MSU LIBRARIES .-:_ RETURNING MATERIALS: Place in book drop to remove this checkout from your record. .FINES will be charged if book is returned after the date stamped be10w. l l - . -_... WI“ :55 Lite COST-EFFECTIVE STRATEGIES FOR IMPLEMENTING A COMMUNITY GROUP HOME CURRICULUM TRAINING By Franklyn Giampa A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements. for the degree of DOCTOR OF PHILOSOPHY Department of Educational Administration 1987 Copyright by FRANKLYN GIAMPA l987 ABSTRACT COST-EFFECTIVE STRATEGIES FOR IMPLEMENTING A COMMUNITY GROUP HOME CURRICULUM TRAINING BY Franklyn Giampa Public education and the public mental health system, as a part of their mission, have a responsibility to provide services that meet community needs. The~ costs involved in operating organizational units that deliver these services to the public have been continuing to increase. This researcher investigated how collaboration and planning between higher education and the public mental health system can provide cost-effective training for paraprofessionals who work in community group homes for the developmentally disabled and mentally ill. Costs, staff turnover, and content retention over time were variables analyzed at two instruction sites: a community college and a Michigan Department of Mental Health (MDMH) facility. Sixty individuals were randomly selected from a waiting list to be trained as group home workers. Thirty received their instruction at an MDMH facility, the other 30 at Schoolcraft Community College. The tool used to assess the trainee’s mastery of the course material was the assessment package for the Group Home Curriculum. After four Franklyn Giampa months, individuals still working in the group homes were asked to return to the facility for retesting on the same materials. The results were analyzed as a 2 x 2 factorial experiment. There were two factors of interest (instructional site and time of testing) and two levels of each factor (the MDMH facility versus community college for instructional site, original test versus retesting for the time factor). Analysis of variance was used to test hypotheses concerning quality-of—treatment means and to check for interactions between the two factors (instruction site and time). Test scores at both sites were fairly similar over time. In Module 48, test scores dropped over time for both instructional sites. For Module 5, scores were lower on the retest at the MDMH, whereas scores actually rose at the community college. Turnover rates differed between instructional sites. Total cost per trainee for the Schoolcraft Community College instructional site was $199. The cost per trainee at the MDMH site, as computed by the retrospective method, was 5135, whereas the cost per trainee, as computed by the prospective method, was $l22. DEDICATION To Mother and Dad: Thanks for being there through all the difficult times, from Shaw Avenue to all the other schools in between, and finally, it happened!!! To Sandy: Without your support, love, and even temper, this most definitely would never have happened. ACKNOWLEDGMENTS Appreciation and thanks to the members of my doctoral committee, who stood by me through this long, long endeavor. They are Drs. Howard Hickey, Lawrence Lezotte, Halter Johnson, and van Johnson. Dr. Walter Johnson’s wisdom and insight kept the research alive, and Dr. Lezotte’s patience, expertise, and understanding made it happen. To employees of the Michigan public mental health system, your cooperation is greatly appreciated, especially Jerry Provencal, Marcia Tessiser, and the 0RD staff. Without support from the National Institute of Mental Health, HRD Division, this research would not have occurred. Sue Cooley’s expertise saved me more hours and aggravation than she will ever realize. A special thanks to Dr. Owen Anderson, who worked with me many hours trying to explain, explain, and re-explain. Above all, one person put the time in when I needed it the most. He used his expertise unselfishly, and I owe him a great deal. Thanks, Dr. James M. Gardner. vi TABLE OF CONTENTS LIST OF TABLES ......................... LIST OF FIGURES ........................ LIST OF APPENDICES ....................... Chapter I. INTRODUCTION ...................... The Problem ..................... The Higher Education System ............. Background of the Research .............. Human Services in Michigan .............. II. LITERATURE REVIEW ................... Generalist Training Programs ............. Specialist Training Programs ............. Research Evidence .................. III. DESIGN AND METHODOLOGY ................. Method ........................ Limitations and Delimitations ............ Subjects ....................... MDMH Training .................... Community College Training .............. Materials ...................... IV. RESULTS ........................ Turnover ....................... Effectiveness .................... Costs ........................ vii Page ix xi CONN-d IS 16 T9 22 26 SUMMARY, CONCLUSIONS, RECOMMENDATIONS, AND REFLECTIONS ..................... 43 Summary ....................... 43 Findings and Conclusions ............... 45 Turnover ...................... 45 Effectiveness ................... 46 Cost ........................ 48 Recommendations ................... 50 Recommendations for Further Research ........ 50 Recommendations for Change and Future Direction . . 51 Reflections ..................... 52 APPENDICES ........................... 55 BIBLIOGRAPHY .......................... l4l viii Table #00“) H acumen-boom LIST OF TABLES Training Sequence ................... Number of Individuals Hho Completed Testing ...... Fate of Individuals Not Retested ............ Comparison of Test Scores Between Schoolcraft fiE'STIR‘T’F‘."Teie.“'.“‘.“.‘°.°?"?"?'“‘?"’f it"??? . . . . Comparison of Test Scores Between Original Test Date and Retest Dates for Each Module ........... Mean Test Scores for Each Treatment Combination and Module ........................ Total Cost Report, by Program ............. ANOVA for Module l ................... ANOVA for Module 2 ................... ANOVA for Module 3 ................... ANOVA for Module 4A .................. ANOVA for Module 48 .................. ANOVA for Module 5 ................... ANOVA for Module 6 .............. ‘ ..... ANOVA for Module 7 ................... ix 37 37 38 4l l35 l35 I36 137 I37 Figure LIST OF FIGURES Page Effects of Inflation on Department Budget, Using Detroit Consumer Price Index Inflation Rate ..... l0 Organizational Chart: Michigan Department of Mental Health .................... ll Organizational Chart: Office of Resource Development . l3 Module 48 Test Scores ................. 39 Module 5 Test Scores .................. 39 Appendix A. Immmcnw LIST OF APPENDICES GONGHER NEWS SERVICE EXCERPT REGARDING HUMAN SERVICE AND FINAL 1983-84 GENERAL FUND BUDGET ..... PORTION OF POSITION DESCRIPTION FOR ORD DIRECTOR . . . . GROUP HOME CURRICULUM, PART I ............. MDMH PRIORITY SERVICE GOAL STATEMENT .......... LETTERS OF AGREEMENT WITH COMMUNITY COLLEGE ...... APPLICATION OF GRANT SUBMITTED TO NIMH ......... ASSESSMENT PACKAGE OF THE GROUP HOME CURRICULUM CALCULATION OF RETROSPECTIVE AND PROSPECTIVE METHODS . . ANOVA TABLES: MODULES l, 2, 3, 4A, 4B, 5, 6, 7 DEFINITION OF TERMS .................. xi Page 55 57 59 7l 72 79 8O l30 l35 CHAPTER I INTRODUCTION Public education and the public mental health system, as a part of their mission, have a responsibility to provide services that meet community needs, yet the costs, both direct and indirect, continue to increase. As the tax base diminishes and state and federal deficits increase, the competition for limited resources escalates among service providers. To complicate matters, the legal constraints concerning rights to an education, freedom of access, and so on, add to the cost of delivering services to an already stressed public system. This study was undertaken to investigate how collaboration and planning between higher education and the public mental health system can provide cost-effective services and enhance both systems. The Problem The Michigan Department of Mental Health (MDMH) has an increasing need to develop cost-effective» methods for delivering services to its clients. As state and federal dollars diminish, delivered units of service have decreased, and the quality of those services has also been reduced. The MDMH is continually making a concerted effort to reduce administrative costs as an alternative to program reduction. These efforts were successful by reducing the MDMH original 1981-82 state- general-fund-supported administrative expenses of $31.6 million to the 1982-83 appropriated level of $25 million--a reduction of 21%. Central administrative staff, located for the most part in Lansing, was reduced from a high of 750 staff to 373 in l982-83. This reduction in staff greatly decreased the MDMH’s ability to manage a complex institutional and comunity-based system. This reduction most directly affected the Office of Resource Development (0RD), which coordinates the education and training efforts for the MDMH. The Higher Education System Higher education has gone through a similar shortfall of operating funds. As of June 7, 1983, the Michigan Senate rejected a $763.5 million spending measure for higher education. The Governor recommended $706.6 million, which seems to be a considerable increase over the l984 budget of $652.2 million until an inflation factor is considered. Over the past decade, state support for higher education in Michigan has declined by ll percent in real terms, according to a survey by the Chronicle of Higher Education. Only in Illinois have state funds declined as steeply. As a result, Michigan, which was once one of the top four or five states in the country in terms of per capita support for higher education, now ranks 38th. Most of the reductions, moreover, have come in the past few years and they have often been made in the form of emergency recissions with virtually no advance notice. ("Downsizing at the University of Michigan," 1983, p. 283) Background of the Researeh Although resources continue to diminish and the fiscal crisis in Michigan is still present, the MDMH has determined that community residential placement and deinstitutionalization are, and will con- tinue to be, a top priority as referenced in the department’s goal and mission statement contained in the l981-1986 Program Policy Guidelines. If this public policy is to be successfully implemented, more effective use of existing resources is needed. The purposes of this field test were to (a) determine whether two public agencies with public service missions can provide quality training in a cost-effective manner and (b) improve the management of resources by the MDMH. This researcher field tested the training of staff working in group homes and identified the strengths and weaknesses of training sites: a state facility and a community college. The results of this research could influence the planning cycle and the distribution of resources managed by the MDMH. More specifically, for the first time, the ORD could formally evaluate different strategies for disseminating training curricula (state facility and community college). Collaborative efforts from all public agencies are necessary if one is to (a) identify the most cost-effective delivery system to carry out the necessary public services, (b) ensure appropriate deployment and use of resources, and (c) ensure competency of service providers. The researcher specifically addressed the first and second items and touched tangentially on the third. The cost of higher education is also not a new issue. Controlling rapidly rising costs has had an increasing influence on parents, students, unemployed/dislocated workers, and public officials. The evolution of cost issues and cost-containment strategies in the health care sector serves as an instructive precedent for evaluating cost-containment strategies for higher education. Anthony H. Morgan (1983), Assistant Professor of Educa- tional Administration at the University of Utah, stated: Drawing lessons from the health care sector appears to have prima facie validity. Each enterprise is built around a core of semi-autonomous professionals who traditionally have determined the conditions of and evaluated their work. The role of administrator is equally tenuous in educational and health institutions, at least from the perspective of the core professionals. Quality of service is a central, albeit cloudy, issue in both sectors as is duplication of services and the allocation of resources. Incentives prevalent in health care institutions, particularly hospitals, and institutions of higher education are strikingly similar. Each generally strives for a comprehensive array of programs or services, for the latest technology, and for general organizational goals of excellence, quality,’ and prestige. As Howard Bowen has pointed out, the cost incentives inherent in such organizational goals are not counteracted by incentives leading to parsimony or efficiency; that is, each institutional type tends to spend up to the very limit of its means. While these and other parallels are perhaps more clearly drawn for universities than for community colleges and some types of four-year institutions, the author believes that the possible unfolding of events and policies outlined will develop at sectors of higher education, regardless of real or perceived differences. (pp. 279-83) One of the factors affecting health and higher educational costs over the past decades has been an increased demand for those services. In 1978, almost 1.5 million students were enrolled in 6,813 private and 812 public vocational schools in the United States, offering programs in such areas as business and commerce, cosmetology, flight training, art and design, and health-related occupations. Between 1976 and 1978, postsecondary occupational school enrollment increased at a rate of 6.9%, compared to a 2.3% growth in college enrollment ("Post Secondary Enrollment Responses," 1983). Also, community/junior colleges in 1983 enrolled a significant 53% of all freshmen and sophomores in the nation (Montessi, 1983). Another factor influencing cost of higher education is the enrollment decline projected for the 19805 and 19905, which also increases unit costs for institutions of higher education. Costs generally cannot be cut, at least in the short run, in direct propor- tion to enrollment declines. Morgan (1983) proposed three general policy strategies that should be considered by higher education policy' makers when dealing with these problems: (a) a central coordination or planning approach, (b) a variety of regulatory strategies, and (c) market strategies. This researcher addressed the responsibility of higher education to fulfill its mission by meeting the training demands of other markets, namely the public human service agencies serving the community. Traditionally, departments of state government have not been one of the arenas with which public education has been involved to any great extent. It should be the responsibility of leadership in higher' education and state government to investigate! ways of delivering service units in the most cost-effective manner. Their collaboration and sharing of resources may result in direct benefits to both participating organizations and the communities they serve. Some experts in higher education have been bold enough to state that "the job of the admission office should be to identify a large number of prospects within the primary market" (Hofford & Timmerman, 1982, p. 49). This market may include state government’s staff development and training needs. A marketing—information system is a continuing and interacting structure of people, equipment, and procedures designed to gather, sort, analyze, evaluate, and distribute pertinent, timely, and accurate information for use by the marketing decision makers to improve their marketing planning, execution, and control (Kotter, 1980). Continuing to provide a high quality of service in a highly competitive, limited-resource environment, marketing strategies must be employed in conjunction with cost-containment approaches. Lolli and Scannell (1983) stated that: Commitment, resources, and technical skill represent the neces- sary components of a meaningful market research endeavor. The extent to which each is present has an impact on both the quan- tity and quality' of the outcome. Those institutions which recognize the need for marketing research and plan accordingly will be in the best position for dealing proactively with the changing environment of the ’805. (p. 150) Recognizing the need to examine additional markets, i.e., public mental health employees, may increase the support for higher education dollars. Human Services in Michigan Since this research involved collaboration and networking with postsecondary and human service agencies, a description of the human service system and its confronting crises/problems is necessary. The recession and resultant nearly 36 months of double-digit unemployment have had a dual effect on Michigan’s human service system. At precisely the same time that families and individuals bearing the brunt of the recession turned to public human services, often for the first time, federal and state revenues necessary to support these services have drastically declined. What resulted was a reduction in the state’s ability to respond to legitimate requests for mental health services, income maintenance, and other social services. Neither the 300 percent increase in the caseload receiving Aid to Dependent Children in the last three years (1980-83) nor the dramatic change in the proportion of involuntary commitments to state institutions for the mentally ill should be a surprise to anyone. During 1982-83, the unemployment rate increased to 17.2 percent. Perhaps the best study of the relationship of the economy’s health to the social and mental health of the popula- tion was completed by Dr. Harry Brenner of Johns Hopkins Univer- sity. Dr. Brenner found that, for every one percent rise in the national unemployment, 39,000 deaths could be expected, includ- ing: --20,000 from cardiovascular problems --900 from suicides --500 from cirrhosis of the liver --650 from homicides. (Report of the Financial Crisis Council Expenditure Committee, 1982) While Michigan’s Department of Mental Health state facilities have not had the resources to conduct research concerning the social effect of the economic crises during the past three years, evidence has indicated that the loss of employment for thousands of workers has had serious effects. For example: 1. While overall admissions to state psychiatric facilities have decreased slightly since 1979-80 due to restrictive admission policies, the facilities serving high-unemployment areas have had increased admissions (i.e., Northville--5% in 1982). 2. The Michigan Department of Public Health reports that infant mortality has increased from 12.8 deaths per 1,000 in 1980 to 13.2 per 1,000 in 1981, reversing a 30-year trend. 3. The Cabinet Level Hunger Task Force, which was co-chaired by William Long (Director of the Michigan Department of Labor) and C. Patrick Babcock (Director of the MDMH), received reports from numerous public and private agencies regarding increased demand. Probablythe most dramatic were from (a) Detroit, where the Mayor estimates as many as 8,000 individuals are homeless, and a growing number of people are turning to soup lines or emergency food supplies; and (b) Lansing, where the emergency food bank has experienced an increase in demand from 2,550 people in 1981 to 9,681 in l982--a 279% increase. Eighty percent of the monthly participants are requesting assistance for the first time. Unfortunately, these are not isolated examples. A survey of any sample of public or private human service agency will provide more indicators that unemployed and low-income citizens in Michigan have experienced, and are continuing to experience, serious income, social, and health problems. Given the increased demand for services coupled with the state’s general-fund budget-reduction plan, budget reductions for mental health, education (including higher education and school aid payments), and social services were inevitable. Together, these departments comprise 76% of the general-fund budget. The legislature and executive branch have tried to protect basic services during the years. In spite of these efforts, of the over $770 million in Executive Order reductions in 1981-82, approximately 65% was from mental health, education, and social services. The experience for the MDMH clearly indicates the effect of reduced state revenues. As outlined in Figure 1, the data indicated that the 1982-83 MDMH budget has increased by $130.9 million (32.7%) since 1978-79. However, had the 1978-79 budget been adjusted purely for inflation, the MDMH’s budget in l982-83 would have totaled $593.3 million. The result is a $57.6 million shortfall from the actual purchasing power four years ago. (See Appendix A for the most recent update.) On the positive side, this reduction in purchasing power has forced the MDMH to prioritize its service better and to find more cost-effective alternatives to traditional services (e.g., psychiat- ric, continuing education, training, and so on). However, the quality of these services is influenced significantly by the caliber of employees the MDMH can attract and the continuing staff- development efforts offered through its training programs. The educational and training programs are coordinated by the 0RD. Organizationally, the Director of the MDMH is appointed by the Governor of Michigan. The Director’s responsibility is to develop a system of public mental health delivery that is consistent with the legislative mandate as prescribed in the Michigan Mental Health Code. As the organizational chart indicates (Figure 2), the management goals and service-delivery priorities are implemented through the four Deputy Directors, with primary operational responsibility assigned to the Chief Deputy Director. 10 Million GF-GP 600 ‘ .///’Inflat10n 550 8 500 v 450 i 400 v 350 o 300 i , . . : FY FY FY FY - FY 78-79 79-80 80-81 81-82 82-83 Figure 1.--Effects of inflation on department budget, using Detroit consumer price index inflation rate. 11 .cppwm: Fmpcmz mo “coaucmawo cmmwcovz .acmsu chowumecmmgo--.m oczmwm mcLEF< vco mwmaa .mmuw>cmm gupowx Co mu_wwo .uumWOLa ugocgam >u_c:EEou. Luzos>_a .Usz muum>aceo all... .1111“ AYYYYYYYYYYI . auYYYYYYYYYj a c all . a w u _ . L . . . . .. _ _ m auwcmo _ fl . . _ fl . - L . . use M . mmu_>cmm _ H mmpu,_ruco M _ co_m>mo “ .wo_;umm” u ” mpcupcmo: h _ coF¢>wc ” _mwrmmox . womano ccwmm . m .I.:.u . . 8.. o . outscwmm , _ .a . 8.:atup_;u* “ .mcaum a I o o p wmm_um_wucu mo “mo ammgam pmwuummpw - - mo muwmmo *-. maphmwmam _ co ammcsm ’ .coca mo " Co amocam _ wcouumcwo “ Legumcmo u .o H Louumgvo . . Louumcpo _ Louumcwo _ cowmw>wa " couumcwo r .Y . All _ P _ . . _ _ r, . H L L m p L ., _ . _ . . z _ . _ _ _ w _ . a _ _ _ _ . _ . n u . 4 . c . w “.m>m .Lomcw .omcmxw .- - -L «mzc .amu s w n » mc_mmm< mmuw>gwm m>wumgum_cmsv< “ fl. mwuw>me _ wow—nan can m>_ua_m_om4 Lo» H “_mu_vmz new Fou_:_pu _ Louumxmo muaamo Cmrzum w Low Louumcwo ausamo _ Louumcwo zuaawo W _ Low Louuwcwo xuaamo m R r r . i. M . _ M mogcrm u=m_acoma co moLLCOb . u . - - - L __u==ou so _ . _ . r _ m m=o_ua_mm Lona; m _ m uco mmuw>cmm pwccomcma mo mummwo b . “ co_uu< m>_uweLwLL< use mcowuo_wau . _ I _ . . . m _oucmEch>oacmuc_ mo mu_mwodlllllL. YIIIILIIIIII. _ _ mmpuwszoc >Lom_>u< * Lauumc_o W _ W mmur>cmm _mu_vmz w pmuwcwku. . _ _ W . .. w mmur>me_ “ _ _ m . _ _o_ucmcwmmm >u_c:ECOU mo muwmwom . _ _ I + . . . . Loccm>oc M IIYIII. Frucaou acomw>umo umuvzm Lo movemo filllllt YYYYYYTYYIYYY " ;u_owx _aucm: .mcmu_u_u. . . . _ 12 The MDMH is responsible for planning, program development, management, evaluation, and necessary support functions on a statewide level. Administrative responsibility for state facilities and community mental health service boards (CMHSBs) is managed through this bureau structure. The 0RD, which is also Michigan’s State Manpower Development (SMD) Office, reports to the Office of the Chief Deputy Director (Figure 3). As mandated by the National Institute of Mental Health, Health and Human Services (NIMH, HHS), a major emphasis of the SMD Office is to network with higher education agencies for the purpose of developing and disseminating successful manpower strategies to the field. A number of joint planning, program, and curriculum- development projects with the University of Michigan, Western Michigan University, and Michigan State University are in process. The 0RD is also responsible for the development, retention, distribution, and use of a competent work force (Appendix B). 0RD staff work closely with decision makers at all levels to establish these capacities throughout the system. Primary attention is being given to converting validated information and usable alternatives into the year-to-year budget, decision model, program prioritization, and planning process. The 0RD offers consultation and support to training and resource-development staff in the field regarding specialized curriculum preparation and training of trainers, using materials from a variety of resources. The 0RD engages in planning, standards development, and continuing education for a variety of persons within l3 .uom m. H114 oao .pcwsao—m>mo mugzommm mo mummwo wkm — P up; P mpg — N fiIIIIIIIL mucaumwmm< acuczum ocwc_mgp mo mw_;u H acoEWmomm< _. _ a ucoEooacmz _ umm_ mucacw .0 ca_=u omo _ mpg — ”acmsu _mcomum~wcmmgouu.m mgamwu wku F c mku p um,_a_uacv _occomgma L meowum_ma gone; a mouw>cwm _ Louumcwo xuaaoo _ .ummwlll e Louumcwo ucosao_o>mo mUL30mom mo muwwwo . zzo L Louumcwo xuaaoo um_cu :zo a Louumcwo _mccomcma Lo mowmwc “ ISO _ ;u_eoz —mu=wz mo “coaucoomo . b all coccm>ou e=_=uLLL=u Lo Co_;u .oum ago m. - cowmvsvo _ mmuw>gum cowumscomcm zzo 14 the public mental health system. Assessing the work force has been a function of the 0RD since its designation as the SMD Office. For example, the 0RD has been given responsibility to implement a program to facilitate re-employment of MDMH employees who have been displaced due to deinstitutionalization and the poor economy. Recently, the 0RD published the Community Group Home Curriculum, Part I (Appendix C), which has been used to train paraprofessionals (residential program aides) providing services to MDMH clients in community group homes. These newly published materials support the MDMH’s major program direction and specifically support MDMH goal statements 6, 7, 11, and 12 (Appendix 0). These goal statements support quality of service delivery in community residential settings and the downsizing of state psychiatric hospitals and developmental- disabilities centers. The success with which the MDMH meets these goals is contingent on the quality of the training effort put forth by the 0RD. With this as the impetus for the study, the researcher proposed to answer the following questions: 1. What is the difference in cost when offering the new Community Group Home Curriculum at a community college versus an MDMH facility? 2. What are the differences in mean assessment scores for each training module for each site (i.e., Schoolcraft Community College and an MDMH facility)? 3. How do the original assessment scores for each paraprofes- sional compare with a four-month follow-up assessment score for each training site and each training module? 15 4. What is the turnover rate for trainees at each training site over a four-month period? CHAPTER II LITERATURE REVIEW This chapter initially provides the reader with a summary of general training programs for mental health workers enrolled in community colleges and then identifies specific programs that community colleges have designed to meet identifiable community needs. Finally, there is a discussion concerning research evidence for different instructional methods. Generelist Treining Proggeme The training of mental health workers at colleges began at Purdue University in 1966, as a result of a National Institute of Mental Health (NIMH) grant (Simon, 1970). By 1970, there were 45 programs for training mental health workers at community colleges around the nation, and new programs were being added at the rate of one per week (Simon, 1970). These programs, generalized in nature, stressed the use of sensitivity techniques to develop confidence, leadership, and insight for the trainees (Krauss, 1970). Emphasis was placed on the social—cultural aspects of mental health problems and the value of humanistic, caring, remotivating relationships with patients in therapeutic settings (Krauss, 1970). Students in the programs ordinarily took a wide variety of courses, including 16 17 English, speech, arts and crafts, philosophy, psychology, anthropology, and first-aid in addition to more specialized training in communication theory, group dynamics, and behavior modification (Atty, 1970). Students also spent. up 'to 1,200 hours in actual contact with clients (Atty, 1970). The two-year curriculum in mental health developed at the Community College of Philadelphia in the late 19605 was typical of the programs that were created. The objective of the program was to produce mental health "generalists" who could function in almost any mental health setting (Sippel, 1971). In their first year, students took general education courses to develop communication skills and an attitude of "open-minded inquiry into all phases of human behavior" (Sippel, 1971, p. 9). In the second year, more specialized courses in human growth and development and normal and abnormal adjustment were offered. Specialized mental health courses were designed to give trainees a broad background in the field of mental retardation, to show trainees how to apply their knowledge of the social and behavioral sciences to the understanding of people, and to develop skills in communication, observation, group membership and leadership, and the "therapeutic use of self" (Sippel, 1971, p. 9). No fewer than 20 separate organizations cooperated with the Community College of Philadelphia by providing field-work placements for the students. As a corollary to the development of the new college programs, major conferences were held to determine the role two-year colleges should play in supplying the burgeoning need for paraprofessional 18 health workers (Gillie, 1972; Penningrowth, 1966). Furthermore, representatives of organizations that might eventually hire the trainees (psychiatric institutes, mental health centers, schools for exceptional children, geriatric centers, hospitals, and halfway houses) frequently formed advisory committees to help the community colleges plan their curricula (Sippel, 1971). The organizations were also willing to allow the trainees to do their clinical training in their facilities. Despite these efforts, there is currently an inadequate number of trained paraprofessionals to provide adequate care for group home residents (Ashburn, 1982). Community colleges and even large univer- sities have attempted to respond to this void by developing job- relevant curricula for training paraprofessionals in ammtal health (Ashburn, 1982; McPheeters, 1973). Given the tremendous interest that was focused on the training of mental health workers, why is there still a shortage of competent mental health paraprofessionals? Basically, the community colleges were establishing two-year mental health technology programs at the associate of arts level. Small numbers of students actually enrolled in the programs, possibly because they were not convinced that a job would be forthcoming after graduation (Krauss, 1970; Sippel, 1971). At the Community College of Philadelphia, which had developed a model program with full institu- tional cooperation (see above), Iless than 50% of the trainees actually graduated (Sippel, 1971). 19 In addition, while the response of community colleges in establishing the programs had been admirable, the process of using a degree program to obtain qualified workers was a lengthy one for state departments of mental health in need of immediate staffing. Also, graduates of the programs were usually not specifically trained to function in a particular setting and had to be retrained by departments of mental health in order for them to function well at their facilities (Ashburn, 1982). Thus community colleges provided general education programs for state departments; however, they did not establish specific, much-needed training programs. Specialist Training Programs Community service has long been one of the important missions of community colleges (Frank, 1980). Furthermore, community colleges are currently giving greater attention to community needs and contributing more to community improvement than was true in the past (Mase & Wattenbarger, 1979). Thus, it is reasonable to believe that community colleges could respond positively to the need for better training programs for health paraprofessionals. Several recent reports have pointed to the need for community colleges to work more closely with local governments, businesses, unions, hospitals, and/or community organizations (Frank, 1980) and to determine the best way of meeting the training and educational needs of adults in their service districts, including experimentation with new delivery systems and better coordination with government agencies (Cuyahoga Community College, 1980). 20 An example of a quick reaction by a two-year college to a pressing community need can be observed in the response by Kellogg Community College to the problem of deinstitutionalizing elderly citizens. A national survey conducted in 1968 pointed out that 39% of the elderly institutionalized residents did not need the specialized care of the institution in which they were living (Andrews, Zinn, & Rae, 1978). Furthermore, a study at the Benjamin Rose Institute in Cleveland, Ohio, found that home-aide services could reduce institutionalization to a great extent and were much cheaper than hospitalization (Beggs, 1970). Shortly thereafter, an analysis completed by the Adult Services Unit of the Calhoun County, Michigan, Department of Social Services found that 33% of the 728 institutionalized senior citizens in that county could have returned to their homes if home-aide and supportive services had been available (Rae, 1975). Accepting a leadership role in its service area, Kellogg Community College officials compiled a list of agencies that were providing some sort of home health care service and surveyed the agencies to determine whether they would offer employment to trained home health care aides. They were shocked to find that there were no trained aides available, that such aides could find immediate employment in the county once trained, and-that several social agencies were in the process of obtaining funds to improve home health care services (Andrews et a1., 1978).. Kellogg Community College immediately established a home health care aid training program. The program was suited to the immediate needs of the 21 community, providing intensive training over a short span of time and thus quickly producing a batch of competent home health care aides (Andrews et a1., 1978). Developed by Kellogg Community College in cooperation with the Michigan Department of Social Services and an organization entitled Services to Seniors of Calhoun County, the program comprised 150 clock hours, about equally divided among the topics of home-management skills, basic nutrition, psychology, physical therapy, and field placement. Newspaper advertisements to entice individuals to enroll in the program stressed the excellent job outlook for home aides and the flexibility and satisfaction inherent in the job. Out of 30 original enrollees, 26 graduated from the short program, a high success ratio when compared with two-year programs (Frank, 1980). All 26 found immediate employment, and many senior citizens returned to their homes. Thus, Kellogg Community College made a very effective response to the problem of unnecessary institutionalization of the elderly. An example of a fine-tuned program that directly attacked a problem at a specific state-operated facility was the one described by Kimball et a1. (1980). In this instance, a local university provided an intensive combination of tutorial and direct training for the staff at a residential center for boys with learning disabilities and/or emotional problems. Morale at the center had been a problem. The university trainers instructed the group home staff in specific techniques to be used in the center, and pretests and posttests of 22 staff perceptions of the group home climate were used to determine the effectiveness of the training. In Florida, four community colleges serving the geographic areas in which the four largest state mental hospitals were located devel- oped degree programs to train staff members for the group homes that would be used as part of Florida’s emphasis on deinstitutionali- zation. Realizing that the degree process was too lengthy, the colleges instead set up inservice training programs (Ashburn, 1982). These programs involved 290 actual hours of instruction and included training in specific job skills, medication administration, approaches to crisis intervention and control, and psychosocial treatment methods. Also emphasized was the need for trainees to work as part of a multidisciplinary treatment team. Trainees who became actual staff members appeared to do well on the job. First reports indicated that they had positive attitudes toward their jobs and used a team approach more than did staff members who had not undergone the inservice training. Patients also seemed to respond more positively than they had to the traditional staff. Research Evidence There is a need not only to develop useful staff-training programs, but also to evaluate empirically the merits of different instructional methods. How does one determine whether a training program has real value or compare the worthiness of different training techniques? Schinke and Wong (1977) described the extreme importance of posttraining testing. In their research, trainers from 23 the University of Washington provided 12 hours of instruction in techniques of behavior modification to staff members in six group homes for mentally retarded persons. Operant techniques corresponding to specific interactional patterns and the physical structure of small familial settings were taught to staff members. In six other (control) group homes, the trainers provided no direct instruction. Instead, they made themselves available to answer staff questions and provide help with problems as they arose. Pretesting was accomplished in four ways: (a) a knowledge test dealing with concepts, (b) an attitude checklist to determine how staff felt about the group home residents, (c) an evaluation of changes in job satisfaction by the staff, and (d) naturalistic observation done by the trainers to determine whether the quality of interactions between staff and residents had changed as a result of the training. The results were striking. Trained staff realized significantly more positive gains than untrained staff in knowledge of behavioral techniques and in attitudes toward the residents. While job satisfaction declined in both groups, the decrease was significantly less for the trained staff. Furthermore, trained staff were more efficient in responding positively to positive resident behavior and in responding neutrally or negatively to negative resident behavior. Response latencies also decreased. The frequency and duration of positive interactions between staff and residents increased; the reverse was true for negative encounters. Thus, posttraining testing validated the techniques of instruction used and pointed to the merits of active instruction of staff versus merely providing support 24 when needed (Schinke & Wong, 1977). As deinstitutionalization continues, there must be some way to evaluate quantitatively the effectiveness of differing group home staff-training programs. In 1972, Gardner compared role-playing and lecture methods in evaluating the effectiveness of teaching behavior modification to 20 female institutional attendants. Pre- and posttest measures were obtained for two major outcome variables--knowledge of behavior modification principles and ratings of ability to apply behavior modification techniques-- using the Training Proficiency Scale and the Behavior Modifica- tion Test. Results indicate that role playing was more effective in teaching behavior modification skills while the lectures were more effective in teaching principles of behavior modification. In addition, cost factors associated with different training programs must be considered. Unfortunately, this is seldom done (Mase & Wattenbarger, 1979). With the tightening of funds for extensive and unnecessary programs in allied health education, administrators in community colleges must give careful attention to the employability of graduates and the costs of their programs. State agencies would like to obtain quality training for their staff at the lowest possible price (Mase a Wattenbarger, 1979). Overall, the paucity of examples in the literature wherein state government and community colleges have actually worked closely together indicates that state government has simply not used public postsecondary education to meet its needs. Community colleges have struggled to anticipate those needs by setting up generalized pro- grams. The successful results produced in those cases in which gov- ernment agencies and community colleges have worked closely on 25 specific programs indicate that closer linkages need to be developed and used. CHAPTER III DESIGN AND METHODOLOGY This chapter provides the reader with the details concerning implementation of the training program, a description of the trainees who participated at both sites, and a list of the limitations and delimitations of the research. Given that the Michigan Department of Mental Health had limited access. to historical data related to delivering training at its facilities, the Office of Resource Development (0RD) proposed to evaluate the cost effectiveness of training dissemination, turnover, and trainees’ content retention over time. As a part of the National Institute of Mental Health, Health and Human Services l982-83 grants program, the 0RD requested funds to evaluate training dissemination in two types of training sites: (a) an MDMH facility and (b) a higher education facility. More than two sites were recommended for evaluation, but limited resources necessitated ‘that the study be restricted to the two sites. The facilities selected were Wayne Community Living Services (WCLS), one of the MDMH’s largest training operations; and Schoolcraft Community College, located less than one mile from that same facility. Both facilities were in close proximity to 0RD’s satellite office, located in Northville, Michigan. 26 27 Method A meeting was held with the director of the facility selected by the 0RD director (the researcher) to complete the training for the MDMH. The research, its purpose, and its benefits to the public mental health system were discussed. The rules for participation were as follows. The facility director’s training staff would be requested to train 30 trainees (paraprofessionals) selected by simple random selection, using the MDMH/0RD newly published Group Home Curriculum, Part I. Trainees would be selected from the MDMH’s facility trainee waiting list (the same list was used for both sites). The staff conducting the training would be qualified trainers meeting the ORD and Michigan Department of Civil Service standards/qualifications. The training would take approximately three weeks, with approximately 30 hours of classroom training per week. At the end of each training module (Part I included seven modules), an assessment would be administered; scores would be tabulated and recorded. At the end of the three-week period, all tests taken by each trainee would be sent (by mail) to the 0RD. Trainees who passed the seven assessments would return to the group home, where they could work independently with clients. At the end of four months, the group home staff originally tested would be asked to return to a testing site to be retested on the same materials used in the initial training. This retesting was to be coordinated by 0RD staff. Tests and scores would be submitted to the 0RD director. A meeting was also scheduled at the second training site, the community college. This meeting was held with a program 28 representative of the comunity college. Thirty randomly selected trainees were chosen from the MDMH’s facility waiting list (the same list was used for both training sites). The researcher explained the study, discussing purpose and responsibility. Both the 0RD and the community college signed agreements (Appendix E) reflecting contract obligations. The contract outlined the following: 1. Training would be done by paid community college staff. 2. The 0RD would pay a negotiated fee for each class going through the four-week course. 3. Tests would be administered at and by the college. Follow- up testing was supervised by ORD staff. 4. Community college staff had to meet the trainer qualifications or participate in a one-day training of trainers session offered by the 0RD. 5. Use of the 0RD’s trainer manual for the Community Group Home Curriculum was optional. The reimbursement mechanism used was developed by the MDMH’s Accounting Office. The procedure was as follows: (a) a bill would be submitted to the 0RD by the community college for the services rendered; (b) a justification memorandum would be sent to the MDMH’s Accounting Office with the account number to be billed; and finally (c) a check would be issued within 30 days. The MDMH billed the Federal Manpower Grant, issued by the NIMH, HHS (Appendix F). Both the community college and MDMH facility representatives agreed to provide the 0RD access to information/data relating to cost 29 variables. The community college agreed to a flat fee for training. Thus, the variables that determined their cost were not necessary. The facility costs were more involved; the following variables were used: --Salaries/wages of trainers --Fringes (vacation, hospitalization, sick leave, etc.) --Travel --Meals --Equipment --Indirect costs (maintenance, housekeeping, general administra tion, depreciation, utilities) --Contract services, supplies, and materials (CSS&M)--telephone, printing, contract with Red Cross Limitations end Delimitations The following limitations applied to this research project: 1. The limited state and federal dollars allocated for this research forced the sample size to be small for the community college training site. 2. No comparison to previous costs for training was available due to the lack of historical data. 3. Due to limited funds, only two training sites were used in the design. 4. The research was time limited due to the work-plan restric- tion for the federal grant. 5. Because of estimated high turnover rates of paraprofes- sionals, a four-month retest interval was used. 30 This research compared quality and cost variables at two different instructional sites, using the Group Home Curriculum. This curriculum has been published under the direction of the investigator and is the longest curriculum-development activity undertaken by the MDMH in the past 12 years. It is also the first curriculum that has a complete set of policies and standards for trainer practice. Although this research design may be relevant to a number of MDMH curricula materials, the investigator selected the one that focused on the MDMH’s major goa1--deinstitutionalization and community placement of its clients. Subjects Sixty individuals were randomly selected from the MDMH’s facility trainee waiting list to be trained as group home workers. Thirty received their instruction at a MDMH facility, the other 30 at Schoolcraft Community College. All subjects had to be over 18 years of age and agree to sign consent forms. MDMH Trejninq At the MDMH facility, the staff conducting the training consisted of qualified trainers meeting MDMH/0RD and Michigan Department of Civil Service standards and qualifications. The MDMH’s newly published Group Home Curriculum, Part I, was used as the basis for training. The training took approximately three weeks, with 30 hours of classroom instruction per week. Also, trainees had to attend the training for cardiopulmonary resuscitation (CPR) and first aid, which was subcontracted to the local Red Cross. 31 Trainees were tested at the end of each training module (there were seven modules in all). At the end of three weeks, trainees passing all seven assessments went to newly developed group homes to work with clients. After four months, individuals still working in the group homes were asked to *return to 'the ‘facility for retesting on the same training materials. This retesting was coordinated by 0RD staff. In an attempt to maximize the number of people taking the retest, the test was offered on three separate dates, all within the span of one week. Each person taking the retest qualified for a $15 expense voucher. This training sequence is summarized in Table 1. Table l.--Training sequence. Module Number of Questions on Exam NO‘U‘I-h-FWN-fl W) U" o mmuni l e Trai n ' At the community college, training was done by community college staff. Community college instructors either met qualifications or participated in a one-day training-of-trainers session. Before the instruction, the 0RD director met with a representative of 32 Schoolcraft Community College to discuss the purpose and responsibilities of the research. An agreement signed by the ORD and the community college provided for tests to be administered at and by the college, with results reported directly to the 0RD. The ORD was required to pay a negotiated fee for each trainee going through the course. (This course was offered after 5:00 p.m.) ‘The agreement also indicated that use of the 0RD trainer manual was optional. Like the facility trainees, individuals tutored at the community college who had passed the seven module assessments began working at newly developed group homes and were called in for retesting after four months. Materials The tool used to assess the trainee’s mastery of the course material was Form A of the Assessment Package for the Group Home Curriculum. (See Appendix G.) CHAPTER IV RESULTS The purpose of this chapter is to report the results of the data analysis. The researcher will make no attempt at drawing any conclusions, making suggestions, or interpreting data in this chapter. In Chapter V the major conclusions that can be drawn from these analyses are reported. This chapter is organized around the themes of turnover, effectiveness, and costs. The Turnover section reports the results of data collected on trainees immediately posttested and then retested four months later. This section answers the question, ”What is the turnover rate fOr trainees at each training site over a four-month period?" The Effectiveness section reports the results of the data collected and analyzed to answer the questions, (a) "What are the differences in mean assessment scores for each training module fer each site?" and (b) "How do original assessment scores for each paraprofessional compare with a four-month follow-up assessment score for each training module and each training site?" The Costs section reports the data analysis designed to answer the question, "What is the difference in cost when offering the new Community Group Home Curriculum at a community college versus an MDMH facility?” 33 34 114mm: Turnover rates differed between dissemination strategies (Table 2). At the community college, of the original 30 trainees, 20 individuals (67%) completed the instructional program and took the original tests. At the MDMH facility, 29 persons (97%) completed the course and were tested. Only five individuals (17%) from the commu- nity college were retested, whereas ten (33%) MDMH trainees took the retest. Table 2.--Number of individuals who completed testing. Number Beginning Number Completing Number Instruction Original Test Retested Community college 30 20 5 Department of Mental Health 30 29a 10 aEight of the 29 were not originally tested on Modules 4A and 4B. Thus, out of 20 individuals completing the original test at the community college, 15 were not retested (Table 3). Ten of these quit working at the group homes (a 50% dropout rate), whereas the other five were still employed but did not show up for retesting. Out of 29 individuals completing the original test at the MDMH facility, 19 were not retested (Table 3). Only six of these had ter- minated employment (6/29 - 21% dropout rate). The other 13 simply failed to appear for retesting. 35 Table 3.--Fate of individuals not retested. No. No. Still Completing No. Not No. Employed But Original Test Retested Dropouts Not Retested Community college 20 15 10 5 Department of Mental Health 29 19 6 13 Effectiveness Statistically, the results (number of correct answers for each person for each module) were analyzed as a 2 x 2 factorial experi- ment. There were two factors of interest (instructional site and time of testing) and two levels of each factor (the MDMH facility versus community college for instructional sites, original test versus retesting for the time factor). Analysis of variance was used to test research questions concerning quality-of—treatment means and to check for interactions between the two factors (instruction site and time). The actual number of correct answers by each person for each module was used as the variable of interest rather than the percentage score to avoid problems associated with the use of ratios (i.e., non-normality). Scores obtained for each treatment combination for each module were analyzed for normality using a test developed by Shapiro and Wilks (cited in Gill, 1978). No departures from normality were found. Use of Hartley’s fmax test (cited in 36 Gill, 1978) ruled out heterogeneity of variance between treatment groups. Since replication was unbalanced (as mentioned, at Schoolcraft 20 people took the first test and 5 took the retest; at the MDMH facility from 21 to 29 people were tested on various modules on the first date and 10 people were retested), a technique developed by Federer and Zelen (cited in Gill, 1978) to deal with unbalanced data was used. Sums of squares (and thus mean squares) were calculated for main effects (instructional method and time), interaction, and experimental error, and F-tests were carried out to determine statistical significance. Test scores obtained by trainees at the MDMH facility and the community college were very similar (see Table 4). No significant difference was found between test scores at the two instructional sites for Modules 1, 2, 3, 4B, 5, 6, and 7. For Module 4A, scores were significantly higher at the community college (42.45) than at the MDMH (39.45). This was the only case in which instruction site had a significant effect on test results. Test scores were fairly similar over time (see Table 5). For Modules 1, 2, 3, 5, 6, and 7, time of testing had no significant effect on test scores. For Module 4A, test scores dropped from 47.9 on the first test date to 34.0 on the retest. For Module 48, test scores declined from 47.28 to 41.70. 37 Table 4.--Comparison of test scores between Schoolcraft Community College and the Department of Mental Health. Mean No. Correct Mean No. Correct Module No. of Questions Answers: Community Answers: MDMH on Exam College (1 1 SE) (1 1 SE) 1 50 44.68 1 1 02 42.85 1 .81 2 25 22.51 1 52 21.65 1 .41 3 50 42 73 1 88 43.15 1 .70 4A so 42 4s 1 71a 39.45 1 .63 4B 50 45 20 1 64 43.78 1 .57 5 50 44 43 1 69 44.15 1 .55 6 25 21 80 1 .35 22.02 1 .28 7 20 19 90 + .08 19.80 1 .07 aIndicates mean score for Schoolcraft was significantly higher than MDMH mean score. Table 5.--Comparison of test scores between original test date and retest dates for each module. Mean No. Correct Mean No. Correct Module No. of Questions Answers: Original Answers: Retest on Exam Test (1 1 SE) (1 1 SE) 1 50 42.83 1 .73 44. 70 1 1. 31 2 25 21.71 1 .37 22. 45 1 .67 3 50 42.73 1 .81 43.15 1 1.13 4A so 47.90 1 .ssal 34. oo 1 .92 48 so 47.28 1 .50a 41.70 1 .83 5 50 45.32 1 .49 43.25 1 .89 6 25 21.67 1 .25 - 22.15 1 .45 7 20 19.85 1 .06 19.85 1 .ll aIndicates score was significantly higher on original test date. 38 Two significant interactions were fOund between test site and time--on Modules 48 and 5 (see Table 6). In Module 48, test scores dropped over time for both instructional sites, but the decline was more precipitous (from 47.76 to 39.8) at the MDMH than at Schoolcraft (46.8 to 43.6) (Figure 4). For Module 5, scores were lower on the retest at the MDMH (the decline was 46.79 to 41.5), while scores actually rose at the community college (from 43.85 to 45.0) (Figure 5). Table 6.--Mean test scores (1 1 SE) for each treatment combination and module) Community College MDMH No. of Module Questions Original Original on Exam Test Retest Test Retest 1 50 43.351.92 46.0011.70 42.3011.20 43.401 .79 2 25 21.831.49 23.201 .37 21.591 .55 21.701 .82 3 50 41.451.92 44.001l.41 44.001 .89 42.3011.25 4A 50 49.501.28 35.4011.94 46.301 .70 32.601l.93 4B 50 46.801.71 43.6011.21 47.761 .64 39.8011.25 5 50 43.851.90 45.0011.38 46.791 .53 41.5011.25 6 25 21.201.41 22.401 .40 22.141 .35 21.901 .43 7 20 19.801.12 20.0010.00 19.901 .06 19.701 .15 Sums of squares, mean squares, and F-ratios are shown for each module in Appendix H. To summarize, in the seven modules, only five significant effects were found, all in Modules 4A, 4B, and 5. One significant effect (Module 4A) was found for instruction site, two effects were obtained for time of testing, and there were two inter- actions. Correct Answers Correct Answers 48 4s 44 42 4o 48 46 42 4O 39 Community college ‘\MDMH I I Original Retest Test Figure 4.--Module 4B test scores. \\ \ \ \ o ________—:1-