 99.0999 99.909 9H9 9.9 . . . . . . . . . . . . . . . . . . . . 99999>9909 909>999 99299929 99909 9 99.9999 99.999 9.9 0.9 . . . . . . . . . . . . . . . . . . . . . . . 9990099 929 9990999 909>999 9.9 99.999 99.99 9.0 9.0 . . . 20999999929 929 999309 99909999 909>09\909>999 992099999099 9999290 9.9 NOINOOM BOvaN wIV MIm I I I I I I I I I I I a I I I I I I I I I I I I IMUHEMW QmH‘zmHKOI/Hufigw HID mOIMHmm QNIOFN NIV DIM. I I I I I I I I I I I I I I I I I I I I I 8HHH>HHMU< symmw 058nm \Hmmm ZOHWZMme NIU 99.0999 09.999 9.9 9.9 . . 20999999929 929 9099990 99909999 909>99\9909>999 992099999099 9999290 9.0 99.99999 99.9999 9.99 9.99 . . . . . . . . 99999>9909 2903 9>999990 929 .99299990209 .20999999 99909 9 90.9999 90 999 9 99 9 9 . . . . . . . . . . . . .92929099>99 992099999099 929 99299990209 9999290 9.9 VOIthVH HMHOMHH OImH OIMfi I I I I I I I I I I I I I I I I I I I I I I I I I I I I guaomm UHamHummm Num RIO-mac omnmm 090 000 I I I I I I I I I I I I I I I I I I I I I Immmlfifio amfimmm 02¢. mH‘BHPmZH HIm 0N OFOOM GK. ONQN momfl QIHM o I I I u I 0 I o I I I I I I I I o o I I 0 I I WMHHH>H8< QZHIUEH \HGHHQH < 99H9999 90H999 9H9 9H9 . . . . . . . . . . . . . .92929099999 929 20999999 2999099990 929 999900 9.9 99.0999 99 999 9 9 9 9 . . . . . . . . . . . . . . . . . . . . . . . . 02999>99 2990099 09299909 9.0 mm boom OVI¢mv VI“ HIm I I I I I I I I I I I I I I I I I I I I I I I I I I I ozgugfi QBBEUE NI d 99.90999 99.9999 9.99 9.99 . . . . . . . . . . . . . . . . . . . . . . . . . . . .02920999 999999209 9.9 99999999 999999 99909 90 2993 999 >99>9909 99909 90999>9 9290 999 9099999 99900 99902 92299 999 99u2 00.09909 99.0099 000.009 0.09 00.00909 09.9009 000.009 0.99 99.9990 00.0000 000.009 9.09 00.0 00.0 000.0 0.0 99909 99.9999 99.909 999.9 0.9 90.0009 00.090 999.0 9.9 09.9009 99.999 999.99 0.99 00.0 00.0 000.0 0.0 9 90909 90.9999 00.990 099.0 9.0 90.909 09.009 990.0 0.9 00.9909 99.999 999.99 0.0 00.0 00.0 000.0 0.0 9.9 09.9099 99.990 909.0 0.0 09.090 09.099 099.0 0.0 09.090 90.999 990.0 0.0 00.0 00.0 000.0 0.0 9.9 90.9090 09.099 909.9 9.9 00.090 99.999 909.9 0.9 90.9099 99.999 909.09 0.9 00.0 00.0 000.0 0.0 0 9990» 99.9909 99.999 009.0 0.9 00.0 00.0 000.0 0.0 99.999 00.909 009.9 0.9 00.0 00.0 000.0 0.0 0.0 90.990 00.99 000. 9. 00.0 00.0 000.0 0.0 99.99 00.90 009. 9. 00.0 00.0 000.0 0.0 9.0 99.0909 90.999 000.0 9.0 00.090 99.999 909.9 0.9 99.090 09.990 099.9 9.9 00.0 00.0 000.0 0.0 9.0 09.0999 90.900 990.9 9.0 90.999 90.999 999.0 0.9 99.99 00.90 009. 9. 00.0 00.0 000.0 0.0 0 99909 99.009 00.90 009. 0. 00.0 00.0 000.0 0.0 00.0 00.0 000.0 0.0 00.0 00.0 000.0 0.0 9.0 00.9009 99.000 999.0 9.0 90.999 90.999 099.0 9.9 99.90 00.90 009. 9. 00.0 00.0 000.0 0.0 9.0 99.00999 99.0909 090.00 0.99 09.0999 00.0999 009.90 0.99 00.0999 99.0009 099.99 9.09 00.0 00.0 000.0 0.0 9 9990» 99.9099 09.000 009.99 9.0 99.0909 90.099 999.09 9.9 99.090 09.990 099.9 9.9 00.0 00.0 000.0 0.0 0.9 90.9090 99.9909 000.99 9.99 99.9990 09.9909 999.09 0.09 09.0999 90.900 009.99 0.09 00.0 00.0 000.0 0.0 9.9 00.099 09.90 009.9 9. 00.0 00.0 000.0 0.0 00.90 90.99 090. 0. 00.0 00.0 000.0 0.0 9.9 00.09909 00.9000 909.90 9.90 99.9990 90.9999 990.00 9.00 09.9990 90.0009 099.00 9.99 00.0 00.0 000.0 0.0 0 90909 99.9999 90.999 999.0 0.9 99.999 99.999 999.0 9.9 00.909 99.99 990.9 0.9 00.0 00.0 000.0 0.0 0.0 00.0999 90.009 900.9 0.9 09.9099 99.990 009.9 0.0 90.090 99.009 099.0 9.9 00.0 00.0 000.0 0.0 0.9 09.9000 90.000 000.9 9.0 00.9009 90.999 900.99 9.9 00.990 99.999 099.9 0.0 00.0 00.0 000.0 0.0 9.0 90.00999 90.0099 999.00 0.09 00.9990 99.9999 900.99 9.99 99.0999 09.0009 900.00 9.09 00.0 00.0 000.0 0.0 9.0 99909 9099 90909 90 000 90909 9090 90909 90 0>0 90909 9099 90909 90 000 90909 9099 99909 90 000 909 000 9290 999 909 000 9290 999 909 900 9290 «me 909 000 9290 999 99900 9990: 99900 99909 99900 9009: 99900 9990: 909999099 909999099 999900999 909909099 920999990 9090999929 9u2 9u2 9u2 0u2 9 9209990990 m unmeuummwv 90m 9999 an 90909>9009 ou 00909n990990 0900 0:9 0E99|I.0 94998 mm.mmoom 102(3 bflIvmov OIOOH vIno I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I \H<0H\8 99.9999 99H999 9H99 9.9 . . . . . . . . . . . . . . 99999>9909 999992200 929 9>990999929299 90909 9 09.9% 9.8.. N... m... .. H 9 H H H H 9 9 H H H H H H H H . . H . . . . . 222.550.9508 3 00.9999 99.M99 0.9 9.9 . . . . . . . . . . . . . . . H H . . . . . . . 999999 9>999999929299 9.9 99.9909 99.9mm 9.9 9.9 . . . . . . . . . . . . . . . . . . m9m9m>9909 909>999.99299929 99909 9 00.0 00.0 9 9.9 9.9 . . . 9990099 929 9990999 909>999 9.9 99.0999 99.999 0.0 0.0 . . . 20m9¢9m9m2m 92m 999309.999w9m9m m0m>99x909>999 992099999099 9999290 9.9 NV I mmm mv . vv N I w w I N I C I I C O I I I . I O I I I I I O I I I . IMUHgmm QmHzm HMOIHEgH‘m H .0 OOIO OOIO OIH. “In I I I I I I I I I I I I I I I I I I H I H ImeIHH.>H0HIU< agmmm UH‘HmDhm IH¢HOIHI U 99.99 . 0.0 0.0 . . . . . . . 909>999 209929929 9.0 -. 9 99.99 0 9 9 0 . 20990999929 929 9999900 99909999 909>99\9909>999 992099999099 9999290 9.0 99.9999 99.999 9H99 9H99 . . . . . . . . 99999>9909 2903 9>999990 929 .99299990209 .20909999 99909 9 ww.mmww wm.mmm 9.99 9.9 H H H H H H H H H H H H H9mmmmm9m>w9 992099999099 929 99299990209 9999290 9.9 00.0 00.0 9.9 9.9 . . . . . . . . . . . . 99090099 09990999 9.9 . 0. 0 o 0 0 . . . . . . . . . . . . . . . . . . . .9999290 20999999 929 990999929 9.9 NmINIQOMN VOIOCON WIOO wIOV I I I I I I I I I I I I I I I I I I I I I I I I $HHH>HIHU¢ OZEUEH \HJNIHOH‘ G9 mm.wmm9 Mw.mm9 9.9 9mm H H . . . . . . . . . . . .92929099>99 929 20999999 2090099900 929 999000 9.0 - I H I H KI m m 0 I I I I I I I I I I I I I I I I I I I I I I OZngQ< Z§8~mm UHEQ9 9290 999 9009900 >99>9909 99900 99902 92299 999 9u2 00.0099 00.0090 000.009 0.99 00.00999 99.9909 000.009 0.09 00.0000 00.0000 000.009 0.09 90.90099 09.9990 000.009 0.99 00909 99.9990 99.9090 099.99 0.90 90.090 99.900 009.0 0.9 90.0909 00.909 909.99 0.0 90.909 00.900 090.9 9.0 9 90909 00.999 90.999 999.0 0.9 99.900 99.999 990.0 0.9 00.999 09.000 900.0 0.9 09.900 00.999 000.0 9.9 0.9 09.9999 09.9999 900.00 0.00 99.900 99.990 990.0 9.9 99.009 00.009 090.0 0.9 00.909 99.90 000.9 0.9 9.0 99.090 99.090 990.0 0.9 09.909 09.000 900.0 9.0 09.999 99.009 999.9 0.9 09.009 99.090 900.9 9.0 0 9090» 09.090 99.090 000.0 0.9 00.90 99.00 090. 9. 99.009 00.009 900.0 9.9 00.900 99.90 000.9 0.9 0.0 00.0 00.0 000.0 0.0 00.0 00.0 000.0 0.0 00.0 00.0 000.0 0.0 00.0 00.0 000.0 0.0 9.0 00.0 00.0 000.0 0.0 99.909 99.900 999.0 0.0 90.000 00.009 990.0 9.9 90.909 00.009 090.9 990 9.0 00.0 00.0 000.0 0.0 00.0 00.0 000.0 0.0 00.90 09.00 900.9 0. 00.000 99.90 099.9 0.9 0 90000 00.0 00.0 000.0 0.0 00.0 00.0 000.0 0.0 00.0 00.0 000.0 0.0 00.0 00.0 000.0 0.0 9.0 00.0 00.0 000.0 0.0 00.0 00.0 000.0 0.0 00.90 90.00 900.9 0. 00.090 90.90 099.9 0.9 9.0 99.999 99.099 900.0 0.0 90.9000 90.0009 090.09 9.9 99.0009 09.0909 090.00 9.99 90.9099 90.909 009.09 9.0 0 90009 99.09 99.09 099.9 0.9 90.0009 09.9909 990.09 9.9 99.9009 90.009 090.09 0.9 00.0099 09.999 009.99 9.9 0.0 00.999 90.999 999.0 0.0 00.0009 99.999 000.0 0.0 99.909 09.000 900.99 0.0 09.900 99.90 009.9 0.9 9.0 00.0 00.0 000.0 0.0 00.0 00.0 000.0 0.0 00.0 00.0 000.0 0.0 00.0 00.0 000.0 0.0 9.9 00.9999 00.9990 000.90 0.99 90.0999 09.9090 909.00 9.9 09.9900 99.0009 999.09 0.00 99.0900 00.0909 909.09 0.00 0 90909 09.09 99.09 090.9 0.9 90.909 90.990 009.0 0.0 09.090 00.900 909.9 0.0 09.990 00.090 000.9 0.0 0.0 90.999 90.999 999.0 0.9 99.900 90.990 990.0 0.9 99.99 90.00 990. 9. 99.900 00.090 090.0 0.0 0.0 00.0 00.0 000.0 0.0 09.990 90.090 000.9 0.0 90.000 90.909 900.9 0.9 00.9900 09.990 000.09 0.09 0.0 90.9909 99.9909 900.00 0.09 99.9090 00.0900 099.00 9.9 90.0090 90.0000 909.90 9.90 09.9999 00.9099 000.90 9.09 9.0 90900 0090 90909 90 000 90909 9090 90909 90 000 90909 9000 00909 00 000 00909 0090 00909 00 000 909 000 9200 090 900 000 9290 990 009 000 9290 090 900 0>0 9290 000 09000 09002 09000 0000: 09000 90002 09000 0900: 909090090 909009090 090900000 000000000 920909000 0090099029 Huz muz Nuz muz 9 hzwzhm9909 ou 009050990990 0900 099 08991:.9 mqmdfi 121 ~0.m~no« mo.oo«o «n.no:m n~.omo oN.:m:: o«.o::a :~.omm No.00 mo.m:s s«.mcm 99.9 Na.o¢m ::.om- :o.hmoa no.9mou 99.9 mo.mn¢h mo.omo mo.m:h on.ooc noowmhm thOb m VMOMthm mv-oomm COOOH Oovm O Q Q I I O O O O O O O O I I O O O O O O O O C O O I I U I O O C I \HcHB 3.9.30 55.03. m.: v.0 . . . . . . . . . . . . . . wwwHH>HHU<. mmaeszoo 02¢ m>Hu§emHszo< .398. m OH.NBGH mb.HwN H.m m.N . . . . . . . . . . . . . . . . ZOHHHH¢memHszo< H.m vm.bOHN NH.HOM h.m H.m . . . . . . . . . . . . . . . . . . . . mMHHH>HHU< MUH>mmm Jmmm m.o bw.oo hN.VH N.n H.n . . . ZOHHDHHHmZH MIR Qm<3OH QMHUMKHO mUH>Q<\mUH>mmm Ammm QMHZMHNCIHZMQDPw H.D Hm.how 0N.mNH 0.N v.H . . . . . . . . . . . . . . . . . . . . . mMHHH>HHU< mUH>mmm UHJmDm Aommw 4Hao< amoz m>Hpma .qaonwmmmOmm az< mammchOIUm gHPU< OZHIUo< zm9H>Heu< AiHOH m0¢mm>< Hsz «mm wU¢mw>c meOO wmnc: mxz0h mm.mmm :oN.o ~.m MN.nm«« «H.m¢m «~:.o« n. m o~.can mn.cmn mn«.o 9.: 99.9 nn.n nnn.o o.o n.m oN.ooM moo.m N.» h0.Nmo :M.o:: nm9.~ c. n n:.nHNH N~.mno mm:.ma 9.9 99.9 99.9 999.9 9.9 N.m nn.n 999.9 9.9 nn.n no.9 nnn.o 9. o no.9 no.9 nnn.o o.o no.9 99.9 999.9 o.o «.m no.0:oN :Nn.n: m.o~ on.mm«9 Mm.obn: oNn.«w N.:n nn.nnom n:.m«mN «m~.:m N.mn 99.9 99.9 999.9 9.9 « nakOk No.mmN mom.: n.n «H.mmm om.~mN onN.m m.N 00.0NN 9:.mua moo.N m.« 99.9 99.9 nnn.o o.o ¢.< no.0:N soo.: m.~ wa.mno om.o«: nnn.s m.m n:.o:r NN.oN« o:m.: M.N nn.n nn.n nnn.o o.o n.< ma.ocu mNh.N h.« mo.m~n No.09m 099.0 o.o os.:oN wM.Nn« Non.» o.« 99.9 nn.n nnn.o o.o N.< nn.nan onw.an ”.ma aw.«mmm sm.m~o~ omn.Nm 9.mN o~.9mH: mn.mm9N nNo.mm n.~N nn.n nn.n nnn.o o.o «.4 moan 4< nqbo» nova adhob mo o>< nahOh moan nqhop mo o>< 44POF moan 4<»OF mo o>< mom o>q hzwo «ma «on o>< hzwo «ma non o>¢ hzmo owe you o>< hzmo awn hwoc wmnox wkmoo manor mwmoo manor mhmoo manor wommwuoma mommmuoma mhdHooqu mommwuoaa FqumHMMc neponyhmzH muz Nuz an Ouz m h2mtbuqamo m unmeuummmv How xcmu Mn moflufl>fluom og coausafluumflw umoo flaw mEHBII.m mqm mo manumfla omumfl>oubnwc s as MOM vma mama 0mm. .tniillltltillllllilillli :om-»moo kzou-»moo wa»-»woo umm. oqmo oxH» roam» zomnamma «H k. on m. .H . osowocom. an owuo-hmoo fig-nsmo. kawwwnom. mooomKNo. oocmowmn. mwmcouom. . ca :omcsmoo . . . . . nonsnom. madam... - oo sausages H co¢m~aak ocmmmnru omfioommm . monoNnuo. s“ pzoo pm soonesoa.u om-«.n¢. owsoo¢::. oon~n~m~. womNN. o“ uxahusmoo assesses.“ dowo.:m.. amkmm.«o Nat . . ma pmmq aqua ccsooooo.l n...o.odnu “memwsm. .« wap xoaw. accesses gasses...“ mu :0. awe. «so: amxbo a: hoahzoo Iom muse: sumo u~wm mmcno >ouuaawuu Na 3 3 m ., .- «oémoo oomaammm. .Nofimmma. «sucksmn. omnksn~.. ookom.o.. ommmNcns.- MN mwm-.woo Nmooonam. onwooknw. ommo~o~k. «sucmqfiw. «owooaoo. oNdnwsma. ha pzoo-pmoo noonmnm. mmwooaso. oawommcs. smsmnonn. whence... kosocomd.- oa qu»-pmou ~wn«.«n:. kmNNdoom. .«m30msm. “condofifl. amonamm.. mNoMHma«.- ma hmmq oqmo Nmmoomsa. mmflookma. mwkmkmcm. mnamsa~fl. omomdosm. mmmo°~.«.- «a eta» roams macnosdo. wmmomnom. snoosomw. om~daaoo. nmmmo.a~. d~¢-n-.- ma zoa amen o~:oo~m~. skmwomon. kmmamkkm. nono.co«. o:~ssfinm. o.osmmod.- « «no: worse ooaacsoo.d konosm~. .mnomouo. nanomoMO. onsssdfis. nmommnad - N x: soahzoo accesses.“ Hwncswvo. m..oos~a.- nkao.mmo. mononnm«.- «a sosocseo.~ .kmNOAOH. Naomcmko. :nssoouo.- an Ion osoaaooo.a ~m-ow:s.- «mawkwao. m mane: ammo sacoocsa.« 0.5m.oos.- o mNHw mmqno Doocoosa.a k >ozwsowau 4u>mu «so: nape. onpomw oz ouyu omXHu uoszamoxm xzqy o m : n N « ammonoo«.- ~0~¢-=«.- mmnnmm~..- moo.~.wo.- o~mmd~oo. .mkaomom. on ammo-hmoo «stmmma.- «seam50a.- okomowo~.- “ammosws.- mm:.sdma. amounted. on zom-pmoo o.o:on:~.- mmmosoko.- oaoms~cm.- ~oo~odmo.- omoanomo. soaokawd. ks ono-»mou 004~onm~.- .aaodmm=.- m.:«:m¢m.- commokn=.- mammomoo. :kmamomn. we mthukmoo nomdndm~.- moonoows. w.mdmomc.- owNQNamo.- ssflok:o~.- smcmonda. m" pmmq came n¢«¢~m~«.- nooaa::a. ssososss.u smdmmsoa.- mmHoN~m«.- nnsmmond.- :« mzH» :oqmb mwomn=m«.- nomacsna. .Nmmomo~.- Nacoawmo. mmo:omm«.- Homeosos.- ms zoq amen moocomso.- onanmnsm. somca¢m~.- nm~ao-~.- .:~o:ooo. o.m:-co.- ~H «so: amzho masoncdd.u o3~o4~mo.- sammfiaam.o nomonsss.- sumam«~«.- mmooohsa.- «a «I poahzoo oooomfinm.- o.mmo”~«.- mesmNNoa.- Nasce~mo. ~koc.«oo. mssokssa. on row «mumsosd. mnsknpoa. m3mnsamfi.- «Na.o:oo.- HQNONJJO. oNKNmmoH. m muse: ammo nao~suomu- somckkfia.- son:momfl.- somaanoo. o~.o~.oo. Hm~o¢~m«. o MNHm mmm4 accesses.“ “mmoH~.H.- :«adammfi.- :ma~«:m°.- ”Noowonn. m «no: Jakob assesses.” mksmngw. endomNKfi. oa«~mom~.- . onpuwm oz assesses.“ .kdoeoon. somQNood.u n ouau ouXHt escapees.“ mssoamdm. m mozmHmmaxm assesses.“ fl gzaa .oz ya) «woaumfluwuomumno nmonuoz amcowuosuumcH Mo maoflbmHmuuoo mHmEHmII.oa mamfie 126 Unabbreviated Listing of Variables Abbreviated (as in Table 10) l. 2. 12. l3. 14. 15. l6. 17. 18. 19. Rank Experience Fixed Cred No Section Total Hour Level Frequency Class Size Cred Hours SCH Contact Hr Other Hour Prep Adm Teach Time Grad Asst Cost—Time Cost—Cont Cost—SCH Cost-Cred Related to Table 10 Faculty Years of teaching experience of the faculty member Fixed credit courses as opposed to variable creditS' Number of sections and/or courses taught by the faculty member Total work week expressed in hours Course level of instruction Number of times the professor has taught the course Number of students enrolled in the course Course credit hours Student credit hours Formal scheduled contact hours or class hours Informal contact outside the classroom Preparation and administration for the course Total course time - time devoted to a course Whether or not a graduate assistant is assigned to the course Course costs allocated on actual time devoted to the course Course costs allocated on formal contact hours Course costs allocated on student credit hours Course costs allocated on course credit hours 127 There did appear to be some flexibility in the use of time. There was a slightly positive relationship between total time per course and class size (.215) as well as student credit hours (.239). Also, as enrollment increased the time spent in preparation and adminis- tration increased slightly (.331). There was also a greater possibility for graduate assistant help (.500). Therefore as the class size increased there may have been a slight increase in total time devoted to that class, but the total workweek was not likely to change. It also appears that if a faculty member taught more sections or courses he may have spent less time per course section. That is, the higher the number of sections taught, the fewer hours spent in formal con- tact (-.344), other contact (—.258), preparation and administration (-.264), and total teaching time (-.400) per section. This is meaningful in analyzing workloads. As there was an increase in the number of times the professor taught the course (frequency), the total number of teaching hours related to that course decreased only slightly (—.224) and the total workweek changed little (-.123). As expected, preparation and adminis- trative time devoted to the course decreased slightly (-.200) as did other contact hours (-.l43) when fre- quency was increased. However, formal contact hours also decreased with an increase in frequency (-.153), a phenomena difficult to explain. 128 As the level of instruction increased there was a decrease in class size (—.561) and SCH (-.532). Yet there was little correlation between level and total teaching hours (—.l30), contact hours (-.110), and other contact hours (-.006). That is, as the level of instruction increased the size of the classes got smaller while the total teaching hours worked per course and the contact hours per course may have changed very little. It is interesting that higher level courses did not have a higher amount of other contact hours reported since at the graduate level a premium is placed on informal contact with students. Although total teaching hours per course increased when formal contact hours increased (.803) the total workweek changed little (.144). In addition, as formal contact hours increased there was a slight increase in other contact hours (.284) as well as preparation and administrative time (.400). Therefore, an increase in contact hours may increase the teaching activities but it will not necessarily increase the total workweek. Cost Relationships The interrelationships among the four costing methodologies are noteworthy. Table 10 shows that costs developed on formal contact hours correlated with costs developed on total course time at a ratio of .934. Costs based on time correlated with costs based on 129 student credit hours at .734 and with costs based on credit hours at .628. Costs based on contact hours and student credit hours correlated at .701, between contact hours and credit hours at .563, and between student credit hours and credits at .654. Costs developed on total course time reported by the faculty correlated higher than any other method with each of the others except credit hours, which correlated at .654 with student credit hours compared to .628 with time. However, because costs develOped on formal con- tact hours correlated so very highly with the total course time as reported by the faculty (.934, see Table 10), one should weigh quite closely the advan~ tages of faculty activity survey over simply costing on formal contact hours. It was apparent from the information in Table 11 that the methodology selected for allocating costs to courses would to some extent determine that cost. Allocating costs based on total course time and on formal contact hours paralleled each other quite closely. Although costing on credit hours, to a limited extent, resembled total course time and contact hours, costing on student credit hours varied a great deal from the other three patterns. This might be expected since student credit hours are really a subset of course credit hours. That is, the other three methods focus 130 on the course, whereas this method considers both the class size and the course credit hours. In interpreting Table 11, it is imperative that caution be exercised. The base unit was the individual course, thus to compare between levels it would be preferable to reduce the average course cost to a common unit. TABLE ll.--Course cost patterns by level of instruction for four costing methodologies (in dollars) Average Course Allocation Methodology Section Cost by "N Course Level Student Course nggze ngfiigt Credit Credit Hour Hour Lower Undergraduate 19 1365 1388 1849 1194 Upper Undergraduate 71 731 730 696 747 Graduate 34 772 750 562 841 Rather than reporting per course costs by level, in Table 12 the course costs were reduced to per hour of total course time, per contact hour, per student credit hour, and per course credit hour. For each allocation methodology the unit basis for that allo- cation could not be used as the unit of measurement in the per unit cost since there would be no differential in the cost by level. This accounts for the blank spaces in Table 12. 131 EOHM ©m#mummmm on uoc UHSOSm mHQMp mflflu .M#MU can mo 0 .m3OHHom SOHSB scammsomflp msn usumn map mo mmsmomm an W. moa III mwa mma mom 9m evadomuw . mom III mom mom now an wamscwnmnmccn momma o H mow III omm own man ma mpmscmnmumpan meson usom uflpmuu mmusoc Hem mum ma om III 9H 9H 9m mundpmuc N.N HH Ha III Ha HH an mumnpmnmumUCD HmmmD o H m m III m m ma muMSUmnmquGD HmSOA .Hm pflpmnu pcm©3pm Mom 5. 99m Nov mom III mom vm mpmsomuc 4. saw omm mom sun mam as mumzwmumsmeqo “was: o.H owe omm woe nun was an mumsemumumwcs umzoq usom pUMpGOU Mom 5. no moa on an III em mpmscmuw m. mm mm on no III on msmsnmumuwpcn Hogan o.H OMH moa 46H HNH nu. ma mumsemumumwas nmaoq mEflB mmusoc nsom uflpmuo H503 #Homnu musom mafia Ofipmm “W00 mmHSOU “Gmwggm “OMUFHOU mmHSOU z umoo m0 was: mmmum>< mmoaoconpwz coaumooaac emfifle mmnsoo Hmuou paw .HDOS uomucoo tusoz uflpmuo pchme .HSOQ uflpwno mwnsoo Hem umoo “Hz: or Umbum>coo Hm>®a ha wumoo mmnsocll.ma mqmmq mmusoo men. Hmo. omo. moa. ecmm shadows Umosaoxm umou mucmumflmmm meadowuw mse.u mom.) mom.) mmm.- assume mmmHSOU .HO\ Uflm mQOHubwm MO HGQESZ mew. Hem. Hme. owe. mNHm mmmao omH.u 064.: evN.n va.- Hm>mq mmusoo «om. moa. NNH. Hma. xcmm muasomm UwUSHOQH #moo m.#QM#mHmm< wessomuo deom mHsom GMWMW uflwmuo mom uompaoo o . fiance meQMHHm> mo coflbmooaam mmmuqmonmm so Ummmm memOHOUOSpmfi mafiomoo mumoo mmnsoo tam mmHQMHsm> macaw mQOmeHwHMoo Umpowaomll.ma mqmfie 135 Because method of instruction is on a nominal scale, the Pearson product moment correlation is inappropriate. Therefore a one-way analysis of variance was used for this method in answering this question. a. Rank had only a slight correlation with costs regardless of the costing methodology employed. The highest correlation was with costs developed on credits (.204); the lowest, with costs based on student credit hours (.108); with contact hour costs at .122 and costs based on total course time at .151. b. Leyel_had a low negative correlation with course costs based on total course time (-.264) and contact hours (-.244). The definite but small relation- ship may be attributed to a slightly negative relation- ship between the course level and the teaching time (-.130) and contact hours (-.110), see Table 10. A moderate correlation existed between level and costs developed on student credit hours (-.460). This was due in part to the negative correlation between the level of instruction and class size (-.561), see Table 10. Because costs were allocated to the course based on stu- dent credit hours, the impact was such that the larger classes cost the most per class. Therefore, it follows that the level of instruction would have a negative relationship with course costs developed on student credit hours. 136 c. Class size correlated moderately with costs based on total course time (.460), contact hours (.431), and student credit hours (.681). It had a low cor- relation with costs based on credit hours (.285). The positive correlation with course costs based on student credit hours was not surprising since class size cor- related at .975 with student credit hours, the basis of allocating the costs, see Table 10. However, the substantial relationship between class size and course costs based on total course time (.460) was higher than might be expected since the correlation between class size and the total time devoted to a course was .215, see Table 10. This discrepancy can be explained by the fact that class size and the possibility of graduate assistant help was correlated at .500. Graduate assistants tended to be used on larger classes increas- ing the costs for those classes by the proportional allocation of the graduate assistant's salary. As reflected in Table 13, graduate assistants only effected a meaningful difference in correlations dealing with class size. With the exception of costs based on student credit hours, the correlations with class size would be lowered approximately ten points if graduate assistants were excluded from cost. d. Number of sections taught was split as a variable between fixed credit sections and total sections. 137 There was very little correlation between fixed credit sections and costs regardless of methodology employed. The ratio ranged from —.028 to -.065. However, there was a negative relationship between the total sections taught and costs developed on the basis of total course time (-.363), formal contact hours (-.302), SCH (-.268), and course credits (-.476). That is, as the total number of sections taught increased, there was a slight decrease in costs. It has already been reported that as the total number of sections taught increased, the total time devoted to a course section decreased (-.400) as did also the contact hours (-.344). Therefore, each course got a smaller percentage of the total contact hours and teaching hours which resulted in an allocation of costs over more courses. The same logic held for SCH's and course credits. The four variables used to allocate costs to the courses were correlated with the course costs before and after allocating the graduate assistant's salary. The correlations are reported in Table 14. It was evi- dent from this table that all of the correlations were higher when the graduate assistant costs were included. With the exception of the variable of total course time, the variables used for allocating the costs only cor- related moderately with the course costs. Other factors such as the salary schedule must account for the dif- ference. 138 vNN. Nvo. NHO. moo. mHDOE #HUOHU mmo. 4se.. flea. was. mom mNo. moo. own. Hmw. mnsom poopcou oaa. HNN. Hon. 9mm. mmHSOU 0p Umuo>mo mafia HopOB UmUDHUNm omoo n.pGMpmHmm¢ oposomuo who. Ham. omH. mad. mnsom “Homho mmm. ems. mew. pom. mom ova. wmm. woo. woo. mnsom pomwcou mom. mmo. mmn. mam. mmmsoo 0p omu0>mo mEHB Hosea omUSHOaH pmoo n.pcmumflmmfl momsomuo made musom musom . mow mmusoc uflomnu om so u p U Hobos mmaomflum> mo Goflpmooaafl womocmouom so Ummmm mpmoo wmusoo mmflmoaoooopmfi snow :0 ommoam>mo memoo mmnsoo tam mumoo mcflpmooaam mo momma mGOEd mGOHuMHmHHOOII.vH mqmfia 139 e. Method of instruction was analyzed next to determine whether or not it had a relationship with costs based on the four methodologies. For this analysis it was necessary to use the one-way analysis of variance technique for a correlation analysis would be inappro- priate. The one-way analysis of variance was testing the hypothesis that the mean course costs were equal for all of the six methods of instruction. If this hypothesis were true then method of instruction would not have a relationship with cost. However, if the mean costs of the methods are different then method of instruction may make a difference in cost. To determine where the differences were located another statistical test was necessary. Unit cost data are in the appendices. Analysis of variance for method of instruction was conducted using course costs for each of the four costing methodologies with costs including and excluding graduate assistants. The analysis was also made using course costs, including graduate assistant cost, and converted to per credit cost for the three costing methodologies using total course time, formal contact hours, and SCH. The methodology employing credit hours was omitted since there would have been no difference because it was the base unit used on the other three methodologies. Therefore, the units of study were 140 both the course costs and the course cost per credit. The F statistic, F table value, and the level of sig- nificance for each of the analysis of variance tables are listed below. The complete analysis of variance tables for each test are included in the appendices. As Table 15 highlights, the hypothesis of equal mean costs for the methods of instruction must be rejected. There were statistically significant dif- ferences in the mean costs of the methods for each of the four costing methodologies whether or not graduate assistant costs were included, or if the costs were based on per course credits instead of total course costs. In every case the level of significance was at the .0005 level except for course costs allocated on credit hours, but excluding graduate assistants. In this case the .008 level was still a very high level of significance. This statistical test answered the research question by demonstrating that there was a relationship between method of instruction and costs under each of the four costing methodologies. However, where the differences were located has not yet been discussed. To identify the differences the Scheffé method of post hoc comparisons was utilized. Table 32 in the appendices shows that in only five contrasts were there significant differences, yet they reflect a grouping pattern. __‘n...__-_.- ‘44-:- -_ .. ~ 141 TABLE 15.--Analysis of variance F statistic for method of instruction . . . F Table Level of Costing Methodology F Statistic Value Sig. Without Graduate Assistant Costs: Total Course Time 13.04 4.79 .0005 Formal Contact Hrs. 15.55 4.79 .0005 Student Credit Hrs. 17.23 4.79 .0005 Course Credit Hrs. 3.30 4.79 .008 With Graduate Assistant Costs: Total Course Time 15.47 4.79 .0005 Formal Contact Hrs. 17.36 4.79 .0005 Student Credit Hrs. 17.99 4.79 .0005 Course Credit Hrs. 5.90 4.79 .0005 Per Credit Costs: Total Course Time 7.37 4.79 .0005 Formal Contact Hrs. 9.29 4.79 .0005 Student Credit Hrs. 12.84 4.79 .0005 Inspection of this data suggested that costs of lecture and discussion were similar as were the costs of seminar and independent study. Laboratory and tutorial may form another group, Table 32. Upon reflection of the defi— nitions of each of these methods of instruction and the tendency in the literature to group instruction into laboratory, nonlaboratory, and independent study,3 it was decided to re-group instruction and test using Scheffé for similar mean costs. The three groups of instruction became (1) lecture and recitation/discussion, (2) laboratory and tutorial, and (3) seminar and inde- pendent study. 142 These three groups were contrasted against each other both on course costs and on course costs per credit. The complex contrasts were calculated by both weighting the contrasts with the number of units in the given method and without weighting the contrasts. Identical results were found with both the weighted and unweighted approaches. The results reported in Table 33 in the appendices demonstrate that lecture and discussion were definitely different in cost from seminar and inde- pendent study under all four costing methodologies whether determined on a per—course basis or on a per- credit basis. Laboratory and tutorial differ from the other two groups part of the time depending upon the costing methodology employed. It was apparent that method of instruction did have a relationship as a variable of cost with each of the four costing methodologies. Furthermore, the methods seemed to group into the three categories as discussed in the literature,4 from a methodological point of View as well as from a cost point of View as demonstrated in this research. Because method of instruction is on a nominal scale it was not possible to include it as a variable in the remainder of the research. The statistical test performed on method of instruction demonstrated that it was certainly a major factor and highly related to 143 cost. The issue must remain unsettled as to whether or not class size determines the method of instruction. However, the evidence suggested a strong relationship between cost and method of instruction under all four of the costing methodologies. 2. What is the relative importance of the variables within each costing methodology? This question was answered by utilizing multiple regression stepwise addition and deletion with course costs developed on the four methodologies as the depen- dent variable. The results of the multiple regression analysis are reported in Tables 17 - 20. The equations were calculated setting the significance level at .05 for both approaches in addition to in Tables 16 - 29. All variables were included to obtain a rank ordering, however, conclusions will be drawn only on those that are significant. The equations were calculated setting alpha at .05 for both approaches in addition to including all variables regardless of their level of significance. A complete set of data for the multiple regression equations including only those significant at the .05 level are included in the appendices. Identical results were obtained by the stepwise addition and deletion approaches. 144 Amflmmamcm Mow macaocmmmm memo modemsnm> umnuo mm. om. mcHoSHoxm c053 .mHm mo. pmmma om sues moanmflum> >HGO* mm. om. mmaomfium> flee m mm mpcmHOwaooo cospoamnuoc mamfiuasz mom. man. no. mo. No.l ma.mo MH.NN I H®>mq meSOU mam. new. no. mo. mo. «$.44 mo.ma scam seasomm mam. mooo. as.ma mo. mm.) Gm.mm se.mma . .pnmsme macawomm mo .02 Ham. moon. NH.nH 0H. mm. mo.a mn.o mNHm mmmau u.» 09.noaa pampmcoo m muOB cu moawsnauucoc domed mm5Hm> mmwwmnmo mpnmflmz pcmflcwwmmoo mucmHOmeoOU ®HQMHH6> w>HpMHSEDO Hm m cumccmpm whom mHOHHm .opw GOHmmmHmmm pcmocmmoocH Am®HQMHHM> Msomv oopnommu mfiflp mwusoo Amoco so omumooaam mpmoo How mpasmmn coflmmmmmmu wamfluaszll.oa mqmfle 145 AmHmMHmcm Mom mmOHUGmmmm memo mmHQMHHm> Hmepo mo. om. manosaoxw can: .oflm mo. ummma on runs mmmnmanm> some an mo. om. mmanmflnm> Ham .m mm mucmHOHmmmOU coapmamuuou cameras: mom. How. mo. on. Ho.n ms.ooH om.ma n Hm>mq mmusoo mom. smo. oo. oo. mo. oo.om om.aa xqmm spasomm mom. moo. om.o oo. om.n mm.mo om.oma n «unosme mCOHfiomm MO . OZ moo. mooo. mo.oH oH. om. oo.m os.s .mwam mmmao mm.hooa pcmwmcoo mmeMMMWhmwo Hm>mq mmsam> mmwwmumo whooflmz unmounwwwoo mucmaoammmoo manuaum> m>Hoo+sEDU mam m onmoamum mumm muounm .obm QOHmmemmm osmocmmmocH AmmHQMHHm> usomv musos oomucoc Hmfiuom co Umpoooaam mpwoo mom mpasmcn COHmmemmH mamfluaszll.na mqmfie 146 Amammamcm How mmoflocmmmm . memo mmHQMHHm> umeuo mcflodauxo GGSB .mflm mo. on. on. a m pmmma pm spHB mmHQMflHc> H 0* Ho. om. mOHQMHHm> HH4 .m mm mpcmfioflmmmou coaumawumou mamflpasz xcmm muasomm oom. mmm. mo. no. Ho.| wm.om ma.m 1 com. mea. ma.m on. ma.) mo.om vo.awa I Hm>oq mmusou ooo. moo. oo.o so. an.) sm.ao oH.moH : .uooomm chHuocm mo .02 eon. mooo. mh.mm mo. mm. mm.H oo.ma mNHm mmoao 9% om.omaa peopmcoo mm Hmuoe 0p mopcm mo cmaoa coaosnflnucoo waww mmMHm> Hounm mpSmHmB p . .wwmou mucmHOmemOO oaowfium> ®>Hpmasfido . cumocmum comm muomnm .cpm GOHmmmnmmm ocmocmmmocH mason AmmHQMflHm> Hcomv #Homno unmosum so ompmcoHHm mpmoo How muasmon scammmumwu mamauaszll.ma mqm Hoboo . . mnaUSHoxc Gm£3 .mHm mo. mm mm . w c ummma um npflz mmHQMHum> a 0* mm. mm. moaooflum> HH< m mm mommfloommooo coopmamsnoo meanness Hm>mq cmnsoo mom. omo. mo. on. no.1 om.mo mm.mm I mam. omm. so.a oo. on. mm.om so.ao name soaoomo Hum. moa. o>.m oa. ma. mm.a mo.m vswim mmmHo omm. mooo. Ho.om oo. mo.) mm.om mo.oma n .oooome mcoflocmm mo .02 no.9hma ocmomcoo mm Hence 0a mopmm o cmaoa mo coaucnflupcou wamq mmDHm> Houum mpcofloz p . .WW 0 mucmHOHmmmOO maomflnm> m>flooacaso Hm m osmocmom cpmm mnouum .cum scammwummm occocommccH AmmHQMHum> Hcomo mHSO£ aflomuo mmnsoo co UmpmooHHo mumoo How measmmn QOHmmemou mHmeHDZII.mH mqm scape mGHoSchm aces .mflm mo. ummwm pm suns moanmaum> sane. mmaomflum> Ham mason monocoo mam. ooo. so. so. so. om.HH os.H oao. oos. so. oo. No. oH.mm sm.o muses “Homno mmHSOO mam. mmm. no. mo. mo.| so.ov mm.vo Hm>mq mmHSOO moo. mam. sm.H mo. oo. ms.HN m~.s~ unmade mQOHpocm mo .02 oao. mma. mo.~ mm. mm.) Hm.o mo.m : mufim mmmao moo. mooo. mm.mm oo. mm. om.om oo.mma «some summons sos. moo. Ho.m mm. Hm. mm.H ma.m .musom umomuo unmosum ooo. mooo. om.oHH so. os. No.o oo.om mass mmudooaamuoe mm.omm: unmumcoo cowudnfluocoo HWqu mosam> womumo muomfimz #cmHOWWmmOU mprHOHmwwOU magmanm> m>euma5fisc Hm m oumocmum mumm muonnm .cpm scammmummm unmocmmmocH Ammanmaum> pooaoo cmouommu mEHp mmusoo Hmuoo co cmumooaao mumoo How muasmmu coemmmummu mamfluaszll.am mqm<9 RM“ c -M}&.—mfi 153 faculty rank, made significant contributions to the total explained variance in costs (R2). Faculty rank was significant in this analysis whereas without including the bases of allocating costs it was not significant (see Table 16). The highest multiple correlation coefficient of R2 (.82) was achieved by this costing methodology. By adding the bases of allocating costs as variables in the multiple regression analysis, the costing methodolo— gies were put on a more comparable basis. This was the case because the variable of class size had, in effect, the impact of one of the bases of allocating costs, namely student credit hours. However, the high R2 is a result of adding the variable used in allocating the costs. Thus, caution needs to be exercised in inter- preting the high R2. The multiple regression results for allocating costs on contact hours was reported in Table 22. A total R2 of .72 resulted from the analysis, .70 of which was attributed to total course time, student credit hours, faculty rank, and class size. It was noteworthy that contact hours did not make a significant contri- bution to the total R2, yet it was the basis of allocat- ing the costs. When considered independently of the other variables, contact hours correlated at .604 with costs allocated on contact hours (see Table 10). 154 on No mm mquHOHmmmOO newsmammhoo mamwuasz Amflmmamcm mom mmofltcmmmo mmmv mMHQMHHm> Hmnuo mchSHoxo cons .mom mo. pmmma pm sues mmHQmHMm> Nance mmanomum> Ham mas. mom. sm. mo. oo.s oo.mm om.oon Hm>mq mmasoc was. mow. om. no. mo. nm.Hm Hm.mm mhdom poncho mmucoo sas. moo. mm.m mo. an. oo.am oo.om pomsme mcofluomm mo .02 oHs. ooo. oo.m mo. sH. oo.sH ms.Hm muses powpcoo mos. ova. HN.N Hm. ov.I oa.o oa.o I «onwm mmmau How. Hoo. sm.NH mo. ma. ms.mm om.mma exemm mpasomm oom. omo. om.m Hm. ms. om.a so.o .muoom unomuo uaoosum mom. mooo. No.mo mo. Ho. mo.o mm.om mama mmusoo«HMpoe mo.mmml pampwcou m ku09 on moonsnfluucoo wamq mmdam> mmwwmumo munmflmz ucmHOWWMmou mucmflcfimmmou magmaum> m>fludessu Hm m pumccmum comm muouum .cum scammmumom “cmchQmoGH Ammmnmmum> voodoo mHSOfi #OMULHOO HMEHOM GO UOQMOOHHM mumOO MOM... m#HDmOH GOHmm mummu wHoHuHSEII.NN memes 155 Student credit hours was the basis for allocating costs and used as the dependent variable in the multiple regression analysis reported in Table 23. Whereas this costing methodology had the highest R2 when just the four variables were included (.50, see Table 18), it 2 had next to the lowest with an R of .65 in this analy- sis. Class size, which accounted for .46 of the R2 reported in Table 18, contributed no significant addition to the total R2 when all eight variables were considered. Student credit hours had a level of sig- nificance of .0005 in the equation that included only the four significant variables (see appendices). In the analysis reported in Table 23 it lost that level of significance when it was considered along with all eight variables. The difference was a result of a major shift in the standard error of the regression coefficient for the variable. Therefore, student credit hours made the major contribution to the explained variance as reflected in R2, but its level of significance was altered by the inclusion of all eight variables. Table 24 reflects the results of the multiple regression analysis with costs allocated to courses based on course credit hours as the dependent variable. An R2 of .51 resulted from the analysis, the lowest of the four costing methodologies. This was the one analysis where a different result was obtained from 156 “mamMamsm How mwoaocmmmm memo mmaamaum> stuo 09. mo. mcaodaoxw GGSB .mam mo. ummma um spas mmanmaum> saco .5 am. mo. mwanmaum> aa< m mm mpcmaoammooo Goaumamnnoo mamapasz «mo. mom. on. no. no.1 om.mm mo.a I w£o5m9 mcoauoom mo .02 «mm. oms. Na. om. ma. os.s nn.n muam mmwao «mo. Now. ma. mo. No. oh.vv hm.ma Mcmm haacomm mmo. ova. aN.N oa. ma.1 om.am mm.am 1 mucom uomucoo woo. mao. «o.o so. sa.1 om.~m mm.so~1 l.awtqu wwusoo 9mm. «mo. mm.v no. sa. No.oo sm.mma ammaowuo wmusoo son. mooo. oo.ma oa. we. so.aa om.mw *mEaB mmusoo amuoe mam. avm. mm.a am. no. sm.m ms.m summon . uaowuo uaoosum mm.om unopmcoo m o o mMawswawacwo am>ma mmsam> mmwwwumo muzoawz unmaowwmooo mucwaoamwmoo magmauo> 0>auMassdo mam m onmpcmum mowm muouum .oum ceammmummm pcwccmmwoca Ammaomaum> unmaoo mason uaomno uncoupm co ooumooaam mumoo Mom muacmwu coammmummn mamaua3211.mm mamas 157 “mammawcm How mmoaoawmmm . memo moanmaum> Heapo on so. wcaoSaoxw Ewes .mam mo. . ummma um spas wwanmaum> saqo am am. mwanmanm> aam mvcwaoammmoo coaHMawHuoo mamauadz HHM. 05h. mo. 0%. NH. NN-m mW-H GNHm www.mHU Cam. mmN. wN.H no. mo. HN.om vm.vm Mflmm %#H50wm mom. woo. om.m Na. NN.1 om.oa oo.mm 1 Moon penance moo. omo. om.o mo. oa.- oo.mm om.oaa1 am>mq amusoo Nov. Hoo. Hm.HH NH. Hv. hv.h m¢.mN ¥OEHB omudoo amuoa mmv. mvm. oo. aw. mo.1 oo.a aa. 1 *deom uaowno unwoohm mhm. mooo. mm.hH OH. 0v. Ho.mv Hh.NwH *mhdom paomuo mmusoo QNN. H00. Hh.NH mo. hN.I hm.mN mH.mm I ##fimDMB macapowm Mo .02 mm.oom unmumcoo mm amuoe on aw>mq mosam> mmumm mo muomamz ucwaoawmmoo magmaoamwwou magmaum> COaUDQaHpcoo .mam m Houum muwm m . coamwmuomm unwocwmopsa m>apmadeso . onmccmum mnouum cum . Amwanmanm> unmamo mnsos pamouo mmhdoo :0 cmumooaam mumoo How muazmmn scammwummh mamaUaDEII.vm mamfia 158 stepwise addition and deletion (see appendices). The variables denoted by an asterisk in Table 24 are the ones considered significant under stepwise addition. A phenomena similar to that reported in Table 23 regarding student credit hours also occurred with this analysis reported in Table 24. Under stepwise addition student credit hours had a significant level of .002 (see appendices), but when all the variables were included as reported in Table 24 the level of sig- nificance was excessively high (.945). This difference was also traced to the shift in the standard error of the regression coefficient. The variable, number of sections taught, made the greatest contribution to the total R2, both when all eight variables were considered (Table 24), and when just the original four variables were considered (Table 19). Total course time reported and the number of student credit hours were considered significant in all four costing methodologies. Of the original four variables, faculty rank was considered significant twice; and each of the other three variables, class size, course level, and number of sections taught, were each significant once. The variables considered significant with costs based on total course time were also con- sidered significant with costs based on contact hours and in the same relative order of importance. This was 159 true for all eight variables as reported in Tables 21 and 22, and also with the four original variables as reported in Tables 16 and 17. The question of the influence of graduate assistant costs that were included in the course costs was answered by the information in Table 25 which com- pared the multiple correlation coefficients, where the dependent variable excluded the graduate assistant costs, with those where graduate assistant costs were included. Identical results were obtained by the stepwise addition and deletion techniques where graduate assistant costs were excluded. TABLE 25.--Comparison of multiple correlation coefficients including and excluding graduate assistants costs Costing Methodology R2 R Total Course Time: With Graduate Assistant Cost .82 .90 Without Graduate Assistant Cost .83 .91 Contact Hours:' With Graduate Assistant Cost .72 .85 Without Graduate Assistant Cost .71 .85 Student Credit Hours: With Graduate Assistant Cost .65 .81 Without Graduate Assistant Cost .62 .79 Course Credit Hours: With Graduate Assistant Cost .51 .71 Without Graduate Assistant Cost .44 .66 _¥ 160 Table 25 reflects the fact that graduate assistant costs made very little difference in the multiple correlation coefficients when costs were allocated on total course time and on contact hours. As might be expected, the biggest differences resulted in costs allocated on student credit hours and on course credit hours. The difference in R2 for costs allocated on course credit hours was explained in part by the fact that student credit hours made a significant contribution to the R2 when graduate assistant costs were included. Graduate assistant costs correlated highly with large classes, thus with student credit hours. The same principle held for costing on student credit hours. 4. To what extent is there agreement in the rank ordering of independent variables across the costing methodologies? The Kendall Coefficient of Concordance was used to assess the degree to which there was agreement in the rank ordering of variables between the costing methodologies. This approach provided another way of comparing the costing methodologies. The formula for calculating the Kendall Coefficient of Concordance yielded a coefficient between 0 and l where l repre- sented perfect agreement among categories. 161 The rank ordering of the four variables among the four costing methodologies yielded a coefficient of concordance of .85. This coefficient was tested for significance at alpha .05 and found to be significant. That is, the probability was that the result was not a function of chance. Details of the calculations are in the appendices. The variables were ranked by their relative 2 in the importance in contributing to the total R following order: (1) class size, (2) number of sections taught by the faculty member, (3) faculty rank, and (4) course level. Each of the variables received its rank three out of four times. Summary The profile analysis reflected a large portion of faculty time (45%) devoted to noninstructional activities. The distribution of time over activities is consistent with the mission of the departments and representative of other similar studies. The relation- ships between workload factors provided some noteworthy correlations. It appeared that the total workweek was relatively fixed as instructional workload factors increased or decreased. The total time devoted to a particular course by the faculty member correlated highly with contact hours, but it lacked a high cor— relation with other factors such as class size, course 162 credit hours, student credit hours, or level of instruction. As the level of instruction increased, the size of the class tended to get smaller while the time devoted to the class changed very little. As formal contact hours increased the total teaching time increased, but the total workweek changed little. The correlation between course costs allocated on total course time and course costs allocated on formal contact hours was .934. Cost developed on total course time correlated higher than any other method with each of the other costs except credit hours. By inspection of the data it was evident that course costs and per unit costs were highly similar for the two costing methodologies based on total course time and on contact hours. Allocations based on course credits only partially resembled these two, and as eXpected, allocations based on student credit hour deviated the most. It was determined that the method of allocating the costs to the courses and the method of reducing that course cost to a common base would, to some extent, determine that cost. Although previous research suggested that graduate instruction was four times the cost of lower undergraduate instruction when deter- mined on a per-student credit hour, this research, while supporting that conclusion for that one methodology, 163 found lower undergraduate instruction more costly under the other methodologies. However the other methodologies develop the cost based on some unit of the course, whereas allocating on student credit hours is calculat- ing a cost based on the student. Student credit hour costs, as discussed earlier, are a subset of costs allocated to course credit hours. Method of instruction was found to be a signifi- cant variable of costs. Differences in the mean costs for methods suggested a grouping of lecture and recitation/ discussion, laboratory and tutorial, and seminar and independent study. Class size and the number of sections taught by the faculty member correlated the highest with costs. When the four original variables were considered, class size and number of sections taught made the only sig- nificant contribution to the total explained variance in costs under all four costing methodologies. When only the four original variables were considered, costs allocated on student credit hours provided the highest R2. This was due to the variable class size. When the four bases of allocating costs were included in the multiple regression analysis, which put the costing methodologies on a more comparable basis, the methodology which had the highest R2 was costs allocated on the basis of the reported total course time. Therefore, the 164 equation which provided the best model for predicting costs was based on the total time devoted to the course as reported in the faculty activity survey. The rank ordering of relative importance of variables was: (1) class size, (2) number of sections taught by the faculty member, (3) faculty rank, and (4) course level. Although there was high agreement between costing methodologies in the rank ordering of these variables, considering both the original four variables and the four bases of allocating costs, the highest explained variance is achieved with costs allocated on total time devoted to the course as reported in the faculty activity survey. NOTES--CHAPTER IV lWaldo Keith Anderson, "Factors Associated with Instructional Costs in Kansas Public Higher Education“ (unpublished Ph.D. dissertation, University of Minnesota, 1963), Dissertation Abstracts, Vol. 2507, p. 3908. 2Ruth E. Eckert, "College and University-- Programs," in Encyclopedia of Educational Research, ed. by Chester W. Harris (New York: The MacMillan Company, 1960), p. 279; Stanley Ikenberry, "Instructional Cost and Quality," College and University, XXLVII, No. 3 (Spring, 1962), 50; Paul L. Dressel) "Measurement and Evaluation'of Instructional Objectives," The 17th Year- book, National Council on Measurements Used in EducaEIOn (Ames, Iowa: National Council on Measurements Used in Education, 1960), p. 5. 3Anderson,o _p. cit., p. 3908; John M. Evans and John W. Hicks, An Approach to Higher Educational Costs Analysis, Studies in Higher Education (Lafayette, Indiana: Purdue University, 1961), p. 20. 41bid. 165 CHAPTER V THE PROBLEM, FINDINGS, CONCLUSIONS, AND RECOMMENDATIONS FOR FUTURE RESEARCH The Problem Efficient resource projection and allocation requires a thorough analysis of the faculty activities and their related costs. Effective utilization of faculty is a major responsibility Of college and uni- versity administrators. Even in light of this many institutions have little idea of how the faculty allocate their time or how this distribution effects costs. "One has only to raise the question as to what professors in a given department actually do, to learn that in most departments and most universities only the professor can provide an answer."1 Few institutions have devised a meaningful method of gathering this information, analyzing the data, or costing out these activities. Only now is the concept of faculty activity analysis emerging with general interest and acceptance. 166 167 To manage the cost of instruction requires the identification Of the major variables Of costs. Yet research on costing methodology in higher education has lacked a thorough examination assessing what results when different costing methodologies are employed. To generate a cost figure is not sufficient when it is realized that the costing methodology will to a degree, determine that cost figure. A need for research which inquires into the nature of the costing methodology used for allocating costs to the individual units chosen was reflected in the literature. Research Objectives The objectives Of this dissertation were divided into two parts. The first was to examine the inter- relationships Of selected instructional workload factors through determining the distribution of faculty time among several activities and assigning costs to these activities based on the salaries of the faculty. The profile analysis was developed by department and by rank. The second Objective was to compare four costing methodologies used in allocating costs to courses. The bases of allocating costs were: (1) the total course time, (2) formal contact hours, (3) student credit hours, and (4) course credit hours. This comparison was not only based on the costs generated per course, but it also considered the differences in selected variables Of 168 costs and the respective costing methodology. The five variables of costs which were examined were: (1) faculty rank, (2) course level, (3) class size, (4) number of course and/or sections taught by the faculty member, and (5) the method Of instruction. The comparison Of costing methodologies as related to these variables centered on answering four research questions which addressed the comparability of these methodologies. Parameters of the Data The population from which the data were drawn for this study were the faculty and course offerings of one college at Michigan State University. This college offers both undergraduate and doctoral level programs, and its five departments are diversified in both the nature of their activities and course offerings. The entire faculty were surveyed and a return rate of 94.1 per cent was realized. All but one faculty member not responding were contacted in a follow—up. The results of the follow-up were only used to determine their com- parability with the rest of the college. All of the data, except salaries, were collected for the fall team of 1972 through a faculty activity survey instrument designed by the National Center for Higher Education Management Systems. 169 Findings The findings must be interpreted in light of limitations of the study. The costs were developed from only the instructional portion Of the faculty's time and from one college at one university. 10 A large portion Of faculty time (45%) was devoted to noninstructional activities. The distribution of time spread over activities was consistent with the mission Of the departments and repre- sentative of other similar studies. It appeared that the total workweek was relatively fixed as instructional workload factors increased or decreased. The total time devoted to a par- ticular course correlated highly with formal con— tact hours, but not with other factors such as class size, course credit hours, student credit hours, or level of instruction. As the level of instruction increased, the size of the classes tended to get smaller, while the total time devoted to the class changed little. As faculty rank increased there was a slight decrease in total course time, formal contact hours, and the number of sections taught. A very high correlation of .934 was realized in comparing costs allocated on total course time 170 and costs allocated on faculty reported formal contact hours. Costs developed on total course time correlated higher than any other.method with each of the other costs except those developed on credits. Costs develOped on the methods, total course time, and faculty reported formal contact hours were highly similar in relation to the level of instruction. It was determined that the method of allocating the costs to the courses and the method of reducing the course costs to a common base did, to some extent, determine that cost. Although previous research suggested that graduate instruction was four times the cost of lower- division undergraduate instruction when deter- mined on a per-student credit hour, this research, while supporting that conclusion for that one methodology, found lower-division undergraduate instruction more costly under the other method- ologies.. However, these other methodologies were costing on the basis of a unit related to the course, whereas the student credit hour is really a subset of course credit hours. This finding supports the contention that the costing methodology will to some extent determine the cost. 171 8. Class size, the number of sections taught by the faculty member, and method Of instruction were found to be significant variables of course costs. 9. The dependent variable used in the multiple regression analysis which explained the greatest amount of variance in costs was the method for which costs were developed on total course time as reported in the faculty activity survey. Conclusions The value of a faculty activity survey has been demonstrated in this research. When such a large portion Of faculty time is devoted to noninstructional activi- ties, it is important that the faculty time be utilized efficiently and effectively. Work assignments for faculty should be individualized to take advantage of the strengths of each faculty member. A perusal of the individual survey instruments suggests that workload needs to be more evenly distributed, for some faculty carry an excessively heavy load. Teaching assignments should never be based solely on student credit hours, and certainly consideration should be given to contact hours. To achieve maximum utilization Of faculty, a program of self-renewal through a faculty development program may increase their productivity. Faculty should be given tangible incentives for assisting in finding new methods for reducing costs 172 without reducing the quality of instruction. Since a large portion of faculty time is not under the influence Of administration, it may be desirable, as suggested by others2 to define a teaching load which represents the full-time involvement of faculty and adjust that load when the faculty member is involved in research, service, or other activity. Although there appears to be some rationale for allowing a small increase in contact hours, the major areas where economy can be achieved would be in reducing the number of courses and increasing the class size. These decisions need to be made in light Of the instruc- tional strategies and purposes. A variety Of classes and class sizes might be desirable and possible if a thorough analysis were made Of existing courses. Further research would be indicated; however, the areas of noninstructional activity may provide an additional area where economy can be achieved. As demonstrated in this research, method of instruction is definitely related to course costs regard- less Of the costing methodology employed. However, its importance, as noted by others, is likely to be related to its influence on class size and teaching load.3 The true cost of instruction is related to the salary paid the professor and how he allocates his time to instruction. A faculty activity survey is the only 173 acceptable method for allocating costs to courses Of the four considered in this research, because time does not correlate very highly with the bases Of allocating costs under the other methodologies with the possible exception of contact hours. To develop a costing methodology on the presumption that a relationship exists between time and some other basis of allocation would be erroneous. Only formal contact hours had a high correlation with time. Because a large portion of instruction is based on variable credit courses and the formal contact hours for these courses are not listed in the schedule of courses, it may be necessary to survey the individual faculty members for this information. If a faculty activity survey is reasonably reliable and valid, as a growing body of research seems to indicate, then it can be used as an effective and efficient methodology for allocating costs to courses. If the data cannot be considered reasonably reliable and valid then costs developed on one of the alternative costing methodologies still will not provide a true reflection of how resources are really being allocated. It is imperative that costs be understood in relationship to the purposes for which they are to be used. The limitations Of cost analysis should be kept clearly in mind when a cost study is being planned. 174 Recommendations for Future Research Several recommendations emerge from this research. The issue of the reliability and validity of faculty activity surveys needs additional study. Some method for assessing the high number of hours worked per week, as reported through faculty surveys, would make a sig- nificant contribution to existing research. This study was conducted for one term in one college. It would be desirable to have this study duplicated using an entire year and crossing several colleges within the university. To extend the research to a larger sample with several institutions partici- pating would strengthen the argument for the use of faculty activity surveys.l There is still a need for more research on com— paring different costing methodologies. Not only is there need for examining the results Of several methods of allocating costs to courses and methods of converting those course costs to unit costs, but also for conducting this kind of comparative study over a wider range of institutions which will have different emphases. Definitive research is needed to assess the impact of an increase in Class size and an increase in the number of sections taught by the faculty on the quality of instruction. It appears that the greatest 175 opportunity for economy is in class size, yet little is known about its impact on the quality of learning. Since such a large portion of faculty time is spent in noninstructional functions, research to explore the cost-benefit Of such activities could make a major contribution to workload studies. As higher education continues tO adopt management information systems, it is important that cost-benefit be researched. The problems in defining outcomes are many, yet without more research in this area, the value of management information systems is limited. NOTES--CHAP TER V 1 Marcus, The Confidence Crisis (San Francisco: Jossey- Bass, 1970), p. 186. 2Reeves, et a1., op. cit., p. 277; Floyd W. Reeves, Nelson B. Henry, 33d EEEn Dale Russell, Class Size and University Costs (Chicago: University Of Chicago Press, 1933), p. 149; Paul L. Dressel, private discussions on this research (East Lansing: Michigan State University, July, 1973). 3California and Western Conference Cost and Statistical Study for the Year 1954-55 (Berkeley: Uni- versity of California, n.d:), pp. 30-31; Beardsley Ruml and Donald H. Morrison, Memo to a College Trustee (New York: McGraw-Hill, 1959), pp. 19-22. 176 Paul L. Dressel, F. Craig Johnson, and Philip M. APP END ICE S APPENDIX A UNIT COSTS FOR METHOD OF INSTRUCTION BY COURSE COSTS, COURSE CREDIT HOUR, AND STUDENT CREDIT HOUR aNm ovv 00w Com m how Nma moa VBN 5N 00V hha wmm mom Oa moo ONOa mmaa vaa mm mmu Coo vao 0mm ma maOa coma OONa awaa av MHSOI WHDOE WHsom @BHH Z paomuo mwnsoo pacwuo PCOUSwm Homecoo amauom amusoo amuOH am omeNUOaa< memou Annmaaoo can coauOsHMMCa mo cosmos an amaHOuSH AUDwm ucwonowcca amcaEww coawmsomao Cam ceapmvaomm mucumuooma THDPONJ coaponuvmca MM pocpwz UwOO meDOUll.mN mamdfi 177 50m 006 00% m aMHHOPDH ow mm mo so sospw pcmocmomoca N5 mma mma Oa Hmcasww msm dam vow om coawmsowao Ucm GOMHMpaowm m Goa aom ONM Ma AHOHQHODMA mom Nvm vmm av manpoma muso: manor OBaH [ml coaewwuuwca mo coopwz wacmuo sconcem Homecoo amsuom mwusoo adeH am OOHMOOaa< uwou noon pacwuo wwusoo Hem H50: uaomno mmnsoo Ham cu omuum>aoo new moonpmfi mouse ma commooaam pmoo mmHsoo coaocsuamca CO oonamfi mo mpm0011.sm mamfie 'I , - 179 Nv.mh ©©.ON ao.ha vo.aa Oh.vN mh.© Haom «mowuo osmospm Ham 5N Oa wN Ma aw amaHOuSH atopw Pcmccmowcca Hmcaswm coamwsowaa can coavwwacom >HOpmuonmq gunpowq coaposupmca mo UOcHMS GmHDOO umoo noon #aomno Dcmcsum Ham on omuum>soo cam mfiap amuou so omenooaao mumoo mmudoo coaocsuumca mo oospmfi mo MDMOUII .mm mamfiB APPENDIX B ANALYSIS OF VARIANCE FOR METHOD OF INSTRUCTION mma mm.moo.oom.vm Haves om.ooo.voa was oo.smm.vmv.oa mooopme cornea mooo. oo.m oa.o0a.aso m o .ovm.mmw.o moccpme cows»mm mam wnam> m mmmmmWICMNZ Eocwoum wwum: m mo 65m mocmaum> MO wousom MO aw>wq mo mwmumwo maaomuo mwunoo co cmmmfl uwoo mmusoo "manmaum> wcwccwowo .v mma vo.mmo.omv.ms Haves oo.mwo.mom maa om.mom.oos.ma moocuoe carve; mooo. oo.sa as.ooo.mmm.o m mm.mvv.omo.mm moocpws cmwzpmm mam wnam> m wumzmw cmmz Goowmum mmumswm mo saw wocmaum> Ho mousom Mo am>wa mo wwwummo wHSOL #fiUwHO waUSHw CO UGMGD PWOO meDOU uwflflmfiHmsr vaUCwQWQ .m mma vo.wom.ooa.om Hmpoe sm.aao.oom maa am.avm.m0a.am moocpws canvas mooo. om.sa mm.aav.sao.m m oo.smo.smo.mm moonpme cmospwm mam moam>.m whosom com: Bocwwum wmumsmw mo 55m wocmanm> mo wousow mo am>wq mo wwwnmwo WHDOr— POMPCOO C0 UOWMD PWOO wwHDOU "OHDMflHMK/ PCOUCOQOQ .N mma oo.mam.amm.mv Haves mm.sos.oam mam oo.mma.asm.mm woospwe canvas mooo. 1s4.mH mm. so.amm.m m oo.ooo.oms.o moooome cmmapwm 04mm 03Hm> h OHND w EMT: EOUQOHE wWHGD w MO 55% wocmHHm> MO OOHflom wo am>mq mo mmmnowa wEav meDOO amwOH :0 Conan uwoo mmHDOU "maflmanm> wcmccwowo .a “mumoo osmawamwm mumccmnm msaosaocao coauosnumsa mo UOSDOE How mocmanm> mo mamhamcm4 mooo. mam mo am>ma mooo. mam Ho aO>QA mooo. mam Ho am>wq pcmumammm mumsomnm mcaUSaoxmo coaposuumsd mo conumfi mom mocmano> mo mamMaMG¢11.om mamfie mma ma.moo.amo.oa amupe mm.mam.ava oaa vs.moa.ooo.oa mooopma ensue: om.m o¢.omm.oov m om.mvo.amm.m mooopwe cmwsumm Ozam> m mumamm cam: accumum mmuaoom mo saw wocmaum> mo mousow Hc.m0mumma wpacwuo wmnsoo :0 Conan awoo wwusou mma oo.oom.mmo.oo ma. ooH som oaa ma.smm.omo.mm IMNLmHI. s. o o :oaa m MI; am.mvo.vom.mm msam> a mum: w cow: socwwum mmumaow Mo 89m mo mmwumwa maso£ wacmuo wcwcsvm co comma uwoo wwusoo "mannaHmD vcmocwoma aMpOH mooopma sesame mwmrvma cmosvwm wocmaum> mo mounmm «manmaum> vcmOcmomQ ampOH mOOcuma caspaz wmonuma comemm cocoaum> Mo moHsow «wanmaum> economowa amwOH wcoruws casvaz moccammlcmwzemm mma mo.¢mm.omm.ov av. soa mmm mam om.mos.mso.om mm.ma o. m :smo m m ma.oos.moo.oa 03am> m can: m cow: aocwwuu mwumsmw mo 83m Ho mmmumwo wH30£ PUQHCOO CO OQwMD PmOO meSOU mma ov.mmo.moo.mm om.oms.moa oaa ow. ova .omo am vo.mH No.4av.oom.m m mo. mso. moo an mSam> m mumaow cwmz socmwum wmnmsmw no sow Ho mwmnooo weaw me500 ampr so comma pmoo wmusou Humoo mucoaum> mo wousow "manmaum> vCOOCOQOQ .a 182 IMbbthI. 1mmLMHI .mam osam> E mo am>wq Innuen. mooo. om.o .oam moam> m Ho am>wq IMbbthI IthbII .oam msam> m HO am>wq mcacsaoca umoo paomno ammo mma as.oao.vmo.o amoue om.mam.sm maa ov.ooo.omv.v moomums canvas mw.mom.amo .ImI mm.oam.oov.m wooopms cmmspwm mumsow cmwz Bocmmum wwumsmw we saw wocmaum> Ho OUHDom mo wwmuowo pacmuo acmcsvm co comma uwoo pacwuo Hmm "wannaumD wcmccwowo .m mma ms.voo.mms.0a ampoe om.oom.mo was os.mvo.moo.s moospws canvas Iom.mos.moo m mo.aoo.mmo.m mooopms cwosvmm mumamw com: socowum wwumsow we saw wocmaum> mo condom mo wmuomo muse: woopcoo co comma uwoo vacwuo Hmm "manoaum> vcwccwowo .N mma oo.mmo.aoa.o Hence sv.mom.om was mmnaam.oom.o. mooouma caseaz sw.voa.sm¢ 111.m mm.msa.moa.m moospme cwmzpmm mumaom com: cocoons mmummmm mo saw wocmaum> mo cowamw mo mmmummo co comma emoo vaowuu Rom "wanoaum> acwccwomo .a mumoo snowmammm mumsooum coauosnumca mo eczema Rom moamanm> mo mammaoc¢11.am mamfie APPENDIX C SCHEFFE CONTRASTS ON METHOD OF INSTRUCTION TABLE 32.--Scheffé simple contrasts (six methods - course costs) Where mean cost for method of 10 2. instruction is noted by: lecture laboratory discussion seminar independent study tutorial Basis of Costing Methodology —.————-—————— Contact Egmparison Time Hours U1=U2 N N U1=U3 N N U1=U4 Y V U1=U5 Y Y U1=U6 N N U2=U3 N N (12:114 N N U2=U5 Y Y U2=U6 N N U3=U4 Y Y [13:05 Y Y U3=U6 N N U4=U5 N N U4=U6 N N U5=U6 N N Where: Y N = yes: = no, Credit SCH Hours N N N N Y Y2 Y Y2 N N N N N N N1 N N N Y N Y Y N N N N N N N N the comparison can be rejected the comparison cannot be rejected Alpha level .05 more Cannot reject Cannot reject Can reject Can reject Cannot reject Cannot reject Cannot reject Can reject Cannot reject Can reject Can reject Cannot reject Cannot reject Cannot reject Cannot reject 1When graduate assistant costs are excluded this becomes a yes. 2When graduate assistant costs are excluded this becomes a no. 183 184 TABLE 33.—-Scheffé complex contrasts (three methods - course and credit costs) Where mean cost for method of instruction is noted by: H Costing Methodology Egst per course: Total Course Time Formal Contact Hours Student Credit Hours Course Credit Hours Cost per credit: Total Course Time Formal Contact Hours Student Credit Hours Unweighted lecture and recitation/discussion laboratory and tutorial seminar and independent study Where: Y = yes, the compariSQn can be rejected N = no, the comparison cannot be rejected Alpha level = .05 Weighted UJ-U3 UJ—UZ Y N Y N Y Y Y N Y N Y N Y N 92:2321i’31119222ila N Y N Y Y N Y Y Y N Y N N Y N Y Y N Y Y N N APPENDIX D MULTIPLE REGRESSION RESULTS FOR ADDITION AND DELETION ANALYSIS WITH A MINIMUM OF .05 LEVEL OF SIGNIFICANCE (CONSIDERING FOUR VARIABLES) mm» om. a ma mpgoaoammooooa .mam m Hosam mpom mo scammwemom mapmaem> Psoccomooea o>aamasaso onwocmpm macehm .oam w m ocmoamac aw Ho aw>mq mo. Mo ESEacaz m seas weanmaum> How wamNamc< coapwamn can coauaco< Ammaomaum> HDOM mcaumoamcoco ompuommn mfiap meSOO amuou so cmumooaam hence you unannon doammmummn mamauaszI1.vm mamas 0v. INVN. m D N mpmoaoaMMooo coaemaossoo maeapadfi New. moo. mo.w mo. em.1 mm.vv om.mma 1 sconce m:0apoww M0 .02 % oma. mooo. so.mm mo. om. mo.a ao.b wnam monau l mv.m00a ecmpmcou mm awPOH op mooam> mmpom mo mpomaos Pcoaoamoooo mscoaoammooo seasonaapnoo aosoa .mam a gonna opom mo scammoHMom oaooano> Pmoocommona o>aamadsdo osmocmpw whoehm .opm etcataMaccam wo aw>wq mo. Molfimmwmma.m spas cocmanm> now mammamc< coawmacm can coawaco< Ammanmanm> snow OCH Inwcamcooo MHSOQ womucoo so Umamooaam mamoo MOM moadmmn scammmnmwn mamaua5211.mm mamma i i 1%“ i i 187 05. av. m m N mpcwaoaMMooo soapmaohsoo maaapaefi cow. wao. oe.m 1 so. oa.- so.ov mo.a0a1 usosms wcoavoww mo .02 vow. mooo. Hm.oo so. oo. om.a oo.¢a mnam wmmao mo.owh acmvmcoo mm aopoe op maoaos mopmm oo magmas; pummomooooo mpaomoaooaoo coapdbahpcoo aoooa .maw m Hosea meow mo scammoamom wapmahm> Pamocomocca o>apmadsoo ohmcflmem macehm .opm otcmtamaccaw mo aw>wa mo. Mo essacaz m spa? mocmanm> How wam>amc< coavwawm com coauaoc< Amwaomanm> HDOM mcahmoamcooo mason uacmuo ucwosum so ombmcoaam mumoo How mmadmdu scammmnowh maoaua5211.om mamma 188 Nm. 5N. memoaoammooo coaemaohaoo oaoaPaSE abN. boo. NN.N mo. aN. No.a us.N wnam mmmau CNN. mooo. am.am mo. vw.1 Ow.oN ¢O.oma1 prance wcoauowm mo .02 hN.mwNa Peopmcou mm acpoa op modao> mweom mo mpmmaoa pcoaoammooo mpsoaoammooo cOaPSQHNPcoo ao>oa .mam m Hoshm opom mo coamwohmom oanmaam> Pamocomooca o>apoadsdo ohmocum macasm .opm mocmoawacoam mo aw>ma mo. w mHDOQ #HUGHU GWHSOO o EDEacaz m sea: mocmaum> How wawaamc< cOaumamD cam coapapo< Ammanmaam> Hoom meanmcamcooo so OOHMOOaam mpmoo How muasmmh scammmnomu mamapa5211.sm mamma APPENDIX E MULTIPLE REGRESSION RESULTS FOR ADDITION AND DELETION ANALYSIS WITH A MINIMUM OF .05 LEVEL OF SIGNIFICANCE WITHOUT GRADUATE ASSISTANT COSTS (CONSIDERING EIGHT VARIABLES) a0. mw. NB. mpcwaoaMMoooma meDOC amw. ONV. mo. 50. mo. ao.o om.h mHDOI vompcct . psosme omw oma. VO.N v0. 00. om.wa mm.0N wcoapowm Mo .02 me. 5N0. OO.m mm. mm.1 O©.m mo.w 1 wmam mmmaU ONw. v00. mv.m mm. 00. Oa.a aN.m wHSOI “MCOHU vchSPm moo. mooo. wm.mN v0. aN. mm.ON mm.m0a Mcmm heasomm mmo. mooo. ma.OMa to. am. ma.m Ca.oo wsafi wwusou amvoe oo.ooa1 pcmpmcoo mm Homes on mmoams moomm oo magmaae oaoaoaoomoo mpaoaomoowoo scapdnahpcoo aoswa .wam m homam mpmm mo coammmemom oanmaam> anecdomooca o>aaoansco osmocmpm maonnm .opm mumoo asmpmammm mumsomuo udoaaas mfiaa wwhdoo amu0p so cwumooaam mamoo now mua5mmn scammmnmwn mamaaadzll .wm mammd meDOC man. met. 00. 50. NO. mm.hv VN.va mHSOI PHUwHU meSOU womant vac. omo. ah.m 00. aa. N¢.©N mc.om mcoavowm mo .02 won. Vac. wN.o 00. MN. w©.ma oN.®m MHDOI wOoPcOU amo. moo. m0.m am. a©.1 Oh.m Nm.aa 1 wNam mwmaU moo. aOO. mv.aa m0. ma. 00.Nm Nm.aaa Mcmm Apasomu amo. an. mw.o am. am. ms.a hm.m mucom wacwuo ucwcsem Noo. mooo. mo.mv oo. 00. ma.w hm.mm mSHH condom amPOB am.ah¢1 wcmpwcoo mm Hopes op 83$ 93% mo animate paomoaoowoo opaoaoaooooo scapdoanpcoo aosoa .mam m Hosea memo mo scammonmom oaooanm> pnmomomooca w>apoadsco ononmvm whoasm .opm mamoo Damamammm mamsomum “commas undo: uomwcoo amEHom so commooaam momoo How weaSme scammmnmmu mamapaszll.mm mamma 191 on. mpswaoammo001soapmaossoo oaaapadz No. a N sao. Nvo. Ho. om. mo. om.s vm. wwam wmmao sao. ovo. Ho. 60. oo. mv.mv om.m xcmm spasomo vcmomH sac. vow. mo. so. ao.1 as.sm mo.m 1 mcoaaooo so .02 sao. mmm. vv.a 0H. oa.: ca.om ma.vm 1 muse: pompcoc mac. moo. mv.m so. sa.1 ow.ss vm.ama1 aw>wa mmusoo vom. omo. vo.m oo. oa. oo.oo om.oaa mnno: pacmuo mmusoo vsm. mooo. wm.oa aa. mv. mv.oa vm.mv msae wwuaoo ampoe vsv. woe. so.a om. sv. vm.m oo.m manor «acmuo «conspm QF.MBH PCMFMCOL mm aopoe on mooaos mmpom so mpammoe pnoaoaooooo manomoaooooo coapdbahpcoo aoeoa .maw m noshm opom mo scammosmom oapmaem> Pfloocomocca o>apwa5850 ohmoQMPm whoehm .opw mHDO£ paomhc ucoosum so omnmooaam mumoo How maadmmu scammoumou mamauaczll mpmoo acmumammm.®#osoohm “somuas .ov mamdfl co. mm. mm moammoaoomoo-aoapoaoasoo maompaas ovv. ooo. mo. mv. mo. vo.v so. 1 mnam mmmao ovv. omo. mm. so. vo. oo.om ov.ma scam opa50mm omv. mos. o0. mm. ma.1 Nm.a ow. muDOI vacmuu pcmcsam omv. osa. oo.a 00. Na.1 Nh.Nm as.Ns 1 ao>wa monsoo % omv. msa. ow.a ma. sa.1 Ho.ma os.wa 1 - muaor pompcoo 11 omv. coo. om.s ma., om. so.s v5.0a wEfiH mmusou anvOH Nwm. mooo. ao.ma 0a. av. oN.av wo.voa muse: pacouo omnsoo “roams NmN. mooo. Na.ma mo. Nm.1 mm.mN sm.oo 1 mcoaeomw mo .02 ov.Nvm pcmpwcoo mm aopoe OP mosam> mmpom mo mpnwaos Pcoaoammooo mpcoaoamomoo scaponaepsoo aoeoa .mam m Hosea opom mo scammoswom manoasm> Pcocmomoosa o>apmaosso osmosmpm msoahm .opm mumoo adopmammm mumsomum usocpaz mason pacmuo mmnsoo so omumooaam mamoo How muadmmu scammcummn mamama5211.av flames APPENDIX F MULTIPLE REGRESSION RESULTS FOR ADDITION AND DELETION ANALYSIS WITH A MINIMUM OF .05 LEVEL OF SIGNIFICANCE (CONSIDERING EIGHT VARIABLES) 00. am. N mesoaoammooo moapmamaaoo maeapadz wow. mooo. o¢.mN v0. aN. oa.mm oo.caa xcmm Avasomm hos. mooo. ss.mv vo. oN. NN. wv.a wnsos eaOOHU sameneu v00. moon. mw.avm v0. wk. wv.m mN.v© OEMH meSOO ampOH so.vaN1 pcmemcoo mm Hopes 8. moods momma co mammals Emmoaoomoo mpaoaoaooooo soapsbaspcoo aosoa .mam m momma opom mo scammoamom oapoaao> Pmoosooooca o>apoacsdo osmocmpm whoaem .opm cocooamaccaw Ho aw>oq me. no Essacaz m Spas mcsomasm> How wam>amc< COawmawD pom crawaocc Amoaomanm> acmam mcahmoamcooo cmouooch osaa mmucoo ammo“ co commooaam momoo How measmou scammonmmu mamabaczr1.mv Mammy 193 194 Wm. Oh. mpcoaoaMMoooIGOapmaosaoo oaQaPaSE moo. ONO. ao.v MN. Om.1 hm.m ma.Oa 1 ONam wmmau aoo. moo. wv.o mo. ma. ov.vm mo.mOa Mcmm Apasomk voo. aoo. mm.aa MN. ms. av.a oh.v wnsom vacoHU ucwoauw oom. mooo. ov.msa mo. Os. ma.m sN.wo mEaH mmnnoo aNPOH mv.ovm1 ocmomcrt mm amPOH OP mosao> mmpom mo mpgwaos Pnoaoammooo mpcoaoammooo coapcbaapcoo aosoa .mam m Hosea opom mo scammwhwom manwaso> Pmmodooocca o>apmaos50 osmocmpm whoeam .opm wocmoawacmam MO ao>wa mo. MO EJEacaZ m sea: moanmasm> mom mawaamc< ceawoamo 0cm ccapacon Ammanmanm> nomad mcaumoamcooo Munoz pomwcoo amfiuom so UGDMOOaam momoo How headmmu scammwummu mamauaszll.mv mamfie 0m. mo. NB mpsoaoammooowa wmusot wNo. NOO. mm.Oa 00. 0a. ov.mv am.sma mudom pacwnu mMHDOL 5 s00. mooo. ma.om oo. Om. aN.o Oa.vm wEaH ownsou amwoe 9 l mam. nOOO. mh.mo 50. mm. NV. mb.m WHDOI wHUOHU PCOUDPU sm.saa pcmwmc0L mm Enos 8 memos momma so Boone paoaoaooooo mpaomomooooo coapcoaapcoo aosoa .maw m Hosea memo mo scammosmom mammaam> puoomomooca osapoaoeoo osmocmpm msoaem .opm mocmoamacoam mo aw>w4 mo. mo EDEacaZ m spas moasmaum> Mom mamaamc< coawmaom pcm coawaoco Amwaamanm> ammaw.mcanmoamcooo muse: pacmho “cocoon so cmmmooaam mamoo How mbadmou scammoumwu mamaoa5211.vv mamfie 196 00. be. m NB mpcoaoammooo mopom mo menmaoa Paoaoammooo mesoaoamoooo coapobaapcoo ao>oa .mam m Romeo mpom mo soawwoewom oapoaam> Psooeomooca o>aeoassco osmosmpm maosem .opm [Mocmoawaccam MO am>mn mCd mo ESEanE m sea: mmabmaum> How mamawoc< coavaoo< Ammanmaum> unmam meanmoamcooo mason uacmuc omusoc co omumcoaam mamoo How muasmmu scammmummn mamaua5211.mv mamme 197 mpsoaoammooocsoapmaoseoo oaeapaoz mov. amo. ss.v Na. mN.1 ma.va oo.om 1 musom pompcOL who... N00. Om.Oa no. NN.1 NN.o.v vs..ova1 amt/ma meDOC mmv. mooo. mN.ma Na. Nv. vN.s sm.oN maaH wounoo ameOH va. mooo. vo.sa so. cm.1 mo.mm ao.s0a1 asosms mcoeeoom mo .02 m3... mooo. cm.Nm to. av. vole mmivwa wHSOI vacmHU meDOL mN.vms vcmpmcot mm awPOH OP modam> mmpom mo mpsmamg Pcoaoammooo mpcoaoaommoo coapopasecoo ao>oa .wam m momma oeom mo scammoewom manoaam> Psoomooooca o>aewaseso camocmpm mnoaam .opm occmoawaccam mo am>mq mo. mo Essacaz m Leah mwanmauo> How mammamc< coavwawa Ammanmamm> nomad mcaumoamsooo undo: paomno mmusoo so Umumooaam mpmoo mom mpasmmn scammmumon mamauaszll.ov mamme APPENDIX G FACULTY ACTIVITY AND OUTCOME SURVEY INSTRUMENT FACULTY ACTIVITY AND OUTCOME SURVEY Date Academic Term Name Phone Please address any questions to Upon completion, please detach the form and send it to Purpose of Survey This survey instrument has been designed by the National Center for Higher Education Management Systems (NCHEMS) for use in this as well as many other institutions. You are now participating in a pilot test that is being conducted to investigate the practicality of using faculty-generated activity data for costing, budgeting, planning, and reviewing academic programs. If you wish to comment, we encourage you to do so on the last page of this booklet. 198 199 PLEASE READ THE INSTRUCTIONS ON PAGES 2 AND 5 FORM. BEFORE YOU COMPLETE A sample form is included on pages 6 and 7. General Instructions This survey asks you to estimate the hours on spend during an “average week" of this term ed in diff ' asks you to estimate the percents e contri ution of these hours to th engag erent types 9‘ activity. It then plea for each activity as youezomplete the survey. 0 outcomes of the institution. Please read the activnty deiimnons and SECTION A: TEACHING ACTIVITIES A.1. Scheduled Teaching: All activities related to courses (degree and non Mt d credi da the current term. '1th activities would include: m an non t’ y or am) given in mfififl w” W ”mm“ 331% 22%“ game... Supervising these comes 'lutoring Mum” Preparing Meeting scheduled classes Supervising independent 5?an MW” my Evaluating 3° a 8 students Instructions for Columns (a) through (i) (a) Enter de ents, colleges, or other unit desigiatim (0) Enter method of instruction as coded below under ch the course is taught (f) Enter scheduled contact hours/ week (g) Enter hours/ average week of unscheduled contact with (c) Enter section designation. numberofstudents enrolled, and students in course check (~/) if course material is remedial (below college level) or if it is extensiai (windpally directed toward (h) Enter hours/ average week spent in preparing and nonmatriculated students) arranging the activities of the current course (b) Enter number or other designation of course (d) Enter credit hours given for course (i) Enter the total hours/average week (sum of columns(f). (g), and (h) in Section A.1) Method of Instruction Column (e) Code Method Deflnltbn A Lecture Formal presentation - primarily one—way communication B Laboratory instructing, peparing, and supervising student investigations C Recitation/Discussion No-way communication of course materials D Seminar Students carry the major responsibility for preparation E Independent Sudy Students work independently with only minimal faculty direction F Tutorial Students work one-to-one with the imtructor G Programmed Instruction Course contents pesented throng) p‘ogrammed materials A.2 UNSCHEDULE D TEACHING: Teaching not associated with the specific courses listed in A.1. For example: Thesis committee participation Guest lecturing in another faculty member’s course Discussions with colleagues about teaching Giving seminars within the institution A.3 ACADEMIC PROGRAM ADVISIN G: Giving advice to students cmcerning course scheduling and academic programs. Not to be confined with counseling that is included in D. 1 AA COURSE AND CURRICULUM RESEARCH AND DEVELOPMENT: Developing and preparing for future courses. For example: t Preparing course outlines Devising new instructional materials mo :3: 333321;; ts ' Evaluat teaching effectiveness Developing book lists Revising existing materials and planmnging changes Level Codes Column (p) Code Description Code Description A Preparatory E Upper division and graduate B lower division F Graduate c Upper division G Professional D Under-gramme H Other 200 NAME 1. D. on SOCIAL 5!:me N0. DEPARTMENT RANK FOR INS'ITI‘U'I'IONAI. use ‘ PERCENTAGE DISTRIBUTION 10 SECTION A. TEACHING acnvmrs INSTITUTIONAL OUTCOMES (I) (ll 0) (In) in) (o) A.l SCHEDULED TEACHING E ,- 3 E III - Eg gm o+ z m+m+m=m "0‘3” 33 egg E'g E25 5 EU E3 (0 Put 5? go... «:2 z‘E 4:13 '2': 38 ”I" comm (d) (e) (o m WEEK ‘3“ < ir- 23: EFL” 2 . Numb“ Em.“ w (J)! grasp“? oron Ironing.r omen mena- gg Egg 3; E55: 55% fig 2E ..... time 23::- mmuc. Comm” 0335;? 1133:; 33< 52< a: 98}: eemls Se 110N IsraA'rION as % at as a '79 susror/u. ( ) ACTIVITY DESCRIPTION A.2 UNSCHEDULED TEACHING AJ ACADEMIC PROGRAM ADVISINC COURSE AND CURRICULUM RES. & DEVELOP. OUTCOME DEFINITIONS 'lhis section of the form allows ou to indicate what outcomes your activities principally bene t. Please try to make a rough estimate of the percentage distribution for each of your activities to the following outcomes: (3) Student Growth and Development: ‘Results and benefits of activities that contribute to enhancmg personal, social. academic and/or career aspects of students who are registered in the institution. (k) Development of New Knowledge and Art Forms: Results and benefits of activities that contributeto the development, storage. utilization. and/ or appreciation of knowledge and art. (1) Inseparable Combination of (j) and (k): Results and benefits of activities that contribute to both student growth and development and creatim of new knowledge and art forms and cannot be separated. (it is preferable to separate these if possible.) Community Development and Service: Results and benefit: of activities that contribute to educational growth in an provide short- or long-term utility to the community. (II) General Institutional Support: Results and benefits of ac- tivities that contribute to maintaimrg the institution. (0) Personal Professional Growth: Results and benefits of ac- tivities that contribute principally to your professional growth. V (m 2()J. PERCENTAGE DISTRIaImON To INSTmmONAL OUTCOMES U, (I!) (I) (III) (II) (o) I- .8 (I) E In 2+ E AVERAGE. mg 8.4% Bk- >°m 2 go E3 WEEK U... zin- n° :5: o _,_, 3° i-> o°l- <2. Z a. TEE (< 39 Zn! _Zfi KO 9“”0 < 22 i- lsln F: E: :24 I 2 o O9 0° (BO “K 2;: HIE) .. 2E az ivilla-:2 ”.7: ofim El "‘ a H I2< uz< Ea Out: 0E3 a. 85 Acnvrrv AchITV DESCRIPTTON as a» «a as tr. a u INsnTUTEs a RESEARCH CENTERS a': 5 a gig §§, 3 526 8.1 SPECInC PROJECTS ng< < i a E {g a. B m =e mafii 3.3 GENERAL SCHOLARSHIP AND PROFESSIONAL DEVELOPMENT C.l GENERAL PROFESSIONAL SERVICE: ADVICE DIRECTED U '53 OUTSIDE THE INS‘ITI'UTION 32§§ Eéss ‘” " "’ < C.2 EXTENSION SERVICE (NON INSTRUCTIONAL) D.I srt=DENT-ORIENTED SERVICE Id 9. E “.3 0.2 GENERAL PROFESSIONAL IS .13 SERVICE ADVICE DIREtTED z < ; TowARn THE INSTITUTION E25 5; E < DJ SERVICE REPORTS AND RECORDS LEVEL m EI ADMINISTRATIVE ‘2; > DUTIES m Fun 3 a . > “$53 SE 2 Z 25:» gg 5585 E2 COMMITTEE 8 a g < q < PARTICIPATION E LEVEL OF ADMINISTRATIVE AND COMMITTEE ACTIVITIES Code level 1 ............ . . . . ................. Department/ Unit 2 ' ' ......................................... College / School/ Division 3 """ ‘ """"""""""""""" . ............ . . . . ......... . Campuswide ............................................. .Systemwide 202 PLEASE READ TH TO THE LEFT E INSTRUCTIONS ON THIS PAGE AS YOU COMPLETE THE FORM SECTION B: RESEARCH, SCHOLARSHIP, AND CREATIVE WORK ACTIVITIES B.I Institutes and Research Centers: Research scholarshi ' carri ,esea center. (Includes all activities listed in B.2 that are done‘Ioglrnfiggtgt: zomgeflg' if ed on for an institute or rch B.2 Specific Projects: Research. scholarship. and creative work activity related to a specific p'oject. For example: Departmental Research Writing or Develo ' ' rch Propo s g Writing Articles Sponsored Research Admmisterins Research Writing Books Grants Writing Reviews Performm Your PW Giving Recitals Creating New Art Forms fessional ii] Exhibitions 3.3 General Scholarship and Professional Development: All research scholarslii d ' ' current in a professional field. For pie: . p, so creative work activmes related to keeping Reading Articles and Maintaining an Re iewin ' Books Related to Your Artistic S ' Rssleamhgfiosfueagw 5 Profession Attending Profes- Research-Related Discussiai Sionai Meetings with Colleagues SECTION C: PUBLIC SERVICE ACTIVITIES 'fiiis section includes activities that are directed outside the institution [except for those associated with community education (extenswn Instruction) which should be included in A.1.] C-1 General Professional Services/Advice Directed Outside the Institution: Activitia that would not be considered Extension C.2 meant to benefit the community outside the institution. For example: Consulting Community Training Lectures or Seminars for Ad ising Grants the Public v Patient Care C.2 Extension Service: Activities that are directed toward the community outside the institution where fiscal control is shared by the institution and governmait agencies. For example: Agricultural Extaision Urban Extension SECTION D: INTERNAL SERVICE ACTIVITIES This section includes activities related to eneral contact with students. to rmfasional responsibilities within other organizational units Within the institmion, and to fulfilling nstitutionsl requests. For examp e: D.1 Student-Oriented Service: Personal, Career. and Finan- Sponsoring Student Organizations ciai Counseling Preparing Recommendations Meeting with Parents Participation in Social . Interaction Coaching Athletics onadministrative activities related to ormance 0.2 General Professional Service/Advice Directed Toward the institution: All n health clinic, campm architect’s 0 ice. office of of assignments in or consultation to such units as the library. counseling center, the president. Also included are: Interviewing Candidates for Escorting Visitors Positions 0.3 Service Reports and Records: Milling institutional information requests such as: Faculty Service Reports and Preparing Minises Questionnaires Writing and Answering Memoranda Keeping Records SECTION E: ADMINISTRATIVE AND COMMITTEE ACTIVITIES E.1 Administrative Duties. For example: Assigning Faculty Course Load 3:32:333333s3zm m- President or Any cute} inis- Preparing Budgets trative Position Gathering Data Policieis‘mring Personnel Helping During Registration E . 2Committee Participation. For example: Admission Committees Faculty Senate Candidate Selection Planning Committees Committees Budget Committees Code the level of these activities as described at the foot of the form. SELECTED BIBLIOGRAPHY Ad SELECTED BIBLIOGRAPHY Adkins, Chase M., Jr. The Volume and Cost of Instructional Services at Virginia Colleges. Richmond: Virginia State Council of Higher Education, February, 1969. American Council on Education. College and University Business Administration. Revised. Washington, D.C.: American Council on Education, 1968. Anderson, Waldo Keith. "Factors Associated with Instruc— tional Costs in Kansas Public Higher Education.“ Unpublished Ph.D. dissertation, University of Minnesota, 1963. Dissertation Abstracts, Vol. 2507. Anthony, Robert N. "What Should 'Cost' Mean?" Harvard Business Review, XLIX (May-June, 1970), 121. Austin, Philip Edward. "Resource Allocation in Higher Education: A Study of University Costs." Unpub- lished Ph.D. dissertation, Michigan State Uni- versity, 1969. Bailey, Stephen K. "Combating the Efficiency Cultists." Change, V, No. 5 (June, 1973), 9. Balderston, F. E. Cost Analysis in Higher Education. Ford Foundation Program for Research in Uni: versity Administration. Berkeley, Calif.: University of California, July, 1972. Berdahl, Robert O. Statewide Coordination of Higher Education. Washington, D.C.: American Council on Education, 1971. Blackburn, Robert T., and Trowbridge, Keith W. "Faculty Accountability and Faculty WOrkload: A Pre- liminary Cost Analysis of Their Relationship as Revealed by Ph.D. Productivity." Research in Higher Education, I, No. l (1973), 1-12. 203 Bog' 204 Bogue, E. G. "An Inquiry into the Relationship Between Instructional Cost Patterns and Assumptions Influencing Analysis of Basic Data in Unit Cost Studies." Paper presented at the llth Annual Forum of the Association for Institutional Research, 1971. ; Stovall, Thomas F.; and Norman, Brenda. Faculty Time Distribution and Evaluation on Performance, Fall, 1972. A Report Prepared for the Tennessee General Assembly by the Tennessee Higher Education Commission, April, 1973. Bolin, John G., and McMurrain, Tom. Student-Faculty Ratios in Higher Education. Athens: Institute of Higher Education, University of Georgia, 1969. Bolton, Dale L. "Measuring Faculty Load." Improving College and University Teaching. Summer, 1965, pp. 157-580 Bowen, Howard R., and Douglass, Gordon K. Efficiency in Liberal Education. Carnegie Commission on Higher Education. Hightstown, N.J.: McGraw-Hill, 1971. Breneman, David W. Internal Pricing Within the Uni- versity, A Conference Report. Ford Foundation Program for Research in University Administration, P-24. Berkeley, Calif.: University of California, December, 1971. Briner, Conrad. "Administrators and Accountability." Theory Into Practice, October, 1969. Bumpus, Herman C. "Efficiency in the University." School and Society, I, No. 19 (May 8, 1915), 664-67. Business Trends and Progress, 1967 Edition. American Credit Indemnity Company of New York. Toledo, Ohio: Century Press, 1968. California and Western Conference Cost and Statistical Study fOr the Year'l954-55. Berkeley: Uni- versity of California, n.d. Calkins, Ralph Nelson. "The Unit Costs of Programs in Higher Education." Unpublished Ph.D. dissertation, Columbia University, 1963. Dissertation Abstracts, Vol. 24. Ce 205 Campbell, Donald T., and Stanley, Julian C. Experimental and Quasi-Experimental Designs for Research. Chicago: Rand McNally, 1963. Campbell, T. J. Program Cost Allocation in Seven Medical Centers. Washington, D.C.: Association of American Medical Colleges, 1969. Cavanaugh, Alfred D. "A Preliminary Evaluation of Cost Studies in Higher Education." Berkeley: Office of Institutional Research, University of Cali- fornia, October, 1969. Chambers, M. M. Financing Higher Education. Washington, D.C.: The Center for Applied Research in Edu- cation, 1963. Coffelt, John F. Faculty Teaching Loads and Student— Credit-Hour Costs: Oklahoma State System of Higher Education, 1962-63 Academic Year. Oklahoma City: Oklahoma State Regents for Higher Education, n.d. Collier, Douglas J. Higher Education Finance Manual: An Overview. Boulder, Colorado: National Center for Higher Education Management Systems at WICHE, December, 1972. Committee C, American Association of University Professors. "Statement on Faculty WOrkload." AAUP Bulletin, March, 1970, p. 31- Coombs, Philip H., and Hallak, Jacques. Managing Edu- cational Costs. London: Oxford University Press, 1972. Cooper, Lloyd G. "Decision Ability, Not Accountability." Journal of Higher Education, XLIII, No. 8 (November, 1972), 655-60. Cost Study of Dental Education. The American Associ- ation of Dental Schools, 1963-64. Dodds, Harold W. The Academic President: Educator or Caretaker. New York: McGraw—Hill. Doi, James I. "The Analysis of Class Size, Teaching Load and Instructional Salary Costs." College Self Study: Lectures on Institutional Research. Edited by Richard G. Axt and Hall T. Sprague. Boulder, Colorado: Western Interstate Commission for Higher Education, 1960. DO DI 206 Doi, James. "The Use of Faculty Load Data Within an Institution." Faculty WOrk Load: A Conference Report. Edited by Kevin Bunnelil washington, D.C.: American Council on Education, 1960. Dressel, Paul L. "Measurement and Evaluation of Instruc- tional Objectives." The 17th Yearbook, National Council on Measurements Used in EducatiOn. Ames, Iowa: National Council on Measurements Used in Education, 1960. ; Johnson, F. Craig; and Marcus, Philip M. The Confidence Crisis. San Francisco: Jossey-Bass, 1970. Duncan, Merlin George. "Instructional Unit Costs in Selected Central American Universities." Unpub- lished Ph.D. dissertation, Michigan State Uni- versity, 1964. Eckert, Ruth E. "College and Universities-Programs." Encyclopedia of Educational Research. Edited by Chester W. Harris. New York: The Macmillan Company, 1960. , and Williams, Howard Y. College Faculty View Themselves and Their Jobs. College of EducatiOn, Minneapolis: University of Minnesota, 1972. Etzioni, Amitai. Modern Organization. Englewood Cliffs, N.J.: PrentiCe—Hall, 1964. , ed. Readings on Modern Organizations. Eng1e~ wood Cliffs, N.J.: Prentice—Hall, 1969. Evans, John M., and Hicks, John W. An Approach to Higher Educational Cost Analysis. Studies in Higher Education, XCI. Lafayette, Indiana: Purdue University, 1961. Faculty Activity and Outcome Survey. Boulder, Colorado: National Center for Higher Education Management Systems at WICHE, 1972. Falvey, Francis E. Student Participation in College Administration. New York: Bureau of Publi; cations, Teachers College, Columbia University, 1952. A ii“ —‘_ 1 207 Freeman, Thomas Mason. "A Multiple Correlation Analysis of the Supplies and Services General Fund Budgets for Selected Academic Departments at Michigan State University: 1964-65 and 1965-66." Unpub- lished Ph.D. dissertation, Michigan State Uni- versity, 1967. Grady, Paul. Inventory of Generally Accepted Accounting Principles for Business Enterprises. New York: American Institute of Certified Public Accountants, 1965. Guilford, J. P. Fundamental Statistics in Psychology and Education. New York: McGraw—Hill, 1956. Harris, Seymour E. Higher Education: Resources and Finance. New York: McGraw-Hill, 1962. Harris Committee Report. Robert Harris, chairman. Washington, D.C.: Department of Health, Edu- cation, and Welfare, Office of the Assistant Secretary Comptroller, December, 1971. Haskins, Homer Charles. The Rise of Universities. Ithaca, N.Y.: Cornell University Press, 1957. Heffernan, James M. "The Credibility of the Credit Hour: The History, Use and Shortcomings of the Credit System." Journal of Higher Education, XLIV (January, 1973), 61-72. Heilman, J. D. "Methods of Reporting the College Teacher's Load and Administrative Efficiency." Educational Administration and Supervision, XI, No. 3 (March, 1925}, 167-87. Henle, R. J. "To Devise and Test Simplified Adequate Systems of Measuring and Reporting Financial, Manpower, Facilities, Research, and Other Activities in Colleges and Universities, A Final Report." National Science Foundation and National Institute of Health, June, 1965. Henry, David D. "Accountability: To Whom, For What, By What Means?" Educational Record, Fall, 1972. Hicks, John W. "Faculty Workload—-An Overview." Faculty Werk Load: A Conference Report. Edited by KeVin Bunnel. Washington, D.C.: American Council on Education, 1960. H! 208 Hobbs, Mary T. "Teaching Loads in Selected Liberal Arts Colleges." Liberal Education, December, 1966, pp. 419-210 Hodgkinson, Harold L. "How Can We Measure the 'Value Added' to Students by a College Education?" The Chronical of Higher Education, November 13, 1972. Hornagren, Charles T. Cost Accounting: A Managerial Emphasis. Englewood Cliffs, N.J.: Prentice—Hall, 1962. Hubbard, Robert E. "An Approach to Instructional Cost Hull, Analysis." Basis for Decision. Edited by L. J. Lins. Madison, Wisconsin: Dembar Educational Research Service, 1963. L. E. "Pitfalls in the Use of Unit-Cost Studies." Journal of Higher Education, XXXII (October, 1961). , and McWhirter, D. A. Unit Cost Analysis Pro- cedure, Indiana University. Bloomington, Indiana: Bureau of Institutional Research, Indiana Uni- versity and Indiana University Foundation, May, 1964. Ikenberry, Stanley. "Instructional Cost and Quality." College and University, XXLVII, No. 3 (Spring, 1962). Jackson, A. J., ed. Professions and Professionalization. London: Cambridge University Press, 1970. Johnson, Walter F. "Some Comments on Efficient College Management." Review of Efficient College Manage- ment, by William W. Jellema, in Educational Administrative Quarterly, Spring, 1973. Jones, Gardner, and Cunningham, Wayne. "Cost Analysis Judy, of Instruction by Closed Circuit Television." East Lansing: Michigan State University, December 9, 1964. (Mimeographed.) Richard W. A Research Progress Report on: Systems Analysis for Efficient Resource Allocation in Higher Education. Toronto, Canada: Institute for the Quantitative Analysis of Social and Economic Policy, University of Toronto, January, 1970. Kee 209 Keezer, Dexter M., ed. Financing Higher Education 1960- 12. New York: McGraw—Hill, 1959. Keller, John. "Higher Education Objectives: Measures of Performance and Effectiveness." Management Information Systems. Boulder, Colorado: The National Center for Higher Education Management Systems at WICHE, October, 1969. Kettler, Raymond W. "The Analysis of Class Size, Teach- ing Load and Instructional Costs." College Self Study: Lectures on Institutional Research. Edited by Richard G. Axt and Hall T. Sprague. Boulder, Colorado: Western Interstate Com- mission for Higher Education, 1960. Knowles, Asa S., and White, William C. "Educational Research and Statistics: A New Approach to the Evaluation of Faculty Loads." School and Society, XLIX (May 27, 1939), 682-84. Koenig, Herman E.; Keeney, Martin G.; and Zemach, R. A Systems Model for Management, Planning, and Resource Allocation in Institutions of Higher Education. East Lansing: Division of Engineering Research, Michigan State University, September 30, 1968. Koos, Leonard V. "Adjustment of the Teaching Load in a University." U.S. Bureau of Education Bulletin, No. 15, 1919. Cited by Stevens, Edwin B., and Elliott, Edward C. Unit Costs of Higher Edu- cation. New York: The MacMillan Company, I925. Light, D. W., Jr.; Marsden, L. R.; and Corl, T. C. The Impact of the Academic Revolution on Faculty—_” Careers. Washington, D.C.: American Association for Higher Education, February, 1973. Lorimer, Margaret F. "How Much Is a Credit Hour?" Journal of Higher Education, XXLIII (June, 1962), 302-06. Manning, Charles W. Faculty Activity Analysis: Pro- cedures Manual. Boulder, Colorado: National Center for Higher Education Management Systems at WICHE, April 12, 1973. (Mimeographed.) Marascuilo, Leonard A. Statistical Methods for Behavioral Science Research. New York: McGraw—Hill, 1971. 14ch Meh Mi! Mi 210 . McGrath, Earl J. Memo to a College Faculty Member. New i York: Teachers College, Columbia University, 1961. Mehrens, William.A., and Lehmann, Irvin J. Measurement and Evaluation in Education and Psychology. New York: Holt, Rinehart and Winston, 1973. Michigan State University. Salary Budget 1972-73. East Lansing: Michigan State University, 1972. Miller, James L., Jr. State Budgeting for Higher Edu- cation: The Use of Formulas and Cost Analysis. Michigan Governmental Study No. 45, Ann Arbor: Institute of Public Administration, University of Michigan, 1964. Miller, Richard I. Evaluating Faculty Performance. San Francisco: Jossey-Bass, 1972. Millett, John D. Financing Current Operations of American Higher Education. Washington, D.C.: Management Division, Academy for Educational Development, Inc., December, 1972. Minter, John, and Lawrence, Ben, ed. Management Infor- mation Systems: Their Development and Use in Higher Education. Boulder, Colorado: Western Interstate Commission for Higher Education, 1969. Moore, Carl L., and Jaedicke, Robert K. Managerial Accounting. Cincinnati, Ohio: South-Western, 1967. National Committee on Standard Reports for Institutions of Higher Education. Financial Reports for Colleges and UniversitiES. Lloyd Morey, chairman. Chicago: University of Chicago Press, 1935. O'Neill, June. Resource Use in Higher Education: Trends in Output and Inputs, 1930 to 1967. Berkeley: Carnegie Commission of Higher Education, 1971. Peltier, Lynn H. "Faculty Activity Analysis." East Lansing: Office of Institutional Research, Michigan State University, July, 1971. , and Ingall, David L. "Costing Degree Programs." A pilot study for the Michigan Council of State College Presidents, Lansing, Michigan, May, 1971. PrO‘ Ree Re I 211 Provous, Malcolm. Discrepancy Evaluation. Berkeley: McCutchan, 1971. Reeves, Floyd W.; Henry, Nelson B.; Kelley, Frederick J.; Klein, Arthur J.; and Russell, John Dale. The University Faculty. Chicago: The UniversiE§_ of Chicago Press, 1933. Reeves, Floyd W.; Henry, Nelson B.; and Russell, John Dale. Class Size and University Costs. Chicago: University of Chicago Press, 1933. Romney, Leonard C. Faculty Activity Analysis: Overview and Major Issues. Boulder, Colorado: National Center for Higher Education Management Systems at WICHE, 1971. Rourke, Frank, and Brooks, Glen. Managerial Revolution in Higher Education. Baltimore: Johns Hopkins Press, 1966. Rudolph, Frederick. The American College and University: A History. New York: McGraw-Hill, 1959. Russell, John Dale. The Financing of Higher Education. Chicago: The University of Chicago Press, 1954. , and Doi, James I. "Analysis of Institutional Expenditures." A series of twelve articles on finance. College and University Business, XIX, No. 3 (September, 1955), 19-21; No. 4 (October, 1955), 27-29; No. 5 (November, 1955), 44-47; No. 6 (December, 1955), 39-41; XX, No. 1 (January, 1956), 41—45; No. 2 (February, 1956), 47-51; No. 3 (March, 1956), 41-43; No. 4 (April, 1956), 35-37; No. 5 (May, 1956), 47-48; No. 6 (June, 1956), 48-51; XXI, No. 1 (July, 1956), 43-46; No. 2 (August, 1956), 42-47. Samuelson, Paul A. Economics. New York: McGraw-Hill, 1970. Saupe, Joe L. "Relationships Among Various Measures of Instructional Load." East Lansing: Office of Institutional Research, Michigan State University, April 20, 1961. Scheerer, Anne E. "A Formula for Determining Instruc- tional Costs." College Management, November, 1972, pp. 25-29. Se: Sh (I) Sexson, Jack E. "A Method for Computing Faculty Loads.“ Improving College and University Teaching, Autumn, 1967, pp. 219—22. 212 Sherer, Harvey. "U.I.C.'s Are Dangerous." College and University Business, XVIII, No. 2 (February, 1955). Siegel, Barry N. "Costing Students in Higher Education—- A Case Study, A Progress Report.“ Eugene, Oregon: University of Oregon, August, 1967. ERIC Document, ED 014143. Siegel, Sidney. Nonparametric Statistics for the Behavioral Sciences. New York: McGraw-Hill, 1956. Simonds, Rollin H. "To Increase Man-Hour Output in ' Higher Education." The Educational Record, ' XXXIX, No. 4 (October, 19587, 332—39. Snedecor, George W., and Cochran, William G. Statistical Methods. Ames, Iowa: The Iowa State UniverSity Press, 1967. State of Michigan. 76th Legislature Regular Session of 1972. Enrolled Senate Bill No. 1141. Stecklein, John E. "Research on the Faculty." College Self Study: Lectures on Institutional Research. Edited by Richard G. Axt and Hall T. Sprague. Boulder, Colorado: Western Interstate Commission for Higher Education, 1960. Stevens, Edwin B., and Elliott, Edward C. Unit Costs of Higher Education. The Educational Finance / Inquiry (Series), XIII. New York: The Mac- Millan Company, 1925. I Stuart, Douglas Allen. "The Application of Formula and Cost Analysis Procedures to the Budgeting of Academic Departments." Unpublished Ph.D. disser- tation, Michigan State University, 1966. j l ! Sullivan, Patrick H. PA Study on Bias in Faculty Reports / of Time and Effort Expenditure." Paper presented / at the 1973 Convention of the Association for / Institutional Research, Vancouver, Canada, May, / 1973. 213 The Pennsylvania State University. “Report of Enrollment, Student Credit Hours, Course Credits and Faculty Activity, Academic Year 1972-73." University Park: Office of the President, The Pennsylvania State University, March 30, 1973. Toombs, William. Productivity and the Academy: The Current Condition. Center for the Study of Higher Education, Report 16. University Park, Pennsylvania: The Pennsylvania State University, April, 1972. . Productivity: Burden of Success. ERIC/Higher f Education Research Report No. 2. Washington, ‘j D.C.: American Association for Higher Edu- i cation, May, 1973. | I | Tyndall, D. Gordon, and Barnes, Grant A. "Unit Costs of Instruction in Higher Education." Basis for Decision. Edited by L. J. Lins. Madison, WisconSin: Dembar Educational Research Service, 1963. University of Michigan. "Report on the Pilot Test of the NCHEMS Faculty Activity and Outcome Survey at the University of Michigan." Ann Arbor: Office of Institutional Research, University of Michigan, January 29, 1973. Weathersby, George B., and Balderston, Frederick E. PPBS In Higher Education Planning and Management. Ford Foundation Program for Research in Uni- versity Administration, P-31. Berkeley, Calif.: University of California, May, 1972. Western Interstate Commission for Higher Education. Yardsticks and Formulas in University Budgeting. Boulder, Colorado: Western Interstate Commission for Higher Education, February, 1959. Williams, Robert L. The Faculty WOrk Load — Alternate Methods of Evaluation. Report No. 19. Denver, Colorado: Education Commission of the States, December, 1970. Wilson, Logan. "Analyzing and Evaluating Costs in Higher Education." The Educational Record, XXXXII, No. 2 (April, 1961), 99—105. Withey, Howard A. "Financial Reporting for Nonprofit Organizations." The Journal of Accountancy (December, 1967), 40—53. 214 Witmer, David R. Unit Cost Studies. Madison: Wisconsin Board of Regents of State Universities, 1967. E.D. 013492. Young, William L. "Six Criteria Form a Composite Profile ' Chart of Faculty Load." College and University Business, April, 1964, pp. 59—60. Zeimer, Gordon; Young, Michael; and Topping, James. Cost Finding Principles and Procedures. Boulder, Colorado: National Center for Higher Education Management Systems at WICHE, November, 1971. Ill1|HMHHIIHIIHHHWlilwlmlll\llwimllmg ‘5 - ' '