TW Q770/i37 W llllllllllllllillilllWll|||li|||||||||l|illilWHHIHHI 1’ 3129300774 LIBRARY 1* '— Michigan State Univerdty k m This is to certify that the dissertation entitled FACTORS INFLUENCING SUPERVISORS' REFERRALS TO EMPLOYEE ASSISTANCE PROGRAMS presented by ' David Arthur Harley has been accepted towards fulfillment, of the requirements for Ph .D . degree in Counsellng Psychology J@ /J% Major p esso/ DateLW X/770/ / / Ht MS U is an Affirmative Action/Eq ual Opportunity Institution 0-12771 41.—. .4...— uh“ - A... PLACE IN RETURN BOX to remove this checkout from your record. TO AVOID FINES return on or before date due. DATE DUE DATE DUE DATE DUE _o(:124t994 [D »- 30¢ b— MSU Is An Affirmative Action/Equal Opportunity Institution c:\cim\ddodue.pm3—p.1 : FACTORS INFLUENCING SUPERVISORS' REFERRALS TO EMPLOYEE ASSISTANCE PROGRAMS By f David Arthur Harley A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Counseling, Educational Psychology, and Special Education 1990 31:20 1 (0%“ ABSTRACT FACTORS INFLUENCING SUPERVISORS' REFERRALS TO EMPLOYEE ASSISTANCE PROGRAMS By David Arthur Harley This study investigated supervisors' use of "constructive confrontation" with employees having job and personal problems, a central feature of employee assistance programs (EAPs). A questionnaire survey was conducted at 25 sites from seven organizations. Supervisors' (_N_ = [+15) referral decisions with hypothetical employees, reported past EAP referrals, and referral action taken with an employee identified as having job problems in the previous year were examined in relation to previously identified factors and to newly developed attitude toward The approach/avoidance conflict model was proposed as referral scales. relevant to referral hesitancy. Attitude scales were constructed with approach and avoidance dimensions, and they demonstrated concurrent validation in multiple regression tests with the hypothetical, past, and recent referral actions (3’ of .07, .10, and .05, 2'5 < .003). The strongest predictor of supervisory referral was occurrence of a "trigger event," a dramatic worksite incident, which alone accounted for observed referral action taken with the actual employee as well as all the other variables jointly. Other factors found to increase referral likelihood were familiarity with the EAP, amount of EAP-relevant training received, severity of the employee's job performance impairment, greater age or experience of the supervisor, and greater approach attitudes _______‘.A toward referral. Employee's being male and greater span of control also increased referral likelihood, but not as strongly. Greater amounts of formal education and stronger avoidance attitudes toward referral lowered the likelihood of referral, as did higher organizational status of the employee, weakly. Multiple regression models with those variables accounted for 36% of reported past referrals variance and 31% of recent employee referral variance. Hypothetical referral action was not very well predicted. Other factors were also examined and found not associated with supervisors' EAP referrals——supervisor gender, if supervisor and employee are same or different gender, whether the supervisor tends to take a counselor-like role with troubled employees, and the supervisor‘s perception of the degree of alcohol/drug focus of the EAP. The findings suggest a need to modify and increase supervisory training efforts, to decrease reliance on supervisory referral for EAP casefinding, and to develop a more active "outreach" role for EAP counselors. Copyflght by David Arthur Harley I990 To Sandy, Doug, and Susan, whose love sustains me, and in loving memory to Farra G. Harley, Mary E. Morrow, and Mary Claire Hill, who would clap the loudest ACKNOWLEDGMENTS This research project could not have been undertaken without the support and assistance of many others. Some will not receive acknowledgment in print, because they or their organization desire continued anonymity. My deep appreciation is offered to all who assisted in the project, gave freely of their experience and suggestions, and opened their organizations to the survey. To the managers and supervisors at the survey sites, your data was gratefully received, and i hope the results will help your organizations be both efficient and humane. I especially thank the members of my doctoral guidance committee. Each of my professors made a special contribution to the effort, based on their personal expertise: Dr. Michael L. Moore, whose comments on the earliest glimmerings of the project and methodological advice kept me on the path; Dr. Rom Kriauciunas, who helped develop and pilot the instrument and gave editorial guidance; Dr. Donald J. Freeman, who greatly assisted the development of the scales and final survey instrument; and Dr. Gloria S. Smith, my Committee Chairperson, whose wise guidance, encouragement, direction, and well-timed impatience at critical points made all the difference for me. Their expertness, support, and accessibility was a clear factor in the outcome of the project. Colleagues and associates in the mid-Michigan community of psychologists and employee assistance professionals assisted in numerous ways—-reacting to items, sharing their experiences and ideas about supervisory referral, helping to provide access to organizations, and serving on the initial item panel. Tom Helma, Director of the EAP at Michigan State University; Jim Miller, personnel director of MEDNA, Michigan Education Association; Bill Dowling, Ph.D., Director of the Employee Service Program of the State of Michigan; Ken Salzman, Ph.D., psychologist; Tom Fuller, Ph.D., psychologist; Jo Hall, Ph.D., Director of CARES EAP, Lansing; Bob Smith, Ph.D., psychologist; Russ Scabbo, Ph.D., psychologist; Rex Bangs, R.N., Director of Employee Relations, Buick-OIdsmobile-Cadillac Group of GM, Lansing; Dorothy Hudson, also of B.O.C. Group, Lansing; Darrold Oxender, personnel specialist with the Lansing Board of Water and Light; and Linda Hanks, Human Resources specialist, also with the Board of Water and Light--each 1‘ ‘ made a contribution to the total effort, and I acknowlege and thank you again. Special appreciation is given to Teri Fitzpatrick, Director of Human Resources Management of Owens—Illinois, Toledo, Ohio. Her support for the project and for employee assistance efforts in her organization was of immeasurable assistance. In addition, the research was partially supported by a grant from Owens-Illinois, which is gratefully acknowledged. Additional grant assistance was provided by the Employee Assistance Professionals Association (EAPA), Arlington, Virginia. The grant not only meant funding support, but encouragement and respect by colleagues I7———:9~1 __..-_. in the field. My thanks to EAPA and to Dr. Terry Blum, Chairperson of the Research Committee. I hope my work will be of value to the employee assistance field and will stimulate additional research. Thanks, also, to Susan Harley, who aided in the tedium of tables preparation with her clerical assistance and youthful enthusiasm; to Peggy Gallandt, who gave needed secretarial services at a critical time; and to Sue Cooley, who provided editorial guidance which helped the appearance and readability of the final product. viii TABLE OF CONTENTS LIST OF TABLES LIST OF FIGURES . Chapter I INTRODUCTION . Importance of the Topic . Limitations of the Study . 2 REVIEW OF THE LITERATURE Overview History of Employee Assistance Programs . History of the Constructive Confrontation Strategy Factors Affecting Use of Constructive Confrontation . The Empirical Research Base A Conflict Model of Constructive Confrontation A Theoretical Research Base Indications of Conflict in Supervisory Behaviors . Approach/Avoidance Modeled as Response Conflicts Attitudes as Measures of Approach/Avoidance Tendencies 24 37 37 86 86 87 92 98 Chapter 3METHODOLOGY........... Overview . . . . . . The Research Questions . . . . . . . . Definition of Terms . . . Instrumentation . . . . . . . . Development of the Instrument . . . . Variables and Scales . . . . . . . Participating Organizations . Data Collection and Comments Characteristics of the Supervisory Sample 4 RESULTS . Overview . Data Analysis Approach . Findings on the Research Variables . Approach and Avoidance Attitudes Toward Referral Cognitive Variables Group Supervisor Demographic Variables Group Situational Variables Group Employee-related Variables Group . Location Effects . . . Models of Predictor Variables and Supervisory Referral Multiple Prediction Model of Past Referrals Multiple Prediction Model of Recent Supervisory Referral Behavior . . . Summary of Findings . . . . 104 101i 106 108 110 110 115 125 133 135 138 138 141 141 149 156 167 172 174 174 176 179 Chapter 5 SUMMARY AND RECOMMENDATIONS . Overview Methodological Discussion and Recommendations Discussion of Factors Influencing Supervisory Referral Need for Further Research . . . . Recommendations for EAP Practice APPENDIX A THE SURVEY INSTRUMENT . APPENDIX B ITEMS, SCALES, AND SUBSCALES APPENDIX C ATTITUDE SCALE AND SUBSCALE INTERCORRELATIONS AND RELIABILITIES . APPENDIX D INTERCORRELATIONS OF VARIABLES REFERENCES . xi Page 183 183 184 186 190 197 205 213 226 227 228 LIST OF TABLES Table l Supervisor-related Variables Affecting EAP Referral Actions . 2 Employee-related Variables Affecting Supervisory EAP Referral 3 Organizational/Program Factors Affecting Supervisory Referrals 4 Proposed Cognitions Blocking or Pushing Supervisory Referrals 5 Research Variables and Scales . 6 Survey Distributions and Returns By Type Of Organization . 7 Site Descriptions With Survey Distributions and Returns . 8 Descriptive Characteristics of Supervisory Sample . 9 Variable Means and Standard Deviations by Survey Sites and Total Sample . 10 Regression Results for Approach and Avoidance Scales with Hypothetical Referrals Scale as the Criterion 11 Regression Results for Approach and Avoidance Scales with Reported Past Referrals Variable as the Criterion . 12 Regression Results for Approach and Avoidance Scales with Actual Employee Referral Variable as the Criterion 13 Regression Results for Cognitive Variables Group with Hypothetical Referrals Scale as the Criterion 14 Regression Results for Cognitive Variables Group . _ with Reported Past Referrals Variable as the Criterion . l5 Regression Results for Cognitive Variables Group _ . with Actual Employee Referral Variable as the Criterion Page 38 4o 41 90 126 127 128 137 lli2 145 146 147 150 151 152 Table 20 21 22 23 24 25 C.l D.l Regression Results for Supervisor Demographic Variables Group with Hypothetical Referrals Scale as the Criterion . Regression Results for Supervisor Demographic Variables Group with Reported Past Referrals Variable as the Criterion . . Regression Results for Supervisor Demographic Variables Group with Actual Employee Referral Variable as the Criterion Regression Results for Situational Variables Group with Hypothetical Referrals Scale as the Criterion . Regression Results for Situational Variables Group with Reported Past Referrals Variable as the Criterion . Regression Results for Situational Variables Group with Actual Employee Referral Variable as the Criterion Regression Results for Employee Variables Group with Actual Employee Referral Variable as the Criterion Regression Results for Full Multivariate Model of Validated Factors with Reported Past Referrals Variable as the Criterion . . . . . . . . . Regression Results for Full Multivariate Model of Validated Factors with Actual Employee Referral Variable as the Criterion . . . . . . Summary of Findings on Factors Affecting Supervisory Referrals Attitude Scale and Subscale Intercorrelations and Reliabilities Intercorrelations of Variables xiii Page 157 158 159 163 164 165 169 175 177 180 226 227 LIST OF FIGURES Figure Page 1 Graphical representation of approach/avoidance conflict . . . 94 2 Representation of closer proximity to goal resulting from increased approach tendency, accompanied by increased level of subjective fear . . . . . 97 3 Representation of closer proximity to goal resulting from decreased avoidance tendency, accpmpanied by minimal increased subjective fear . . . . . 97 4 ' Observed correlations between EDLEVEL, TRAINING, APPROACH, AVOIDANCE, and the referral variables PASTREFR and EMPREFER. . . . . . I94 CHAPTER 1 INTRODUCTION Employee assistance programs (EAPs) attempt to address a perennial problem in organizations--employees who were once valuable and productive workers but who have begun to perform their jobs poorly or create disruptions because of personal problems. These employees have proven difficult to deal with because they do not or seemingly cannot respond to the usual managerial control methods such as training, job coaching, changed incentives, closer supervision, or even disciplinary actions. Numerous conditions and factors are known to affect job performance (Miner 6r Brewer, 1976). EAPs are intended to remedy the personal ills which contribute to deteriorated job performance. Historically, the primary focus has been upon the alcoholic employee, but because of greater awareness of mental health problems generally, it is now recognized that virtually any personal emotional or behavioral/health impairment can result in reduced worker effectiveness on the job. Furthermore, employers are striving to reduce and avoid certain costs which have been associated with employee problems--absenteeism, greater use of costly health care benefits, increased accident rates, and increased workers' compensation costs. A full listing of impairing conditions would include virtually every personal woe and malady. The principal concerns of EAPs continue to be alcoholism and other drug abuse conditions and the impacts on families, psychological or psychiatric illnesses such as anxiety states, depression, or psychosis, marital and family conflicts, and the somewhat more external but nonetheless troubling issues like legal or financial difficulties. Recent epidemiological studies confirm that many employees may have conditions which impair them on their jobs. During any given six-month period, approximately one in five adults suffers from a mental or substance abuse problem of sufficient severity to meet current diagnostic criteria (Myers et al., 1984). Among women, the most common conditions are phobias and major depressive states, but among men, alcohol abuse/dependence is most common. The majority of them are likely to be employed, for the ages of greatest incidence of disorders (young adulthood through middle-age) are found during that portion of the lifespan with greatest participation in the labor force. 1 To reduce the impact of these conditions on their workforce, employers have provided health care insurance benefits. Coverage of benefits has been broadened until now employees can gain professional assistance for nearly any problem. So many employees do so, that the annual cost of these benefits is a source of growing alarm to employers (Dowling, l980/l98l; Leonards, I981; Tilson, 1989). However, many persons with personal problems do not seek help on their own, especially not during the earliest phases of problem onset when intervention is often most effective and least costly. The findings from an extensive survey of health service utilization (Shapiro et al., 1984) found that a significant plurality (25% to 40%) of persons with moderate to severe problems did not seek any help at all from any source during the six months prior to the survey. It was observed that women were more likely than men to obtain needed help, and persons with substance abuse problems were much less likely to get help than someone with a mental/emotional problem. A further large number had only two or one visits with a care provider regarding their problem, and that was most likely with a general physician, not a specialist. Clearly, providing the means for troubled employees to get needed help does not necessarily ensure their doing so in a timely manner. The supervisory role includes the use of motivational, instructional, and disciplinary tools with workers to meet the set goals. When an EAP is developed in an organization, however, the supervisors' role has been further expanded to include a function of interfacing between the troubled worker with a job—impairing problem and the specialist service providers who might alleviate the problem. Their role is to link the world of work to mental health care through the agency of the EAP. The strategy which has evolved to accomplish this is the “constructive confrontation," which results in referral of an employee who would not seek treatment or assistance voluntarily for EAP assessment and likely service. The reluctance of some persons to obtain needed help. on their own initiative was incorporated into the design of employee assistance programs. From the very earliest beginnings, EAPs have used mandatory or coerced program participation as a way to at least expose troubled orkers to appropriate services toward which they may be very resistant. he supervisor was identified as the key element in EAP operation. The upervisor is seen as the most likely person in the organization to notice the development of job impairment and the problems which cause the impairments. Further, because the troubled worker's supervisor is the one charged with daily management of the worker and his/her outputs, the supervisor is seen as having a legitimate role in intervention whenever those outputs fall below expected levels. The importance of management-initiated program involvements through supervisory referrals has been repeatedly affirmed, most recently by Roman and Blum (1988). The constructive confrontation strategy has been identified as one of the "core technologies" which make EAPs both effective and distinctive from other efforts to maintain the health, well-being, and productivity of persons in various settings. However, research on supervisors' performance of their critical role in EAPs has found a reluctance among the supervisors to use the constructive confrontation technique. Trice and Belasco (I965) termed this the "policy—practice gap," indicating that official creation and sanction of an EAP did not ensure that it would be adopted and used by the supervisors it was intended to support. Riediger (1985) stated, somewhat pessimistically, that effective program utilization rests almost entirely on the shoulders of the supervisors. But a lack of program utilization represents the Achilles heel of the employee assistance movement. In spite of a heavy promotional emphasis, the employee assistance movement has failed to become an integral part of the workplace. It has, typically, received token support, which may well disappear when its lack of effectiveness becomes apparent and discredits the promotional claims which have fueled the movement thus far. (p. 401) Supervisors typically (Trice, 1965) go through four stages before sing an EAP--(a) a "disrupted but normal" stage, during which an mployee has intermittent performance problems; (b) the "blocked-awareness" stage during which the supervisor realizes the pattern of poor job performance probably stems from a personal problem, but is prevented from acting by individual barriers against making a referral or even confronting the employee for his behaviors; (c) a "see-saw stage" during which accumulated performance problems have overcome some of the barriers to referral, but the supervisor wavers in resolve to act and still avoids a clear confrontation with the employee; and (d) the policy use stage, when repeated instances of poor performance finally overcome the supervisor's remaining reluctance towards a referral and his/her avoidance of confrontation and a formal job action. The role of supervisors in EAP operation is recognized as crucial, but the dynamics of their interactions with problem employees is not well understood. Even though it is such a central, core feature of EAPs, it ‘is acknowledged that supervisors will require support and encouragement to confront and refer (e.g., Roman at Blum, 1988). It is known and often decried by program administrators and staff that supervisors typically do not perform the casefinding and referral role very readily. Importance of the Topic In order for EAPs to be made more effective and able to serve the troubled worker population to the greatest possible extent, a better understanding of the supervisors' referral decision and action is needed. More research is needed on the topic of constructive confrontation. here have been numerous calls for research into the supervisors‘ use of AP and their use (mostly nonuse) of the constructive confrontation trategy (Bayer, I987; Googins, 1979; Hoffman 6c Roman, I984; Kurtz, 1982; Roman, 1982; Roman 6c Trice, 1976; Trice, 1980; Trice 6c Beyer, l984a). Evidence is accumulating that constructive confrontation is simple and attractive in concept, but rather complex, poorly understood in its operational effectiveness, and frequently disused by the supervisors it is intended to support. Efforts to identify the forces and barriers which operate within and upon supervisors have been marginally successful. Very little of the research has attempted to replicate the findings of other researchers and none has combined all known factors in a multivariate study of referral action. Efforts to relate the individual supervisor‘s personality factors or motivational states to referral action are just beginning to be seen (Bayer, 1987; Bayer a Gerstein, 1988; Krucher, 1986/1987; Young, Reichman, 6t .Levy, 1987; Sonnenstuhl 6r Trice, 1989). The importance of this research is twofold. It consists of a replication effort, retesting much of the previous research on supervisory referrals to EAPs. Secondly, attitude scales have been developed which measure motivational factors of the individual supervisors using a theoretical model of approach/avoidance. The attitudes scales are tested as predictors/correlates of referral behavior with both actual and hypothetical employees with job performance problems in an attempt to validate their usefulness in understanding and predicting supervisory referral action. Unlike previous research in this area, it is proposed that attitudes having approach and avoidance components are related to and predictive of supervisors' handling of a troubled employee in their charge. The approach/avoidance conflict theory of Miller (1944) and Dollard and Miller (1950) is used as a framework for hypothesis generation. Findings of this research may be able to guide EAP administrators in a modification of their supervisory training efforts, in order to increase appropriate use of and referral to EAPs. A potentially useful psychometric attitude survey tool for use with supervisors has been developed. The attitudinal and other scales developed for the instrument may find application in future research efforts in the employee assistance field and in training activities. Limitations of the Study A major limitation of this study is due to its nonexperimental survey methodology. Causative status for the variables of interest cannot be asserted, yet in some respects it would be difficult to link supervisory behavior and some of the observed variables (such as age and Span of control), as no explanatory theory has developed supporting such links. Rather, this study attempts to develop a prediction model which may aid our understanding of the multiply determined nature of the supervisor‘s response to a troubled employee. The effort to develop and partially validate new psychological scales and to further study previously identified factors seems well-suited to the exploratory nature of the survey method. The nonrandom and nonsystematic constitution of the survey sample represents a further source of possible invalidity of the study. Samples of convenience at organizations voluntarily participating in the study create the possibility of selection bias in the findings. This combines with the problem of all surveys, unknown bias resulting from refusal to respond (mortality from nonparticipation), leaving doubts about both the internal and external validity of the study. It is hoped that these have been minimized by recruiting respondent organizations from a variety of types and settings, and using high-level organizational sponsorship of the study to increase response rates. Absence of bias cannot be adequately claimed or proven, but the eventual sample size, large number and type of participating organizations, eventual survey return rate (142%), and analyses which test for location effects argue against severe limitations due to presence of biases. The other principal limitation, after sampling and mortality biasing, might be the selection of variables studied. Those replicated here are the ones given prominence in the EAP literature up to now, with the innovation of new motivational variables, attitudes toward referral. A prediction model can only be as strong as the factors in the model, and other, better predictors might (would) improve the power and utility of the model. Another researcher may very possibly choose to attend to other aspects of supervisors' personality or other situational or organizational attributes in relation to EAP referrals. It is h0ped that this study, with its acknowledged limitations, will suggest future similar efforts to identify psychosocial, organizational, and personality factors which better explain and predict supervisory referral behaviors, an important component of EAP operations. CHAPTER 2 REVIEW OF THE LITERATURE Overview This literature review consists of four sections. First, a brief history of employee assistance programming is given. This will enable the reader to gain a perspective on the central role of the supervisor in EAPs since their inception. Such a history also places into a perspective the research issue of program focus. Then, the strategy of constructive confrontation is discussed at length. A history of the technique and how it was developed and has evolved is given from the theoretical literature of EAPs. The research base regarding the factors/variables which have been found to affect supervisory use of the constructive confrontation strategy is explored next. Seven major and approximately a dozen lesser variables are discussed in relation to supervisory referral behavior. Nearly all of those variables are replicated in this research. The reluctance of supervisors to engage in constructive confrontation and to make referrals to EAPs is next discussed, within the context of a theoretical model of conflict. The approach/avoidance Conflict model of Dollard and Miller (1950) is introduced and applied to the observed behavior of supervisors when they must deal with an employee having job problems. It is proposed that supervisors' referral behaviors 9 10 (especially their vacillation and hesitancy in making the referrals) is very similar to the behaviors described by Miller's approach/avoidance conflict model. A means of measuring that conflict by means of attitudes comprised of both approach and avoidance components has been proposed by Kelman (1974, 1979), and this technique will be used in this research. History of Employee Assistance Programs Modern employee assistance programs evolved from efforts of industrial organizations to cope with alcoholism. This condition has been and continues to be widely prevalent in our society. The effects of alcohol abuse create severe problems in complex organizations -- errors of judgement and inefficiencies, deteriorating personal health, and interpersonal conflicts. In more recent times, the mission of EAPs has evolved to address any of the problems which workers might have which adversely affect their productive capacities. Early in American history, alcohol consumption on the job was considered normal, the custom having been brought over from Europe. Well into this century the practice persisted, even though it was disapproved of by many. Farmers and industrial employers first began consistently to attempt to ban alcohol from the workplace around the turn of the century. Warnings to businessmen of the dangers of "industrial alcoholism" in the form of injuries, alcohol-related pathologies, and low productivity began to appear. The "temperance movement" accelerated concern about workplace abstinence; employers became committed to eliminating alcohol from the orkplace. According to Trice and Beyer (1982b) alcohol came to be seen 11 as "anathema to efficiency" and was considered one of the main problems of socializing a dependable and productive workforce. The legitimate basis for intervention with impaired employees stems from the right of employers to expect performance in exchange for wages or salary. The employer's claims for role performance at a specified standard is inconsistent with an impaired or substandard performance. "Practically all job-based alcoholism policies use the fact that job performance declines among developing alcoholics as the starting point and justification for their intervention“ (Trice 6C Beyer, 1982b, p. 26). Historical background on the development of "occupational programming" responses to alcoholism and mental health problems have been provided by Lotterhos (1975), Roman (1981), Roman and Trice (1976), Trice and Beyer (1982a, 1984a), and Trice and Schonbrunn (1981). Companies developed their own ways of handling problem employees through the 19405. The typical organizational response to alcoholics was to fire them when their problems reached an intolerable stage. Some evolved more positive approachs; these became embodied as the Yale Plan. This early employee assistance approach was fairly widely disseminated, and by 1960 an estimated 50 companies had an alcoholism program. In 1969, the Hughes Act mandated the creation of the National Institute on Alcohol Abuse and Alcoholism. Roman and Trice (1976) called this the "major event“ for occupational programming, for the whole alcoholism field. Since then, the federal government has taken an active role in dissemination of EAP program models and the stimulation of program adoption in both government agencies and in private business and industry. In 1972, states were given grants to train and employ 12 "occupational program consultants" whose mission was to promote the institution of alcoholism programming in organizations within their states. Other factors and organizations also contributed to the early development and spread of occupational alcoholism programs. The Yale Center of Alcohol Studies, the General Services Office of Alcoholics Anonymous, the Christopher D. Smithers Foundation, the National Council on Alcoholism, and the Committee on Alcoholism of the American Association of Industrial Physicians and Surgeons were noted by Trice and Schonbrunn (1981) as instrumental in the efforts at education and influence in the forties and fifties. Early research efforts began to dispel the myths and stereotypes about alcoholism at the same time that they were demonstrating the prevalence of the problem and the immense costs to business. The "billion dollar hangover" and the legendary "half—man" were beginning to convince business leaders of the importance of the problem of alcohol-impaired workers, and more importantly that rehabilitation was possible and worth pursuing. The role of universities in making occupational alcoholism studies legitimate and in promoting program concepts was another major factor in their gradual acceptance. The first efforts in dealing with problem drinking among the workforce were the institution of company rules and policies forbidding the posession or use of intoxicants on the job or coming to work in an intoxicated state. When it soon became apparent that such rules had little impact on actual worker behavior, some responsible leaders realized that more constructive solutions were needed. The earliest 13 programs which addressed the problems of drinking and worklife were both informal and unwritten, largely operating with the support and resources solely of the medical director. Examples of such early programs were those at DuPont under Dr. George Gehrmann, at Eastman Kodak under Dr. John Norris, and at New England Telephone Company, thought to be the first ever, under Dr. Daniel Lynch. Trice and Schonbrunn (1981) saw three forces combining to foster the innovation and spread of industrial alcoholism programs: the birth and growth of Alcoholics Anonymous (AA), emergence of the role and influence of company medical directors, and the labor conditions during World War 11. The desperate need of companies for workers during the war was an incentive to provide rehabilitative programs which permitted the hiring and keeping of employees whose alcohol problems would previously have resulted in rejection or firing. ln-plant meetings of AA and active outreach by recovering AA members was central to those early programs. What appears to have been the first "formal program“ was at Caterpillar Tractor Company, which in 1945 developed a comprehensive company-wide mental health program. Interestingly, this was an explicitly EAP approach similar to today's broadbrush programs, for though alcoholism treatment was included, that was not the main focus of the program. A different, parallel development took place in the "employee counseling programs" in other organizations of the time. These were typically based in personnel departments, as exemplified by Western Electric's Hawthorne Plant and Kennecott Copper Company's Chino Mines Division. Later years saw the appearance of industrial and l4 organizational (I/O) psychology and the industrial social welfare movement, both of which share some of the concerns of the EAP movement but have different foci and methods. One of the most important contributions to the initial efforts to gain acceptance of EAP approaches was the Yale Plan, developed over a period of years by Henderson and Bacon (1953), and disseminated by them to business executives and managers. They introduced the concept of the "half-man" who is at work but unproductive and unsafe. They also described another ”half-person," the woman whose husband is alcoholic and drives her to distraction with worry and repeated domestic crises. The plan was presented as both attitude changes and institution of policies and role expansions. The Yale Plan had many of the elements of current EAPs. The major steps were: 1. Change attitudes and educate top management about the problem of alcoholism and gain their support. 2. Assign responsibility for addressing the problem to a department of the company, typically medical or personnel or employee relations. 3. Identify an individual to carry out the program, an alcoholism rogram supervisor. 4. Mobilize plant facilities and company resources to provide eatment, including use of AA members and groups. 5. Use supervisors to develop plant policy and practices regarding :ipline, severance, retirement, and so on of problem employees. 6. Establish services of counseling and referral. The supervisor seen as able to fill that role initially. The duties are to identify Jcoholic employee, to interpret his problem to him, and to refer him 15 to appropriate means of recovery. A trained counselor may be needed as the caseload develops and the program gains acceptance. 7. Give a "leading role" to supervisors and administrators. "It is not suggested that the supervisor assume the role of a practiced counselor. Rather, because of his closeness to the employee and his knowledge of his habits and temperament, he should act as liaison between the employee and the responsible division within the plant" (Henderson 6: Bacon, 1953, p. 260). Conduct an educational campaign in concert with development of The goal is to 8. the alcoholism program and the availability of treatment. increase awareness and accurate knowledge of alcoholism and foster an acceptant attitude towards treatment. 9. Conduct a survey in the plant to get an estimate of the prevalence of the problem in the plant. Remarkably, the Yale Plan's precepts of program design continue to the present virtually unchanged, though the EAP focus has generally been broadened to include all "problem employees" whether the problem is alcoholism or some other impairment. Lotterhos (1981) noted that the early programs erred in conception by placing the supervisor in the role of identifying alcoholic behavior 1 the job and basically trying to turn him into a diagnostician of ness, a chore which he both rejected and for which he was not apared." These early programs were not very successful and developed a Iative image of being "witch-hunts." "The end result of this approach that company programs wound up relating only to the chronic cases of holism" (p. 30). 16 Job-based programs initially tended to be seen as purely management- oriented, but union involvement and interest developed during the forties and fifties. Leo Perlis, Director of the AFL-CIO‘s Department of Community Services supported a nationwide alcoholism program under the auspices of the National CIO Community Services Committee. A union policy was reached of keeping the controversial area of alcoholism out of collective bargaining. However, unions were very suspicious of alcoholism programs, and their support and consultation during the initiation of a program became imperative. In 1950, workers at Allis Chalmers threatened to strike when an alcoholism program was started. Walter Reuther was asked to mediate the matter; when he convinced management to start program development over again with full involvement of the union, the conflict ceased and the program achieved acceptance ‘ (Trice 6t Schonbrunn, I981). The historical response of unions to alcoholic employees has been to represent their job security interests. When workers' behavior, absenteeism, or inadequate performance finally resulted in disciplinary action which threatened their job, the union would defend them through the grievance process and try to get the discipline or termination reduced or reversed. Very often, the union did succeed in gaining reinstatement for alcoholic workers, but as Trice and Belasco (1966) showed, such "last last chances" rarely benefitted the employee, as their alcoholism and poor work performance unremittingly advanced. Eventually, since the cause of the problem had not been addressed, successful grievances and arbitrations were to no avail. Taking the usual route eant that the union would ultimately fail in its intent of saving the l7 member's job. For this reason, unions have become supportive of EAPs, as rehabilitative solutions to the underlying problems are what is needed by troubled workers for true job security. Miner and Brewer (1976) thought that ineffective performance and its management should be considered in the context of a "control model." They listed alcoholism, drug addiction, environmental stress, emotional disorders, and physical and cultural handicaps as frequent causes of employees' substandard performance. Though EAPs are not intended only to "control" absenteeism, inefficiency, and other personnel problems, the implicit or direct claim of EAP advocates is that EAP implementation will impact those problems. Because an EAP operates as a parallel process to disciplinary action, its status in the labor relations process can be ambiguous or confusing. Union leaders have sometimes been suspicious of EAPs, fearing that they would be misused to "label" workers as troublesome and to harass them while circumventing union advocacy and the due process of grievance procedures. As Trice and Beyer (1984a) noted, EAPs were brought into organizations from the domain of the social sciences and rely on an implicit blending of humanitarianism and concern for productivity to gain the attention and support of corporations and managers. "EAP programs differ from earlier programs in that their efforts to improve poor performance involve helping employees with their emotional problems" (p. 246). Also, in contrast with other innovations intended to improve performance, such as scientific management, the human relations movement, and Quality of Worklife (QWL) programs, the EAP focuses attention on the marginal employee, not upon employees in general. Differences are 18 clearly seen between the "human relations" efforts to improve morale and productivity and the early job-based alcohol programs with their rehabilitative focus. A major difference was the lack of open management support, for there was a strong sense of stigma attached to alcoholism, and managers' public position was still a denial of the very existence of such problems in their companies. Cairo (1983) observed that many companies identified themselves as offering "counseling" to their employees, but the actual service provided was little more than a brief discussion between supervisor and employee leading to a referral to an outside agency. Other forms of counseling, career guidance and pre-retirement counseling, are also available in some companies. Little is known of the effectiveness of these and other counseling efforts, as there is a general absence of research on the effectiveness of personal counseling programs in industry. One of the major difficulties with research in this area is the non-specificity of the generic term "counseling," which has been applied to very diverse activities of supervisors and professionals. Out of the early program efforts the "job performance-based" program concepts evolved. The role of the supervisor was considered key to the operation and success of the program, but the role of the supervisor was no longer to spot and confront symptoms of an impairing condition; they were simply to monitor job performance and react to poor performance with confrontation and referral. Use of an employee assistance program is, in theory, only a slight elaboration of the supervisor's usual role. According to Beyer and Trice (l982) one of the factors aiding the spread of EAPs was that implementation of a job-based alcoholism program 19 required only modest resources of an organization, many already present--supervisors whose job expectations are consistent with the policy, benefits and services including health care insurance and paid sick leave, company medical staff, and a training component. "Policy-specific resources needed are usually quite modest relative to the savings the policy could effect" (p. 219). Sadler and Horst (1975) made one of the first efforts at identifying elements of EAP programs thought essential to success: 1. Policy statement affirming disease concepts and potential for treatment; 2. Joint labor/management cooperation and administration of the program; 3. Information and education to workforce; 4. Acceptance of role of identification and referral by first-line supervisors and support for that role by all company managers; 5. Training for supervisors to recognize substandard job performance and to see that the troubled worker is encouraged to seek needed help; and 6. Professional staff services of a specialist who can diagnose, understand, and treat the troubled worker. The "ideal model" for EAPs which came to be recognized by practitioners in the late 19705 was described by Shain and Groeneveld (1980). Several of the elements of the ideal programs involve supervisory action towards the impaired worker. According to them, a program‘s foundations are in corporate policy, with union's concurrence when present. An ideal EAP is "performance—based" and has as its purpose 20 the early identification of problem workers, provision of open, well—known pathways to help for impaired workers willing to self-identify and self—refer, but also having the mechanisms for induction of a job crisis through supervisory confrontation if they do not self-refer. Shain and Groeneveld note that EAPs are based upon an implicit theory of impairment which considers the dysfunctional aspects and problems as stemming from the individual impaired worker, a condition, disease, maladjustment, and so forth. An alternative conception could place the locus of impairment and need for intervention within the environment, societal or corporate. That approach is embodied in various organizational development (O.D.) efforts and in primary prevention programs such as "employee wellness" campaigns. Managers generally accept the performance impairment assumptions of EAPs that personal problems cause poor performance (Trice 6r Beyer, 1984a). In actuality, though, consistent relationships between personal problems and poor performance have not been demonstrated. Also, the converse may sometimes be true-~personal problems may result in improved performance. Attention has increasingly been given to the adverse health effects of "workaholism" and to obsessional traits and anxiety conditions that may find expression in intensified work efforts. Managers generally accept the humanitarian aspect of EAPs, though not as strongly. They believe that an organization has some responsiblility to help employees with their personal problems. A broader, all—inclusive perspective of concern began to predominate in the early 19705. The position was officially taken by the National Institute on Alcoholism and Alcohol Abuse (NIAAA) that occupational 21 programs should move from an exclusively alcoholism focus to a "broadbrush" concern for all troubled workers. This shift in focus was believed to offer two principal advantages-~occupational programming would find greater acceptance when all types of problems were addressed and programs would be more effective since other problems besides alcoholism could cause job performance impairments. This transition has been a source of continuing controversy and conflict among EAP practitioners. A5 a primary force behind adoption of EAPs, the NIAAA position had wide and powerful influence (Roman, 1981c, 1983; Trice 6r Schonbrunn, 1981). Roman (1983) contended that this very significant change from an exclusive alcoholism concern to a broadly mental health concern was based upon very limited rational justifications and little or no empirical basis. The increased awareness of other problems also coincided with the rapid growth of the community mental health movement and a growing involvement of human service professionals, including counseling psychologists, in EAP development and operation. Shain and Groeneveld (1980) compared early programs which had a largely alcohol focus with current EAPs. They note major differences in the approaches. Formerly, supervisors were taught the signs and symptoms of alcoholism and were specifically urged to attempt to identify alcoholics in their work units. Now, at least in theory, the basis for EAP referral should solely be job performance deficits, the behavioral evidence of which are detected by the supervisors. Secondly, modern EAPs make a greater reliance upon self-referrals rather than mandatory referrals. Trice and Beyer (l982a) distinguished between two extreme types of EAPs--the older employee alcoholism programs which focused 22 primarily on alcoholism but tried to treat the accompanying personal problems, and newer EAPs that treat alcoholism as but one of the many problems of living which can adversely affect the job performance of workers. Programs usually find their basis in company policy statements indicating a desire to help employees with assistance when performance is impaired because of a personal problem. The values from which job-based policies were developed (Trice 6r Beyer, 1982a) capitalize on social controls present in the workplace and inherent in the employer-employee relationship. Those social controls are based upon internalized values regarding work, social distance between supervisors and their subordinates, and group membership bonds and expectations. Several factors have contributed to the recent rapid diffusion of EAPs into public and private organizations. It is widely believed that there is a "productivity crisis" in the United States. As one of the interventions which promises to increase productivity, and since there is some good evidence now that EAPs do in fact assist in rehabilitation of impaired workers, the appeal and acceptance of EAPs is well-founded. It has been estimated (Roman, 1983) that there are now over 8,000 work organizations with EAPs in some stage of development. The broadening of the approach has also opened the field to numerous human services workers and counselors who now find professional interest and employment in the employee assistance movement (Forrest, 1983; Leonards, 1981; Roman, 1981c; Toomer, 1980). Furthermore, the "treatment industry" for alcoholism, drug addictions, and emotional and mental health problems have created pressures of their own for greater utilization of their 23 services, and EAPs are often seen as an effective means of case-finding. Those somewhat self-serving motives are thought to exert pressures on the program design and operations themselves (Roman, 1983; Trice 6c Beyer, 1984a; Trice 6: Roman, I972). The increased numbers of providers, EAP practitioners, and treatment facilities have contributed to a proliferation of programs through aggressive "marketing" of EAPs to organizations. The basic employee assistance "program" has been largely unmodified in theoretical conception since its description by Henderson and Bacon (1953). The activities of "programs" are not standardized, however, and sometimes the adapting and incorporation of elements of an "ideal EAP" within a particular organizational setting defy precise description beyond the general enumeration of those elements claimed to be present and their relative emphasis, officially and in actual practice. As Roman (1983) and Trice (1980) have pointed out, the employee assistance field has developed and spread largely without a solid empirical or theoretical underpinning. It remains for employee assistance practitioners, using the methodologies and theoretical contributions of the social sciences to demonstrate the validity, necessity, and utility of each of the purported elements of EAPs, including constructive confrontation, notwithstanding their unquestioned acceptance and widespread adoption. This section has traced the development of the employee assistance movement. Organizations now have an alternative and parallel process to support progressive discipline in managing the problem employee. Efforts to rehabilitate the impaired substandard employee very frequently succeed 24 following a timely EAP intervention. The next section describes the "crucial role" assigned supervisors in the typical EAP. History of the Constructive Confrontation Strategy As EAPs have gradually evolved in their conception and operation, so too has the role of supervisors with respect to them. Supervisors have always been seen as having an important role in EAPs. In fact, EAPs were "invented" specifically to provide first-line supervisors with an additional tool to use with "problem employees" who do not respond to "the usual" supervisory interventions. A central feature of the typical, or ideal, EAP is the supervisory referral of a troubled worker for diagnostic, referral, or treatment functions by EAP staff or service provider. The timing and method by which this supervisory action occurs has come to be known as the "constructive confrontation." Constructive confrontation is essentially a motivating strategy consisting of threats of job loss if a worker with a problem fails to get needed help and restore job performance to an acceptable level. Since job and income are very important to the worker, the recommendation to seek help is very persuasive to the impaired worker, and the convincing threat of job loss begins to break down psychological denial of a problem and resistance to treatment. Constructive confrontation, as it is now called, is thought by most EAP practitioners (Roman, 1982) an effective and necessary element in an EAP. The development and history of the strategy have been recorded by Bayer (I987), Riediger (1979), Sonnenstuhl and Trice (1986), and Trice and Beyer (l982b, 1984b). Constructive confrontation continues to be a 25 central element in EAPs, though the supervisors' role in offering specific treatments has been shifted to an EAP staff role; the supervisor now refers to the EAP rather than to specific resources for help (Trice 6c Beyer, 1982b). The constructive confrontation strategy had its origin in one of the earliest alcoholism programs. Dave M., a recovering alcoholic, was hired in 1943 by Dr. George H. Gehrman, the medical director of DuPont, specifically to help with problem drinkers. The program functioned with a rather primitive "constructive coercion" untempered by current labor relations practices which insure workers' rights to due process and grievances. Their approach, which was formalized into DuPont company policy, began with a supervisor's confronting a worker about poor job performance. The supervisor made an effort to determine the cause of poor performance, and if alcohol was felt to be a factor, the worker's name was sent to the medical department. Dave M. would then quietly approach the employee and inform him that he had been identified as having a drinking problem. He would tell the employee about Alcoholics Anonymous and the group meetings available in the Wilmington, Delaware area. Dave M. would make it clear that it was entirely up to the employee whether to join AA or not, but unless steps were taken to completely overcome the job and personal problems, he would be fired. The employee was placed on probationary status for the next three months, and a decision was made in the medical department whether to terminate. Despite its crudeness, or perhaps because of it, the coercive approach at DuPont was seen as effective. Dr. Gehrman claimed that in his previous 28 years he had had no success with alcoholics, but after hiring Dave M., 26 over 65% of the referred alcoholics were restored to full productivity (Trice 6r Schonbrunn, 1981). In the early 19605, Harrison Trice began to advocate a strategy of causing a job security crisis for alcohol-impaired workers, while simultaneously offering them an alternative to drinking. The concept of confronting the drinker with negative consequences balanced by positive acceptable alternatives was supported by authorities on alcoholism, since it attempted to deal realistically with the primary dynamics of alcoholism--denial of the problem and loss of self-control as the condition advances. Also, the strategy was thought generally supported by applied social science research and accepted labor relations practices whereby people are motivated to change their behavior when given appropriate feedback and reinforcements. Constructive confrontation was partly derived from a principle of Alcoholics Anonymous, "hitting bottom." Roman and Trice (1976) described it as "a rather strong presentation to the employee of the facts of his deteriorated performance due to drinking, coupled with the offer of treatment assistance" (p. 463). During the confrontation, the supervisor is to state that failure to take action about his problem and improve his performance would lead to discipline or even dismissal. "Thus, constructive confrontation or coercion is used to precipitate a crisis that parallels 'hitting bottom' but it occurs at a much earlier point than a natural devolution to 'bottom'" (p. 463). Roman and Trice (1968) first used the term "constructive coercion" and described its use in an occupational alcoholism program. The confrontation of the employee who allows alcohol to enter in any way into his work role involves a simple statement that repetition 27 of this act will lead to termination. There is no referral because such referrals are not necessary if confrontation occurs at this It should be stressed that this is not a policy where an point. individual is confronted after alcohol interferes with his work performance; rather, the simple presence of alcohol in the form of drinking on the job or hangover is regarded inherently as impairment of performance. (p. 249) This approach was typical in the early programs, and may be seen as an extension of the workrules forbidding alcohol. The "hard-line approach" of intolerance of alcohol was proposed to reduce the ambiguity of the situation for the supervisors. Otherwise, extensive training would be needed to allow the supervisors' making fine distinctions of The use of this whether or at what point the employee was "impaired." early version of constructive coercion was seen as "breaking up" an inevitable progression into alcoholism. In the early EAP literature, questions began to be raised as to why Leggo (1967) found that supervisors did not deal with known alcoholics. many supervisors believe alcoholism provides a "built-in excuse" for any antisocial act or inadequate performance and that discipline would be As a general principle of "transferring the unjustified with them. responsibility to the employee," Leggo exhorted supervisors to confront the impaired employee and set limits and expectations for improved He encouraged the supervisor to discuss the employee's performance. problems, and "is advised to recommend that the employee refer himself to Leggo Alcoholics Anonymous or a local alcoholic [sic] clinic" (p. 98). also recommended that supervisors become extensively involved with the employee’s problem, by encouraging the workers to disclose the nature of their problems, even going so far as discussing treatment and prognosis As the strategies of supervisory with their physician or help provider. 28 referral developed, such intimate involvement with workers' problems came to be seen as very undesirable. Another problem had to do with the terminology used in programs. Heyman (1976) noted that the term, "constructive coercion" antagonized many groups, particularly the labor unions, to whom it implied negative connotations and consequences. She recommended "constructive confrontation" as a more palatable term, as it suggested a "more positive and structured strategy in which everyone's best interests could be served" (p. 901). In 1972, in their important collaboration Spirits and Demons at Work, Trice and Roman articulated the constructive confrontation strategy as it They termed it "the central is currently and most widely applied. The principles of EAP program rationale strategy" of job-based programs. and the use of constructive confrontation were succinctly restated as: 1. Drinking behaviors that forewarn of chronic problem drinking tend to manifest on the job relatively early in the development of the condition; 2. Impaired performance breaches the fundamental contract between employer and employee, giving the employer legitimate right to intervene in the life and problems of impaired employees and offer constructive help; 3. The impact and effectiveness of the employers' intervention lies in the emotional and material investment of workers in their jobs, for security and meaning to their lives; 4. Intervention consists of a series of confrontations of the impaired employee directed toward the creation of an eventual "crisis" for the employee, who then resolves to "do something" in order to restore job performance to acceptable levels. (Trice 6: Roman, 1978, pp. xiv-xv) In one of the early books on EAPs which contributed to dissemination of program methodology, Wrich (I980) outlined his recommended procedures for supervisors to follow in using an EAP: (a) document with specific dates and time changes in the employee's attendance and performance; 29 (b) if in doubt whether substandard performance is due to a poor job fit or lack of skills or a personal problem, the supervisor should consult with EAP staff; (c) discuss the situation and the job problems with the affected employee; ((1) resist the impulse to diagnose the nature of the employee's personal problems and recommend no solutions besides referral to the EAP; (e) supervisor should review the performance problem situation with OWn higher level supervisor; (f) confront continuing performance problems by again discussing problems with affected employee (and union representative if indicated) and threaten disciplinary action; firmly recommend that employee use EAP services and make appointment for employee leaving option of attendance at session with the employee; (g) Check whether the employee did, in fact, attend EAP interview; (h) continue to monitor employee's job performance, but be sensitive to employee's needs while treatment, recovery, or job re-entry occur. The mode of supervisory intervention which Wrich (1980) described assumes a high level of concern and awareness of employees, resulting in a "gentle persuasion" of employees towards an acceptance of EAP referral. He recognized that there are some conditions under which this type of intervention may be difficult or impossible, and he believes that its success depends upon the prior relationship between the supervisor and the employee. Wrich recommended that supervisors be trained in "pre-job performance problem interventions" (p. 55) and that equal emphasis on other motivational techniques should be given. In Wrich's experience, the usual supervisory techniques of additional training, closer monitoring and providing feedback on performance, or job reassignment will handle 90% of employees' performance problems. For the other 10%, 30 those approaches will be ineffective or counterproductive. The difficulty for supervisors is "knowing their own limitations" and when to adopt another stance with a poor-performing employee. Constructive confrontation is "constructive" because it (a) expresses emotional support and group concern for the employee's welfare; (b) emphasizes that group membership can be maintained if the employee can be more conforming; (c) suggests alternative behaviors available to regain satisfactory work performance and provides access to possibly necessary assistance. It is "confrontive" because it (a) reinforces the organizational norms and values for adequate, productive performance,; (b) reminds the employee that he/she is not meeting those expectations adequately and that sanctions will follow if the expectations continue to be violated; (c) establishes social distance between the employee and the conforming group members, setting the stage for application of punitive sanctions if needed (Sonnenstuhl 6C Trice, 1986; Trice 6t Roman, 1978). The role of the supervisor in relation to the troubled worker and the EAP has been defined as involving four functions: (a) identification of a problem employee; (b) documentation of employee‘s inadequate role performance; (c) confrontation of the employee with the record of poor job performance; and (d) referral of the employee to the EAP. (Trice 6c Roman, 1978) Kurtz, Googins, and Williams (1980) added another function to the supervisor's role: (e) re-integration of the troubled employee back into the work group following treatment absence. Each of these functions has proven problematic and presents potential barriers to -—-«- 4.4.1; _. 31 apprOpriate and timely use of the EAP (Googins, I979; Kurtz, 1982; Trice 6: Roman, 1978; Wilcox, 1984/1985). Trice and Roman (1978) asserted that use of the constructive confrontation strategy does not require any new or additional skills of the supervisory personnel. Beyer and Trice (1982) stated that the constructive confrontation strategy could be easily and readily introduced into organizations, since it capitalizes on processes that already exist--behaviors of direct supervision, actions and roles of union stewards and officers, supervisory training, performance expectations, employer concern for employee health and welfare, and utilization of health care benefits and services by employees. Trice and Beyer (l982a) claimed that basing EAP referral on job performance and delegating a referral role to supervisors through an EAP policy and constructive confrontation are thoroughly consistent with the general dynamics of supervision and organizational leadership. Constructive confrontation requires two types of behavior from the supervisor -- attentiveness to workers' outputs and a willingness to apply sanctions when employees do not perform up to the standards, and a willingness to offer supportive help to an employee with a personal problem. _ Supervisory behaviors tend to cluster into two catagories that parallel these requirements: one task-oriented and concerned about productivity, and the other employee-oriented and concerned about social and emotional factors. A central issue in research on general supervisory behaviors and also on implementation of job-based policies is to what degree most supervisors can successfully combine these two types of behavior. (p. 961) Sadler and Horst (1975) described supervisors as production oriented, and readily able to accept an alcohol control program once they 32 see that it will help them deal effectively with the impaired problem worker. They may have an initial resistance because of feeling already overburdened or expected to fulfill a counselor role for which they are unprepared. They may also have considerable difficulty with the documentation aspects of an EAP. "Judging when an employee's work slips below the standard is something else. Everyone has bad days and makes mistakes. Standards of acceptable performance simply do not exist for many jobs" (p. 30). The practices prescribed by the constructive confrontation strategy are not unique to it or to job—based occupational alcoholism and EAP efforts. The constructive confrontation discussion is similar to the informal, oral discussions and warnings recommended in the labor relations and personnel literature when an employee‘s work is substandard. Confrontations of unacceptable job behaviors are to occur in a "progressive" manner, in which increasingly severe disciplinary action that might follow is discussed and positive offers of help for remedying the situation are given. When employees do not respond satisfactorily to such discussions, progressively severe discipline is given. The contribution of EAPs to progressive discipline as an employee relations and behavioral control technique is the prediction that personal problems are likely to be present or the cause of many job performance problems and that positive offers of assistance are emphasized. In this manner, employees can address and remediate the bases for their poor performance and respond in a lasting way to disciplinary action. 33 Trice and Roman (1978) described the constructive confrontation as involving several separate encounters between supervisor and substandard worker. The first discussion mentions poor work and offers constructive help without stigma if the worker wants it. If the poor work continues, a second confrontation meeting is held, showing the employee documented evidence of continued unacceptable performance and reminding the employee of available health benefits with which help for a possible problem may be obtained. If the employee cannot improve performance and keep it at an acceptable level, another confrontation meeting occurs; the supervisor refers the employee to the medical department (or EAP staff) who suggests that some problem is very likely to be the reason for continued job performance difficulty and a warning is given that continued problems will place the employee's job in jeopardy. The availability of resources for help is again stressed to the worker. A clear threat of future negative consequences of continued poor job performance is made, while help is offered. If problems continue, the threat is carried out. Trice and Beyer (l982b) indicated that the constructive confrontation strategy is “consistent with the prevailing management ideologies and cultural values regarding discipline. The strategy utilizes elements of active intervention which make the wrongdoer accountable for the deviant behavior" in combination with offers of support for changes in order to preserve group membership (pp. 28-29). However, "it is crucial to note that the goal of constructive confrontation is employee action that returns his or her performance to an acceptable level. The use of confrontation to pressure an employee 34 into treatment is inappropriate, as is an attempt to pressure him to cease drinking" (Trice 6r Roman, 1978, p. xv). The series of increasingly formal engagements between supervisor and troubled worker was investigated by Kurtz, Googins, and Williams (1980) in a series of interviews with supervisors. "Coffee talks" were informal discussions where a supervisor noted poor job performance or problems and expressed concern; "encounters" were still informal, but longer and in stronger language which warned of formal action; and “confrontations" were formal meetings, usually including the union steward, and often the next-level supervisor, where the employee was presented documented inadequate job performance. The documentation was typically poor, though, as the supervisor belatedly realized that more and better documentation was needed since the employee's behavior had not shown the hoped-for improvement. Sonnenstuhl and Trice (I986) advised patience and repetition in doing constructive confrontation. "A supervisor may need to hold a number of discussions with an employee whose performance is unacceptable. In the confrontational part of the initial discussion, the employee is given the specifics of unacceptable performance and warned that continued unacceptable performance is likely to lead to formal discipline. In the constructive part, supervisors remind employees that practical assistance is available through the EAP" (p. 25). Kaden (1977) also described supervisory confrontation of troubled workers as "a series of warnings of increasing severity, to the point of final confrontation" (p. 368). He also used two new terms for the strategy: "benevolent coercion" and "supportive confrontation." 35 Roman and Trice (1976) commented that it is the "persistent and accumulative impact" of poor job performances on a supervisor which will eventually lead to exercise of supervisory authority through the labeling of deviant behavior and confrontation of the problem employee. They believe that a "simplistic formula" for appraisal or documentation of job performance is impossible. They observed (1976, p. 498) that "impaired performance is what supervision defines it to be," and supervisors must be relied upon to use the EAP appropriately. Dancey (1973) was one of the first to speculate openly why such a technique offering "miraculous benefits" is not more widely used in the actual settings. He proposed that persons in authority may have drinking problems themselves, and they scuttle the EAP to cover~up their own impaired behavior. He also suggested that the existence of a company alcoholism program demands strong support and numerous resources of a host organization--staff, training, special counselors, and a "top-flight health department." He suggested that the constructive confrontation technique which works with alcoholics should work equally well with abusers of other drugs. Dancey implied, in his hearty recommendation of the strategy, that supervisors will readily and effectively use the power of their relationship to push an employee towards needed help. But as numerous other researchers have found, there is a serious "policy-practice gap" between the institution of an EAP and the formal policy supporting constructive confrontation and the acceptance of that program and routine use of the policy in the form of program referrals by the supervisors. Evidence continues to accumulate (e.g., Bayer 6t Gerstein, 1989; Foote 6t Erfurt, 1989) that acceptance of the constructive 36 confrontation strategy by supervisors has not been rapid or complete by any means. The technique, though simple in conception, has proven complex and quite problematic in practice. As Riediger (1979) commented, the use of constructive confrontation is a very "high risk" activity for the supervisor, who often feels that the employee's welfare, the employer's interests, and his/her own authority and credibility are at stake in a delicately timed complex judgment. Use of the approach calls for a strong exercise of power and authority, while the workplace generally has been becoming less authoritatian and is encouraging egalitarian values. Those general trends in the social atmosphere of work combine with a host of specific barriers and resistances, resulting in the observed failure of supervisors to apply constructive confrontation as "theory" has recommended. The constructive confrontation strategy has evolved over the years from the relatively stand-alone technique of its beginnings, altering an impaired worker's job performance largely by means of threatened actions without any formal treatment, into primarily a casefinding activity in support of the EAP staff. Though it is accorded high status in the social technology of EAP, being considered a "core element" of practice and program design, the constructive confrontation strategy merits closer attention and study if it is ever to achieve full effectiveness within the typical organizational setting. This section has described the tool supervisors are encouraged to use with their employees showing job problems--constructive confrontation. Those in the EAP field created a distinct referral task 37 for the supervisors, but evidence from several lines indicates that they have been chary in its performance. The next section will explore the attempts thus far to empirically differentiate the factors which facillitate and impair the use of that tool by the supervisors. Factors Affecting Use of Constructive Confrontation The Empirical Research Base There is a limited, but developing literature related to the factors which might affect supervisory referral actions to an EAP. Analysis of program data, surveys, and correlational studies have been the means of investigating EAP. Of the factors examined thus far, some are intrinsic to the troubled employee, some characteristics of the supervisor, and others relate to the organizational context and the EAP focus. Conceptually, the factors are broad. They have been addressed repeatedly in theoretical and expository articles about establishing and operating EAPs. The number of empirical studies is relatively few, and their methodological rigor quite variable, but the factors have shown a fair degree of consistency in their effects upon referral actions. This investigator has grouped into three tables the known and proposed variables which affect supervisors‘ referring to EAPs, with an estimate of the strength of effect as described in the cited studies. (See Tables 1, 2, and 3.) The estimates are impressionistic only, as the studies considered empirical ranged from large, rigorously drawn extensive surveys with specific hypotheses, to interviews with a small group of supervisors from a single setting. An understanding of the multiple influences on supervisory use of 38 Table l Supervisor-related Variables Affecting EAP Referral Actions Factor affecting referral Empirical study Effect Familiarity/knowledge Beyer 6c Trice (1978) ++ of EAP Googins (I979) ++ Riediger (1979) ? Demographic Variables Age (older) Daghestani et a1. (1976) Beyer 6t Trice (1978) Riediger (1979) Googins 6t Kurtz (1981) Sonnenstuhl 6t Trice (1989) Experience (increased) Beyer 6c Trice (1978) Googins 6t Kurtz (1981) Krucher (1986/1987) Young, Reichman, 6c Levy (1987) Education (increased) Beyer 6r Trice (1978) + Iol++-Q++++ Gender effects Status (male) Beyer 6c Trice (1978) '? Riediger (1979) + Cahill, Volicer, 6c Neuburger (1982) + Cahill (1983) + Krucher (1986/1987) 0 Young, Reichman, 6c Levy (1987) +7 Status (cross-gender) Cahill (I983) - Training Belasco 6t Trice (1969) +? Beyer 6t Trice (1978) +? Googins (1979) +? Shain 6c Groeneveld (I980) ? Cahill,VoIicer, 6t Neuburger (1982) +? Reichman (1983) '? Young, Reichman, 6t Levy (1987) + Gerstein et a1. (1989) Googins, Schneider, 6: Colan (l989) +'? Sonnenstuhl 6r Trice (1989) +'? 39 Table l (cont'd.) Factor affecting referral Empirical study Effect Identification with the role Trice (1965) - of the counselor Trice 6c Belasco (1965) - Trice 6c Beyer (1977) - Beyer 6c Trice (1978) - Cahill (1978) ‘ Wilcox (1984/1985) - Outright rejection of the Riediger (1979) - constructive confrontation Googins (I979) - strategy Kurtz 6c Googins (I979) - Googins, Kurtz, 6: Williams (1980) - Wilcox (1984/1985) - Previous successful referral Sonnenstuhl 6c Trice (1989) +? Personality/Attitudinal Factors Attitudes toward drinking Googins (1979) 0 Attitudes toward women Young, Reichman, 6c Levy (1987) ? Attitudes on women drinking Young, Reichman, 6r Levy (1987) O Expectancies of intervention Googins (1979) + Positive expectancies Wilcox (1984/1985) + Young, Reichman, a Levy (1987) +? Negative expectancies Sonnenstuhl 6c Trice (l989) -'? Sympathetic attitudes Belasco 6c Trice (1969) -- toward employees Googins (1979) - with problems Trice dc Beyer (l982b) - Use of confrontational Krucher (1986/1987) + behaviors Expect "trouble" if refer Sonnenstuhl 6c Trice (l989) -? Psychometrically measured personality dimension Krucher (1986/1987) +2 Key = Strong increase (++) through no effect (0) to strong decrease (--) ? indicates Uncertain or non-significant trend effect 40 Table 2 Employee-related Variables Affecting Supervisory EAP Referral Factor affecting referral Empirical study Effect Severity/duration of Daghestani et a1. (1976) ++ impairment Beyer 6c Trice (1978) ++ Riediger (1979) ? Trice 6r Roman (1978) + Shain 6c Groeneveld (1980) ? Googins 6c Kurtz (1984) '7 Beyer 6c Trice (1984b) ++ Bayer (1987) ++ Bayer dc Gerstein (1989) ++ Sonnenstuhl dc Trice (1989) + Triggering Incident Heyman (1976) + committed by employee Kurtz, Googins, 6: Williams (1980) + Googins 6c Kurtz (1984) + Beyer 6r Trice (I978) ++ Gender effects Status (employee male) Beyer 6r Trice (1981) +? Cahill, Volicer, 6r Neuburger (1982) + Cahill (1983) Mastrich (1985/1986) Sonnenstuhl (1986) Brodzinski 6c Goyer (1987) Status (employee female) Reichman (1983) Kleeman 6c Googins (1983) Occupational level Trice 6c Beyer (1977) of employee (higher) Beyer 6r Trice (I978) Riediger (1979) Sonnenstuhl (1986) Key = Strong increase (++) through no effect (0) to strong decrease (--) ? indicates Uncertain or non-significant trend effect 41 Table 3 Organizational/Program Factors Affecting Supervisory Referrals Factor affecting referral Empirical study Effect Program focus on Cam et a1. (1983) + alcoholism Shain (1985) ? Sense of organizational Trice 6: Roman (1978) + support Sonnenstuhl (1986) + Beyer 6r Trice (1978) ? Sonnenstuhl 6t Trice (1989) + Presence of labor union Beyer 6c Trice (1978) + Beyer, Trice, 6: Hunt (1980) + Large span of control Beyer 6c Trice (1978) + + Beyer 6c Trice (1984b) Young, Reichman, 6c Levy (1987) + Good line/staff relations Daghestani et al. (1976) + Wilcox, (1984/1985) + Key = Strong increase (++) through no effect (0) to strong decrease (--) ? indicates Uncertain or non-significant trend effect 42 constructive confrontation is accumulating, but is not yet sufficient to guide EAP program operation or training efforts. Thus far, no study has explored the simultaneous action of all or even most of the factors, so it is not yet possible to estimate their combined or relative capacity to account for the desired referral actions by supervisors. Familiarity/knowledge of EAP. A first factor to consider relevent to supervisory use of an EAP is program knowledge or familiarity. It seems self-evident that supervisors will not constructively confront and refer employees if unaware of a program's existence or how they or the employee can access the EAP services. Trice and Roman (1978) observed in their original exposition of the strategies of supervisory referral that there are several ways in which supervisors must be supported in their use of EAP and the constructive confrontation. There must be a written organizational policy covering all medical-behavioral problems, and that policy must be widely circulated and known to all members of the organization. As Kurtz (1982) stated, "neither supervisors nor employees are likely to use a plan or program that they do not know about" (p. 276). The problem of program awareness and publicity is very practical and elemental, but for many reasons supervisors' knowledge of the EAP cannot be simply taken for granted. Googins (1979) found in his supervisory sample that 21% of the first-line supervisors did not know of the existence of the company EAP, even though it had been operating for over 12 years! Riediger (1979) reported that 23% of the supervisors he interviewed felt that they "didn't know how to use the EAP," and only 27% indicated that they do "know how to use it." j l \ Illliilllllllll A 43 The importance of this variable has been generally supported by research findings. In their study of implementation of occupational alcoholism policy and programs in federal agencies, Beyer and Trice found that "for past use, familiarity is the most important predictor, or equally important with assessed need" (1978, p. 64). They found that supervisors must be thoroughly familiar with the EAP and its supporting policy and referral strategies if they are to use them routinely and effectively. A strong association between EAP referral and familiarity with it and knowlege of the company policy regarding alcoholism was found by Googins (1979). Referring supervisors had a greater awareness of employees who went on their own initiative to the EAP or upon referral by another supervisor. "It is clear from these data that absence of familiarity acts as a powerful barrier, limiting the alternatives that supervisors consider in handling an alcoholic employee" (p. 84). Googins also found that supervisors' personal drinking habits and knowledge of alcoholism and attitudes towards drinking were not significantly associated with their referral actions. He also reported that knowledge of the physiological effects of alcohol and of alcoholism on health or of the symptoms of alcoholism did not affect the supervisor's referral actions. Program—specific and program-relevant knowledge is important, but apparently knowledge related to the problems EAPs are intended to impact is not (Googins, 1979; Googins 6c Kurtz, 1980). The findings of Riediger (1979) were ambiguous as to the importance of knowledge as a factor. "Knowledge of the program does not appear to be associated with program utilization" (p. 177). He found an 44 association between program knowledge and number of employees supervised, however, suggesting that supervisors who had greater potential need for the program and who might be expected to utilize it more had been given less opportunity to learn about it and become comfortable with EAP concepts and procedures. He reported a discrepancy between actual use of constructive coercion and attitudes about it, affected by program knowledge. "Knowledgeable supervisors endorsed the use of constructive coercion with alcoholic employees much more strongly than did the other groups. In addition, they were less opposed to the use of this technique with other problem employees. They disagreed more strongly as well with the concept that most problem employees are willing to accept treatment voluntarily" (p. 184). This factor might well be nonlinear in affecting supervisory action. Presence of a certain quantum of program awareness might be necessary for supervisors even to consider a job—based EAP referral when dealing with a problem employee. Beyond a certain level, however, additional program knowledge might not influence a supervisor's readiness to confront and refer. Training. A component in any typical EAP is supervisory training. The training effort takes a number of forms. Sometimes it consists of a brief overview of company policy on troubled workers or alcoholism, how to make a referral, and an introduction to the EAP staff. Other efforts are much more lengthy and involving for the supervisor and may cover extensively materials about addictions, mental illnesses, and treatment resources. The intent of supervisory training, in theory, is not only to inform the supervisors about the existence of the EAP and their role in 45 it, but to dissipate any resistances they may feel, gaining their active Support for the program. Gam, Sauser, Evans, and Lair (1983) advocated training of supervisors to enable them to assess performance accurately, to become aware of the signs of stress disorders, and to use the appropriate procedures of confrontation with ineffectively performing employees. Myers (1984) identified lack of training as a main reason for supervisors' reluctance to confront and refer employees. Roman (1982) stated that the link between the EAP's goals and a supervisory readiness to confront and refer employees was only assumed to be closed by supervisory training. He thinks this may be based more upon faith in the "deeply respected American panacea of education" (p. 372) than upon sound evidence. He commented on the absence of rigorous study of training outcomes other than that done by Belasco and Trice (1969). Roman stated that EAP's assumptions about supervisors‘ being tough-minded, task-oriented, and eager to identify and refer deviants in their charge are wishful fallacies, and he opined that "the implementation of programs which involve substantial use of constructive confrontation requires an orientation to organizational change rather than individual change" (1982, p. 372). As such, supervisory training must be effective in creating and maintaining organizational suppport for the supervisor in the routine use of the EAP through referrals. The implied desired outcome of the training is that supervisors will become able and willing to conduct constructive confrontation interviews. Evaluation of training as it affects supervisors' formal and informal referral actions is virtually never done. Overall, the research which 46 has been reported shows a mixed or absent effect from training as it is currently conducted. The largest, most rigorous, and most influential study of supervisory training related to EAP was by Belasco and Trice (1969). Their evaluation research determined that the training sessions did, to a degree, cause supervisors to feel more prepared to confront workers over their job problems. "The differences, however, between the trained and untrained groups were small” (p. 91). After training, the supervisors had increased knowledge, more favorable evaluations of employees, a view of problem employees as more potent and active attitude objects, and a more constructive orientation towards them. There was a definite interaction between training and pretesting. They concluded that the research testing "sensitized" the supervisors to the training content, which made the training more potent for the pretested group. "In short, training is relatively impotent in producing change, while the change it does produce can be traced to the content of the training experience and the visibility of the change sought" (p. 110). A rather counter—intuitive finding was also reported by Belasco and Trice. They began their investigation with an expectation that negative attitudes. towards alcoholics and alcoholism were associated with supervisory reluctance to confront and refer troubled employees. Their data on the evaluative studies was surprising. Firstly, the strongly negative attitudes did not exist, and secondly, positive, tolerant attitudes were the barrier to confrontation and referral. When training served to lower the supervisor's tolerant attitudes towards deviant behavior, there was a significant increase in declared intention to 47 refer. "Our data suggest that the supervisor's reluctance to exert sanctions against the alcoholic stems from his favorable attitude toward him and not from the stigma at all" (Belasco 6r Trice, 1969, p. 157). It appears that effective training, made more potent with some form of pretesting, forces the supervisor to become more realistic and abandon previous willingness to "string along" with poor job performance and take a positive action. This results partly from the instillation of more negative attitudes towards troubled workers as a by-product of that training. This investigation of the effects of training suggested that the emphasis in EAP supervisory training should be on general management and supervisory issues and strategies, not on the problem employees and their conditions. Training which refines supervisors‘ job performance expectations and strengthens their orientation to actively manage their workers, using all available staff resources, should result in increased use of the EAP by supervisors (Belasco 8r Trice, 1969; Trice 6t Belasco, 1968). In their study of EAP implementation in the federal sector, Beyer and Trice (1978) found only a slight relationship between hours of supervisory training given and past use of the program by those supervisors. Training hours were not predictive of anticipated future use or of assessed need for the program. They concluded that training impacted in a general sense "familiarity" with the program, and was an essential, but not sufficient factor in program use. After their training, supervisors did not remember the alcoholism policy accurately, but only in terms of the implicit general social norms against impairment. In spite of their training, supervisors frequently endorsed as possible job actions quite beyond the intent of the EAP policy and unsupported by it. Beyer and Trice (1982) endorsed the approach of EAP training as general supervisory skills training. They propose that one of the main attractions of EAP implementation is the opportunity to upgrade and refine supervisory skills. "Supervisory training about job-based intervention strategies provides an excellent opportunity to train line managers how to be more generally effective, by being concerned about both task performance and employee welfare" (p. 198). They also identied ingredients in EAP-specific training which might possibly affect supervisory use of the program-~familiarity with the underlying policy statements and specific knowledge of the existence of the EAP and its entry portals. In their study of over 300 Canadian organizations, Shain and Groeneveld (1980) found that the number of supervisors trained had no effect on reported rates of referral. They thought that this was because of a typical mistaken focus in the supervisory training on attitude change rather than upon specific behavioral changes in supervisors and upon dealing with organizational constraints against program use. Shain and Groeneveld observed that "although the EAP consultant may be frustrated, the client industry may be getting as much as it wanted out of the policy [and EAP]. That is, the consultant may want a policy that gets down to early identification [of substance abusers]; the industry may simply want to clean up a backlog of multiproblem employees" (pp. 18-19). Certainly their findings remind us that "training" is not a magic carpet of program acceptance as some might wish. 49 As a population often underserved by EAPs, women employees have received special attention in projects intended to increase their participation and referral. Those projects have also added indirectly to the negative evidence about supervisory training effects. Cahill (1983) reported on the results of a special training effort directed to both supervisors and the women employees who comprised about half the workforces of twelve federal agencies hosting the research. The training and educational efforts were concluded to have had some impact, as female employees' use of the EAP was significantly increased, and the proportion entering treatment before being formally disciplined decreased. A specific effect of supervisory training could not be determined, however. It was observed that the proportion of supervisory referrals to self-referrals did not change, and neither did the proportion of alcohol—related referrals. Reporting on that extensive project, Cahill, Volicer, and Neuburger (1982) concluded that "for all types of problems taken together, the intervention [training and program information for supervisors and female employees] resulted in an increase in referral of females from the experimental agencies to the EAP. There were no measured changes with respect to alcohol related referrals" (p. 231). In a similar effort to increase the referral rates of women employees, Reichman (1983) conducted training in four EAPs intended to change supervisors' attitudes and perceptions of problem drinking among women. Supervisors were given both "conventional" EAP-related training and specific training experiences intended to sensitize them to women‘s drinking problems and to change stereotyping tendencies. The desired training outcomes of increased program knowledge and greater imagined 50 likelihood of referral of a hypothetical case were seen. An interrupted time series analysis of actual program referral data from the sites did not demonstrate a training effect, however. Actual referrals declined overall during the three months of post-training observation. Other approaches to supervisory training have been tried besides dissemination of program mechanisms and underlying policy. Hawthorne and Davidson (1983) saw training as instrumental in disentangling the supervisor from an "enabling process" which is part of the personal and family dynamic of alcoholism. They saw a parallel between the stresses felt by an alcoholic's family and the supervisor. All struggle with a confusion of feelings about the alcoholic, which makes them feel frustrated and trapped—~a mixture of hope, compassion, and loyalty in conflict with feelings of anger, resentment, failure, and disappointment. Hawthorne and Davidson believe that supervisors are only receptive to training about the EAP when they have failed repeatedly to solve the situation on their own and have reached an emotional crisis themselves. They recommended training for only those supervisors feeling the immediate need for EAP assistance, consisting of a period of "catharsis" followed by re-direction of their efforts into positive behaviors of confrontation and referral. A somewhat similar approach was suggested by Older and Phillips (1978), who asserted that most of what is done in supervisory training programs to increase referrals is either useless or counter-productive. They asserted that the wrong materials are covered, and the lack of immediacy of need results in a lack of motivation in the trainees to assimilate and later use the training inputs. Though earlier they were 51 advocates of training supervisors in taking the intervention roles of identification, confrontation, and referral (Phillips 6: Older, 1977), they later directed their efforts away from defining an immediate role for the supervisors. They advocate a more limited training goal of having supervisors accept and label their own feelings of distress about problem employees. A second modest training goal is that supervisors would consult with EAP staff whenever faced with struggles of dealing with an employee showing job problems. Similarly, O'Connell (1987) recommended EAPs go beyond ”one-shot training" experiences. He suggested followup contacts with each supervisor, or in small groups, to re—cover the training issues and to rehearse supervisors in how to make referrals. He sees this as an opportunity to answer individual's questions about the program and to deal with their resistances. Googins (1979) observed that program awareness and familiarity develops in a number of ways, including contact with and knowledge of other supervisors and employees using the EAP, voluntarily or otherwise. Training and program orientation for supervisors is a means of generating this factor. In attempting to apply those findings and their implications, Googins and Kurtz (1979) suggested a modification of training content to make it more effective in generating supervisory referrals. They proposed building the training program around a core of supervisors who have successfully used the EAP to handle a problem employee. They conceived that a "supervisory network" of peer models and satisfied users would influence fellow supervisors to accept and use the EAP. A degree of "peer pressure" among supervisors, along with the 52 effects of behavior modelling and the creation of a supportive reference group of referring supervisors was proposed to increase the impact of program information. Implementation of such an approach and an evaluation of its effectiveness has not yet been reported. Young, Reichman, and Levy (1988) reported on a multivariate study of program referrals by 120 supervisors. A difficulty in interpreting their findings is due to their combining "EAP knowledge" and "training" into a single scalar variable. Awareness and familiarity were scored at a low value, and training exposure was assigned a higher value on the variable. Even so, the compounded variable of knowledge/training was statistically associated with the referral of male employees, but not of females, and it was not associated with the identification of either male or female problem drinkers in their work units. Reporting on an extensive telephone survey of EAP practitioners, Googins, Schneider, and Colan (1989) provide evidence that the training currently being provided in organizations is "fairly homogenous" (p. 96). It is typically a stand-up lecture/presentation about identification and referral of troubled workers, a film depicting an EAP referral, and a brief discussion period about program use and benefits. They reported that EAP providers are largely satisfied with their training efforts and see lack of time and opportunity to train as the primary obstacle to maximal impact. They compared EAPs which conducted some supervisory training with others reporting no training activities in the previous year; those doing even some training had a reliably greater "referral rate" (5.2% compared to 2.1%), though the variable of referral rate was poorly defined and inconsistently measured. The authors qualified their L ._- avg; 53 findings as possibly the result of uncontrolled and confounded variables and not a true training effect (pp. 94-95). To the extent that referral action may be facilitated by more accurate and more confident identification of job-impaired employees, Gerstein, Eichenhofer, Bayer, Valutis, and Jankowski (1989) report that EAP-related training is beneficial. Supervisors who had or had not seen a one to two hour lecture on their company's EAP were compared on their identification of employee characteristics as indicators of job impairment. The trained supervisors were slightly more likely to identify the problemmatic behaviors and attitudes constituting job impairment. In the case of "lndustriousness," one of the four impairment factors studied, a strong interaction effect was noted, in which favorable attitudes towards the EAP in combination with the brief training accounted for the accurate identification of the job impairments. In summary, the evidence for training as a factor affecting supervisory referrals is still somewhat equivocal. It is likely that "training" ought not be thought a generic drug, compounded of the same experiences in each organization and having similar effect or lack of effect on all supervisors —— even though it is typically administered that way. Descriptions of training efforts in the literature are quite general and nonspecific, but the findings of Googins, Schneider, and Colan (1989) indicate that the content and process of training is, in fact, fairly homogenous across organizations and the benefits taken for granted rather than carefully evaluated. In order to better assess training's impacts on supervisors and their referral behaviors, it will 54 be necessary in future research, as recommended by Wood and Steiner (1989), to much more precisely define and measure the training inputs and desired short- and long-term effects on their cognitions, attitudes, and actual behaviors. The observable effectiveness of "training" may, at this time, be synonymous with "familiarity" or program awareness, as Beyer and Trice (1982) suggested and as Young, Reichman, and Levy treated them. They are surely not identical conceptually or operationally, however, and need to be more cleanly differentiated from one another and the effects of various components of the training better evaluated. Severity and Duration of the Job Performance Impairment. In his landmark study of the job behaviors of recovering alcoholics, Trice (1962) found the rationale for basing alcoholism intervention programs on job performance. Through the recollections of recovering alcoholics, he found that as drinking problems worsen, employees' work efficiency steadily declined. Procrastination, loss of initiative, compromised standards of work, and erratic performance were highly characteristic of the problem drinker moving into middle-stage alcoholism. Trice concluded "there can be little doubt that work efficiency declines as alcoholism develops" (p. 496). He also made an important and counter-intuitive finding that the presumed relationship between advancing alcoholism and on-the-job accidents is only folklore. By the use of absenteeism and lowered effort and output as compensation for their impaired state, alcoholics had the same or lower accident rates as other workers. Dramatically increased rates of absenteeism, especially for low-status workers, and sporatic work effeciency were consistently related with worsening status. 55 A questionnaire survey of supervisors at Illinois Bell Telephone was conducted and reported by Daghestani, Barlow, Hilker, and Asma (1976). Supervisors' responses (134, a 77% response rate) indicated that absenteeism and deteriorated quality of work were the primary bases for their identification and referral of the alcoholic employee. The supervisors expressed a general sense of concern and responsibility for the well-being of the employees under them. The severity and continuing nature of their job problems had finally prompted them to take action. Beyer and Trice (1978) found that among the supervisors studied in the federal agencies, "perceived need" was the most important predictor of anticipated future use of the EAP. This was measured by supervisors' global indication of whether they felt there was a need for the alcoholism policy and EAP at their agency. Many cognitions, no doubt, affected that response, but followup interviews disclosed that specific awareness of employee situations “fitting" the policy influenced that sense of need. Some have thought characteristics of the drinker or of the job setting might account for supervisory inaction with troubled employees. Kaden (1977) attributed the failure to refer to the "likeability" of alcoholics and to their ability to offer endless, usually plausible excuses for their lapses on the job. Kaden asserted that only when the situation reaches such severity that a "conspiracy of silence" can no longer be maintained will confrontation and referral occur. Riediger (1979) conducted interviews with supervisors which disclosed a number of hypotheses which they held to explain (really, explain away) poor job performances--environmental pressures, attitudinal 56 factors, specific job stresses such as shiftwork or travel requirements, unsuitability for their current job, or a health problem. The supervisors said that they, in fact, made an effort to diagnose the employee's problem before referring (against program design and training) in order "to make sure the referral was appropriate." Supervisors typically were aware of a problem for several years before taking any corrective action. In many cases, the supervisor had attempted to counsel the employee himself and indicated one of the principal goals with the employee was to avoid a confrontation. A supervisor's own performance is judged by the work outputs of the department, so there are ultimate limits of tolerance towards nonperformance. Trice and Roman (1978) identified several factors thought to pressure a supervisor to take action. Other employees and the union expect not to have to bear the financial burden of high absenteeism, shoddy work, or an increased risk of accidents or injury. Also, the supervisor may have undergone training to accept identification and confrontation of deviancy in the workplace as part of the supervisory role. An important factor, they thought, in successful confrontation is an expectation of support by the supervisor's own supervisor and by the union, if present. Routine monitoring of work outputs (ongoing job performance appraisal) and intimate awareness of the subordinates were proposed by Roman and Trice (1979) as the stimuli for supervisory referral. However, problematic aspects of those supervisory elements are little discussed in the literature. One reason is that supervisory relationships with subordinates does not always involve the close performance monitoring 57 assumed to be present. Complexity of the task and dispersal of the workforce are just two reasons why supervisors may be unable to spot poor job performance and promptly intervene. Also, performance evaluation typically includes built-in ratings distortions which are inherent in any subjective observational process. "In short, performance appraisal is a human process that contains the error, biases, and stereotyping present in any decision-making that people make about other people" (p. 497). In spite of those difficulties, Roman and Trice asserted that "it is the persistent and accumulative impact of poor job performance on the supervisor which will eventually lead to exercise of supervisory authority through the labeling of deviant behavior and confrontation of the problem employee" (p. 497). Not all programs have such a low rate of problem identification and intervention. Beyer and Trice (1984b) reported that "the lower the employee's work performance and the greater the employee's work dysfunction, the more they [supervisors] tended to use several forms of discipline short of discharge. Employees seen as having poor social adjustment provoked only confrontive topics in informal discussions," but "those for whom the alcoholism policy was deemed applicable experienced more constructive topics as well as formal discipline short of termination" (p. 754). A severe job impairment is likely to attract supervisory attention and response. Beyer and Trice found their research on the use of discipline and referral to EAPs "reassuring." They found supervisors most often use discipline when they consider the behavior "a relatively serious and disruptive problem and the employee has relatively poor and deteriorating work performance including absenteeism, unexplaine apparent . was, even supervisor performar As 1 employee (or accep 39 interv within 51) drinking. They rep! years! " drinking job-relate (p. 163). admitted insisted ' absentee COmplain mentions PIOblems having 5 said the} Sir Strategy 58 unexplained disappearances, changes in physical appearance, and the apparent inability to Work effectively" (p. 759). Referral to the EAP was, even if often long delayed, based upon a clear judgement by supervisor that the severity and duration of the worker's impaired performance merited special attention and an unusual response. As Googins and Kurtz (1984) reported, the lapse of time between the employee's own recognition of a problem and the supervisor's recognition (or acceptance) of the problem can be considerable. "Only three [of the 39 interviewed] employees reported that their supervisors took action within six months of the development of serious work impairment due to drinking. Most experienced a much longer period [of delay]" (p. 163). They reported that the average delay for supervisory action was 4.9 years! "The occupational [programming] model presumes that problem drinking will manifest itself over time in the workplace through job-related behaviors. The data from this study confirm this hypothesis" (p. 163). They found that about two-thirds of their interviewees admitted job impairments due to their drinking, but the other third insisted that they had had no job impairments. Poor quality of work, absenteeism, drinking on the job, tardiness, wasting time, customer complaints, safety violations, and complaints from co-workers were mentioned as job problems that developed along with the drinking problems. Though 24 of the 39 EAP clients identified themselves as having significant job performance problems, only 15 of the employees said their supervisors thought they had a drinking problem. Similar difficulties in employing the constructive confrontation strategy were documented by Foote and Erfurt (1989). They examined the intake to company. serious jc typically clients he concludec periorma widesprez Findings can ever superviso troubled In ' supervisc Gerstein stimulate that "jot make E; (1989, p. likelihoo SUpervis< referral a 30% c 50% cha In re: re' we 59 intake records and referral sources of M4 EAP clients from a large company. Over two-thirds of the clients reported that they were having serious job performance problems at the time of their coming to the EAP, typically for an alcohol or other drug abuse problem. Only 10% of the clients had been referred by their superviors, however. Foote and Erfurt concluded that "the union and medical department produced more work performance referrals than did supervisors," and "there was evidence of widespread supervisory coverup“ of employees' job problems (p. 33). Findings such as these cast doubt on the degree of effectiveness which can ever be attained in companies whose EAPs rely mostly or wholly on supervisory referrals for their penetration into the population of troubled workers. In their study of employee characteristics indicating impairment to supervisors and likely to result in imagined referrals, Bayer and Gerstein (1989) reported that presence of one, two, or more impairments stimulated referral, though inconsistencies were evident. They reported that "job productivity played a critical role in helping supervisors to make EAP referrals, even though other work-related problems were present" (1989, p. 59). Their analysis of the imagined referral action likelihoods with hypothetical employees gave further evidence that supervisors do not readily or enthusiastically carry out their intended referral role. Hypothetical employees depicted with one job problem had a 30% chance of referral; those with two or more sets of problems had a 50% chance of referral. In fact, the most disturbed employee scenarios generated an average referral rate of only 62%. It appears . . . supervisors were reticent to engage in EAP referrals. It also seems that supervisors were surprisingly tolerant of work adjustment and productivity prol diff supe The occasions until sorr "do some "Th bet con per fit whl ofti Grc A 1 complete between such an the moti (p. 15). Th 10b pert perform; available eventual also abr inordina deficien the lacl 60 problems even when employees were shown to exhibit multiple difficulties related to job performance, attitude, and co-worker and supervisor interactions. (p. 59) The operation of the severity factor may be nonlinear, combining occasional incidents over time into an accelerating curve of awareness, until some critical value of alarm results in a supervisor's felt need to “do something" before getting negative attention himself. "The problem remains, however, that there is a very fuzzy area between the point at which job performance is clearly affected and constructive coercion is warranted, and the point at which job performance is so barely affected that only a self—referral would fit ethical requirements. . . . This, in fact, may be an area which gives supervisors the greatest difficulty in practice. They often simply do not know what to do in marginal cases. (Shain 6r Groeneveld, 1980, p. 15) A possible resolution of the difficulty would be for EAPs to completely rely upon self-referrals in order to create the linkages between impaired employees and needed help. Shain and Groeneveld think such an approach "would create quite a strain on our assumptions about the motivation of employers to be interested and involved in EAP" (P- 15). The available evidence to date rather strongly supports the impaired job performance rationale for EAPs from the perspective that poor job performance and occurrence of personal problems often co-exist. The available research also supports the expectation that job problems are eventually a primary stimulus for supervisory referral action. It is also abundantly established, however, that there is typically an inordinate timelag between the onset of job problems, even serious job deficiencies, and the supervisory response. The causes of that lag and the lack of remedies for it at the present time present real obstacles to EAP cour programs big the first industry. are inten personal in approa primary : emphasis (Roman t Sonnenstr historical involvem offset by and volu'. ROI Which w; Prominer transforr SUpervisr GStablish ‘0 it, se identific Th s”Wort 61 EAP counselors and administrators in improving the effectiveness of programs to reach and assist workers in need. Alcoholism Program Focus. Historically, modern EAPs evolved from the first "occupational alcoholism programs" which targeted alcoholism in industry. “Broadbrush” programs are employee assistance programs which are intended to address the fullest range of possibly job-impairing personal problems of workers. The EAPs of today are typically broadbrush in approach, though some organizations have maintained an exclusive or primary focus on alcohol and drug abuse. This programmatic shift in emphasis has led to several contrasts in EAP design and implementation (Roman 6c Trice, 1976; Roman, 1981a, 1982; Trice or Beyer, 1982a; Sonnenstuhl, 1986). Some have seen in this shift away from the historical alcoholism focus a relative de-emphasis on coercive involvement of problem employees in self-change or treatment. This is offset by a greater reliance on self-identification by troubled workers and voluntary help-seeking in the broadbrush programs. Roman and Trice (1976) asserted that constructive confrontation, which was a "key element" in the earlier programs, could become less prominent and less used now. There are a variety of ways in which the transformation of EAPs into broad-focus programs can be a factor in supervisory referrals. "The assumption has been stated that if a program established its credibility through successfully handling cases referred to it, self-referrals will increase and the need for supervisory identification of problem employees will decrease" (p. 488). The shift to broadbrush was, at the time of its proposal (and strong support by NIAAA) seen as a desirable advance from the negative potentials Johnston supervisor drinking 1 basis of 1 ‘The and refer meemph the orgar mandator involvem: (1981a) s inEAPs confronta The by Roman which are by superv (bl'inedi "Professic Professior ConstrUCt supervisor Claims to same tim because 1 62 potentials of overzealous "witch-hunts" for alcoholics. Freedberg and Johnston (1978) saw the shift to job-based broad-focus EAPs as helpful to supervisors, who were no longer having to confront employees over drinking patterns, but were now confronting and referring solely on the basis of poor work. The shift in work-based programs away from supervisory confrontation and referral is accompanied by an encouragement of direct links between the employees and the EAP counselors and treatment resources external to the organization. This leads, inevitably, to less program reliance on mandatory referrals from supervisors for program existence and less involvement and interest by them in program operation and success. Roman (1981a) saw this as minimizing the role and importance of the supervisor in EAPs, and he thought it would quickly lead to disuse of constructive confrontation by them. The evolution away from an exclusive alcoholism focus was proposed by Roman (1982) to be among the three main trends in EAP development which are undermining the use of the constructive confrontation strategy by supervisors--(a) a de-emphasis on alcoholism as the focus of programs, (b) "medicalization" of deviance in the workforce, and (c) the "professionalization" of EAP staffing. Roman concluded that the professionals working in EAPs indirectly discourage use of the constructive confrontation strategy because involvement of the supervisors with the employee‘s problems and recovery undermines their claims to expert knowledge and necessary credentialed skills. At the same time, supervisors feel much more "exposed" in their referral efforts because they do not have a "cloak of confidentiality" and privacy that n 5’“ the profr close scr co-worke "experts preferrer terms of needed 1 The professio EAP mis of super highly vi ‘save far subordin; AlCOhOlll outcome relations Supervisr construc external SUpervis. expertis Tr. PEFCeive the maj. 01 const 63 the professionals have, and their activities and decisions must pass a close scrutiny of union, higher management, and the employee's co-workers. The values of the providers also are a factor, because "experts look most favorably upon self-referral as the alternative preferred over coercion, both in terms of its ideological fit and in terms of self-referrals' confirmation that the expert's assistance is needed in the work setting" (p. 374). The overall resultant of the forces of "medicalization," professionalization, and the broadened inclusion of all problems into the EAP mission is, according to Roman, "a steady reduction of the likelihood of supervisory use of constructive confrontation except in late stage, highly visible circumstances where such a confrontation may serve to 'save face' for the supervisor who has long harbored a nonproductive subordinate" (1982, p. 374). He also asserted that the identification of Alcoholics Anonymous principles of total abstinence with desirable EAP outcomes, an incorporation of the values of the doctor-patient relationship, and the approaches of "depth psychology" are seen by both supervisors and EAP staff as inconsistent with the coercive elements of constructive confrontation. Furthermore, the increased reliance upon external service providers in the broad-focus programs can reinforce supervisors' belief that they lack the necessary understanding and expertise to deal with problem employees. Trice and Beyer (l982a) described some of the differences they perceived between alcohol-focused programs and broadbrush EAPs. One of the major differences, they believe, is the lesser importance of the use of constructive confrontation and the use of crisis precipitation to get employees be more ' they beco "self-refer Trict played ou alcohol-f0 the helpir into jobs organizati into EAP getting e coercion differ cor occupatio identifier the EAP cooperate confident importan troubled Thr Upon the pmgfams health" 1 describer 64 employees to accept treatment. There is a tendency for EAPs to claim to be more "professional" and to rely less on supervisors for casefinding as they become broadbrush. They show a decided preference for "self-referrals" over coerced referrals. Trice and Beyer (1984a) perceived an ideological conflict being played out as occupational programs were transformed from an alcohol-focused into broadbrush programs. This occurred as members of the helping professions (psychology, psychiatry, and social work) moved into jobs and leadership roles in the EAP movement and supporting organizations. Those authors asserted that these more recent entrants into EAP roles are less concerned about "motivation" as an issue in getting employees to accept help and are less comfortable with the use of coercion as a program strategy. The role of the supervisor will then differ considerably under the new EAPs as compared to the older occupational alcoholism programs. Supervisors become, at most, identifiers and confronters of poor performers who are then referred to the EAP staff. The supervisors are seen as "passive functionaries who cooperate with counselors and outside treatment personnel in maintaining confidentiality" (p. 287). They regret this change, pointing out the important role of the supervisor in maintaining social controls with troubled employees. Three types of EAPs where be distinguished by Riediger (1979), based upon their relative emphasis on alcoholism--"narrow scope" alcoholism programs, mixed model or "broadbrush" EAPs, and comprehensive "behavioral health" programs with a strong prevention and "wellness" orientation. He described the emphasis given to alcoholism as the "central conflict" that must be in the We Wri Society c Roman ti are move supervise Long ex; job may basic jol: frequent workers' personal In confront alcohol-e conclude possibly controls the EAP Th has beer field, be PTOgrarr rePortee 65 must be addressed as an EAP is starting and sets its mission and identity in the workplace (p. 29). Wrich, who became the first President of the Employee Assistance Society of North America (EASNA), "strongly disagrees" (1980, p. 47) with Roman that supervisory referrals become less important in the EAPs which are moved to a broadbrush approach. Wrich sees two main barriers to supervisory referral in all programs which are based on job performance. Long experience or being over-qualified or exceptionally talented for a job may permit a worker with very serious problems to continue to satisfy basic job demands, thereby avoiding supervisory confrontation. Also, frequent changes in supervision creates less opportunity to observe workers' performance and less investment in handling an employee with personal and job problems. In his review of the issue, Kurtz (1982) concluded that confrontation was about equally emphasized and equally effective in the alcohol-only and in the more general assistance programs. He also concluded that the theorized advantage of a constructive confrontation to possibly avoid an eventual referral by effectively re-establishing social controls on problem drinking was not supported in his experience or in the EAP literature. This factor, the necessary degree of program focus on alcoholism, has been the center of much discord and theoretical wrangling in the EAP field, but it has been little researched as to its actual effect on program effectiveness or referrals. Gam, Sauser, Evans, and Lair (1983) reported on a program implementation of EAP services for a private company which were contracted for and provided within a community mental health ce an occup unreporte that the emotiona mechanis referrals participa populatic other dr was not figures 5 alcohol-: types se Sh. alcohol- particip; years. method hipothe broadbri Occupat broadbr "Sugges initiatie alcohol. 66 health center. This program evolved in its identity and focus away from an occupational alcoholism program towards a broad-focus EAP. Other unreported transitions may also have been taking place, but it was noted that the program began to rely primarily upon self-referrals for emotional and family/marital problems, though it still had the formal mechanism of supervisory referrals available. Supervisor-initiated referrals decreased from 39% to 13% of total referrals. Female participation in the EAP services increased from 35% to 60% of the client population. The proportion of clients identified as having alcohol or other drug-related problems decreased from 36% to 18% of the total. It was not reported how overall participation rates changed, but these figures strongly suggest that as an EAP makes a transition from alcohol-focused to broadbrush, changes in the referral source and problem types served will likely change as well. Shain (1985) compared the broad-based EAP approach with the alcohol-focused one fairly directly. He gathered the program participation data from ten broadbrush Canadian EAPs during over two years. The study was limited in meaningfulness by the nonstandardized method of generating and reporting program statistics. It was hypothesized that the "identification rate" of alcoholics in the broadbrush programs would equal or exceed the rate in earlier occupational alcoholism programs. it was also posited that in the broadbrush programs there would be a greater reliance on "voluntary" or "suggested" referrals because employees would take greater personal initiative in dealing with their impairing conditions. Also, alcohol-impaired employees were expected to enter treatment at earlier, rather it between alcoholis alcoholic The into bro; with any EAP, bu services leading our unde first of entering program which e typical]: The oth sufferer into tre 67 rather than later stages of the condition. Shain found no differences between the ten broad-based EAPs and their predecessor occupational alcoholism programs in the rate of identification and direction of alcoholics into treatment. The available evidence, then, suggests that transformation of EAPs into broadbrush programs which are performance-based and prepared to deal with any personal problem may affect how supervisors perceive and use the EAP, but changes in overall program effectiveness, even in delivering services to alcoholics, do not necessarily occur. Triggering Incidents. The concept of a "triggering incident" leading to program referral was proposed by Heyman (1976) as crucial to our understanding of help-seeking. She defined the "trigger" as the first of a series of events which eventually led to the inpaired worker's entering treatment. About half the patients in an alcoholism treatment program who were interviewed by her stated that the "triggering incident" which eventuated in their treatment was impaired job performance, typically absenteeism or poor quality work confronted by their employer. The other half, however, claimed that their work performance had not suffered, but some other event or interpersonal conflict propelled them into treatment. The strong associations related to the impaired-work trigger support the purpose of the [EAP] program and the use of coercion by the company as leverage. They also imply that the identification of these individuals as alcoholics was not made in the early course of their problem drinking. Furthermore, it is highly doubtful that an employee ever comes to an alcoholism program in industry on a truly voluntary basis. The triggering incident appears to be a more differentiating variable than the source of referral. (Heyman, 1976, pp. 906-907) The recent re event whi person in' of worker result of clients, C leading te result of which the control 5' In 1 Beyer an was a "ti frequent performa (l3%) of event de supervise Bas ihlhanns gFOUps a Sensitizir Particula though, SUPErvisr 68 The concept of a triggering incident has been somewhat changed in recent research to mean the occurence of a dramatic, or "last straw" event which is decisive to a referral agent in urging or coercing a person into evaluation or treatment. Mastrich (1985) found in his study of workers entering treatment involuntarily, most were referred as the result of a dramatic on-the-job incident. In their interviews of EAP clients, Googins and Kurtz (1984) reported similarly -- confrontations leading to treatment, when they finally occured, were typically the result of a "last straw" gross violation of company rules or decorum which threatened to expose the supervisor as incompetent or unable to control subordinates. In their study of federal employees' and supervisors‘ use of EAPs, Beyer and Trice (1978) found the most frequent basis for referral action was a "triggering event" of rapidly deteriorated performance. The most frequent were excessive absenteeism or tardiness, strikingly poor work performance, drinking on the job, or a combination of the above. Some (13%) of the employees referred to the EAP had no specific triggering event described, and for another group (6%), the event attracting the supervisor's attention was something happening away from the job. Based on their interviews with supervisors, Kurtz, Googins, and Williams (1980) found that there was a "collective awareness" in work groups as to who had problems. Workgroup gossip played a major role in sensitizing the supervisor to problem employees, raising awareness of a particular employee and poising him, perhaps, to react. They found, though, that triggering incidents were typically the basis for supervisors' referrals to the EAP. Problems attracting the supervisors‘ attention (23/37), if problems with othe refer. Tl referred « finally m. "the supe found out there wa: Thu occur, en decision . all cases, dramatic intervent that som alcoholisr not have The defined j manifest; Chronic j aPPGaran marker 5 69 attention were absenteeism (35/37), poor quality work (33/37), tardiness (23/37), wasting time (16/37), and drinking on the job (15/37). Other problems were mentioned (safety violations, customer complaints, problems with other employees) but were not seen as influencing their decisions to refer. The majority of supervisors indicated that job performance of the referred employees deteriorated gradually over a long period, so they finally made a program referral. In over a third of the cases, however, "the supervisors never susPected the employee had a problem and had not found out about it until the employee came and asked for help or until there was a dramatic episode" forcing action by the supervisor (p. ‘i6). Thus far, research suggests that triggering incidents do sometimes occur, ending the residuals of doubt for a supervisor struggling with the decision of whether to refer a worker to the EAP. They are not seen in all cases, however, and it is very likely that waiting for such a dramatic "sign" operates as a negative program factor impeding earliest intervention. Further complicating the matter is the consistent finding that some workers with severe, advanced conditions (such as late-stage alcoholism or emotional illness) show little or no job impairment or do not have a rapid decline in job performance. The concept of the triggering incident is still rather unclearly defined in the EAP literature. Sometimes, it implies extreme manifestations of dysfunction during an otherwise gradual slide into chronic impairment. Other times, it denotes a dramatic symptom appearance in an acute-onset condition; or, it is said to be a dramatic marker symptom in an otherwise "silent" disease process. Yet other useage implies that it means supervisory reaction to a record of poor job perform: events it exist me construc factors needs tc in EAP. g supervis of the e formal high vis that the signifier the risk perforrr by the organiz TI was do relatlor take jc higher. theme that in more 1 70 performance generally. At any rate, supervisors do report that certain events which would usually result in severe discipline if the EAP did not exist may be interpreted by them as a "cry for help." resulting in constructive confrontation. A clearer differentiation between the factors of severity/duration of impairment and a trigger incident perhaps needs to be made for the concept to have continuing utility to research in EAP. Occupational Level of Employee. Another factor believed to affect supervisory use of constructive confrontation is the occupational level of the employee and supervisor. Trice (1965) noted the possibility that formal alcoholism programs would diagnose only lower-status employees in high visibility jobs. Alternatively, the definite possiblility exists that the lower-status strata of an organization actually has a significantly higher prevalence of problem drinking employees. Further, the risk factors in the development of alcoholism or other ‘ performance-impairing problems may actually be inherent in and fostered by the job environment itself, especially at the lower levels of the organization. The most extensive study of level and status barriers to referral was done by Trice and Beyer (1977). They found a fairly strong relationship between organizational level and supervisors' readiness to take job action with an alcohol-impaired employee. More than 60% of the higher-level supervisors indicated they would likely "handle the problem themselves," whereas t40-50% of the two lower-level supervisors indicated that intention. Also, the lower-level supervisors were significantly more likely to apply formal discipline to workers when dealing with a (hypothetl occupatio actions. were bet' more pes employee In sum, ‘ lower-sta such em; more sec facilitati supervise perceivee My from job status at to a rele confront and high tradition Special I It Perform; °Fganiza epidemie eslélblisl 71 (hypothetical) absenteeism problem. There was a possible confounding of occupational level with other factors influencing supervisors' referral actions. They reported that the supervisors of lower-level employees were better informed about the EAP and its supporting policy and were more positive about the policy in general, but they also had more problem employees and correspondingly greater opportunity to confront and refer. In sum, "the higher prevalence and visibility of problem drinkers at the lower-status job levels makes for policy salience among supervisors of such employees. At the same time, these managers probably experience more social distance between themselves and their subordinates, facilitating policy use. This relative readiness to use the policy among supervisors of low—status employees is further reinforced by few perceived risks in its use and by relatively more rewards" (p. 65). Myers (1984) described "confrontation immunity" as the insulation from job action which appears to be confirmed as one achieves higher status and level in an organization. He saw several factors contributing to a relative safety from supervisory observation and confrontation--frequent changes of job duties, conditions, or locations and high occupational level. The freedom of higher-level workers from traditional job performance appraisals or progressive discipline presents special barriers to EAP use with high-level employees. It is likely that workplace triggers and evidence of poor performance function differently at the different levels of an organization and that this difference cannot be fully explained by the epidemiology of impairing conditions. In the early study which established much of the job performance basis for the modern EAP, Trice (1962) lo from the alcoholisr up but "e handling. work, bu Sen referral with an cases we to lead 1 different publicity entry po directly The is not e the fede awarene associat. Riediger relation would 5.. Th rePreset the Con 72 (1962) found that lower-level employees are likely to absent themselves from their job when they feel hungover or ill from their developing alcoholism. Employees higher in status and responsibility tend to show up but "shrink" the job to a level of effort they feel capable of handling. In other words, impaired higher-level employees tend to be at work, but are more successful in covering-up. Sonnenstuhl (1986) found some evidence for differential supervisory referral tendencies by level. In six cases of supervisory discussion with an employee leading to involvement in the health program, all of the cases were low-level employees. Routine medical exams were more likely to lead higher—level employees to help-seeking. Sonnenstuhl thought the difference might be due, however, to differing degrees of program publicity, as he saw a general trend of client employees using whatever entry portals they knew about, supervisory, the medical department, or directly through self-referral. The evidence for how this factor affects supervisory referral action is not entirely consistent. In another analysis of EAP utilization in the federal sector, Beyer and Trice (1978) found that supervisors' awareness of the union's position on the EAP removed the statistical association between employee skill level and supervisors' use of the EAP. Riediger (1979) found some slight effects as to level of supervision in relation to program utilization, "but not as striking as the literature would suggest" (p. 196). There is some indication that status effects may be present. Labor representatives to in-plant EAP committees and EAP administrators share the concern about whether a program is "for the hourly workers only." From the 1 operation. 9% affect use continue it greater ris resulting i (b) women men's, res to the EA through St relate to granting t Trice, 195 1983; Mas Beyer, 19 reasons, 1 referred 1 a less-Ser Particular Thu affects u “978) 101 Supbrvlso altitudes 73 From the existing evidence, this is a potential problem for program operation. Gender Effects. Three separate gender-related forces seem to affect use of constructive confrontation by supervisors--(a) males continue to have greater incidence of alcoholism, which places them at greater risk of developing work-impairments as the condition worsens, resulting in greater likelihood for men to be referred to the EAP; (b) women have attitudes towards help-seeking which are different than men's, resulting in their greater use of the self-referral entry portal to the EAP, while men, when they finally come, enter involuntarily through supervisory referral;(c) supervisors, who are predominately male, relate to female employees differently than to their male counterparts, granting them relative freedom from confrontation and referral (Beyer 6r Trice, I981; Brodzinski 6c Goyer, I987; Cahill, I983; Kleeman 6r Googins, 1983; Mastrich, 1985/1986; Reichman, 1983; Sonnenstuhl, I986; Trice 6r Beyer, 1979; Young, Reichman, 6c Levy, 1988). For these, and likely other reasons, women are preportionally underserved by EAPs and less often referred by their supervisors. In absolute numbers, women are frequently a less-served population because they may be a lesser proportion of a particular workforce. Thus far, evidence as to whether the gender of the supervisor affects use of constructive confrontation is unclear. Beyer and Trice (1978) found that among the federal agency supervisors they studied, male supervisors were more likely than female supervisors to have negative attitudes towards the alcoholism policy. There was no relationship found, he policy. Rice on how te the EAP Female SI benefits, to them . referrals. job perfe But they workers 1 Cat job actie that whil alcoholic the idem either se The factor in in part, alcoholis seek hel1 ‘0 Confri male the EAP) ES] 74 found, however, between sex of the supervisor and actual use of the policy. Riediger (1979) found that female supervisors felt less knowlegeable on how to use the EAP. Male supervisors made more voluntary referrals to the EAP and used constructive confrontation more than female supervisors. Female supervisors had similar beliefs regarding program purpose and benefits, but they looked to the medical departments as a greater support to them in identification of troubled employees and in initiation of EAP referrals. Also, they disagreed even more than men with the premise that job performance problems were likely to be related to alcohol problems. But they also expressed greater disagreement with the idea that troubled workers would seek help voluntarily if given the chance. Cahill (1983) found that male supervisors were more likely to take job action with an employee. But Young, Reichman, and Levy (1988) found that while male supervisors were more likely to identify more male alcoholic employees, the gender of the supervisor was not associated with the identification of female problem drinkers or referrals to the EAP of either sex employee. The gender of the impaired employee has frequently been found a factor in supervisors' constructive confrontation. This is due, at least in part, to different patterns of job impairment among women with alcoholism or other problems and to a combination of women's tendency to seek help voluntarily and sooner and of supervisors‘ greater reluctance to confront and refer them. The net result would be proportionally more male than female clients constructively confronted and referred to the EAP, especially in alcohol-focused programs. Beye from a ct nearly all were fem The bases work-imp: physical 2 problems and other detected attendanc significan in raw ne clients he the clien‘ problems Mas to be ref self~refe1 Pregram, coerced ' Were typ findings, SGilletimr a PFObler manner ( 75 Beyer and Trice (1981) analyzed the referral data from 1965 to 1977 from a corporate alcoholism program. Of about 1000 cases, which were nearly all supervisory referrals, only about 11% of the program clients were female, and yet the workforce was 32%-38% female during those years. The bases on which the supervisors identified the presence of a work-impairing drinking problem were no different for men or women: physical appearance (15%), poor work performance (12%), attendance problems (25%), intoxication at work (13%), during a medical exam (12%), and other non—work evidence (23%). "Males were more often than females detected on the basis of work performance (10% vs. 5% for females), or attendance (27% vs. 20% for females)" (p. 244). They also found significantly more women than men were recorded as self-referrals, both in raw numbers and case-years of involvement in the program. Male clients had been more often disciplined, but about equal proportions of the clients (70%) indicated that their supervisor had "discussed" their problems with them. Mastrich (1985/1986) found that men were significantly more likely to be referred to EAP by their supervisors. Women are more likely to self—refer. In his sample of over 300 persons in an alcoholism treatment program, he found that 56% of the men, but only 37% of the women were coerced to seek help, a highly reliable difference. The men in the study were typically confronted by the work supervisor. In reviewing those findings, Mastrich and Farmer (1985) commented that "male supervisors sometimes have difficulties confronting female workers. They will ignore a problem, create obstacles for the woman, or treat her in a social manner (e.g., protecting her, enabling, or taking care of her)" (p. 11). A simila believed As a res women 1 was a w troubled formal 1 whereas pressure: Pre reported an expee er inter1 self-refe more lil EAP by was rare personal likely tr men am in men': Ba 1iteratu1 suPervis ftémale hOWEVeI 76 A similar observation was made by Cahill (1983) that "supervisors believed the females were particularly defensive and highly emotional. As a result, they tended to overlook problems and avoid confonting the women for poor performance" (p. 19). Sonnenstuhl (1986) found that there was a wide range of intensity and depth to the discussions between troubled workers and their supervisors, from casually informal to a formal performance appraisal. Females got less formal "suggestions," whereas the men were more likely subject to formal evaluation and pointed pressures to seek help. Program utilization statistics on about 1200 EAP clients were reported by Brodzinski and Goyer (1987). They found further support for an expectation that women will self-refer for problems of a psychological or interpersonal nature. Men, however, were much less likely to self-refer, more likely to have an alcohol or other drug problem, and more likely to have beeen constructively confronted and referred to the EAP by their supervisor. They observed that severe psychiatric illness was rarely the cause for referral action. Also, men who had family or personal problems rather than chemical dependency problems were more likely to seek help on their own. Their data on the modal problems of men and women clients are additional evidence of an expected difference in men's and women's self- or mandatory-referral status. Based upon their review of the management and administration literature, Beyer and Trice (1981) could not conclude "whether or not supervisors would generally tend to be more lenient or demanding with female than with male subordinates. If there is a consistent difference, however, it could certainly affect their willingness to confront and to offer a: (p. 238) male SL "This rr recogni subordir toward R supervi: supervi: identify reasons stay av and we Alterne relucta K to ider women below poor p. Presen' 0f imp the 101 lenden COIIfro‘ 77 offer assistance to female problem drinkers under their supervision" (p. 238). In an earlier article, Trice and Beyer (1979) suggested that male supervisors may have a tendency to "protect" their female workers. "This misplaced 'chivalry' can lead male supervisors to avoid the recognition and confrontation of poor job performance among female subordinates, giving them the opportunity to manipulate the supervisor toward tolerance of their shortcomings" (p. 381). Reichman (1983) reported the results of a questionnaire survey of supervisors and employees in four organizations. Both male and female supervisors indicated that it was more difficult for them to both identify and confront female employees with drinking problems. Various reasons were hypothesized for that difference. Females were thought to stay away from their jobs more than men when impaired by abusive drinking and were therefore less likely to arouse concern of their supervisors. Alternatively, biasing attitudes and stereotypes could make management reluctant to use the EAP as a resource with the female employees. Kleeman and Googins (1983) proposed that work impairment approaches to identifying problem drinking employees may not be as effective with women alcoholics because they, like other women, tend to occupy jobs below their educational level, and supervisors may be more tolerant of poor performance in women. They also believe that women alcoholics present greater difficulty in identification for supervisors. The signs of impairment are more subtle and actually less likely to be present on the job compared to male alcoholics. Further, they found evidence of a tendency to protect the woman alcoholic from identification and confrontation. A readines assertier training question reportin females. job perf and am supervis (p. 229) It that su1 will res a power (1984/ 1'. held an Tl substan own cle Researr comple. While c is Whei I0 engr less ilk 78 A training project especially designed to increase supervisors' readiness to deal with female problem drinkers lends some support to that assertion. Cahill, Volicer, and Neuburger (1982) reported on special training completed by 180 supervisors. A three-month followup using a questionnaire method disclosed that the proportion of supervisors reporting job performance problems in male employees was higher than for females. "The majority of [supervisor] respondents claimed to have no job performance problems or none severe enough to warrant confrontation, and among those reporting female job performance problems, male supervisors were more likely to take action than female supervisors" (p. 229). It was proposed by Trice and Beyer (1979) and Trice and Roman (1978) that supervisors fear confronting women because of the belief that they will respond with suicidal behavior. Such a belief, if present, might be a powerful barrier to use of constructive confrontation. Wilcox (1984/1985) found some evidence of that fear, but it was not very widely held among the supervisors he surveyed. The greater likelihood of women to have emotional rather than substance abuse problems (Myers et al., 198#) and to seek help on their own clearly affects their referral status when they become EAP clients. Research on the gender effect, thus far, has not been sufficiently complex in design or analysis to tease out the other simultaneous factors while clearly demonstrating the gender effect. What remains to be seen is whether, other things being equal, women supervisors are less likely to engage in constructive confrontation, troubled women employees are less likely to be confronted and referred than male employees for the same ce likely te themselv of facte other fa behavior (Beyer d Kurtz, l supervi5r mandate more 111 job perf alcoholis thought Krucher signifies associat (1988) r identifie In With grr Gbegins levels 0 greater TriCe (1 79 same conditions or performance problems, and whether supervisors are less likely to confront and refer an employee of the other sex than themselves. Miscellaneous Factors Affecting Supervisory Referrals. A variety of factors rather unrelated to one another and difficult to classify with other factors have also been found associated with supervisors' referral behaviors. The first of these is age of the supervisor. Several studies (Beyer dc Trice, 1978; Daghestani, Barlow, Hilker, 6c Asma, 1976; Googins 6c Kurtz, 1981; Riediger, 1979; Sonnenstuhl 6c Trice, 1989) found that older supervisors were more likely to utilize constructive confrontation and a mandatory EAP referral. Beyer and Trice found older supervisors not only more likely to refer, but also significantly more in agreement with the job performance basis for referral and stronger in the belief that an alcoholism policy was easy to administer. The younger supervisors thought that there was greater need for the policy, but used it less. Krucher (l986/l987) found that younger, less experienced supervisors made significantly more referrals, however. In another study which tested the association of age and referral behavior, Young, Reichman, and Levy (1988) found no relationship between either age or experience and the identification or referral of employees with drinking problems. In related findings on demographic variables of supervisors, those with greater experience were more likely to refer (Beyer 6t Trice, 1978; Googins 6r Kurtz, 1981). Beyer and Trice (1978) found that increased levels of education were associated with non-referral, less education to greater tendency to refer. Of all the variables measured, Beyer and Trice (1978) found that the demographic variables of age, supervisory .9- experier. the alcc Pe underste supervis operate worker referral knocks" more re (1983) e develop thnely Ti (1988) r identifi- fhufing experie because not one A badc c with ti for the is that Counsel 80 experience, and education were the best predictors of actual past use of the alcoholism policy in the federal agencies they studied. Perhaps the relationship of these intrinsic variables can be understood as representing the accumulation of "negative experiences" supervisors attain as they spend time in their role. Education might operate negatively by leading the supervisor to believe the troubled worker can be handled without recourse to constructive confrontation and referral. A more experienced supervisor coming from the "school of hard knocks" realizes sooner that EAP is the needed supervisory tool and is more ready (on a relative basis) to use it. As Hawthorne and Davidson (1983) commented, supervisors' effectiveness in dealing with alcoholics develops out of a growing awareness of "what doesn't work" reinforced by timely support from the EAP and its staff. Those findings contrast with those of Young, Reichman, and Levy (1988) who reported that age and experience were not predictive of identification and referral of problem drinkers by supervisors. The finding by Krucher (1986/1987) that relative youth of the supervisor, not experience, is highly predictive of EAP referral may be an anomaly, because his data were collected during the startup phases of a new EAP, not one well-established and fully operational. Another important factor is the relationship and integration of the basic constructive confrontation strategy and related EAP role demands with the general role perception and expectancies that supervisors have for themselves. The principal negative role belief frustrating EAP use is that supervisors should handle all the problems of employees and counsel troubled workers themselves. This issue was first raised by li’ Trice ( myself" 110% of supervi: EAP w« supervi: problen of the myself‘ pervasi difficul B supervi propen: official connec rob. r percep they h they h withou self-ex Succes SUppor S referre 81 Trice (1965) and Trice and Belasco (1965). They observed a "do it myself" attitude strongly held by supervisors. Trice found that 30% to 40% of his supervisory sample endorsed the idea that it was the supervisor's job to handle a problem employee. To refer someone to the EAP would be for them a public admission that they cannot do the job as a supervisor. Trice and Beyer (1977) found that when posed a hypothetical problem of a troubled employee, over 60% of the high—level and about half of the lower-level supervisors responded that they would "handle it myself" rather than refer. This role barrier to EAP use is rather pervasive (Beyer dc Trice, I978; Cahill, 1978; Wilcox, 19811/1985) and difficult to overcome. Beyer and Trice (1978) noted the "obvious tendency of the supervisors to handle the matter themselves, as evidenced by their propensity to counsel and their tendency to refrain from doing anything official about the problem" (p. 80). Trice and Roman (1978) saw a connection with "indoctrination in the human relations aspect of their job. It is no surprise to find them incorporating into their role perceptions a helping, ‘do it myself' note which reflects the training they have received" (p. 163). Some supervisors have the attitude that they must be able to control the behavior of their workers at all times, without any assistance from others (Kurtz, 1982). Clearly, that self-expectation is unrealistic, as is the idea that they can successfully counsel without adequate training or the necessary role supports. Some supervisors specifically reject constructive confrontation and referral as part of their supervisory role (Googins, 1979; Googins, Kurtz, 1984/15 coercio needed privacy and use standin; On the enhance manage 0 identifi formal He or referra general action unsuita Roman standar cenfror for the referra respom SUpervi 19801 82 Kurtz, 6: Williams, 1980; Kurtz 6r Googins, I979; Riediger, 1979; Wilcox, 198le/l985). Some verbalize personal values antithetical to the use of coercion to effect the positive goal of getting the troubled worker into needed treatment. They feel so strongly that they should respect the privacy and individuality of a worker that they are unwilling to accept and use the strategy, even when clearly needed. In such a case, they are standing on a principle while killing the troubled worker with kindness. On the other hand, previous successful referral of an employee to the EAP enhances supervisors' comfort with and expectancies of management-initiated EAP contact (Sonnenstuhl 6r Trice, 1989). Other role-related barriers to supervisory referral have been identified. In some instances a supervisor may lack a commitment to the formal aspects of the role within an organization (Wilcox, 19814/1985). He or she may be unwilling to document poor performance or make a referral (Riediger, 1979). In some other instances, there may be a general unwillingness to confront poor work or to apply disciplinary action when appropriate (Googins, 1979). Such a supervisor might be unsuitable for the responsibilities of the role, but such do exist. Roman and Trice (1976) found supervisors with clear and definite standards of work performance more likely to use constructive confrontation; conversely, supervisors lacking a clear sense of standards for their workers were hesitant and uncertain in relation to EAP referrals. In some settings, there may be a lack of clarity just whose responsibility it is to intervene with a troubled employee, the supervisor, the medical department, or personnel (Shain 6c Groeneveld, 1980). [i 'supper that a expects ltice a Trice ( In their lnconsh EAP ar "eversh implem Sonnen: supervi pohcy (I supervi supervi adutra rehabh SUpervi labor c sUpperr when 1 Hunt T Confror 83 Another aspect of role-definition is whether the supervisor feels "supported" in a formal job action and referral. Trice and Belasco noted that a group norm for or against referral can develop (I965). The expectation of support and agreement was considered important to use by Trice and Roman (1978) and by Sonnenstuhl (1986) and by Sonnenstuhl and Trice (1989). Empirical evidence for this factor is not strong, however. In their study of federal agencies, Beyer and Trice (1978) noted inconsistent relationship between upper-level support and enthusiasm for EAP and lower—level use. They concluded that it would be "oversimplistic" to equate top-level program support with policy implementation and program acceptance and use by supervisors. Sonnenstuhl and Trice (1989) reported that an equal number (60%) of supervisors who had and had not made a referral under an alcoholism policy expected such action to "please management." Certain organizational factors have been found which affect supervisory referrals. The presence of a labor union increases supervisory use of the EAP. There is an increasing trend of labor arbitrators to reverse disciplinary actions and discharge if rehabilitative options have not been offered a troubled employee. Supervisors are more likely to do confrontation and make referrals in a labor organized setting, especially when they know that the union supports the EAP and related poliCy and expects them to refer workers when job performance begins to slip (Beyer 6r Trice, 1978; Beyer, Trice, 6: Hunt, 1980). The supervisor's span of control appears related to readiness to confront and refer and to discipline (Beyer 6c Trice, 1978, 1984b). Being in char though larger ' betweel adminis Levy (1 identifi of com obvious frequer A the org to con: resulte. and pr: to doul develo; constrL On the and th. referra Wilcox E TEIatec Superv EAP ll 81+ in charge of a larger workforce leads to more routine use of the EAP, though why that is so is not known. Those authors hypothesize that in a larger workgroup there is less group cohesion and greater social distance between supervisor and worker, so there is a greater readiness to administer sanctions for substandard performance. Young, Reichman, and Levy (1988) report that supervising more employees is associated with the identification and referral of more problem- drinking employees, but span of control is not introduced as a control or blocking factor, but in the obvious sense that greater opportunity to refer is related to more frequent referral events. Another group of factors relate to the line-staff relationships in the organization. Trice (1965) reported that a reluctance of supervisors to consult with and interact with the staff (including EAP staff) resulted in a decrease of referrals by supervisors. Increased visibility and prestige of the EAP staff, though, are believed to cause supervisors to doubt their competence to deal with the troubled employee and to develop a greater reliance on the "experts," reducing their use of constructive confrontation (Roman 6c Trice, 1976; Trice 6: Roman, 1978). On the positive side, a close relationship between supervisors, the EAP, and the medical department have been shown associated with increased referrals by supervisors (Daghestani, Barlow, Hilker, 6c Asma, 1976; Wilcox, 198li/l985). Expectations about the possible outcomes of referral action appear related to the supervisor's readiness to refer an employee to the EAP. Supervisors who have a strong belief in the benefit and effectiveness of EAP intervention and subsequent treatment are more likely to confront and refer tl Wilcox, St specific actions, U969) : toward: and rel relucta aversio oi harr problen supervi attitudt reinfort himself intentit G were n Young, and at‘ Woman Probler T Person; by Kru 85 refer the impaired worker (Googins, 1979; Sonnenstuhl é: Trice, 1989; Wilcox, 1984/1985; Young, Reichman, 6t Levy, 1988). Several studies have attempted to examine the relationship between specific supervisory attitudes towards the troubled worker and referral actions. As noted above in relation to training, Belasco and Trice (1969) found that negative, not supportive and understanding, attitudes towards impaired employees are much more associated with confrontation and referral. Positive, concerned, kindly attitudes are associated with reluctance to refer (Googins, 1979). Respect for individual rights and aversion to control or manipulation of others (Riediger, 1979) and fear of harming someone‘s career (Wilcox, 1984/1985) or of creating further problems for them (Trice 6t Beyer, 1982b) all operate to hinder supervisors in use of constructive confrontation. These personal attitudes, which we would usually consider commendable, appear to reinforce in the supervisor a tendency to try to handle the problem himself without formal job action. Experience has shown such good intentions do the impaired worker little good for very long. Googins (1979) found that attitudes towards drinking, in general, were not related to supervisors' use of constructive confrontation. Young, Reichman, and Levy (1987) reported that attitudes towards women and attitudes towards women drinking (prejudice or stigma directed at the woman alcoholic) were not predictive of the identification or referral of problem-drinking employees, male or female. The sole piece of research which has attempted to relate a personality trait of supervisors to their referral behavior was reported by Krucher (l986/l987). He administered a psychometric scale (the FHiCll interpe dhnens conclui respon EAIK prediC‘ coniro nature span 0 intnns Exphc. SUperv woukl differe introve investi Psychc a prht Unders 86 FlRO-B, authored by Schutz) which is used to describe an individual's interpersonal needs structure. He found that only one of six personality dimensions measured by FlRO-B covaried with referral behavior. Krucher concluded that supervisors who tend to desire and accept more responsibility are more likely to confront and refer employees to the EAP. Supervisors' use of confrontational behaviors generally was highly predictive of their making referrals to the EAP. A Conflict Model of Constructive Confrontation A Theoretical Research Base The variables studied thus far in relation to constructive confrontation and supervisory referral have been "sociological“ in nature, reflecting global or structural aspects of the setting (e.g., span of control and organizational status of the employee) or immutable intrinsic variables such as gender or supervisor age or experience. Explicitly psychological variables and processes could also be related to supervisors' referral behaviors. Examples of such intrapsychic variables would be needs (e.g., achievement, dominance, affiliation), trait/state differences such as anxiety level, or personality dimensions such as introversion and extraversion. Yet another avenue which can be investigated with respect to motivation of supervisors is specific psychological processes such as values, attitudes, or cognitive style. As described above, the hesitation of supervisors to use the EAP is a primary factor in its limited effectiveness in organizations. To understand the motivational factors within individual supervisors calls for investigation of psychological variables which could potentially explain contribt of supe] approac Miller, the the deal wi Robinso Belasco howeve: the mo referral 1n supervi survey is mad attitudt T and co] situatic at has action] The ap hESltan —7———_fifi a“, g 87 explain the complex referral behavior. A psychological model which might contribute to our motivational understanding of the subjective experience of supervisors and of their (non)referral behavior is the approach/avoidance conflict model (Dollard 6c Miller, 1950; Epstein, 1982; Miller, 19%). A number of researchers and observers of EAPs have used the metaphor and language of conflict in describing supervisors who must deal with an impaired employee (Kurtz, Googins, 6: Williams, 1980; Robinson 6c Robinson, 1963; Sonnenstuhl, 1986; Trice, 1965; Trice 6c Belasco, 1965; Trice dc Roman, 1978; Wilcox, 198(4/1985). No investigator, however, has explicitly applied a conflict model of approach/avoidance to the motivation of the supervisor in a prediction or postdiction of referral action. In this section, the metaphor of conflict, which has been applied to supervisors needing to handle a troubled worker, is explored. A brief survey of the theory and research evidence of approach/avoidance conflict \ is made and applied to EAPs and supervisors. The rationale for using an attitude measure to assay the approach and avoidance tendencies is given. Indications of Conflict in SupervisorLBehaviors The behaviors and self-reports of supervisors suggest that anxiety and conflict are terms applicable to their experience of the referral situation. The existence of an EAP and its underlying corporate policy, at least for many supervisors, does not resolve the indecision between action/intervention or a range of accepting and accomodating behaviors. The alternatives are now expanded to include referral to the EAP, but hesitancy and indecision remain. That the EAP is intended to be a more benevol reduces distress and oft begin ti feelings with th reporter l98i/15 R cognitit inactior awaren‘ ll Conflic Worker CODSldt 88 benevolent alternative intervention than disciplinary action only partly reduces the often-observed period of supervisory vacillation and Increased levels of tension and stress, heightened emotion, distress. and often painful indecision are reported by many supervisors as they Supervisors report intense begin to engage a problem employee situation. feelings both for referral and for temporizing and avoiding confrontation with the employee. A subjective experience of intense conflict has been reported (Riediger, 1979; Trice, 1965; Trice 6t Roman, 1978; Wilcox, 1984/1985). Behaviors of vacillation and indecision are common. Robinson and Robinson (1963) were the first to begin detailing the cognitions and reactions which might help us account for supervisory inaction. They identified cognitions which alternately dominate their awareness as the situation worsens: a) The other men filled in for the impaired worker and didn't seem to mind. b) Supervisor had known the impaired worker for years and thought "he is a nice guy with a family." c) He had "warned him" about his excessive drinking several times. d) Supervisor has feelings of aversion towards applying discipline and is trying to avoid it. e) Feels sorry for the wife and family and imagines them destitute if he disciplines and eventually fires impaired worker. f) Engages in wishful thinking that "some way he‘ll leave my department." g) Is protecting a self-image of being kind and understanding. h) Is taking a less difficult course than confronting the problem. 1') May be over-identified with the worker‘s problems. 1') Supervisor is blaming himself unrealistically for the worker‘s problems. k) Believes he can reform the impaired worker by appealing to him to change. (p. 15) In his interviews of supervisors, Trice (1965) found evidence of conflict and vacillation which underlies their reactions to troubled They tended to dichotomize workers in their perceptions, workers. Even as the supervisors considering them "good" or "bad" employees. began t intense worker prominr used su torn be to 'do genuine "blockir EAP re affectiv engage T: was eit and Spe employ greater with er within "factor the art how re A the "pi and W1 emPIOy 89 began to move a particular employee into the "bad" category, there was an intense feeling of ambivalence which clouded their perception of the worker and interferred with referral action. "Indecision emerged as a prominent theme. Regarding action about 'problem' cases, supervisors used such phrases as 'I see-sawed back and forth about him,‘ or 'I was torn both ways.‘ At the same time, there was a distinct note of pressure to 'do something'" (p. 178). Trice recognized their ambivalence as genuine and intense, and he introduced the concept of "pushing" or "blocking" factors (see Table 1+) which incline or deter supervisors in EAP referral action. Those factors are attitudinal, perceptual, and affective in nature and were hypothesized to affect their motivation to engage in constructive confrontation. Trice reported that endorsement of the pushing or blocking attitudes was either low (around 5%) or fairly widespread (around 20% agreement), and specific factors (attitudes) were associated with the various employee problem types. The supervisors of alcoholics experienced greater conflict and ambivalence than did the supervisors of employees with emotional problems or major mental illness. The supervisory sample within which Trice found a concurrent type of validation for his "factors" had all ultimately referred a troubled worker to an EAP. Thus, the attitudes were not explicitly validated as predictive of whether or how readily supervisors will refer. A partially supportive effort to establish concurrent validation for the "pushing/blocking factors" of Trice was reported by Kurtz, Googins, and Williams (1980). They interviewed 37 supervisors of troubled employees who had entered treatment. They found a general agreement 90 Table 4 Proposed Cognitions Blocking or Pushing Supervisory Referrals Blocking factors He might be "bugged." He had often been a capable worker. Referral would hurt his family. Fellow workers accepted him, aided him. I had come up through the ranks with him. Referral would mean separation from the Company. It'was my duty to handle the problem. He would soon be shifted. He "snapped out of it" frequently. Medical and Personnel might not back me up. Referral would get me mixed up with the union. Refferal would involve me in inconvenience. Pushing factors He was often surly, arrogant, even defiant. He put me in a "bad light." It would help out both him and his family. He had a bad effect on fellow employees. He believed he was "putting one over on me.” I had to watch him closer. He wasn't doing his part of the work. He was out a good deal. 1 never knew what to expect. He was lacking in self-respect. I was afraid he was a safety risk. My boss would back me up. Needed to do something as soon as possible. Source: Trice, 1965, p. 179 among ‘ them if his fact non-refr barriers of harm for the course < supervis staff. 1 involved and "prc supervis referral A and Bela Supervisl contrary themseli 0f alcoh EAP. T as a per rElerral. allproact Ultimate ——i—:——-~ , 91 among the supervisors regarding the "blocking" factors which discouraged them from action. The experiences of those supervisors did not validate his factors hypothesized to encourage referral, however. Referring and non-referring supervisors were not compared in the study. Numerous barriers were mentioned--fear of involvement in employee's problems, fear of harming the employee's families, avoiding a burden of responsibility for the employee, and concern for their own welfare and safety in the course of confrontation or discipline. They also found that many of the supervisors avoided referring because of poor relations with program staff. Role conflicts between the supervisors' felt need to "stay involved with the problem" and EAP staffs' need to maintain confidences and "protect" the workers resulted in a poor working relationship between supervisors and program staff and generated tension and resistance to referral actions. A description of the "see-saw" vacillation was described by Trice and Belasco (1965). They rejected as inaccurate portrayals of supervisors as actively trying to cover-up for impaired workers. On the contrary, feelings of group cohesion and loyalty, role expectations of themselves as having to handle all problems, and inaccurate stereotypes of alcoholics result in their being "torn“ between using or not using the EAP. Trice and Roman (1978) described the "see-saw stage" of indecision as a period of intense ambivalence which a supervisor experiences towards referral. It is that "see-saw stage" which may be explained by approach/avoidance conflict model(s) and which practitioners in EAPs ultimately hope to eliminate or greatly compress. l and 0 progra feel a typica about and in theh't need i resulu ( progre F.lle. inteHe of Pia in psy Psyche incoml ( finch combh most 1 approa FOUr 1 92 Based upon their experience in EAPs in the federal sector, Phillips and Older (1977) designed their supervisory orientation and training program around the intense feelings of conflict supervisors typically feel as a problem employee situation develops. They described the typical supervisor as devoting an inordinate amount of time to "fretting about the situation," troubled all the while by feelings of guilt, fear, and inadequacy over not being able to solve the employee's problems or their own conflict. The presence of a troubled worker and the impelling need for confrontation is "highly unpleasant," according to supervisors, resulting in their avoidance of confrontation if possible. Approach/Avoidance Modeled as Response Conflicts Conflict is one of the innovative learning theory concepts which was progressively developed by Neal Miller and John Dollard and others (e.g., F. Heilizer and S. Epstein). As Epstein (1982) points out, the intellectual roots of conflict can be found in the early Greek philosophy of Plato and Aristotle. The phenomenon of conflict has found application in psychodynamic, behavioral, physiological, and cognitive approaches in psychology. The central idea of conflict is competition of mutually incompatible responses, cognitions, or impulses. Conflicts have been grouped and analyzed in various ways, but one which has attracted considerable attention is based upon various combinations of approach and avoidance (escape) tendencies. One of the most interesting and heuristic of the conflict models has been the approach/avoidance model of Miller (19%) and Dollard and Miller (1950). Four fundamental principles are the basis for analysis of conffi subkx is str avoww proxh a ten- dnvei the n goalj respoi tende the s avokl ways. devel comb reSpe stren depic gradi Some aPPM tende 93 conflict—-(a) tendency to approach a goal is stronger the closer a subject is to that goal; (b) tendency to escape or avoid an object goal is stronger the nearer a subject is to that goal; (c) the strength of the avoidance tendency or response increases more rapidly with closer proximity to the goal than does the tendency to approach; (d) strength of a tendency to either approach or avoid is based upon the strength of the drive(s) upon which it is based; increased drive may be said to increase the intensity of approach or avoidance motive across the entire range of goal proximity. The utility of these principles is the insight that the response seen will be the summative resultant of the competing tendencies. If the drive to approach is greater than the drive to avoid, the subject will approach the goal, or atttempt to avoid or escape if the avoidance drive is greater. These principles have been graphically represented in a number of ways. Those representations depict a conflict situation which may have developed in the learning history of an individual or through a combination of drives which may be operating at a given time or with respect to a certain goal object. The linear representation of drive strengths is termed a gradient. The approach/avoidance conflict model is depicted in Figure 1. An important assumption is that the avoidance gradient is steeper than approach and that the gradients intersect somewhere before actual goal attainment. A distinctive feature of approach/avoidance conflict is a region of vacillation, a point where the tendencies to approach and to avoid are equally strong. confl be Ct have TGSEE 94 Gradient of Avoidance Area of vacillation Approach Drive Strength Proximity to goal 4 or Stimulus Similarity Figure I. Graphical representation of approach/avoidance conflict. (Adapted from Dottard 8. Miller, 1950) Epstein (1982) described the basic mechanism of approach/avoidance conflict. It is readily apparent how the dilemma of the supervisor can be conceptualized in these terms. The individual will approach the goal when at a distance, but avoid it when close. A stable equilibrium is thus established in which the individual vacillates around the point of intersection of the gradients. The model not only predicts symptoms of conflict, such as vacillation and sustained high levels of fear and arousal in the absence of coercion, but also explains how conflicts can be maintained without reinforcement. According to the model, the individual, in effect, is imprisoned by his own desires, as he can neither relinquish nor obtain the goal object. (pp. 52-53) The principal assumptions of the approach/avoidance conflict model have generally been confirmed by experimental, largely animal laboratory, research (Heilizer, 1977a, 1977b). The condition of conflict can be readi apprc a for proxi learn avok apprr and . goal Alter Stunt phys sever steer treat cone and unde deal P- 51 toh as w Fenz belo 95 readily induced in experimental animals by first training them to approach a goal under the drive of a basic need such as hunger (e.g., to a foodbox), then associating that same goal with punishments such that proximity to the goal induces a fear reaction. Fear is a powerful learned drive and can also serve as an internal cue to engage in avoidance (escape) behaviors. The animal will then be seen to rapidly approach the foodbox, then stop and withdraw before actually reaching it and eating. It will then take up a position intermediate to the goal-~unable to approach and eat, but unable to leave completely either. Alternating movements to approach then escape may be seen. The basic conflict model has also been reinterpreted so that stimulus and response displacements are the theoretical equivalent of physical proximity to the goal. Epstein (1982) notes that there are several problems with the model. Avoidance gradients are not always steeper than approach as claimed by Miller, and the concept of "drive" is treated in a different manner than usual by learning theorists. He concludes, though, that those are not critical limitations of the model, and it "remains an impressive tool that not only has enhanced our understanding of an important conflict but also has stimulated a great deal of research and has received a fair degree of confirmation" (1982, p. 58). There have been a number of efforts to apply Miller's conflict model to humans, but as Epstein observed "even simple drives are not as simple as we thought" (1962, p. 1‘13). He reported (Epstein, 1962; Epstein (Sr Fenz, 1962) a study of the arousal level of parachutists on several days before and on the day of a jump. A stimulus list of words from low to lugh (GSR. as cc suppc and ' hlucl lfflle inhib vehw supe' nnigt coni the pote the alon stro befc driv aPP' 96 high content relevance to jumping was given and galvanic skin response (GSRs) were recorded as indicative of arousal. The findings were taken as confirmatory of a theory of perceptual defense, but it also gave support to the use of a written or verbal list to assay the drive levels and tendencies of subjects with respect to a conflict-producing action. Dollard and Miller (1950) applied learning theory to psychotherapy. Much of their theory of neurosis and of therapeutic change derives from Miller's conflict model. Their analysis of how persons overcome inhibitions and faulty learning through therapy, concieved as a special vehicle for learning and conditioning, is relevant, perhaps, to supervisory referral actions. There are two different strategies which might be used to move someone closer to a goal state or action which is conflicted. The level of the approach drive might be increased, moving the person towards a new equilibrium position closer to the goal, potentially all the way to the goal (see Figure 2). However, this has the side-effect of dramatically increasing the level of arousal and fear along the way. The unpleasant subjective experience of fear will strongly motivate the person to engage in distancing (escape) behaviors before goal attainment. The other strategy is to lower the avoidance drive level, allowing the person to move towards the goal under existing approach motivational drive levels (see Figure 3). Drive Strength Figure 2. from in level of Drive Strength Figure 3 from a increg 97 avoidance .i§§.f.l§!?l.. gradient approach e————- low Drive Strength Proximity to goal a or Stimulus Similarity Figure 2. Representation of closer proximity to goal resulting from increased approach tendency, accompanied by increased level of subjective fear. (Adapted from Ballard 8. Miller, 1950) strong Widance _t_ea_r_]_e_v.e_lm : high GDDFOGCh Drive Strength avoidance Proximity to goal 4 or Stimulus Similarity Figure 3. Representation of closer proximity to goal resulting from decreased avoidance tendency, accompanied by minimal increased subjective fear. (Adapted from Dollard & Miller, 1950) This 1 tension and latter metl change. There on re increz stage: reduc the g would likely In thi including 1: attitudes a behavioral behavior, e Miller (195 are seen, c a person is troubled er entirely le: motivating, Trice troUbled er 10 altering termed by the barrier 98 This latter strategy results in only slightly higher levels of tension and subjective fear/arousal. Dollard and Miller recommended this latter method of helping persons resolve conflicts and achieve personal change. Therapists have found that the first thing to do is to concentrate on reducing fears motivating avoidance . . . rather than to try to increase the motivation to approach the feared goal. In later stages of the treatment, the fear motivating avoidance may be reduced to the point where the subject is able to approach near to the goal without undue fear or conflict. When this has occurred, we would expect increasing the . . . motivation to approach to be more likely to produce a good effect. (1950, p. 359) In this model of learning and change, higher-order mental processes including learned drives, language, and discriminated responses such as attitudes are very important. These learned responses may serve as behavioral cues, and have an important role in mediating and motivating behavior, especially through the powerful drive of fear. Dollard and Miller (1950) suggested that whenever unexplained indecision and conflict are seen, concealed sources of avoidance are highly likely. Likewise, if a person is faced with a fearful stimulus or task, such as managing a troubled employee in the department, the fact that the situation is not entirely left suggests strongly that approach factors are partially motivating, since someone free to do so will escape fear—evoking stimuli. Attitudes as Measures of Approach/Avoidance Tendencies Trice and Roman (1978) alluded to the role of attitudes towards the troubled employee as a possible factor in referral actions. Resistance to altering one's coherent perception of "a good employee" has been termed by them the "normalization" process. According to them, many of the barriers to referral are in the form of overlearned rules and expectatio reactions, social rela troubled a work outpr begin to a towards la interventio those prob. referral ac considerabl inappropria reaction, 1: deviancy rj (p. 168). attributes r theorists--t With furthe inconsistent Examj approach/at given varior Construct is evaluative, concews or a Social obj 99 expectations of social relationships. Associated cognitions, evaluative reactions, and behaviors of toleration help preserve and maintain our social relationships. During the onset and progression stages of the troubled employee's problem, various changes in on-the—job behaviors and work outputs are noticed. Incidents and changes in personal qualities begin to accumulate in the awareness of a supervisor which incline towards labeling the impaired employee as deviant and requiring intervention. There is a continuing tendency, though, to "balance out" those problems with favorable perceptions and attitudes which block a referral action. "Supervisors and work associates apparently go to considerable lengths to 'normalize' behaviors that are clearly inappropriate and disruptive. This type of reaction, or absence of reaction, is common throughout social life because accusing others of deviancy risks considerable disruption of interpersonal relationships" (p. 168). The concept of normalization captures several of the attributes of attitudes which are emphasized by current attitude theorists——their stability, their potential for active and passive change with further experiences of the actor, and their typically conflicted, inconsistent nature. Examination of attitudes held is one means of measuring the approach/avoidance tendencies within an individual. Attitudes have been given various definitions. A generally agreeable definition of the construct is that attitudes are "a relatively enduring system of valuative, affective reactions based upon and reflecting the evaluative oncepts or beliefs which have been learned about the characteristics of social object or class of social objects“ (Shaw 6: Wright, 1967, p. 3). In this stu: organizatio definition 4 them - at degree of ' attitudinal taken to it only infere the persons individuals true for Sp of stateme Kelmt attitudes a as a motiv in predictic mill predit multiply-de strongest d demands m 5P€Ciiic as altitudes u lhEm Tower employee c cillllt‘adicto 0f attitude, 100 In this study, confrontation and referral of troubled workers to the organizational EAP resource is the social object of interest. The definition of attitudes is closely linked to the methods used to measure them -- attitude scales consist of a collection of statements of varying degree of positive or negative affective content regarding the attitudinal referent. Endorsement of those statements by subjects is taken to indicate their evaluative reaction to the attitude object. "The only inferential step involved is the assumption that the evaluations of the persons involved in scale construction correspond to those of the individuals whose attitudes are being measured. This may or may not be true for specific items, but the error should be small for the total set of statements" (Shaw 6t Wright, 1967, p. 14). Kelman (1979) notes that the consistently low correlations between attitudes and actions point up limitations of the concept of "attitudes" as a motivational construct, but attitudes do not demonstrate invalidity in prediction either. There are several reasons why attitudes do not fully predict an individual's behavior. Firstly, behavior is multiply-determined; attitude theory does not claim to be the only or strongest determinant. Situational constraints such as social norms and demands must also be taken into account. Next, attitudes can be highly specific as to object and situation. For example, supervisors might have attitudes towards alcoholics drinking on the job which would motivate them toward one behavior, but attitudes towards an emotionally upset employee crying on the job which motivate toward a seemingly contradictory behavior. Lastly, individuals possess a vast collectivity f attitudes, and an investigator cannot be certain that the stimulus which is rt another in: performant needy pers employee 1 attitudes a Kelm range of r In particul. and action is that att approach a attitudes f concluded between at avoidance salience in It is both strong conflicted exPitt the moderate r altitudes. C(lllfrontati deterrent q C"mponent. 101 which is most salient to one person is even perceived as influential to another individual. One supervisor may define an employee's poor performance as impairment and act influenced by attitudes towards ill or needy persons, but another may define the situation as one in which "the employee is being lazy" and follow motivational paths determined by attitudes about "lazy" persons. Kelman (1979) also suggested that attitudes be interpreted as a range of relationship and possible actions towards an attitude object. In particular, attitudes "include both positive and negative associations and actions, both approach and avoidance tendencies. The assumption here is that attitude toward an object very often is resultant of conflicting approach and avoidance components" (p. 133). Examining the effects upon attitudes following a person's participation in induced actions, Kelman concluded that "we can gain a better understanding of the relationship between action and attitude change by separating the approach and avoidance components of the attitude and examining their relative salience in the action situation" (p. 189). It is recognized that persons can endorse attitudes indicative of both strong approach and also strong avoidance, reflecting their conflicted state with respect to the situation. However, we may also expect their attitudes to be somewhat internally consistent, with a moderate negative correlation between their approach and avoidance attitudes. Following Ballard and Miller, the avoidance components of the confrontation/referral attitudes will be hypothesized to have a greater deterrent effect upon a referral action than do the positive approach components motivate a supervisor‘s referral. The attitudes of the referring 1 because ti favorable of a work Since more com attitude is determinar are both c interplay 1 other in a as confron psychologii re-examine with expec handling 0 role perim and avoids ESpecially his judgme more posit 1°55 negati Previ imPasse dL expllCii ap underlyjng 102 referring supervisors may be found to be higher on the average, though, because their approach attitude components increased following a favorable evaluation of the outcome of their confrontation and referral of a worker. Since attitudes are not simply and directly related to action, a more complex view of their role in motivation has been developed. An attitude is not a simple index of action, but simultaneously a determinant, component, and consequence of actions. Stability and change are both characteristic of attitudes. There is a continual, reciprocal interplay between attitudes and actions, such that "each generates the other in an endless chain" (Kelman, 1971+). Enactment of a new role (such as confrontation and referral of a troubled employee) can set into motion psychological forces conducive to attitude change. A conscious re-examination of one's attitudes and an effort to make them consonant with expectations of others (such as the organization's policy on the handling of alcohol-impaired employees) and congruent with new expected role performances would likely lead to shifts in the perceived approach and avoidance valences of referral action. After making a referral, especially if the referral outcome pleased the supervisor and confirmed his judgment in the instance, it is expected in Kelman‘s paradigm, that more positive (approach) attitudes towards referral would be held and less negative (avoidance) attitudes. Previous application of the conflict metaphor to the supervisors' impasse during crisis buildup has suggested the rationale for this explicit application of the conflict model to their behavior and underlying attitudes. This study will use Kelman's method of separating the approa and will e and hypotl attitude cl towards rt that inter1 determine the super» consistent 103 the approach and avoidance components of attitudes about the referral act and will examine the relationship between those attitudes and the actual and hypothetical referral behaviors of supervisors. In this study, attitude changes over time are not the object of interest. Attitudes towards referral may be, as described above, both cause and effect, and that intertwined nature is accepted. The research effort here is to determine whether referral attitudes form predictive scales and whether the supervisors‘ attitudes and behavior fit in a manner which is consistent with the Dollard and Miller approach/avoidance conflict model. This and descrii especially program-pt focus of g means-pas of an actu hypothetica problems. which teSt thought to four vigne' The : research, 1 the previoi Second obj measuring emillily'ees. have “so CHAPTER 3 METHODOLOGY Overview This research utilizes the observational survey method to measure and describe supervisors' attitudes about various aspects of EAP, especially constructive confrontation, and various organizational, program—perception, and exogenous factors. Referral behavior is the focus of greatest interest, and it will be explored through multimodal means--past use of EAP through employee referral, referral or nonreferral of an actual employee under supervision, and referral actions with hypothetical employees described as having job performance and personal problems. Instrumentation consists of attitude and informational items which test for effects of most of the previously identified factors thought to affect supervisory referral behavior and imagined response to four vignettes of problem employees. The first research objective is replication of much of the prior research, thereby supporting or weakening the claims of association of the previously identified variables with supervisory referrals. The second objective is construction and initial validation of an instrument measuring supervisory attitudes towards constructive confrontation of employees. Previous researchers in motivation and attitude measurement have suggested that the supervisors' attitudes towards referral are 101+ comprised this researt scaleable, t referral be avoidance t predicts. The r hypotheses inquiries m to individu; next, with supervisory characteris participatir Speck next Chapit variables a will be 101, Analysis of Strength of Examinatio Variables ir Strength of 105 comprised of both approach and avoidance components. A major goal of this research is to determine whether attitudes towards referral are scaleable, are associated with and predictive of supervisors' actual referral behavior, and whether the two dimensions of approach and avoidance attitudes are related to that behavior as conflict theory predicts. The research questions which form the basis of the testable hypotheses of the study will be stated. They will guide statistical inquiries made of the data, moving from questions about variable groups to individual factors. The data collection methods will be addressed next, with some comments about problems encountered in studying supervisory behavior. The latter portion of the chapter describes the characteristics of the participating organizations and the supervisors participating in the study. Specific findings on the research questions will be covered in the next chapter, following a description of the data analysis approach. All variables associated statistically with supervisors' referral actions will be joined into a multiple correlation/regression prediction model. Analysis of the proportions of variance accounted for will indicate the strength of the variables in joint prediction of supervisory action. Examination of the squared semipartial correlations of the predictor variables in the regression models permits a comparison of their relative strength of prediction. The confront: These set construct research (Que (Que (Que (Que (Que (Que (Que. (Que: (Que: .05), but maintained a significant negative effect with PASTREFR (g’ = -019, _t_ = -3-42, p<.001). Although 154 somewhat inconsistent, these findings support the assertion of AVOIDANCE attitudes as valid predictors of referral actions by supervisors. Familiarity with EAP. The next cognitive factor to be examined is Familiarity with EAP, measured by the four-item scale FAMILIAR. This factor is supported as being positively related to referral action. In tests of its association with each of the measures of referral, FAMILIAR showed significant positive relationship. This factor had 4% of its variance in common with HYPREFER (_r_ = .197, p<.001), nearly 30% of common variance with PASTREFR (.5 = .542, p<.001), and 4% common variance with EMPREFER (r_ = .197, p<.001). In the multiple regression models of the cognitive variables with PASTREFR and EMPREFER, FAMILIAR showed consistent positive effects, accounting for over 10% of past referrals variance and 3% of EMPREFER variance, both highly reliable evidence of effect. With HYPREFER, the scale did not achieve the prescribed level of significance for the two-tailed test, but had the test been one-tailed in the direction suggested by previous research (greater familiarity associated with greater referral behavior), significance would have been found (£1 = .006, _t = 1.697, p = .09). The very high correlation with PASTREFR suggests some problem in interpreting the meaning and significance of this factor as a predictor. Referral and "program familiarity" are rather redundant in a real sense, as a supervisor cannot very well have made one or more referrals if he/she doesn't know how to make a referral. The consistent relationship, even with hypothetical referrals and action taken with a specific, single 155 problem employee greatly strengthens this factor‘s claim to be a valid predictor of supervisory referral. TRAINING Factor. Findings on this factor, a component of most EAPs, supports training as an influence on supervisory referrals. Research Question 6 is, by these findings, answered in the affirmative. TRAINING demonstrated significant positive relationship to past and recent employee referrals. In bivariate correlation with HYPREFER, it approached, but did not quite achieve demanded significance level (L = .08, p = .056). The correlation of TRAINING with PASTREFR was .386 (B<.001) and was .182 with EMPREFER (B<'Ol)' In multiple correlations with the other cognitive variables, TRAINING showed significant positive effect with PASTREFR (g2 = .024, t = 3.827, p_<.001), but its effect, after controlling for the other variables, was nonsignificant with EMPREFER. TRAINING is, therefore, affirmed a valid 1 predictor of supervisory referral action. COUNROLE Scale. Research Question 7 asked whether supervisors‘ tendency to see themselves in the role of a counselor with employees was related to their EAP referral action. The COUNROLE scale was designed to measure that tendency to "identify" with a counselor-like role. As can be seen in Tables 13, 14, and 15, it did not demonstrate reliable association with any of the measures of referral. lts correlations with hypothetical, past, and recent referral were -.055, .027, and .09, respectively (each p>.05). This scale is, therefore, not retained as a valid predictor. 156 Supervisor Demographic Variables Group The next analyses (see Tables 16, 17, and 18) provide the answer to Research Question 8--whether demographic characteristics of the supervisors show relationship to their referral actions. The regression results with PASTREFR and EMPREFER, but not with HYPREFER support the valid association of one or more the demographic variables with referral. The group was a valid, but not very strong predictor of reported past referrals (PASTREFR) (5’ = .06, _F_(4, 405) = 6.61, B<.001). Entrance of dummy-coded variables for survey locations increased 3’ to .151, an increment significantly greater than O (3132 = .09, 5(24, 381) = 1.68, p = .025), showing a location effect for the regression of the demographic variables with PASTREFR. The demographic variables modeled with EMPREFER as the criterion also showed significant association (3’ = .099, 5(4, 209) = 5.77, p<.001). Entrance of dummy-coded variables for survey locations increased 3’ to .203, an increment not reliably greater than 0 (sR’= .104, E(24, 185) = 1.01, p = .46), this time not showing a location effect. This finding supports examination of each of the demographic variables with those dependent measures. — 7 157 Table 16 Regression Results for Supervisor Demographic Variables Group with Hypothetical Referrals Scale as the Criterion a) Analysis of variance of the multiple regression Source df Sum of Squares Mean Square F Regression 4 103.233 25.81 1.89 Residual 410 5603.034 13.67 R2 = .018 p = .112 b) Semipartial multiple and bivariate correlations Variables Multivariate tests Bivariate tests sr2 t p r r2 p SUPGENDR .002 - .858 .391 -.002 <.01 .482 SUPAGE <.001 — .025 .980 .077 <.01 .080 YRSEXP .012 2.281 .023 .156 .024 .003 EDLEVEL .002 - .860 .391 -.050 (.01 .168 Note. The dependent variable is HYPREFER; the least pairwise l_'l_ with other variables is 336; missing data are mean- p-lugged in multiple regression. Probabilities (p) are two-tailed for t and one-tailed for r and F, sr2 is squared semipartial correlation, the change in R2 when a —variable enters into the regression model last or is removed. 158 Table 17 Regression Results for Supervisor Demographic Variables Group with Reported Past Referrals Variable as the Criterion a) Analysis of variance of the multiple regression Source df Sum of Squares Mean Square F Regression 4 46.010 11.50 6.61 Residual 405 704.468 1.74 R2 : .061 p = <.001 b) Semipartial multiple and bivariate correlations Variables Multivariate tests Bivariate tests sr2 t p r r2 p SUPGENDR .001 — .625 .532 .046 <.01 .178 SUPAGE .001 .550 .582 .168 .028 .001 YRSEXP .034 3.811 <.001 .249 .062 <.001 EDLEVEL .006 -1.595 .112 -.107 .011 .016 Note. The dependent variable is PASTREFR; the least pairwise n with other variables is 337; missing data are mean-plugged in multiple regression. Probabilities (p) are two-tailed for _t_ and one-tailed for r and F; sr2 is squared semipartial correlation, the change in R” When a—va_riable enters into the regression model last or is removed. 159 Table 18 Regression Results for Supervisor Demographic Variables Group with Actual Employee Referral Variable as the Criterion a) Analysis of variance of the multiple regression Source df Sum of Squares Mean Square F Regression 4 95.79 23.95 5.77 Residual 209 868.05 4.15 R2 = .099 p : <.001 b) Semipartial multiple and bivariate correlations Variables Multivariate tests Bivariate tests sr2 t p r r2 p SUPGENDR <.001 - .252 .801 .073 <.Ol .144 SUPAGE .001 - .584 .560 .149 .022 .021 YRSEXP .031 2.684 .008 .192 .037 .004 EDLEVEL .063 -3.818 <.001 —.253 .064 <.001 Note. The dependent variable is EMPREFER; the least pairwise n with other variables is 188; missing data are mean— plugged in multiple- regression. Probabilities (p) are two-tailed for t and one-tailed for r and F, sr is squared semipartial correlation, the change in R2 when a variable enters into the regression model last or is removed. 160 This finding supports examination of each of the demographic variables with those dependent measures. The demographic supervisor variables group did not account for significant amounts of the HYPREFER scale variance (32 = .018, _F_‘(4, 410 = 1.89, p>.10). The location effect was likewise nonsignificant, as entrance of dummy-coded variables for locations increased 3’ to .080, a nonsignificant increment (313} = .075, 5(24, 388) = 1.32, p = .146). Supervisor Gender. The dummy-coded variable SUPGENDR did not show significant relationship with either PASTREFR (p = .046, p>.05) or EMPREFER (p = .073, p>.05). Thus, gender of the supervisor is not supported in this study as having relationship with referral. Research Question 9 is answered in the negative. Supervisor Age. The variable SUPAGE was reliably associated with both PASTREFR (_r_ = .168, p<.01) and EMPREFER (p = .149, p<.05). Research Question 10 is answered in the affirmative. These findings support most of the prior research which found that older supervisors are more likely to make referrals to EAPs. The effect of increased age on referral is not strong, however, accounting for only 2—3% of the observed variances. The age effect is not evident in the multiple regression models of all demographic variables, for the likely reason that age is redundant (collinear) with experience and educational level, the three constituting a "cohort" effect (see Appendix D). Experience of the Supervisor. Variable YRSEXP (years of experience of the supervisor) was found to be a valid predictor of past and recent employee referrals to EAP, thereby answering Research 161 Question 11 in the affirmative. 1n the bivariate correlations, increased experience is seen to be a significant positive predictor of referral (L = .249, p<.001 with PASTREFR; L = .192, p<.01 with EMPREFER). The positive effect of greater experience on referral is also seen in the multiple regressions of the supervisory demographic variables with PASTREFR (35’ = .034, _t_ = 3.81, p<.001) and with EMPREFER (g2 = .031, _t_ = 2.684, p<.01). YRSEXP, therefore, should be included in future prediction models of supervisory referral action. Educational Level of the Supervisor. In answer of Research Question 12, the supervisor's educational level was found to have a significant negative relationship with referral behavior; the more educated the supervisor, the less likely he/she referred to the EAP. EDLEVEL'S correlation with PASTREFR was -.107 (p<.05) and with EMPREFER was -.253 (p<.001). In the multiple regression model of EMPREFER, EDLEVEL marginally accounted for 6% of the variance (_t_ = —3.818, p<.001). In the model of PASTREFR it did not reach significance in the marginal test, accounting for less than 1% of the observed variance of PASTREFR (_t_ = -l.595, p>.10), again because of apparent collinearity with age and experience. On the bivariate findings, though, EDLEVEL is validated as a predictor of referral action. Situational Variables Group Research Question 13 poses the hypothesis that some factors inherent in the supervisory or referral situation are related to the supervisor's behavior. The two variables SPANCNTR and ADAFOCUS are considered situational characteristics. The number of employees supervised and the 162 supervisor's perception of his/her EAP's focus on alcohol and drug abuse were tested together for relationship with the three criterion variables. As seen in Tables 19, 20, and 21, the situational variables accounted for significant amounts of past referrals variance, but not of hypothetical or recent referrals. A location effect was significant with this group of variables with PASTREFR as the criterion. The research question is tentatively answered in the affirmative, and the variables are examined for relationship with the criterion PASTREFR. A location effect was significant, as entrance of dummy-coded variables for survey locations increased R2 to .151, an increment significantly greater than 0 (ER2 _ .090, ~1:(24, 381) = 1.68, p = .025). This variables group accounted for 6% of PASTREFR (32:.063, 5(2, 407) = 13.60, p<.001), but less than 1% of HYPREFER (32 = .005, F(2, 412) = 1.02, p>.05) and only about 2% of EMPREFER (_R’ = .016, 5(2, 211) = 1.74, p>.05) variances. With hypothetical and recent referral as the criteria, location effects tests were nonsignificant (for HYPREFER, s32 = .074, {(24, 386) = 1.31, p = .149; for EMPREFER, £131 : .104, 5(24, 185) : 1.01, B = .46). 163 Table 19 Regression Results for Situational Variables Group with Hypothetical Referrals Scale as the Criterion a) Analysis of variance of the multiple regression Source df Sum of Squares Mean Square F Regression 2 28.006 14.00 1.02 Residual 412 5678.261 13.78 R2 : .005 p = .363 b) Semipartial multiple and bivariate correlations Variables Multivariate tests Bivariate tests sr2 t p r r2 p SPANCNTR <.001 — .113 .91 -.001 (.01 .43 ADAFOCUS .005 -1.415 .158 -.O77 <.01 .077 Note. The dependent variable is HYPREFER; the least pairwise n with other variables is 320; missing data are mean- p-lugged in multiple regression. Probabilities (p) are two—tailed for t and one-tailed for r and F; sr2 is squared semipartial correlation, the change in _R’ when a variable enters into the regression model last or is removed. 164 Table 20 .Regression Results for Situational Variables Group w1th Reported Past Referrals Variable as the Criterion a) Analysis of variance of the multiple regression Source df Sum of Squares Mean Square F Regression 2 47.001 23.50 13.60 Residual 407 703.477 1.73 R2 = .063 p : <.001 b) Semipartial multiple and bivariate correlations Variables Multivariate tests Bivariate tests sr2 t p r r2 p SPANCNTR .062 5.175 (.001 .261 . 068 (.001 ADAFOCUS .002 -.855 .393 -.033 (.01 .265 Note. The dependent variable is PASTREFR; the least pairwise p with other variables is 350; missing data are mean-plugged in multiple regression. Probabilities (p) are two-tailed for _t_ and one-tailed for r and F; srz is squared semipartial correlation, the change in R2 When a—vaFiable enters into the regression model last or is removed. 165 Table 21 ' Regression Results for Situational Variables Group With Actual Employee Referral Variable as the Criterion a) Analysis of variance of the multiple regression Source df Sum of Squares Mean Square F Regression 2 15.61 7.80 1.74 Residual 211 948.23 4.49 R2 = .016 p = .179 b) Semipartial multiple and bivariate correlations Variables Multivariate tests Bivariate tests sr2 t p r r2 p SPANCNTR .015 1.78 .077 .128 .016 .038 ADAFOCUS .001 .531 .596 .040 (.01 .287 Note. The dependent variable is EMPREFER; the least pairwise p with other variables is 193; missing data are mean—plugged in multiple regression. Probabilities (p) are two-tailed for t and one-tailed- for r and F; sr2 is squared semipartial correlation, the change in R2 When a—vaFiable enters into the regression model last or is removed. 166 Span of Control. The number of employees supervised (SPANCNTR) accounted for virtually all the group's predicted variance in PASTREFR (SJ:2 2 .062, t = 5.175, p<.001). This is further evident by examination of the bivariate correlations of the two variables (Table 20). This is a weak basis for retaining SPANCNTR as a final predictor, though, for it is obvious that supervisors who have more employees under their direction and control will have greater opportunity to refer to the EAP. The finding of relationship with past referrals is, therefore, not surprising! Examination of the correlation with EMPREFER (even though the multiple regression model does not clearly justify that) shows a significant bivariate correlation between EMPREFER and SPANCNTR (L = .128, p<.05). With this additional evidence, research Question 14 is answered in the affirmative, and SPANCNTR is tentatively retained as a weak validated positive predictor of referral behavior. Alcohol and Drug Abuse Focus of the EAPs. Based upon average ratings of ADAFOCUS, the 25 locations, were seen by the supervisors as representing a spectrum of alcohol- and drug-focused vs. "broadbrush“ programs. The average ratings (see Table 9) of ADAFOCUS varied among the locations from 3.9 (giving clear emphasis to alcohol and drug intervention) to 2.2 (having no special focus and definately "broadbrush" in nature). A one-way analysis of variance of the supervisors' ratings of the EAPs' ADAFOCUS supports the assertion that the participating locations vary in their degree of intended and perceived drug and alcoholism emphasis (5(24, 352) = 4.83 , p<.001). Supervisors' perceptions of their local EAP and their "typing" of the program by means 167 of the five available options was only fairly consistent, however. Agreement among supervisors within each of the 25 locations was substantial (40% to 60% agreement) with clustering of responses evident by inspection of the data. But this aspect of the EAP does not seem to be highly salient to the respondent supervisors--l3% of the sample chose not to respOnd to the item. Research Question 15 inquires whether the available EAP's focus on alcohol and drug abuse problems affects supervisory referral patterns. Supervisor's perception of their program's focus on those problems was tested for relationship with the criterion variables. The variable, ADAFOCUS, did not demonstrate significant relationship with any of the dependent variables (see Tables 19, 20, and 21). ADAFOCUS was correlated -.08 (p>.05) with HYPREFER, -.03 (p>.05) with PASTREFR, and .04 (p>.05) with EMPREFER. The question of whether supervisors' perception of the available program's focus (alcoholism vs. broadbrush) affects their referral behavior is answered in the negative. Employee—related Variables Group Research Question 16 is answered by examining the multiple correlation of a group of factors to referral handling of the specific employee with job problems. The factors considered related to the employee are SEVERITY of job performance impairment, the commission of a TRIGGER event, the employee‘s occupational level or status in the organization (OCCLEVEL), gender (EMPGENDR), and whether the supervisor and employee were of the same or different gender (CROSSGEN). In this study, those factors could only be examined in relation to one of the 168 mea5ures of supervisory referral behavior, EMPREFER. Results of the regression analysis are shown in Table 22. The variables group was found to be highly reliably associated with referral, accounting for about one-fourth of the observed variance of EMPREFER (5’ = .23, 5(5, 208) = 12.44, p<.001). The test for location effects was nonsignificant. Entrance of dummy-coded variables for survey locations increased _R_2 to .327, an increment not reliably greater than 0 (_sR2 = .096, 5(24, 184) = 1.10, p = .351. The research question is answered affirmatively. The individual employee-related factors and the degree and manner each was related to EAP referral of that employee who was identified by the respondent supervisor as "the one with the most job problems and the one most difficult to supervise in the past year" are considered next. SEVERITY of Job Performance Impairment. Research Question 17 addresses the job-performance based claims of the EAPs in the survey-~is supervisory referral actually influenced by the employee's degree of job impairment? Performance impairment, as measured by SEVERITY, was found to be a significant positive predictor of referral. The EAP policies of the participating organizations, as described above, legitimate supervisors' referrals on the basis of an employee's inadequate job performance. This is in accord with accepted "theoretical" statements of EAP function and operation. It would have been a very disturbing finding, indeed, to have found no relationship whatever between seriousness of the employee's job performance impairment and his/her referral to an EAP. The research question is answered in the affirmative. 169 Table 22 Regression Results for Employee Variables Group with Actual Employee Referral Variable as the Criterion a) Analysis of variance of the multiple regression Source df Sum of Squares Mean Square F Regression 5 221.92 44.38 12.44 Residual 208 741.91 3.57 R2 = .230 p = (.001 b) Semipartial multiple and bivariate correlations Variables Multivariate tests Bivariate tests sr’ t p r r2 p SEVERITY .013 1.886 .061 .193 .037 .002 TRIGGER .165 6.686 (.001 .419 .176 (.001 OCCLEVEL .010 ~1.657 .099 -.129 .017 .030 EMPGENDR .011 1.763 .080 .123 .015 .037 CROSSGEN .002 - .775 .439 -.122 .015 .037 Note. The dependent variable is EMPREFER; the least pairwise n with other variables is 213, missing data are mean-plugged in multiple regression. Probabilities (p) are two- tailed for t and one-tailed for r and F, sr2 is squared semipartial correlation, the change in R2 When a variable enters into the regression model last or is removed. 170 In these data, SEVERITY and EMPREFER had about 4% of their variance in common (1 = .193, :2 = .037, p<.01). In the multiple regression model of EMPREFER, job impairment only accounted for about 1% of the observed variance. In a two-tailed significance test of its marginal (last-entered) predictor status, SEVERITY approached, but did not reach the prescribed critical value (I = 1.886, p>.05). On the basis of the bivariate correlation, SEVERITY of job impairment is affirmed to affect supervisors' referral actions. The more severe the impairment of an employee's job performance, the more likely a supervisor is to refer that employee to an EAP. TRIGGER Incident Committed by the Employee. Research Question 18 hypothesizes that the occurrence of a dramatic attention-getting event, identified by variable TRIGGER, may be related to referral action. This was confirmed by the findings. bivariate (biserial) correlation of TRIGGER with EMPREFER was .419 (p<.001). This strong association is also seen in the multiple regression model of the employee variables and EMPREFER, where TRIGGER accounted for about two-thirds of the total predictable variance (3‘ = .23, and g2 = .165, t = 6.686, p<.001). The research question of whether commission of a dramatic "triggering event" affects the likelihood of the supervisor's referring an employee to the EAP is answered "yes, very strongly." Occupational Level of the Employee. Higher occupational level or status of an employee was again found to serve as a deterrent to the supervisor to refer him or her to the EAP. A statistically significant, though modest negative relationship was observed between OCCLEVEL and 171 EMPREFER (L = —.129, p<.05). The relationship was again observed in the multiple regression model of the data, where OCCLEVEL marginally accounted for 1% of the predicted EMPREFER variance (_t_ = -l.657, p<.10). Research Question 19 of whether higher occupational level or status of the employee makes it less likely that he or she will be referred to the EAP is answered "yes, but only slightly." Gender of the EmploLee. Research Question 20 of whether employee's gender affects the likelihood of supervisory referral is answered by the correlation between EMPGENDR and EMPREFER (5 = .123, p<.05). Employee's gender was found reliably associated with EAP referral, males having greater likelihood of being referred. The relationship is not strong, however, accounting for about 1% of the observed EMPREFER variance both by itself and in the multiple regression model. Gender of the employee (EMPGENDR) is retained as a valid, but not a strong predictor of supervisory referral. Gender Difference of Employee and Supervisor. The research Question 21 of whether gender match or mismatch of supervisor and employee affects referral behaviors is not clearly answered in the results of this study. As shown in the bivariate test of association, CROSSGEN and EMPREFER have a modest significant negative relationship (L = -.122, p<.05). On that evidence, a mismatch of gender would be said to make referral action slightly less likely. However, the multiple regression results make CROSSGEN's predictor status less tenable. Examination of the matrix of intercorrelations (Appendix D) will reveal that CROSSGEN has a large correlation with EMPGENDR (being coded 172 partially by means of that variable), but does not covary with the other employee-related variables. In the multiple regression, however, CROSSGEN was marginally associated with only a trivial amount of EMPREFER variance (_s:2 = .002,1 = -.775, p>.05). The conclusion is reached, balancing the available evidence here, that CROSSGEN's status as a valid predictor of referral is uncertain. It will not be retained as a predictor variable in this study. Gender difference between supervisor and employee may deter referral action, very slightly, but it was not treated as a reliable effect in this study. Location Effects Research Question 22 posed the issue of whether the observed effects of the validated factors differ across the various survey sites, i.e., whether there is a location effect. Those results have been reported piecewise above, and they will be summarized here. The data set included the means to examine the apparent generality of the findings. Availability of numerous survey sites permitted entry of dummy-codes for locations as a variables group, testing for a significant effect due to location. Increment of the coefficient of multiple prediction was tested in each of the multiple regression models with the factor groups (cognitive, employee-related, etc.) for the location effect (see Tables 10—22). In all of the analyses of HYPREFER and EMPREFER, the addition of the locations dummy-code variables set did not make a significant addition to the portion of variance accounted for. In the multiple regression model of cognitive variables with PASTREFR, the location effect was non-significant (31$2 = .05, I_:_(24, 380) = 1.28, 173 p>.05), but a significant increment was observed with the supervisor-related variables group (s_rz : .09, 5(24, 381) = 1.68, p<.05) and with the situational variables group (_s_rz = .087, 1:(24, 383) = 1.64, p<.05). As can be seen from Table 9, there is variation on the PASTREFR variable across the survey sites. A one-way analysis of variance of PASTREFR by locations confirmed that the sites differ in supervisors‘ reported history of past referrals made (5(24, 383) = 1.87, p<.05). In posthoc comparisons of the sites using Scheffe's method, no two were found to reliably differ, however. A variety of competing explanations may be offered to explain the differences between the sites on number of reported past referrals by supervisors. It is speculated that the PASTREFR variations occurred because of the considerable differences in ages of the available EAPs. Another possible reason might be that different emphasis is actually placed on supervisory or self-referral case-finding at the different locations, and that, policy similarities notwithstanding, the observed frequency differences reflect true program emphasis differences. With respect to the research question of whether the factors affecting referral operate differently depending upon location, the conclusion is reached that LOCATION does not, overall, contribute a main effect. Looking at the findings as a whole and considering the tests of the three criteria as independent (which they are not, strictly), the lack of effect for location in the tests with hypothetical and recent referral action is more persuasive than the small locations effect noted with past referrals. Furthermore, those differences may plausibly be 174 explained by other known reasons associated with the surveyed programs (though not tested). Models of Predictor Variables and Supervisory Referral The factors found in this study to validly predict the observed referral behaviors of supervisors were combined to generate full multiple regression models of the past and recent referrals. A full regression model of HYPREFER was not attempted, however, because the referral action taken with the hypothetical employees was reliably predicted only by the cognitive factors. HYPREFER was somewhat deficient as a dependent measure, having less reliability (.35 Cronbach's alpha) than one would desire. HYPREFER did not strongly covary with the other measures of referral action (see Appendix D). The multiple regression model of the cognitive variables with HYPREFER, which accounts for 8% of the observed variance, is considered the limit of reliable prediction of the hypothetical referral scale. Multiple Prediction Model of Past Referrals A full multivariate regression model of PASTREFR is shown in Table 23. The predictor variables accounted for 36% of observed variance of past referrals by supervisors, a highly significant result (3’ : .362, 5(8, 401) = 28.48, p<.001). The semipartial multiple correlations of the variable groups indicate that the cognitive variables are major contributors to the prediction equation; their removal would be a highly significant loss—-two—thirds of the accounted-for variance. 175 Table 23 Regression Results for Full Multivariate Model of Validated Factors with Reported Past Referrals Variable as the Criterion a) Analysis of variance of the multiple regression Source df Sum of Squares Mean Square F Regression 8 271.90 33.99 28.48 Residual 401 478.58 1.19 R2 = .362 p = (.001 b) Semipartial multiple and bivariate correlations Variables Bivariate tests Multivariate tests r r2 p sr2 t p APPROACH .217 .047 (.001 (.001 - .53 .60 AVOIDANCE -.292 .085 (.001 .016 -3.201 .002 FAMILIAR .542 .294 (.001 .090 7.520 (.001 TRAINING .386 .149 (.001 .026 4.024 (.001 SUPAGE .168 .028 .001 .001 - .749 .454 YRSEXP .249 .062 (.001 .001 .937 .349 EDLEVEL -.107 .011 .016 (.001 - .359 .720 SPANCNTR .261 .068 (.001 .022 3.721 (.001 SR2 F p Approach and Avoidance Scales .016 5.13 .006 Cognitive Variables Group .261 40.96 (.001 Supervisor Demographic Variables Group .002 .348 .791 Situational Variable .022 13.846 <.001 Note. The dependent variable is PASTREFR; the least pairwise p with other variables is 315; missing data are mean-plugged in multiple regression. Probabilities (p) are two-tailed for t and one-tailed for r and F; sr2 (sR’) is squared semipartial (multiplé) correlation, the - change 171 RT—when a variable (or group of variables) enters into the regression .model last or is removed. Went—1:17» .;.:L:.A-.:n;r. . . . ,, . 176 Collinearity of the supervisor-related variables creates problems of interpretation, resulting in non-significance of their marginal increments of prediction, and even resulting in sign-reversal of the effect of supervisor‘s age. Familiarity made a large significant contribution to the model, marginally accounting for about 10% of PASTREFR variance, but that may result from logical redundency of the variable with PASTREFR. Approach and avoidance attitudes towards referral, training effects, and span of control (probably present as opportunity to refer) also made similar comparable significant contributions to the regression model, each adding about 2% to the proporton of variance accounted—for when entered last. An effect due to locations was tested and again found absent; the increment to the accounted—for variance was non-significant (3132 = .048, F(24, 377) = 1.26, p>.05). Multiple Prediction Model of Recent Supervisory Referral Behavior A full regression model of EMPREFER with the validated factors affecting supervisory referral is shown in Table 24. This multiple regression model accounts for nearly one-third of the observed variance of supervisory referral action taken with the employee identified as having had the most job problems in the previous year. We may say with a high level of confidence that these factors in combination predict fairly well the recent referral action taken or not taken with an employee showing job problems (3’ = .315, F(12, 201) = 7.71, p<.001). 177 Table 24 Regression Results for Full Multivariate Model of Validated Factors with Actual Employee Referral Variable as the Criterion a) Analysis of variance of the multiple regression Source df Sum of S_quares Mean Square F Regression 12 303.70 25.31 7.71 Residual 201 660.13 3.28 R2 = .315 p = (.001 b) Semipartial multiple and bivariate correlations Variables Bivariate tests Multivariate tests r r2 p sr2 t p APPROACH .189 .036 .003 .003 .96 .338 AVOIDANCE -.184 .034 .003 .010 -1.743 .083 FAMILIAR .314 .099 (.001 (.001 .372 .710 TRAINING .182 .033 .004 .002 .814 .417 SUPAGE .149 .022 .021 (.001 - .244 .808 YRSEXP .192 .037 .004 .009 1.666 .097 EDLEVEL -.253 .064 <.001 .022 -2.538 .012 SPANCNTR .128 .016 .038 (.001 .485 .628 SEVERITY .193 .037 .002 .009 1.649 .101 TRIGGER .419 .176 (.001 .123 6.003 (.001 EMPGENDR .123 .015 .037 .005 1.261 .209 OCCLEVEL -.129 .017 .002 .005 -1.l77 .241 SR1 F p Approach and Avoidance Scales .017 2.51 .084 Cognitive Variables Group .028 2.08 .085 Supervisor Demographic Variables Group .032 3.09 .028 Situational Variable <.001 .236 .628 Employee Variables Group .149 10.92 (.001 Note. The dependent variable is PASTREFR; the least pairwise n with other variables is 188; missing data are mean-plugged in multiple regression. Probabilities (p) are two-tailed for t and one—tailed for L and F; sr2 (5R2) is squared semipartial (multiple) correlation, the change E Rrwhen a variable (or group of variables) enters into the regression Tnodel last or is removed. 178 These results, together with the bivariate correlations of the factor measures with EMPREFER, give an indication of relative influence on referrals in this data set. The most striking finding is that TRIGGER accounts for as much observed variance by itself as do all the other variables combined. This may be seen by comparing its coefficient of determination (1’ = .176) with the coefficient of multiple determination with TRIGGER removed from the full model (32 = .192). Omission of the possible occurrence of a dramatic, attention-getting event from an explanatory model of supervisory referral to an EAP would seriously weaken the model (in this case, s_r’ = .123, l = 6.003, p<.001). In comparison, SEVERITY of job performance impairment accounted for only 4% of observed referral action with the job-problems employee, and contributed less than 1% additional variance to the full model (3’ = .009, _t_ = 1.649, p>.05). Thus, poor job performance, which is relatively steady—state or in gradual decline, is much less likely to result in EAP referral by a supervisor, even though EAPs, by design, claim legitimacy on the basis of dealing with job impairment. Overall, these data indicate that about one—half of the referral action is due to characteristics of the employee (332 = .149) in question and half due to factors associated with the supervisors, themselves--their attitudes towards referral, program familiarity, relevant training received, age and experience and educational level, and their span of control. The changeable cognitive factors and the relatively unchangeable demographic factors appear to be about equally influential upon the observed recent supervisory referral actions. 179 Collinearity problems again appear to prevent straightforward interpretation of the full regression model of the factors and EMPREFER. Many of the factors which show significant relationship in their bivariate association with EMPREFER lose their significance in marginal tests in the full model. The effects of SUPAGE even had a change of sign (direction of effect), pointing up, again, the inadequacy of the multiple regression model looked at in isolation from all other inter-relationships in the data. Span of control showed a very weak influence on referral, accounting for a trivial slice of EMPREFER variance in the full prediction model. Its status as a valid predictor is clearly assailable in this data set. A location effect was also not found with the full regression model of EMPREFER. Addition of the dummy-coded variables for location did not significantly add to the amount of EMPREFER variance accounted for by the factors ($2 = .094, F(24, 177) = 1.18, p>.05). Summary of Findings Many of the variables which had been proposed in the EAP literature or in previous EAP research efforts to influence supervisors‘ use of the constructive confrontation strategy were again confirmed. Some were not validated in this study. The findings from all the analyses are summarized in Table 25. A fair degree of consistency among the findings is apparent. The conclusions reached and the implications drawn from the findings will be discussed in the next chapter. The statistical significance levels observed, number and consistency of findings, and general agreement of the findings with previous research 180 Table 25 Summary of Findings on Factors Affecting Supervisory Referrals Group Variable Recent Past Hypothetical Referral Referrals Referrals (EMPREFER) (PASTREFR) (HYPREFER) Cognitive X X X APPROACH X X AVOIDANCE X X X FAMILIAR X X X TRAINING X X '7 COUNROLE * * * Locations * * * Supervisor X X * SUPGENDR * * * SUPAGE X X * YRSEXP X X " EDLEVEL X X * Locations * x * Situational * X * SPANCNTR ? X * ADAFOCUS * * * Locations * x * Employee X not tested in this study SEVERITY X TRIGGER X OCCLEVEL X EMPGENDR X CROSSGEN ? Locations * Location Effects Summary from above * ? * * * Full Regression Models Note. X marks valid predictors; * "not significant"; ? uncertain 181 all indicate that the study had good statistical power and that the measures of each variable probably had adequate reliability and construct validity. The consistent finding that locations variables did not appreciably increase the performance of the prediction models lends some encouragement that the findings have some external generalizability and may be robust to other settings. The Attitude toward Referral Scales (APPROACH and AVOIDANCE) showed modest, yet consistent significant relationships with the measures of supervisory referral action. Referral attitudes are tentatively established as valid predictors of referral to EAP and are worthy of our concern as factors influencing supervisory action, along with the others previously identified. Gender-related effects do not seem, in this study, to be as strong as previous literature has argued. Occupational level or status of the employee also was found to have weak influence on supervisors' decision to refer. Number of employees supervised also influenced referral, but weakly. The degree of program focus on alcoholism and other drug abuse was found to have no relationship with supervisors‘ use of constructive confrontation with employees. Tendency to see oneself in the role of a counselor (as measured by scale COUNROLE) also showed no effect upon referral action taken. The severity of the employee's job performance impairment, having participated in at least some training about the EAP, or otherwise gaining familiarity with the available program and how to refer employees all increased supervisors' likelihood of referring to some extent. The factor that most influenced the supervisor to refer, however, was found r———=—=——=—=—fii——m‘we—W 182 to be the commission of a relatively public and dramatic "triggering event" at the workplace. In this study, that factor alone influenced supervisory EAP referral action as strongly as all the other known and studied factors combined. CHAPTER 5 SUMMARY AND RECOMMENDATIONS Overview This study examined factors which influence supervisors' use of the constructive confrontation strategy with employees. The key role of the supervisor to refer employees having job problems was developed from the EAP literature. Factors shown by previous EAP research to affect referral action were identified, and the current evidence for their importance detailed. It has been noted that supervisory referrals very commonly are not made in a timely manner. The period of vacillation and hesitation supervisors go through before making an appropriate referral was conceptualized as approach/avoidance conflict. This researcher developed attitude scales having approach and avoidance valences in order to test the possible applicability of the conflict metaphor to the supervisor's dilemma: to refer or not to refer. Through a subscale specification and item selection process which sampled supervisors‘ attitudes toward referral, inclining or deterring action, the Attitudes toward Referral Scales (APPROACH and AVOIDANCE) were developed. These attitude scales were embodied in a questionnaire survey instrument which was completed by 415 supervisors in 25 locations. The survey sites provided variability as to type and size of organization, type of EAP (internal or contracted external), age of EAP, focus on 183 184 alcoholism and other drug abuse, and geographic location. All the organizations had well-established EAPs in operation, and each sanctioned the role of the supervisor to initiate or require EAP involvement for employees having job performance deficiencies. The survey instrument collected data on approach and avoidance attitudes and fourteen other variables that were proposed to influence supervisors' referral behavior. Supervisory referral was measured by past referrals made, recent referral action taken with the employee identified by each supervisor as having had the most job problems in the past year, and imagined referral actions taken with four hypothetical employees typical of EAP clients and portraying various job problems. The study provided additional validation for some of the previously identified factors in affecting supervisory referrals. The results provided initial validation of the Attitudes toward Referral Scales as predictive of referral action and suggeSI their possible usefulness in future EAP operations and research. While the conflict model of approach and avoidance was not directly testable in this study, the results were not inconsistent with a conflict model of supervisory referral behavior. Methodological Discussion and Recommendations Supervisory referral action is, like other real-life decisions, a complex, multiply determined behavior which does not allow of simple specification and prediction. This effort to develop a multiple prediction model was modestly successful, but highlights the need for improvement of both the dependent measures of referral and the explanatory model, itself. 185 The first methodological recommendation was made in Chapter 3. It was noted there that data on referral handling of specific employees may be considered too "sensitive" for supervisors to willingly respond to survey questions requiring them to disclose their action taken or not taken with a problem employee. In order to address this problem, it is recommended that future researchers achieve sample size appropriate to the power requirements of the effects tested, but collect the referral data directly from the supervisors through short, focused, confidential interviews. Biased selection undermines internal validity of the study, and using data collection means to achieve nearly complete sample participation would address this threat to validity of the findings. The second recommendation is to develop, through needed future research, a "standard," reliable dependent measure of supervisory referral action. The precision and aptness of the regression models tested were surely affected by the less than desirable reliability of the dependent variables (criteria) used in this study. The three measures of supervisory referral (HYPREFER, PASTREFR, and EMPREFER) correlated positively among themselves, but not as highly as one would wish, considering that each is intended to be a partial measurement of supervisory referral to EAPs. The final multivariate models accounted for around one-third of the observed variance of past and recent referral actions. It would appear, on the face of it, that the model needs further elaboration, using more and better predictor variables. On the other hand, the model may actually be fairly complete and adequate, and only looks weak because of measurement errors or inadequate definition of the criterion variables. 186 The use of hypothetical referral situations, as in the four vignettes here and the derived scale HYPREFER, needs critical investigation as a suitable proxy for actual referral behavior. The findings of this study suggest that hypothetical referral actions bear only a small, though positive, relationship with actual supervisory behaviors. It is recommended that research be undertaken to develop and validate a Referral Behavior Scale based upon typical and appropriate EAP referral cases. The APPROACH and AVOIDANCE scales may also benefit from further development. They can probably be shortened, at the same time improving their reliability and construct and predictive validity. It is noted that the current version of subscale Discomfort Relief for Supervisor has questionable internal consistency. A factor analytic study of the scales will be helpful to better understand the instrument's structure and psychometric properties. Confirmatory factor analysis, not done before or as part of this study, would give some needed support to the contention that these items do, in fact, constitute two separate attitude valences of approach and avoidance and not a large, nebulous clustering of EAP—related beliefs and sentiments. Discussion of Factors Influencing Sgervisory Referral The findings of this study bring a mixture of "good news and bad news" and a lot of new questions that beg for additional research. First, the study lends further evidence that supervisors neither fully accept nor confidently use the constructive confrontation strategy, even in organizations with well-established EAPs. Clearly, they do make 187 referrals—-two-thirds of the survey's respondents (from tabulation of the PASTREFR variable) said that they have made at least one "referral" to their EAP, however defined by themselves. One-eighth of the 415 supervisors described themselves as having made five or more referrals, a very active level of casefinding and referral. That is the good news. On the other hand, when looking at the supervisors‘ handling of the "problem worker" during the past year (EMPREFER distribution), one finds that 42% of the supervisors chose "not to get the EAP involved" to any degree, while another 10% only "casually mentioned" the program. Recall that about half the supervisors chose not to indicate how they handled that employee "having job problems." But of those, one—third (67 in a cross—tabulation) indicated on PASTREFR that they had never made an EAP referral. So, with minimal interpretation, the responses indicate that half or more of the Supervisors took no EAP referral action whatever with an employee identified in their own minds as having job problems and presenting a supervisory challenge in the previous year. Although this should not be interpreted as evidence that these supervisors, on the whole, do not accept EAP or universally reject the referral role, it is further evidence that supervisors do not engage in “early problem intervention." They do not appear to see EAP referral as a tool which closely supports them in their function, and they will not bring it into play quickly when workers have problems. A finding conspicuous in its nonsignificance was the observed lack of relationship between supervisors‘ referral activity and the EAPs‘ focus on alcohol and other drug abuse, as that emphasis is accurately perceived by the reSpondents. This has been, as noted above in 188 Chapter 2, a highly contentious area of the EAP field, but one which is little researched. This study found the supervisors in the organizations with broadbrush EAPs having similar levels of referral activity and making similar degrees of referral effort with the job problems employee as those in organizations with more traditional alcohol/drug focused EAPs. This represents another instance where the concerns raised by Roman (19813, 1982) do not appear to be substantiated. The program focus was not, among the programs studied, a factor in supervisory acceptance and use of their referral option. Claims for the superiority of the earlier "occupational program“ model must be supported on better grounds than authority and dire predictions of lack of appropriate referrals. In this, as in several other areas of EAP practice, we need "less heat and more light" in the form of empirical findings. The conclusion seems tenable that the central role which is "theoretically" reserved for the supervisor in EAP operations ought better be assigned elsewhere. A principal finding of this study supports this position. The factor which best accounted for referral action, by far, was TRIGGER, the commission by an employee of a public, flagrant act in the workplace which draws attention to the employee and to the supervisor. This event was very powerful in the supervisors' minds, a "sign" not to be ignored, and they didn't. The triggering event accounted for referral action taken as well as all the other valid predictor factors combined. Severity of job performance impairment, as measured by supervisors' brief performance rating of the job problems of their most recent troublesome employee, indicated consistently unsatisfactory or marginal performance, but that factor was only about 189 one-tenth as potent in resulting in referral action. It is as a tool for managing substandard performance that EAPs were conceived and by which they claim legitimacy, but it appears the supervisors do not use them strictly for that purpose. Had that been the case, the relative prediction strengths of SEVERITY and TRIGGER might have been reversed. The question may even be raised, based on these findings, whether supervisors use EAP referral in a punitive manner, not primarily out of benevolent and altruistic motives as has been proposed in the past. Recall that Trice and Belasco (1965) long ago discovered that negative, nonaccepting attitudes of supervisors toward impaired employees were associated with referral action, not sympathetic and benign ones. The implication might be drawn that supervisory referral is, for many, thought of largely in terms of social control and only secondarily as rehabilitative and restorative. The very powerful role of the trigger event suggests that supervisors see EAP referral as a "crisis management" tool and not as a routine procedure which plays a frequent and regular part in their managerial activities. It appears to function as a "special" solution to frustrating "special problem cases" which have proven annoyingly persistent, and having tried most everything else, the supervisor refers as a "last resort" following a public triggering incident. Other writers and researchers have made this inference (e.g., Erfurt 6c Foote, 1989; Googins 6t Kurtz, 1984; Riediger, 1979). There are converging indications that the constructive confrontation strategy, as a primary means of casefinding in EAPs, is not working very well. This conclusion is not easily accepted by many in the field of EAP. 190 There is sound evidence (e.g., Trice 6c Beyer, 1984b) that the impaired workers, once confronted and referred, benefit from intervention by EAP counselors. The cost-saving claims of EAP are well founded; it does not matter so much how the impaired employee arrives for help, more that contact is made and appropriate intervention begun. The practical and philosophical problem continues to be: at what point in the development or emergence of problems, personal and job performance, M the EAP practitioner or counselor become involved and by what means can that be accomplished? The position "as early as possible" makes the most cost-saving sense for a host organization. Trice, Roman, Beyer, and, more recently, Blum have been active and persuasive proponents of the constructive confrontation strategy for supervisors, emplacing it as the “key element" of job performance—based EAPs. It is understandable that they would continue to have a certain fondness for and advocacy of their intellectual offspring. However, as a parent must not be blind to the faults and shortcomings of a child, constructive confrontation does not merit its favored status as an entry to EAP services. The anecdotal reports of those in the field and the empirical findings beginning to accumulate should encourage a greater reserve and, perhaps, less enthusiasm for supervisory referral and its natural role in EAPs. Need for Further Research The employee assistance field needs to become less doctrinal and polemical and much more empirical. Much remains to be learned about constructive confrontation as a program component, and supervisors' referral behavior offers a natural setting to explore decision processes, 191 help-seeking and help-accepting behaviors, and how cognitive processes such as labeling, group norms and deviancy control, altruism, motivational arousal, and attribution of responsibility, as examples, affect a significant behavior. This study found that about one-half of referral action variance was accounted for by factors associated with the troubled employee, another one-fourth by possibly changeable cognitive aspects of the supervisor, and another one-fourth by relatively unchanging attributes of the supervisor. This presents challenge and opportunity to EAP researchers and practitioners-—a challenge to researchers to better understand the internal factors in the supervisor's decision process which lead to referral action, and opportunity for EAP practitioners to use what is already known to influence the referral process in the direction of earlier and more consistent casefinding. A good deal more research is needed and recommended to get a more detailed, fine-grained look at what triggering events lead to referral and what motivational and attitudinal sets at the time of the event contributed to the supervisor's subsequent decision to refer the employee to the EAP. Research is needed to find out just how supervisors do view the referral action-~does it carry punitive connotations? Is EAP referral essentially a form of identification and control of social deviancy in the workplace? Bayer and Gerstein (1988) asserted that the supervisor's motivation to refer is essentially prosocial, applying the Bystander Intervention Model of victim assistance. The attribution of benevolent motives to the supervisor is consistent with how EAP practitioners would like to see themselves and their clients, but there may also be less comfortin motives involved. There a ears to be a "cohort" of older 8 PP 9 192 more experienced, less educated supervisors who are consistently more inclined to refer, but they respond to the crisis presented by a TRIGGER, too. By further study of the psychological processes of the supervisor, better predictors may be available to shed light on their referral behavior. Some potential candidate variables are proposed. Altruism scales exist and have been used to study a variety of helping and prosocial behaviors. It would be interesting to learn whether that motive is related to referral. At the other extreme, there is an extensive literature on dogmatism and authoritarianism; it would also be interesting to learn what role an authoritarian or dogmatic interpersonal stance plays in EAP referral. In a similar vein, other trait—like variables can be examined for possible relationship to supervisory referral—-achievement and affiliation needs, social dominance, A/B personality types, and leadership style and/or organizational climate as the context of referral. Although there is not an interpretive nomological net for possible findings, it would be interesting to find if the Rosenzweig Picture Frustration Test (which measures psychological ways persons handle frustration and anger) can differentiate referring and nonreferring supervisors. Another possible research line is suggested by the research of Cohen (l986/1987) and Roth and Cohen (1986). They suggest looking at approach and avoidance as general tendencies or strategies which people use to handle stressful situations, such as may be present when an employee is having job problems and is difficult for a supervisor to manage. Roth and Cohen suggested (1986, p. 816) that "in some situations people have a 193 measuring/predicting supervisory vacillation and delay of referral. The conjecture awaits some research evidence. The design of this study did not permit direct testing of the approach/avoidance conflict model as pertaining to supervisors facing an EAP referral situation. The attitude scales APPROACH and AVOIDANCE were looked at aggregated across all supervisors, those who had, those who wouldn't, and those who needn't refer. Relationships between the scales and referral were found, and findings were consistent with the predictions of conflict theory. The application of the model to the supervisor's referral situation is not adequately tested, however. One line of research, using improved scales, would be to look at different supervisors in different levels of self-described conflict-—are higher levels of conflict associated with stronger attitudes, as conflict theory suggests? Does avoidance show greater motivational strength in those conflict cases? Do attitudes shift just before and/or after referral? What role does "outcome" of the referral play in attitude changes? What role does pressure to refer "from above" play in attitude and behavior shifts? Numerous aspects of supervisory referral may be explored using the developed scales and applying conflict theory, which need not be set aside yet. Attitudes toward referral, program familiarity, and training effects strong preference for either approach or avoidance responses," though use of either strategy is not mutually exclusive of the other. The presence of such a generalized tendency to deal with stressful situations or avoid them as long as possible may be another way of conceptualizing and Were shown to be affecting supervisors' referral actions. These may all 1 — E mmAM—awvmuaj- ,-«.,. . 194 be impacted by EAPs. There is much that can be done to build awareness and familiarity with the EAP and supervisors' possible role. Simply doing more in the way of program publicity may enhance referral effectiveness. Study of various forms and intensities of program "advertising" would be helpful to EAP administrators. As one means to build familiarity, at least some training for all supervisors is easily justified and recommended. What training does to affect and increase referrals remains to be seen. Some of the aspects of the inter-relationships between level of formal education (another validated referral predictor), EAP-related training, referral attitudes, and referral action are shown in Figure 4. What training could be doing besides increasing familiarity is to change attitudes, which have a demonstrable effect on referral behavior. ,J»3§——””Approach 32 Training ;;‘:;<13\ -11 - - 04' 10 Referral Edlevel -29 ~5273‘-~ Avoidance -J8 Figure 4 Observed correlations between EDLEVEL, TRAINING, APPROACH, AVOIDANCE, and the referral variables PASTREFR and EMPREFER. Note: Correlations with EMPREFER are shown in boldface type. 195 Judging from these findings, though, training currently being conducted does not appear to strongly affect attitudes. To the extent that it does, it appears that favorable APPROACH attitudes are increased more through the training experience than are impeding AVOIDANCE attitudes reduced. This imbalance needs to be studied more closely, as conflict theory suggests that avoidance is stronger motivationally, and that more will be gained by reducing avoidance than by increasing approach in order to gain maximal movement toward the goal and minimize distress within the actor. It is also recommended that supervisory training efforts be evaluated as to their "effective ingredients" and short— and long—term impacts on the supervisors and their actual referral behaviors. APPROACH and AVOIDANCE attitudes scales can play a useful part in that. As Trice and Belasco (1969) showed, in their extensive study of EAP training, pretesting of the participants "sensitizes" them to the training inputs and makes the experience more impactful. The attitude and familiarity scales developed in this study could serve that use admirably. Further, post-testing at various intervals will show the training effects upon two major factors known to have relationship with referral actions. As a line of research, similar pre/post testing of attitudes in relation to various types of training experiences and content will be extremely useful in the development of training which is both effective in supervisory change and cost—effective in requiring minimal training time. The study findings disclosed a puzzling relationship between training and formal education and attitudes and referral (Figure 4). It is evident that EAP-related training and higher educational attainment 196 impact referral attitudes and referral action in exactly opposite ways. This raises some new research questions. What is it about having/getting more education that causes the supervisors to be more avoidant of and less favorable toward referral and less likely to do it? Is this really a product of increased education, or is it indicative of certain attitudes and personal values associated with higher educational levels? Perhaps this is finding should simply be attributed to the fact that more educated supervisors were more likely to be in settings in which they supervised smaller, higher-level workgroups. At any rate, the negative relationship between educational level and willingness to refer is somewhat discouraging, as EAP-referral for employees in need seems quite rational and is even more cost-effective when high-level, professional employees can be restored to full effectiveness on their jobs. In this study, a "gender—effect" did not appear as strongly as the past literature would suggest. Men and women supervisors made EAP referrals similarly. Also, whether the supervisor and employee were of same or different gender was not a clear factor in the referral action. But again it was found that males are more likely to be referred by their supervisors to the EAP. It remains to be discovered why this is so. Is it because their "triggering" events are more public or extreme? Does their job performance trajectory follow a different curve from impaired women employees', causing the supervisors to respond differently to them? Is it because men have a markedly higher incidence of substance abuse problems and sociopathy, manifesting different sorts of misconducts in the workplace? Or is this finding mere artifact, since women accept 197 the "suggestion" of EAP referral, whereas men must be coerced to accept EAP help by more extreme threats? Investigating the differences between impaired men and women in the workplace and their respective help-seeking and help-accepting attitudes and behaviors should be a fertile EAP research area. It does not seem to this researcher that continued demonstration of differences between men's and women's entry to and participation rates in EAP service is needed; now we need to know why. The issue of help-seeking and acceptance is as important to the EAP field and practitioner as to community psychologists and psychotherapists. Recommendations for EAP Practice The overall finding in this study, that the likelihood of EAP referral is influenced by qualities of the supervisor as well as the employee, suggests to this EAP researcher and clincian that the role of supervisory referral in EAP theory and operation should be re-evaluated and changed. Specifically, supervisors do have an important role in linking troubled workers to needed assistance, but too many other factors get in the way of that happening in a routine, timely, and efficient manner. As a system of case-finding and linkage, supervisory referrals have a serious shortfall in facillitating the cost-saving and employee-helping potentials of EAPs. Therefore, recommendation is made to modify the role of the supervisor so that referrals become more routine, even automatic. The role of the EAP and assistance worker would be shifted at the same time to one which makes it more active and initiating--an "outreach" function 198 which meets the troubled worker halfway and lowers the threshold of help—accepting for them. The means proposed to accomplish this role shift for both supervisor and EAP worker is a new form which the author would call "Slob Performance Discussion and Action Document." This is an extension of the "employee incidents file" which is frequently recommended as an effective way for supervisors to document and keep track of employees‘ performance, good or bad. It is helpful at time of doing annual performance appraisal and is implied by the long-standing formula of supervisory EAP referral Observe, Document, Inform, and Refer. The documentation role of supervisors has many supports in regard to accepted labor relations and personnel practice, not just in relation to EAP. If disciplinary actions are taken, documented job counseling and warnings are important to sustain grievances that the employee did not know he/she was doing anything wrong. Promotions and raises are often based upon documented job incidents and performance discussions, especially denials. Documented evidence of superior or marginal job performance is extremely useful in doing regular performance appraisal and helps prevent bias in ratings. The benefits and importance of supervisory documentation of job performance is well-established. The proposed form is an elaboration of the usual scraps of paper which become an incident file. This document would be a carbon-less three-page form which would have three areas. The upper section of the form would have an area for the supervisor to document the job performance problem (attendance, violation of company rules, interpersonal problems with co—workers, and the like) and his/her job 199 action (counsel, admonition, formal warning, reprimand, or formal discipline). The center section of the form would have a reminder that job problems are often, though not always associated with personal problems. The organizational policy on employee assistance would be reprinted, a firm statement made that it is always the employee's responsibility to maintain job performance at an acceptable level else his/her job will be in jeopardy, and reminder given that the employee must recognise and address any personal problem which may cause job performance to be unsatisfactory. Finally, in the bottom portion of the form, the EAP address and phone number is given and the employee invited to seek confidential assistance for any problem or condition which might be a factor in this instance of unsatisfactory job performance. The supervisor would fill out the job problems documentation, taking whatever degree of action with the employee seems indicated. Then, discussion would occur with the employee around the job diffulties, the object of which is to formulate a plan of remedial action and to make the employee aware that further personnel action may take place if the problems are not alleviated. From several aspects, this will improve the labor relations and performance control effectiveness of the supervisors. The immediate presence of the EAP policy and referral information will serve as an effective cue to discuss the possible benefit of EAP intervention. But whether the supervisor makes a concerted effort to make the EAP referral (affected by referral attitudes and the other factors which have been the focus of this study) or not, the referral is automatically made--by the form! The supervisor now separates the form, keeping one sheet for his/her employee files, gives a sheet to the 200 employee (with or without strong encouragement or assistance to see the EAP counselor), and, finally, sends the last sheet off to the EAP. A record now exists in the EAP that employee X in department Y is having job performance problems! Not much new has taken place up to this point. This would be ”good" employee and labor relations practice. Ideally, these transactions between supervisors and job impaired employees are already happening in organizations, but we know that they often do not. Supervisors typically do not document job problems and actions taken as they should; employees (and supervisors) often do not know the policy on employee assistance or even of the existence of the EAP; and supervisors often, as we have seen, do not even suggest (let alone encourage or actively arrange) EAP contact for employees who need and would benefit from intervention. A new outreach role for EAP could now come into play. If the EAP counselor receives two or three job action documents, but the employee involved has not yet called the EAP for assistance, it is appropriate now to "outreach“ to the employee and make an offer of help directly. This can and would be repeated every time another job problem discussion is documented and sent to the EAP. There is an increased likelihood that with job threat (and, hopefully, supervisory encouragement) "pushing" from one side and the EAP counselor "pulling“ with offers of confidential assistance and concern from the other, that employees will enter the EAP system more easily, and earlier, for assessment and treatment referral when needed. From a labor relations perspective, any job action taken, especially termination, will be on rock-solid ground when it can be shown that not only did the employee get several written documents regarding 201 job performance problems which also informed of available assistance through the EAP, but that the EAP even made several entreaties of help to the employee directly, but each time the employee refused! This proposal would strengthen, it is believed, the employer's position regarding adverse job actions taken (a costly and distressing matter in its own right), but it should be more effective in getting employees into needed assistance. The role of the supervisor is hereby strengthened and aided in the areas of directing and managing the workers, but is made less critical in effecting the linkages between impaired workers and the needed assistance. The outreach function also applies to the supervisor. As confidentiality is not an issue until a relationship is entered into by the employee in question, it is quite appropriate to initiate contact with the supervisor about the situation. The supervisor's feelings about and perspective on the situation can be attended to, and both concern for the employee and avoidance of "enabling" behaviors can be reinforced. The opportunity for information exchange generated by the Ilob Performance Discussion and Action Document becomes, in effect, a timely, personalized EAP-training occasion. Information and support offered the supervisor at this time should be highly relevant and make the contact particularly impactful. It is important to emphasize that this proposal cannot be simply grafted into an organization with anticipation of eager adoption and utilization any more than EAP policies and operations were. This recommendation is not put forth as a panacea for EAPs which are not penetrating the impaired employee population. Supervisors can ignore and 202 disuse forms as easily as policies, and soon articles would begin to appear on the serious "forms - practice gap." The role of the supervisor to make referrals using constructive confrontation must continue to exist. It needs support by means of training, full integration of the EAP with the other support staff functions (medical department, personnel and labor relations function, and training and development efforts), and a full labor—management cooperation on EAP's value to the organization and the employees. The recommendation made is to support and enhance the general managerial competencies of the supervisors, as recommended by Trice and Belasco (1968) and by Beyer and Trice (1982). General supervisory skills development and training raises the overall level and sophistication of labor relations conduct by the supervisors. EAP referral would continue to be encouraged, as one of the management tools available. The form serves as a guide and cue for the supervisor to engage in the "constructive" side of job performance discussion. Some supervisors will, undoubtedly, continue to be strong supporters and make timely referrals of troubled employees. In other cases, the employee will respond to the job performance discussion and use the information on the form to make a self-referral, as often happens already. The outreach function of the EAP counselor will simply complement the supervisory, self, union, and family referrals means, especially in those cases where the supervior, for any of the factors or reasons discussed does not make a referral effort. This study of the factors affecting supervisors' referrals has led to a fuller appreciation of the supervisor's decision to refer as a g1 203 complex, multiply determined, and often an emotionally—charged interpersonal transaction. Much remains to be learned about the supervisors' performance of this important role which impacts so directly EAP operation within and for an organization. More research is needed on the matter, and several lines of inquiry are beginning to emerge. At this time, however, this researcher concludes that enough evidence now exists to justify the assertion that assigning the supervisor such a critical role in EAP limits the performance and effectiveness of an employee assistance program. Recommendation has been made to expand the role of the EAP counselor, to include a more proactive outreach role with some employees. This could be accomplished by the introduction of new documentation procedures, supported by continuing supervisory training and development efforts, which cue and support job performance discussions and the need for EAP intervention. This research demonstrates the important need for and role of counseling psychologists in the employee assistance movement. As shown in the literature review, close examination of the motivational forces and personalities of supervisors have only begun to be investigated. There is great need for more research in this and other central areas of EAP operation, such as help-seeking behaviors of EAP clients, indicators of quality care and outcome studies of client intervention, program evaluation, needs assessment studies, methodological and measurement studies, studies of "levels of care" and cost-effective treatment, and study of relationships between job-related factors and stressors and employees' problems. Counseling psychologists possess a theoretical base and empirical approach which is greatly needed in the field of EAP, 204 especially regarding mental health, motivation, and treatment issues. Unfortunately, professional preparation programs in counseling psychology have not routinely addressed issues of intervention from a workplace perspective, nor has the significance of EAP been recognized and supported in the field. In coming years, many counseling psychologists will seek professional identity and employment in employee assistance, and the workplace will become, increasingly, a location where professional counselors will provide their services. It is hoped that much of the needed empirical work regarding troubled employees, dynamics of referral, treatment and outcome, and the like will be done by counseling psychologists. APPENDICES 205 APPENDIX A THE SURVEY INSTRUMENT MICHIGAN STATE UNIVERSITY COLlIcGI'. UI EDUCATION ' I)I.I’AIITM}.N‘I ()I COUNSELING. EAST LANSING - MICHIGAN - {flu-10M EDUCATIONAL PSYCHOLUIJ' AND .‘PHJAL H)l|(.ATION August 8, 1989 Dear Supervisor, There is a great need for better understanding of how supervisors perceive, use, and feel about Employee Assistance Programs (EAPs). Supervisors have an important role with the EAP, yet little is heard from them about how they feel about that role. This survey, which is part of my doctoral dissertation research at Michigan State Universny, will help us better understand the attitudes and beliels of supervisors regarding EAPs and how they make referrals to them. By filling out this confidential questionnaire, you will be making 3 Significant contribution to research in employee assistance programs. Your organization is supportive of this effort and has permitted me to request your help. Your participation is entirely voluntary, however. You are free to choose not to participate by simply returning the survey forms uncompleted. Also, you are free to leave unanswered any question which you consider too personal. You indicate your voluntary agreement to participate by completing and returning the survey materials. Your survey consists of a page with four hypothetical employee situations and three machine- scored answer sheets with questions on both sides. Your completed questionnaire will be confidential and seen only bym e. identify yourself on the sheets, and they will remain anonymous as welln mThere is a code number in the lower left corner of each answer sheet; this is so I can keep the three sheets matched up for analysis. I kno ow how often supervisors are interrupted, so 1 designed the survey so it could be done in ”bits and pieces" during the day. Estimated time to complete the survey is 15 minutes for Part 1 and about 10 minutes each for Parts 2 and 3. Mark all your responses on the special answer sheets provided using a soft-lead pencil. Return your completed survey to the contact person in your location who gave it to you. It will be placed into a large manila envelope to be mailed back to me. I appreciate your help in this research very much, and I would be glad to send you a summary of the final results if you want to know about the findings. 1 would be pleased to send a summary to all who are interested. Just send me a postcard indicating your interest, to: EAP Research Project 2633 Edgebrook Drive Lansing, MI 48906 Dr. Gloria S. Smith is my major professor and is supervising my research. She can be reached at MSU (517) 355-8502 if you have any questions or concerns. I can be called at (517) 321-2860 in the evenings. Again, thanks for giving your time and opinions to the research. David A. Harley Ann 1 .. All . r 206 (c) I989 by David A. Harley SUPERVISORY SURVEY Experience tells us that supervisors typically use some combination of three approaches in dealing with employees' job performance problems -- "handling it myself," getting the employee assistance program (EAP) involved, and/or using some form of discipline, from warnings through firing. Read the four situations below which might confront a supervisor. As you read each one, imagine yourself as the supervisor. Then use the machine-scored answer sheet (Part I-Side I) to indicate how you would handle them. Use the following Key to indicate your position on Questions I — 8. A- Strongly B- Agree C- Neither agree D= Disagree En Strongly Agree nor disagree Disagree A)udrey is a clerk in your department. For the last two months, Audrey has shown a pattern of being late to work (typically l5-30 minutes). You have talked to her about being tardy, but she said, "she has problems in the mornings" which slow her down. She divorced her husband maybe a year ago. Recently, she mentioned that she had begun to date him again, but soon after that she said, "it looks like we still have all the same problems -- arguing about everything." During these two months her work output has been less, but there haven't been any major mistakes. Now, answer questions I and 2. B)ill is a long term employee. Nearly everybody knows that Bill has had drinking problems and went through two treatment programs before you became his supervisor. There were rumors that he Started drinking again, but you haven't seen anything specific. Monday, Bill called in about noon, saying he was sick and wouldn't be coming in. Tuesday, same thing. Wednesday, he missed, but didn't call in at all. Thursday, Bill came in, on time, but in a bad mood and looking sick. Answer questions 3 and 4 now. C)arol has wasted most of the morning, and a large report has to get out. go to Carol's desk to ask her if she understands what she is supposed to be doing. She immediately starts crying. You take her into your office, where she continues to cry. Finally, Carol tells you that she has been going out with a married man from the office. Now she thinks she is pregnant. She says she is really confused and doesn't know what to do. Answer 5 and 6 now. You D)on‘s wife creates quite a problem for him and you. According to him, she has "anxiety attacks," so she calls him at the office for his support and reassurances. She calls him about twice a week, and they talk about 30 minutes. Don sometimes asks to leave work early, "so he can go home and get her calmed down." You have instructed Don to tell his wife not to call him at work. He says that he did, but she still calls. Answer 7 and 8 now. Consider the EAP-related actions in 9-l5 on the answer sheet. Then select the action you would most likely take with each of the four workers above. I the column of A's, indicate the one you would do with A)udrey. Then, mark the one you would most likely do with B)i11 in the 8's column, the one for C)arol in column C, and one for D)on in D. (Choose one for each worker —- four marks in all.) Continue with the remaining questions -- using the three machine—scored answer sheets, that have questions on both sides. 207 I I I — one PASS WILL SCAN son-c snoes SIDE 1 v N ~ _ ,_ ____lhxtu{ pytvieuln1lng_vvg_lrtmlntse‘ _H"E§_§ _____ F? |®®©©® l. ln Audrey's situation, 1 would most likely "handle it myself." : 5A A u n so 0, g C \0 I! z®®©®© 2. I wmld use some form of progressive discipline ln dealing with Audrey. . . . g 3 -< o . 5' a E 5 E". 'n a, ct m a®®©©® 3. 1n Bill's situation, 1 mm most likely "handle it myself." i g g g P o a , 2 E i 2 N e g . " g 1 z :1 E 4®®©®® a. I would use some form of progressive discipline in dealing with Bill. 3 “g g g to .. ' c 9 3 2 g I®®©®® 5. ln Carol's situation, 1 would most likely "handle it myself." g 2 l s®®©©© 6. 1 would use some form of progressive discipline in dealing with Carol. 2 1®®©®© 7. In Don's situation, I would most likely "handle it myself." 3 . < E u®®©®© 8. l wOuld use some form or progressive discipline in dealing with Don. E a®.©©® 9. l wOuld not get the Employee Assistance Program (EAP) involved in the § situation at this time. E U7 m®©©©® lo. I would casually mention the Employee Assistance Program (EAP) to the g worker and suggest he or she make an appointment. g a 1|®.©®® ll. 1 would call the EAP for a consultation to dismss the worker's job '< performance and whether a referral is indicated. 5"? 12®.©©© 12. DlSCuss the job problems with the employee more extensively, and very ’— strongly encourage the employee to go to the EAP let some help. 13®.©®© 13. DlsCuss Job performance with the worker and actively help enployee get to EAP, such as by offering to help set up an appointment or time off to go, M®.©®® la. Offer the enployee a choice between taking disciplinary action or going to the EAP for an evaluation interview and/or help. Is®.©®® 15. Make a mandatory supervisory EAP referral -- go to EAP or else! fl®®©©® 16. How realistic do you think these four situations are? Select the statement below that most closely matches your experience as a supervisor. A: Situations very realistic, similar to those 1 fremently encounter. |7®.©@® <- - B: Sltuatlons are realistic, similar to those I sometimes encounter. --------------- ‘ C: Situations somemat realistic, similar to those I mlgat encounter ‘ D: Situations somewhat unrealistic, unlikely I would enccunter. Il®®©®© ' E: Situations very unrealistic, very unllkely I would ever encounter. I ' 17. During the time you have been a supervisor, haw many enployees have ll®®©®© you referred to the EAP (emloyee assistance program)? A: none 8: l C: 2 or J or 5 E: more than > In Special Codes area, colums a a C enter vaur A e ---> ---> zoooooo ' ——L sunvev O NUMBER SURVEY B [OVER] NATIONAL ~- Form No. 19637 gem}? MS M23 Tmhuc‘ Wim'b I7 nééooo ever? n®0©©® n®®©®© u®®©®© u®®©©© u®®©®© u®®©®® n®®©®® u®0©®© n®®©®© n®0©©® n®®©©© n®®©®® "99995-959. u®®©®® u®®©®© n®®©®® n®®©®® n®®©®© u®®©®® I 4 KEEP ovenmmrwo wrrwm mes: unis i Think of three specific employees -— the worst and best enployees you have ever supervised, and the one most difficult to supervise during the past year. ihat employee would be the one who had the most problems on his/her job and caused you the most problems and concern. To help you keep them in mind, jot do-n their initials. Ihen, please rate the follOwing aspects 0' their Job performance: HORST: E : .nB PRIBLEMS: in}; a; 21-3; ' t3 acclaim ' 3; .;.;.~,.;. ’ c; céw' [a Fa; ’ E; p.10; ;l.- Rate-the worst employee on QZality-of-Hork: - 22. Rate the Horst employee on mantity of Work. 23. Rate the worst employee on Attendance. 2A. Rate the Best employee on mality of work. 25. Rate the Best enployee on Quantity of work. 26. Rate the Best employee on Attendance. 27. Rate the Erroldyee with Job Problems on mality of-Hork.- 28. Rate the BrplOyee with Job Problems on Quantity of work. 29. Rate the Ewloyee with Job Problems on Attendance. )0. Rate the Enployee with Job Problems on Knowledge of the Job. )1. Rate the Employee with Job Problems on Dependability. 32. Rate the Emloyee with Job Problems on initiative. 3}. Rate the Employee with Job Problans on Ability to work without Close Supervision. 34. Rate the Employee with Job Problems on Cooperation. Look again at the EAP-related options (9-15) on the reverse side of this sheet. Select the g action that best describes what you did with the Enployee who had the most Job problems last year. Mark that action (from 9-15) in column E. )5. At what level within the organization is that employee you identified as having the most job problems or difficulties in the past year? A=Production, lOwest level c: 3 Supervisor D:1echnical, middle-level 8=Lead worker, group leader or Foreman E=Managerial, top-level 16. what is the gender of the mployee? A: Male 8: female 37. Has the enployee done anything dramatic or "outrageous" which would bring attention, such as coming to work drunk, having a crying spell at work, etc? A=Yes, and many people heard amut it C:Neither yes or D=N0-don't know B=Ves, but few people learned about it no - uncertain of anything )8. Now, a few questions ab0ut you- what is your gender? A=Male a=remaie . 39. what is your level of formal emcation? A=Some Hig‘rSchooi B=HighSchool grad C=Sane college/Assoc. Degree D: B.A.degree 5=advanced/graduate degree 140. How many hours of training on EAP-related topics have you received? 8 =2-lo D=>-8 EsmorethanB SIDE 2 A: non NATIONAL 9‘ 0. "as. raw Cum sum, m I I - AM "ya. mm mm m u“. gegr‘tggfi I I, 209 I I I - ONE PASS WILL SCAN BOTH SIDES sun: 1 y N _i__xe_rp_ QVEE‘PfllNI‘Ni wu_mm_ mes: u__wr_s i _ A |®®©©© l. i know some workers who were referred to the EAP by their - - - - - - - 9. SA A N D SO supervisors and (in general) how it turned wt for them ' As, 8:, etc ' :5 - v is v 2 8 z®®©®© 2. Referring a problem enployee to the EAP helps me iOwer ' printed ' g .E’ ‘ my own personal stress level. ' at 801104 ' r: :‘ ‘ < 3 1 ' or PAGE - E :—.' E E 5. m 3®©©©® 3. I believe in "tougn love" for people, and that means - — - - - - - - g g e. g c): 0 getting the EAP involved. a § § ‘2' ' E i n L. o g '< O - g 4®®©©© la. Sending someone to the EAP would probably mean the end of a friendship. '2. g (2 ET in 2 = I a 3 g 5®®©®® 5. I wmld have a diffiCUlt time dealing with the emotions of workers men I g :1 refer them to the EAP in O 2 §@.©@@ 6. i am too busy to do the kind of documentation necessary to properly send I smeone to the EAP , (n 7Q®©©© 7. The EAP helps me run an efficient shop. g 2 z |®®©©@ 8. In subtle ways, the organization discourages supervisors from using E the EAP m g®©©®@ 9. If i have worked with someone for a long time, it would be very difficult § to refer that person to the EAP, even if the problems were obvious. 2 G m®®©©© 10. when employees have something bothering them, i can usually get them to S talk to me about it 2;; n®.©®@ ii. If an enployee told me to "mind your own business" when I tried to convince 5 him or her to see the EAP Counselor, i guess I wmld deserve it. 3 ., 12®®©©® 12. if I want to do my own job well, I trust refer enployees with job problems a" to the EAP E O 0 n®®©©© l}. A good reason to refer problem enployees to the EAP is 1 Just don't have '— the time to watch over them. u®®©©© in. It's very hard to refer a worker to the EAP if a previous supervisor did nothing about the prwlem. 15®®©©® l5. it is best to refer enployees only after their co-workers have couplained about them or raised questions about their behavior. 15®®©®® 16. Our EAP has a law rate of success. n®©©®© l7. 1 would hesitate to refer anyone to our EAP because i don't like where it is located II®®©®® la. the procedures for making an EAP referral are clear to me. SURVEY NUMBER in Special Codes sac, enter the Nurber of Errployees Supervised by vou -> KEV: indicate your position on these statements and those in Part 3 using. AGE A: smouv c: EITHER AGREE o: DISA me: e. smmv won msnmrr more SURVEY B (oven) Form No. 19637 NATIONAL COMPUTER . J SYSTEMS flO we; haw ”1.3L '7 uééooo SAANDSD n®®©®© n®®©®® u®®©®® u®0©®© u®©©®® n®®©©® n®®©®® u®©©®© n®0©®® n®®©®® n®®©®® n®0©©© u®®©©® u®0©©® u®®©®® n®®©®® u®O©©® MG®©®® w®®©©® I 40. Our EAP does not keep personal matters totally confidential. i KEEP ovenmiurwd WITHIN THESE units l key: A: SYRCNIY B: ME (I: t£iTr£R AEREE D: DISAGQEE E: STRENIY AGREE MR DISAGREE DISAGREE 21. l knOw who to contact if i want to refer an employee to the EAP. 22. EAP referrals show my determination to deal responsibly with the problem emplOyees l enCOunter. 2}. Most of life's problems will clear up if you Just give them some time. 214. My department 5 level of productivity will be Jeopardlzed by having soane ne gone to the EAP or to treatment 25. it wculd hurt an emlOyee's future career in the conpahy if 1 referred him or her to the assistance program. 26. Being a good supervisor means 1 refer problem enployees to the EAP. 27. Unless I hear from several workers that someone is having problems, 1 would probably hold off on suggesting the EAP 28. A supervisor should refer to the EAP instead of trying to handle a worker's problems on hisIher own. 29. Referring someone to the EAP is a form of "tattling." 50. the EAP helps me manage my workers and get the most out of them. )1. lhe best way 1 can help an enployee keep his (or her) Job is to refer to the EAP men problems begin to shOw up at work. 32. It will be a lot less stress on me in the long run if 1 refer an employee with a problem directly to the EAP. 33. when an enployee starts making problems for my other workers, it's best to refer that person to the EAP before things get out of hand 30. If a worker is having "a bad day," I will not add to the problems by making him or her talk abbot the poor work performance. . Confronting employees just invites them to point out your own shortcomings. V V‘ \II 6. It will help the productivity of the department if 1 refer workers with problems to the EAP . 37. My own supervisor expects me to refer to the EAP when it is indicated. 38. Caunseling and other forms of treatment for the various types of personal problems are usually quite effective. 39. i get quite uncomfortable mehever an emloyee begins to talk to me about things in their personal life. ‘. “IE. Manon-l Carnal-v Syn-ma. he. All raw-Ia ru- nat Pun“ II USA NATIONAL SIDE 2 COMPUTER ® 5' STEM SIDE 1 @< @z © @ 9 > 9 ® @ 2 © 0 ® 86 9 ® @ © ® 4®©©©© s®®©®© s®®©®® ‘ Mouoamo SIHl NI 0513:: - 1®0©®© r®0©®® r®®©©© w®®©©® H®O©®© u®0©®© HGO©@© u®0©®© n®0©®© n®o©®® n®0©©© n®0©®© SURVEY NUMBER ll. - ONE PASS WILL SCAN BOTH SIDES _ __ .__ t_ kite oveapamrmo wlrmw rwsss ones i l. I would lose most of my say about that happe‘ns' — ""‘ ------- - to an enplOyee after he/she goes to the EAP. ' A=, 8:, etc ' ' KEY is ' 2. EAP referral feels rimt to me because it is ' printed at ' “the responsible thing to do." ' BO" 3. i don't have confidence in the ability of our EAP to -------- deal with the really hard prcblems employees may have. a. I would not refer an enoloyee to the EAP if I knew of others (especially supervisors) who had similar problems but were not referred. 5. when someone is missing a lot of work, EAP referral is one of the best ways of getting the person back to good attendance. 6. What people with problems need from me is firmness and strong insistence that they go for help and make any needed changes. 7. My workers do creative or professional Jobs which are really hard to evaluate, so there is no way I can tell when they're having a problem. 8. Because our EAP counselor lacks professional competence, I am very hesitant to refer anyone to him or her. 9. I don't want to "prolong the agony," so I believe in quick EAP referrals so we can find out which employees can be saved and which can't. lo. There is a good chance a worker would punch me or try to get even in some way if I refer him or her to the EAP. when a worker does well on some days but not on others, you really can't tell if they need the EAP 12. A record of frequent absences or an excessive use of sick leave makes it clear to me that a worker needs to be referred to the EAP. i}. If one of my workers went to the EAP for help, I believe much inpmvement, if not conplete success, would usually be the outcome. 114. If I refer an erroloyee to the EAP, it would likely just turn into a grievance. 15. Some of the supervisors here think the EAP is more troutfle than it is worth. l6. It would damage a person's reputation or image if I refer him/her to the EAP. l7. I feel it is my role to try to talk to and help an employee who is worried or upset about something. 18. Ihe EAP counselors are respected and supported by management. In Special Codes, arc, enter Your Years of Experience as a Stpervisor -> KEV: Indicate your position on these statements and those in Part 2 using: A'ueop nu; o Apralduios salary: mg a A: STmV B: AGREE C: EITFER ME D: DIME E: STRmY AMEE KR DISABREE DISACREE SURVEY [mmm NMwNM COMPU1€R Forln NO. 19637 SYSTEMS EWVN A anuns I 9DIS'{ 119d ABAUHS lit-DSIAUECDS SNOIanHlSNI DNIMUVW ABAUnS w lapel-I "V prnea it: 6961 (3) "sires u®©©©® n®®©©® u®©©®® u®0©®© u®0©©® u®®©®© u®0©©© n®0©®© n®°©©® u®®©©© n®0©®© n®0©®© u®®©®© u®®©®® u®0©©® n®®©®© u®®©®® n®®©©© w®®©®® I ”I ”ll-WW WINT- I7 Key: A. srmv mm: i KEEP OVERPRINTINO WITHIN TNESE LNES ‘ B: MREE E: STRMY C: ENTER AmEE D- DISAMEE E DISAGREE DI SAW SIDE 2 21. Most of the problems emplOyees have do not really need professional help. 22. I have conSuIted with the EAP counselors for ideas on how to deal with one or more emplOyees who were having job problems. 23. A main reason I would refer someone to the EAP is because it is part of my official role as a supervisor in the organization. 24. Irying to get someone to go see the EAP counselor would get me too involved in the employee's personal business. 25. I seriously consider an EAP referral when there is a dramatic drop in the job performance of a worker who had a good record previously. 26. Most people are not really helped very much by counseling or "therapy" programs like the EAP 27. It's hard to tell if problems are serious enough that an EAP referral is the right rather than the wrong decision. 28. Ihe people who work under me are intelligent enmgh to knew men to get personal help without any intrusion or assistance from me. 29. The procedures for referring someone to the EAP are too involved and complicated. 30. If I referred someone to the EAP, I would expect co-workers to "take sides" with him or her on the matter. 31. I feel I can help most people with their problems, since I have worked out some tough problems of my own. 32. Prowt referral to the EAP is typically better for an employee with a problem than trying to convince him or her to change. 33. It would probably embarrass us both if I confront a worker and refer him/her to the EAP. 34. It would add to the problems we already have for one or more employees to be gone each week for caunseling appointments at the EAP. 35. I have definite standards for work in my area, and people simply must meet those standards. 36. I believe nearly everyone in my organization, top to bottom, thinks the EAP is a good idea. 37. I could make problems at home for a worker if I told him or her to go see the EAP counselor. 38. Ihe union expects me to refer all problem employees to the EAP. 39. EAP referral is a good way to show the employees that I care about them. 40. which statement below best describes the EAP at your location? A: It has a "health a wellness" fOCus, but handles supervisory referrals. 8: Our EAP has no special focus; it tries to handle all problems equally. C: It deals with all problems, but emphasizes alcohol/drug problems. 0: Our EAP has a definite emphasis on alcohol and drug problems, but people with other kinds of problems can go there or be referred there. E= Our Efifi_qolx_deals_witb_9t99_anq_alc0h0l pr09l9m5- 0, ms. Mal-anal Computer int-ma. he. gSHgETAé-R I I - “nuhlaruuvnd Mind-\USA. SVS‘IGMS @. ‘\ APPENDIX B ITEMS, SCALES, AND SUBSCALES (N = 440) APPROACH Attitudes toward Referral SCALE - 33 items in seven subscales Mean: 74.1 Std Dev: 12.6 coefficient alpha: .86 Subscale CONEMPL (Concern for Employees - 4 items) Mean: 13.6 Std Dev: 2.1 coefficient alpha: .62 Q2.31. The best way I can help an employee keep his (or her) job is to refer to the EAP when problems begin to show up at work. No Ans l/SD Z/Disagree 3/Neither 4/Agree 5/SA Mean Sthev 4.8% 1.4% 22.3% 34.3% 31.4% 5.9% 3.19 .91 Correlation with corrected subscale=.426 Correlation with corrected Scaie=.537 Q3.2. EAP referral feels right to me because it is "the responsible thing to do." No Ans i/SD 2/Disagree 3/Neither 4/Agree 5/SA Mean Sthev 4.1% 0% 1.8% 33.9% 50.5% 9.8% 3.71 .67 Correlation with corrected subscale=.467 Correlation with corrected Scale=.600 Q3.32. Prompt referral to the EAP is typically better for an employee with a problem than trying to convince him or her to change. No Ans l/SD Z/Disagree 3/Neither 4/Agree 5/SA Mean Sthev 5.9% .596 20.9% 40.9% 28.4% 3.4% 3.14 .82 Correlation with corrected subscale:.330 Correlation with corrected Scale=.355 Q3.39. EAP referral is a good way to show the employees that I care about them. No Ans l/SD 2/Disagree 3/Neither 4/Agree 5/SA Mean Sthev 6.6% 0% 4.8% 34.1% 47.3% 7.3% 3.61 .70 Correlation with corrected subscale=.417 Correlation with corrected Scale:.526 214 Subscale CONPROD (Concern for Productivity - 5 items) Mean: 16.3 Std Dev: 2.7 coefficient alpha: .65 Q2.7. The EAP helps me run an efficient shop. No Ans l/SD 2/Disagree 3/Neither 4/Agree 5/SA Mean Sthev 3.6% _3.9%_ 9.5% 48.4% 28.4% 6.1% 3.24 .87 Correlation With corrected subscale=.434 Correlation with corrected Scale=.537 Q2.30. The EAP helps me manage my workers and get the most out of them. No Ans l/SD 2/Disagree 3/Neither 4/Agree 5/SA Mean Sthev 5.2% 4.3% 13.9% 42.7% 28.9% 5.0% 3.17 .90 Correlation with corrected subscale:.494 Correlation with corrected Scale=.508 Q2.36. It will help the productivity of the department if I refer workers with problems to the EAP. No Ans l/SD 2/Disagree 3/Neither 4/Agree 5/SA Mean Sthev 4.5% 1.4% 8.0% 39.8% 40.9% 5.5% 3.43 .78 Correlation with corrected subscale=.450 Correlation with corrected Scale=.565 Q3.5. When someone is missing a lot of work, EAP referral is one of the best ways of getting the person back to good attendance. No Ans l/SD 2/Disagree 3/Neither 4/Agree 5/SA Mean Sthev 3.6% 1.6% 21.8% .8% 18.2% 3.0% 2.99 .78 Correlation with corrected subscale=.354 Correlation with corrected Scale=.45l Q3.25. I seriously consider an EAP referral when there is a dramatic drop in the job performance of a worker who had a good record previously. No Ans l/SD Z/Disagree 3/Neither 4/Agree 5/SA Mean Sthev 5.7% l.l% 12.3% 30.9% 43.0% 7.0% 3.45 .85 Correlation with corrected subscale=.270 Correlation with corrected Scale=.365 Subscale DISCMFRT (Discomfort Relief for Sup - 4 items) Mean: 11.1 Std Dev:2.2 coefficient alpha: .47 Q2.2. Referring a problem employee to the EAP helps me lower my own personal stress level. No Ans l/SD 2/Disagree 3/Neither 4/Agree 5/SA Mean Sthev 4.1% 2.0% 10.9% 5. 38.4% 9.1% 3.43 .89 Correlation with corrected subscale=.218 Correlation with corrected Scale=.377 215 Q2.l3. A good reason to refer problem employees to the EAP is I just don‘t have the time to watch over them. No Ans l/SD 2/Disagree 3/Neither 4/Agree 5/SA Mean Sthev 2.5% .17.7% 55.0% 15.0% 7.5% 2.3% 2.20% .90 Correlation With corrected subscale=.238 Correlation with corrected Scale:.048 Q2.32. It will be a lot less stress on me in the long run if I refer an employee with a problem directly to the EAP. No Ans l/SD Z/Disagree 3/Neither 4/Agree 5/SA Mean Sthev 4.5% 3.4% 23.0% 34.8% 31.1% 3.2% 3.08 .91 Correlation with corrected subscale=.357 Correlation with corrected Scale:.479 Q3.9. I don't want to "prolong the agony," so I believe in quick EAP referrals so we can find out which employees can be saved and which can't. No Ans l/SD 2/Disagree 3/Neither 4/Agree 5/SA Mean Sthev 3.6% 8.2% 51.4% 29.5% 6.6% 0.7% 2.38 .76 Correlation with corrected subscale=.269 Correlation with corrected Scale:.216 Subscale INTOLDEV (Intolerance of Deviance - 5 items) Mean: 16.5 Std Dev: 2.5 coefficient alpha: .51 Q2.3. I believe in "tough love" for people, and that means getting the EAP involved. No Ans i/SD Z/Disagree 3/Neither 4/Agree 5/SA Mean Sthev 4.8% 0.2% 3 40.5% 34.3% 10.9% 3.49 .83 Correlation with corrected subscale=.294 Correlation with corrected Scale:.352 Q2.33. When an employee starts making problems for my other workers, it's best to refer that person to the EAP before things get out of hand. No Ans 1/SD 2/Disagree 3/Neither 4/Agree 5/SA Mean Sthev 4.5% 0.9% 20.5% 35.0% 30.7% 8.4% 3.26 .93 Correlation with corrected subscale=.266 Correlation with corrected Scale:.422 Q3.6. What people with problems need from me is firmness and strong insistence that they go for help and make any needed changes. No Ans l/SD Z/Disagree 3/Neither 4/Agree 5/SA Mean Sthev 3.6% 1.1% 13.6% 32 41.4% 7.5% 3.42 .87 Correlation with corrected subscale=.3l4 Correlation with corrected Scale:.335 216 Q3.12. A record of frequent absences or an excessive use of sick leave makes it clear to me that a worker needs to be referred to the EAP. No Ans 1/SD 2/Disagree 3/Neither 4/Agree 5/SA Mean Sthev 3.9% .1.1% 34.5% 33.6% 23.9% 3.0% 2.93 .88 Correlation with corrected subscale=.323 Correlation with corrected Scale:.418 Q3.35. I have definite standards for work in my area, and people simply must meet those standards. No Ans l/SD 2/Disagree 3/Neither 4/Agree 5/SA Mean Sthev 5.7% 0.2% 14.1% 24.3% 51.4% 4.3% 3.48 .81 Correlation with corrected subscale=.192 Correlation with corrected Scale:.203 Subscale ACCREFRL (Accept the Referral Role - 5 items) Mean: 15.9 Std Dev: 2.7 coefficient alpha: .50 Q2.12. If 1 want to do my own job well, I must refer employees with job problems to the EAP. No Ans l/SD Z/Disagree 3/Neither 4/Agree 5/SA Mean Sthev 2.5% 1.4% 18.6% 33.6% 33.6% 10.2% 3.34 .95 Correlation with corrected subscale=.375 Correlation with corrected Scale:.496 Q2.22. EAP referrals show my determination to deal responsibly with the problem employees I encounter. No Ans 1/SD 2/Disagree 3/Neither 4/Agree 5/SA Mean Sthev 4.5% 0.9% 6.8% 31.8% 44.1% 11.8% 3.62 .83 Correlation with corrected subscale=.296 Correlation with corrected Scale:.504 Q2.26. Being a good supervisor means 1 refer problem employees to the EAP. No Ans l/SD 2/Disagree 3/Neither 4/Agree 5/SA Mean Sthev 4.1% 4.5% 19.5% 36.6% 29.1% 6.1% 3.13 .97 Correlation with corrected subscale=.389 Correlation with corrected Scale:.398 Q3.23. A main reason I would refer someone to the EAP is because it is part of my official role as a supervisor in the organization. No Ans 1/SD 2/Disagree 3/Neither 4/Agree 5/SA Mean Sthev 5.7% 4.3% 28.2% 18.4% 38.0% 5.5% 3.13 1.05 Correlation with corrected subscale=.168 Correlation with corrected Scale:.202 Q3.38. The union expects me to refer all problem employees to the EAP. No Ans l/SD 2/Disagree 3/Neither 4/Agree 5/SA Mean Sthev 9.3% 5.2% 30.0% 43.6% 10.0% 1.8% 2.70 .82 Correlation with corrected subscale=.169 Correlation with corrected Scale:.l75 Subscale ORGSUPRT (Organization Supports EAP - 5 items) Mean: 17.6 Std Dev: 2.8 coefficient alpha: .67 Q2.8. In subtle ways, the organization discourages supervisors from using the EAP. No Ans 5/SD 4/Disagree 3/Neither 2/Agree l/SA Mean Sthev 3.0% 25.5% 42.7% 19.1% 8.0% 1.8% 2.15 .97 Correlation with corrected subscale=.457 Correlation with corrected Scale:.201 Q2.37. My own supervisor expects me to refer to the EAP when it is indicated. No Ans 1/SD Z/Disagree 3/Neither 4/Agree 5/SA Mean Sthev 4.8% 0.7% 8.9% 30.2% 48.2% 7.3% 3.55 .79 Correlation with corrected subscale=.322 Correlation with corrected Scale:.435 Q3.15. Some of the supervisors here think the EAP is more trouble than it is worth. No Ans 5/SD 4/Disagree 3/Neither 2/Agree 1/SA Mean Sthev 4.5% 8.0% 29.8% 40.0% 15.5% 2.3% 2.73 .91 Correlation with corrected subscale=.405 Correlation with corrected Scale:.l34 Q3.18. The EAP counselors are respected and supported by management. No Ans l/SD 2/Disagree 3/Neither 4/Agree 5/SA Mean Sthev 5.5% 0.2% 1.4% 36.4% 45.7% 10.9% 3.69 .70 Correlation with corrected subscale=.535 Correlation with corrected Scale:.39l Q3.36. I believe nearly everyone in my organization, top to bottom, thinks the EAP is a good idea. No Ans l/SD 2/Disagree 3/Neither 4/Agree 5/SA Mean Sthev 5.2% 0.5% 14.3% 40.2% 35.0% 4.8% 3.31 .80 Correlation with corrected subscale=.448 Correlation With corrected Scale:.353 Subscale POSITEXP (Positive Expectancies of EAP — 5 items) Mean: 17.5 Std Dev: 2.5 coefficient alpha: .73 Q2.l6. Our EAP has a low rate of success. No Ans 5/SD 4/Disagree 3/Neither Z/Agree 1/SA Mean Sthev 4.5% 4.1% 21.4% 63.9% 5.5% 0.7% 2.76 .65 Correlation with corrected subscale=.481 Correlation with corrected Scale:.308 Q2.38. Counseling and other forms of treatment for the various types of personal problems are usually quite effective. No Ans i/SD Z/Disagree 3/Neither 4/Agree 5/SA Mean Sthev 5.5% 0.2% 3.0% 33.9% 49.8% 7.7% 3.65 .68 Correlation with corrected subscale=.423 Correlation with corrected Scale:.336 Q3.3. I don't have confidence in the ability of our EAP to deal with the really hard problems employees may have. No Ans 5/SD 4/Disagree 3/Neither Z/Agree 1/SA Mean Sthev 4.3% 9.8% 48.6% 28.9% 7.7% 0.7% 2.38 .80 Correlation with corrected subscale=.558 Correlation with corrected Scale:.407 Q3.13. If one of my workers went to the EAP for help, I believe much improvement, if not complete success, would usually be the outcome. No Ans 1/SD 2/Disagree 3/Neither 4/Agree 5/SA Mean Sthev 4.5% 0.7% 9.3% 48.9% 33.9% 2.7% 3.30 .71 Correlation with corrected subscale=.467 Correlation with corrected Scale:.490 Q3.26. Most people are not really helped very much by counseling or "therapy" programs like the EAP. No Ans 5/SD 4/Disagree 3/Neither 2/Agree l/SA Mean Sthev 5.0% 9.8% 50.5% 31.8% 2.5% 0.5% 2.30 .70 Correlation with corrected subscale=.549 Correlation with corrected Scale:.364 219 AVOIDANCE Attitudes toward Referral Scale - 33 items in five subscales Mean: 74.1 Std Dev: 12.6 coefficient alpha: .90 Subscale DENIAL (Denial and Non-involvement - 9 items) Mean: 20.6 Std Dev: 3.7 coefficient alpha: .66 Q2.9. 'If I have worked with someone for a long time, it would be very difficult to refer that person to the EAP, even if the problems were obvious. No Ans 1/SD 2/Disagree 3/Neither 4/Agree 5/SA Mean Sthev 2.5% 26.1% 53.2% 7.5% 9.1% 1.6% 2.04 .93 Correlation with corrected subscale=.392 Correlation with corrected Scale:.5l4 Q2.ll. If an employee told me to "mind your own business" when I tried to convince him or her to see the EAP counselor, I guess I would deserve it. No Ans l/SD 2/Disagree 3/Neither 4/Agree 5/SA Mean Sthev 2.7% 19.5% 57.5% 15.0% 4.3% 0.9% 2.07 .78 Correlation with corrected subscale=.374 Correlation with corrected Scale:.460 Q2.23. Most of life's problems will clear up if you just give them some time. No Ans l/SD 2/Disagree 3/Neither 4/Agree 5/SA Mean Sthev 4.1% 14.8% 50.9% 16.8% 11.6% 1.8% 2.32 .94 Correlation with corrected subscale=.316 Correlation with corrected Scale:.3l6 Q2.29. Referring someone to the EAP is a form of "tattling." No Ans 1/SD 2/Disagree 3/Neither 4/Agree 5/SA Mean Sthev 4.5% 32.3% 58.2% 3.4% 1.6% 0% 1.73 .61 Correlation with corrected subscale=.45l Correlation with corrected Scale:.553 Q2.34. If a worker is having "a bad day," I will not add to the problems by making him or her talk about the poor work performance. No Ans l/SD 2/Disagree 3/Neither 4/Agree 5/SA Mean Sthev 4.3% 6.8% 48.9% 21.1% 18.2% 0.7% 2.55 .90 Correlation with corrected subscale=.269 Correlation with corrected Scale:.278 Q3.4. I would not refer an employee to the EAP if I knew of others (especially supervisors) who had similar problems but were not referred. No Ans 1/SD 2/Disagree 3/Neither 4/Agree 5/SA Mean Sthev 3.4% 15.0% 69.5% 8.9% 3.0% ~ 0.2% 2.00 .62 Correlation with corrected subscale=.511 Correlation With corrected Scale:.581 Q3.21£1 Most of the problems employees have do not really need professional elp. No Ans 1/SD 2/Disagree 3/Neither 4/Agree 5/SA Mean Sthev 4.8% 1.6% 14.3% 30.9% 40.7% 7.7% 3.41 .90 Correlation with corrected subscale=.140 Correlation with corrected Scale:.174 Q3.24. Trying to get someone to go see the EAP counselor would get me too involved in the employee's personal business. No Ans l/SD 2/Disagree 3/Neither 4/Agree 5/SA Mean Sthev 5.0% 12.3% 69.8% 11.1% . 0.2% 2.03 .58 Correlation with corrected subscale=.401 Correlation with corrected Scale:.521 Q3.28. The people who work under me are intelligent enough to know when to get personal help without any intrusion or assistance from me. No Ans i/SD 2/Disagree 3/Neither 4/Agree 5/SA Mean Sthev ' 31.6% 7. 4.5% 6.1% 50.0% 5% 0.2% 2.43 .94 Correlation with corrected subscale=.316 Correlatior. with corrected Scale:.39l Subscale NEGOUTS (Negative Outcomes Expected - 10 items) Mean: 21.3 Std Dev: 4.2 coefficient alpha: .78 Q2.4. Sending someone to the EAP would probably mean the end of a friendship. No Ans l/SD 2/Disagree 3/Neither 4/Agree 5/SA Mean Sthev 2.7% 20.0% 55.9% 16.8% 4.3% 0.2% 2.06 .76 Correlation with corrected subscale=.511 Correlation with corrected Scale:.496 Q2.5. 1 would have a difficult time dealing with the emotions of workers when I refer them to the EAP. No Ans l/SD 2/Disagree 3/Neither 4/Agree 5/SA Mean Sthev 2.5% 16.8% 58.0% 14.8% 7.0% 0.9% 2.15 .82 Correlation with corrected subscale=.498 Correlation with corrected Scale:.497 Q2.25. It would hurt an employee's future career in the company if I referred him or her to the assistance program. No Ans l/SD 2/Disagree 3/Neither 4/Agree 5/SA Mean Sthev 4.3% 20.7% 54.8% 14.5% 5.0% 0.7% 2.06 .80 Correlation with corrected subscale=.491 Correlation with corrected Scale:.516 221 Q2.35. Confronting employees just invites them to point out your own shortcomings. No Ans l/SD 2/Disagree 3/Neither 4/Agree 5/SA Mean Sthev 4.5% 17.7% 60.2% 11.4% 5.7% 0.5% 2.07 .76 Correlation with corrected subscale=.359 Correlation with corrected Scale:.387 Q3.10. There is a good chance a worker would punch me or try to get even in some way if 1 refer him or her to the EAP. No Ans l/SD 2/Disagree 3/Neither 4/Agree 5/SA Mean Sthev 9.8% 3.6% 25.5% 59.3% 1.6% 0.2% 1.88 .66 Correlation with corrected subscale=.479 Correlation with corrected Scale:.449 Q3.14. If I refer an employee to the EAP, it would likely just turn into a grievance. No Ans l/SD 2/Disagree 3/Neither 4/Agree 5/SA Mean Sthev 4.1% 14.8% 63.0% 15.9% 2.0% 0.2% 2.06 .65 Correlation with corrected subscale=.486 Correlation with corrected Scale:.514 Q3.16. It would damage a person's reputation or image if 1 refer him/her to t he E P No Ans 1/SD 2/Disagree 3/Neither 4/Agree 5/SA Mean Sthev 4.3% 10.7% 62.0% 16.1% 6.4% 0.5% 2.20 .74 Correlation with corrected subscale=.550 Correlation with corrected Scale:.521 Q3.30. If I referred someone to the EAP, 1 would expect co-workers to "take sides" with him or her on the matter. No Ans 1/SD 2/Disagree 3/Neither 4/Agree 5/SA Mean Sthev 5.0% 15.5% 58.9% 16.1% 4.5% 0% 2.10 .72 Correlation with corrected subscale=.356 Correlation with corrected Scale:.432 Q3.33. It would probably embarrass us both if I confront a worker and refer him/her to the EAP. No Ans l/SD 2/Disagree 3/Neither 4/Agree 5/SA Mean Sthev 4.5% 8.6% 63.4% 15.7% 7.7% 0% 2.24 .72 Correlation with corrected subscale=.478 Correlation with corrected Scale:.522 Q3.37. I could make problems at home for a worker if I told him or her to go see the EAP counselor. No Ans 1/SD 2/Disagree 3/Neither 4/Agree 5/SA Mean Sthev 5.5% 1.8% 50.0% 36.4% 6.4% 0% 2.50 .65 Correlation with corrected subscale:.287 Correlation with corrected Scale:.291 Subscale ORGBARRS (Perceived Organizational Barriers - 6 items) Mean: 13.5 Std Dev: 3.0 coefficient alpha: .66 Q2.6. 1 am too busy to do the kind of documentation necessary to properly send someone to the EAP. No Ans 1/SD 2/Disagree 3/Neither 4/Agree 5/SA Mean Sthev 2.5% 19.5% 52.5% 18.0% 6.1% 1.4% 2.15 .86 Correlation with corrected subscale=.394 Correlation with corrected Scale:.483 Q2.l4. It's very hard to refer a worker to the EAP if a previous supervisor did nothing about the problem. No Ans l/SD 2/Disagree 3/Neither 4/Agree 5/SA Mean Sthev 2.5% 15.0% 55.7% 10.0% 14.1% 2.7% 2.32 .99 Correlation with corrected subscale=.37l Correlation with corrected Scale:.436 Q2.24. My department's level of productivity will be jeopardized by having someone gone to the EAP or to treatment. No Ans 1/SD 2/Disagree 3/Neither 4/Agree 5/SA Mean Sthev 4.5% 22.0% 57.5% 10.7% 4.3% 0.9% 2.00 .78 Correlation with corrected subscale:.437 Correlation with corrected Scale:.479 Q3.1. 1 would lose most of my say about what happens to an employee after he/she goes to the EAP. No Ans 1/SD 2/Disagree 3/Neither 4/Agree 5/SA Mean Sthev 3.9% 9.3% 50.5% 25.5% 10.2% 0.7% 2.40 .83 Correlation with corrected subscale:.322 Correlation with corrected Scale:.397 Q3.29. The procedures for referring someone to the EAP are too involved and complicated. No Ans 1/SD 2/Disagree 3/Neither 4/Agree 5/SA Mean Sthev 5.2% 8.9% 54.5% 29.8% 1.4% 0.2% 2.26 .65 Correlation with corrected subscale:.434 Correlation with corrected Scale:.542 223 Q3.34. It would add to the problems we already have for one or more employees to be gone each week for counseling appointments at the EAP. No Ans l/SD 2/Disagree 3/Neither 4/Agree 5/SA Mean Sthev 5.0% 7.3% 56.1% 20.7% 9.5% 1.4% 2.39 .82 Correlation with corrected subscale=.397 Correlation with corrected Scale:.422 Subscale IDENTUNC (Identification Uncertainty - 5 items) Mean: 12.1 Std Dev: 2.4 coefficient alpha=.57 Q2.l5. It is best to refer employees only after their co-workers have complained about them or raised questions about their behavior. No Ans 1/SD 2/Disagree 3/Neither 4/Agree 5/SA Mean Sthev 7.5% 2.5% 20.0% 65.2% 4.1% 0.7% 1.98 .72 Correlation with corrected subscale=.399 Correlation with corrected Scale:.561 Q2.27. Unless I hear from several workers that someone is having problems, I would probably hold off on suggesting the EAP. No Ans l/SD 2/Disagree 3/Neither 4/Agree 5/SA Mean Sthev 4.8% 11.8% 62.3% 12.5% 8.0% 0.7% 2.20 .78 Correlation with corrected subscale:.422 Correlation with corrected Scale:.504 Q3.7. My workers do creative or professional jobs which are really hard to evaluate, so there is no way I can tell when they're having a problem. No Ans l/SD Z/Disagree 3/Neither 4/Agree 5/SA Mean Sthev 3.6% 15.9% 64.5% 12.5% 3.2% 0.2% 2.04 .67 Correlation with corrected subscale:.289 Correlation with corrected Scale:.432 Q3.11. When a worker does well on some days but not on others, you really can't tell if they need the EAP. No Ans 1/SD 2/Disagree 3/Neither 4/Agree 5/SA Mean Sthev 3.9% 1.6% 33.9% 33.2% 25.5% 2.0% 2.92 .87 Correlation with corrected subscale=.310 Correlation with corrected Scale:.282 Q3.27. It's hard to tell if problems are serious enough that an EAP referral is the right rather than the wrong decision. No Ans l/SD 2/Disagree 3/Neither 4/Agree 5/SA Mean Sthev 5.2% 0.5% 36.6% 29.5% 25.7% 2.5% 2.93 .88 Correlation with corrected subscale:.257 Correlation with corrected Scale:.335 224 Subscale EAPSPECS (EAP-specific Concerns - 3 items) Mean: 6.6 Std Dev: 2.0 coefficient alpha: .70 Q2.l7. I would hesitate to refer anyone to our EAP because I don't like where it is located. No Ans l/SD 2/Disagree 3/Neither 4/Agree 5/SA Mean Sthev 3.4% 20.2% 47.3% 25.0% 2.7% 1.4% 2.15 .83 Correlation with corrected subscale:.449 Correlation with corrected Scale:.474 Q2.40. Our EAP does not keep personal matters totally confidential. No Ans I/SD 2/Disagree 3/Neither 4/Agree 5/SA Mean Sthev 5.5% 24.3% 31.6% 33.0% 3.9% 1.8% 2.23 .94 Correlation with corrected subscale:.544 Correlation with corrected Scale:.426 Q3.8. Because our EAP counselor lacks professional competence, 1 am very hesitant to refer anyone to him or her. No Ans 1/SD 2/Disagree 3/Neither 4/Agree 5/SA Mean Sthev 4.5% 14.3% 49.1% 29.5% 2.3% 0.2% 2.21 .73 Correlation with corrected subscale=.563 Correlation with corrected Scale:.571 COUNROLE Scale (Sees self in counselor role - 5 items) Mean: 16.9 Std Dev: 2.4 coefficient alpha: .56 Q2.lO. When employees have something bothering them, I can usually get them to talk to me about it. No Ans 1/SD 2/Disagree 3/Neither 4/Agree 5/SA Mean Sthev 2.5% 0.5% 6.6% 22.0% 57.7% 10.7% 3.73 .76 Correlation with corrected Scale:.356 Q2.28. A supervisor should refer to the EAP instead of trying to handle a worker's problems on his/her own. No Ans 5/SD 4/Disagree 3/Neither 2/Agree l/SA Mean Sthev 4.5% 3.6% 28.6% 36.8% 21.8% 4.5% 2.95 .93 Correlation with corrected Scale:.289 Q2.39. I get quite uncomfortable whenever an employee begins to talk to me about things in their personal life. No Ans 5/SD 4/Disagree 3/Neither Z/Agree l/SA Mean Sthev 5.2% 11.6% 61.1% 16.1% 5.5% 0.5% 2.18 .73 Correlation with corrected Scale:.321 Q3.17. I feel it is my role to try to talk to and help an employee who is worried or upset about something. No Ans l/SD Z/Disagree 3/Neither 4/Agree 5/SA Mean Sthev 3.6% 0.5% 5.2% 18.0% 56.8% 15.9% 3.86 .77 Correlation with corrected Scale:.331 Q3.31. I feel I can help most people with their problems, since 1 have worked out some tough problems of my own. No Ans l/SD 2/Disagree 3/Neither 4/Agree 5/SA Mean Sthev 4.5% 7.5% 45.5% 32.7% 9.1% 0.7% 2.48 .80 Correlation with corrected Scale:.322 FAMILIARity with EAP Scale (4 items) Mean: 13.5 Std Dev: 3.4 coefficient alpha: .74 Q2.1. 1 know some workers who were referred to the EAP by their supervisors. No Ans l/SD 2/Disagree 3/Neither 4/Agree 5/SA Mean Sthev 6.4% 9.3% 13.9% 21.4% 33.4% 15.7% 3.34 1.21 Correlation with corrected Scale:.477 Q2.18. The procedures for making an EAP referral are clear to me. No Ans l/SD 2/Disagree 3/Neither 4/Agree 5/SA Mean Sthev 3.0% 7.7% 22.7% 14 % 41.1% 10.7% 3.25 1.16 Correlation with corrected Scale:.574 Q2.21. I know who to contact if I want to refer an employee to the EAP. No Ans 1/SD 2/Disagree 3/Neither 4/Agree 5/SA Mean Sthev 3.6% 4.8% 8.2% .7% 47.7% 33.0% 4.00 1.08 Correlation with corrected Scale:.574 Q3.22. 1 have consulted with the EAP counselors for ideas on how to deal with one or more employees who were having job problems. No Ans 1/SD Z/Disagree 3/Neither 4/Agree 5/SA Mean Sthev 5.5% 9.8% 28.4% 19.1% 32.3% 5.0% 2.94 1.12 Correlation with corrected Scale:.499 HM. 226 APPENDIX C ATTITUDE SCALE AND SUBSCALE INTERCORRELATIONS AND RELIABILITIES Table C.1 Attitude Scale and Subscale Intercorrelations and Reliabilities 12 3 4 5 6 7 8 91011121314 54 27 32 —06 18 18 67 62 37 44 09 26 21 55 73 -29 -24 -22 25 -08 -03 ~51 -47 9O 10 -26 -27 -20 19 -12 ~06 -37 -38 87 66 11 -23 ~21 -13 22 -05 00 -46 —39 88 68 78 12 —20 -15 -16 25 00 -01 -42 -37 83 66 63 66 13 —17 -12 -17 21 -05 -01 -30 -31 77 64 56 60 57 14 —36 -22 -26 10 -10 ~08 -58 -50 65 47 52 44 37 7O \DOOVONWJTUJNH \l p— U'i 0‘ \n N 4? N U H Note. _N = 415; for absolute values of _r_ > .08, p < .05; for _r_ > .11, < .01; for _r_ > .15, p < .001. Cronbach's alpha reliability estimates lie on the diagonal; decimals are omitted. The scales and subscales are: 1 : APPROACH Scale 9 = AVOIDANCE Scale 2 = CONEMPL subscale 10 = DENIAL subscale 3 = CONPROD subscale 11 = NEGOUTS subscale 4 = DISCMFRT subscale 12 : ORGBARRS subscale 5 = INTOLDEV subscale 13 = IDENTUNC subscale 6 = ACCREFRL subscale 14 = EAPSPECS subscale 7 = ORGSUPRT subscale 8 : POSITEXP subscale 227 Am>MAUUO n no mmUUWE. u so >tmm>Mm u no Am>m30m u Nfi mXMmm> u : wDUOm ops—03m do”. .Onm mU< IU1 .n:. A i... no moo—«.5 8203b boumEioeaom 3 m BEES. _tNMd'h “we 2. c 83:3 ummfl “Boo wEmmME no omamoon >um> Pm “woutEo 2950mm. .302 so mo- ow No- _o o. moi mo so mo no fio mo mo no- so no- mo- mo go- mo mo- oo mo NN- oo _o mo no so w: No. oo no ow go- mg on: no- so _oi om oo no- no mo ~o no- no no- N~i no- Do u— mo me. No NN- so do nn so- no N~ fio no No so no- mo No- moi sN oo- so- mo- oo. oo mo No- no no- so mo n— mo do- fioi mo- Nfii so no no no- n_ Ns n~ oo so oo. oo No No do no N_ wo. wo mg n— oo m— moi nmi no- no. so nqi 0‘ o_- nNi fig- no- N— nn on _~ o_ sm no- so mo nN on _d om so so on so- mo n— no mo 0— no ~_ no No mo- no no oo a mo- no mo. won mo no no- w ms ofii mm mg on mo 5 on: gs _m sn om o ow- moi mm- mg- n mo mm mm s MN no M No N no n— ma n: s_ no No ad oo o m m o n s m N fl 333:; no 303235885 “.0 2an mmqm<_m<> LO mZOEkJm—MMOUNEHZ_ D X_Omem< REFERENCES REFERENCES Asma, F. E. (1975). Long-term experience with rehabilitation of alcoholic employees. In R. L. Williams at G. H. Moffat, (Eds.), Occupational Alcoholism Programs (pp. 175-193). Springfield, IL: Charles C. Thomas. Bayer, G. A. (1987). An investigation of the construct validity of the Behavioral Index of Troubled Employees (BITE). Dissertation Abstracts International, 4__8_(6), 1841B. (Universtiy Microfilms No. DA8720504). Bayer, G. A., 6: Gerstein, L. H. (1988a). An adaptation of models of prosocial behavior to supervisor interventions with troubled employees. Journal of Applied Social Psychology, 1_8(l), 23-37. Bayer, G., or Gerstein, L. (1988b). Supervisory attitudes toward impaired workers: A factor analytic study of the Behavioral Index of Troubled Employees (BITE). Journal of Applied Behavioral Science, 24(4), 413-422. Bayer, G. A., a: Gerstein, L. (1989). Helping troubled employees: Supervisors' decision-making behavior. Research Presentations of the l‘8th Annual Conference of the Employee Assistance Professionals Association (pp. 53—68). Belasco, J. A., 6c Trice, H. M. (1969). The assessment of change in training and therapy. New York: McGraw-Hill. Berg, 1. A. (1970). Employee counseling in a well-rounded personnel program. Public Personnel Review, 31(3), 185-189. Beyer, J. M., or Trice, H. M. (1978). Implementing change: Alcoholism policies in work orggiizations. New York: The Free Press. Beyer, J. M., 6c Trice, H. M. (1981). A retrospective study of similarities and differences between men and women employees in a job-based alcoholism program from 1965-1977. Journal of Drug Issues, _1_1(2), 233-262. 228 229 Beyer, J. M., or Trice, H. M. (1982). Design and implementation of job based alcoholism programs: Constructive confrontation strategies and how they work. In Occupational alcoholism: A review of research issues. Proceedings of a workshop May 22-24, 1980, Reston, Virginia. Research Monograph No. 8, pp. 181-239. (DHHS Publication No. ADM 82-1184). Washington, DC: U.S. Goverment Printing Office. Beyer, J. M., or Trice, H. M. (1984a). The best/worst technique for measuring work performance in organizational research. Journal of Organizational Behavior and Statistics, 1(1), 95-115. Beyer, J. M., or Trice, H. M. (1984b). A field study of the use and perceived effects of discipline in controlling work performance. Academy of Management Journal, QM), 743-764. Beyer, J. M., Trice, H. M., or Hunt, R. E. (1980). The impact of federal sector unions on supervisors' use of personnel policies. Industrial and Labor Relations Review, 3_3_(2), 212-231. Brodzinski, J. D., or Goyer, K. A. (1987). Employee assistance program utilization and client gender. Employee Assistance Quarterly, 2(1), 1'13. Burisch, M. (1984). Approaches to personality inventory construction. American Psychologist, 19(3), 214-227. Burke, R. J. (1982). Disclosure of problems and informal helping in work settings. Psychological Reports, fl, 811-817. Cahill, M. H. (1983). Training employees and supervisors to increase use of EAPs by women. Alcohol Health and Research World, 7(3), 18-22. Cahill, M. H., Volicer, B. J., or Neuburger, E. (1982). Female referral to employee assistance programs: The impact of specialized intervention. Drug and Alcohol Dependence, fl, 223-233. Cairo, P. C. (1983). Counseling in industry: A selected review of the literature. Personnel Psycholog , 3_6(1), 1-18. Cohen, J., or Cohen, P. (1983). Applied multiple regression/correlation analysis for the behavioral sciences. (2nd ed.) Hillsdale, NJ: Lawrence Erlbaum Associates. Cohen, L. J. (1986). Coping with traumatic events: A theoretical model and a study of recovery from rape. Dissertation Abstracts International, 41(9), 3949-5. (Universtiy Microfilms No. DA8700015). Cohen, L. J., or Roth, S. The approach avoidance scale: A measure of coping strategies. Manuscript submitted for publication. 230 Cubelli, G. E. (1970). Some dynamics of the referral process. Rehabilitation Literature, 31(7), 200-202. Daghestani, A. N., Barlow, P., Hilker, R. R. J., or Asma, F. E. (1976). The supervisor's role with the problem employee. Journal of Occupational Medicine, £(2), 85-90. Dancey, T. E. (1973). The constructive coercion technique in alcoholism and drug-dependency programs. In J. M. Scher (Ed.), _Drug Abuse in Industry (pp. 243-257). Springfield, IL: Charles C. Thomas. Dollard, J., 6: Miller, N. E. (1950). Personality and psychotherapy: An analysis in terms of learning, thinking, and culture. New York: McGraw-Hill. Dowling, W. F. (1981). A descriptive study of an occupational program's impact on employees' absenteeism and health care cost in state government. (Doctoral dissertation, Michigan State University, East Lansing, MI, 1980). Dissertation Abstracts International, 4_1(7), 2811B. Epstein, S. (1962). The measurement of drive and conflict in humans: Theory and experiment. In M. R. Jones (Ed.), Nebraska symposium on motivation (Vol. 10, pp. 127-209). Lincoln: University of Nebraska Press. Epstein, S. (1982). Conflict and stress. In L. Goldberger or S. Breznitz (Eds.), Handbook of stress: Theoretical and Clinical aspects (pp. 49-68). New York: The Free Press. Epstein, 5., 6c Fenz, W. D. (1962). Theory and experiment on the measurement of approach-avoidance conflict. Journal of Abnormal and Social Psycholog , §_4(2), 97-112. Finlay, D. G. (1975). Constructive coercion and the alcoholic employee: Problems and prospects. Vancouver, BC: Education Department, Alcohol and Drug Commission of British Columbia. Foote, A., 6: Erfurt, J. C. (1989). Study data on supervisory cover-up. The Almacan, 248), 33-34. Forrest, D. V. (1983). Employee assistance programs in the 1980's: Expanding career options for counselors. Personnel and Guidance Journal, 3(2), 105-107. Freedberg, E. J., or Johnston, W. E. (1978). Effects of various sources of coercion on outcome of treatment of alcoholism. Psychological Reports, 43, 1271-1278. Gam, J., Sauser, W. 1., Jr., Evans, K. L., 6: Lair, C. V. (1983). Implementing an employee assistance program. Journal of Employment Counseling, 29(2), 61-69. 231 Gerstein, L. H., Eichenhofer, D. J., Bayer, G. A., Valutis, W., 6: Jankowski, J. (1989). EAP referral training and supervisors‘ beliefs about troubled workers. Employee Assistance Quarterly, 4(4), 15-30. Googins, B. K. (1979). The use and implementation of occupational alcoholism programs by supervisors: An analysis of barriers. Dissertation Abstracts International, 40(4), 2269A-227OA. (Universtiy Microfilms No. 7942701).— Googins, B., Colan, N. B., 6: Schneider, R. J. (1989). Supervisor training in employee assistance programs. Research Presentations of the 18th Annual Conference of the Employee Assistance Professionals Association (pp. 91-99). Googins, B., 6: Kurtz, N. R. (1979). Supervisory networks: Towards an alternative training model. Labor-Management Alcoholism Journal, 2(1), 35‘40. Googins, B., 6: Kurtz, N. R. (1980). Factors inhibiting supervisory referrals to occupational alcoholism intervention programs. Journal of Studies on Alcohol, flUl), 1196-1208. Googins, B., or Kurtz, N. R. (1981). Discriminating participating and nonparticipating supervisors in occupational alcoholism programs. Journal of Drug Issues, 1_l(2), 199-216. Googins, B., 6: Kurtz, N. R. (1984). The employee as client: Perceptions of a work-based alcohol treatment program. Journal of Psychoactive Drugs, 16(2), 161-168. Hawthorne, W. B., 6: Davidson, B. N. (1983). The alcoholic's supervisor: Another victim in need. Occupational Health and Safety, 5_2(2), 28-29,40. Heilizer, F. (1977a). A review of the theory and research on the assumptions of Miller's response competition (conflict) models: Response gradients. Journal of General Psychology, _91, l7-71. Heilizer, F. (1977b). A review of theory and research on Miller's response competition (conflict) models. Journal of General Psychology, fl, 227-280. Henderson, R. M., 6: Bacon, 5. E. (1953). Problem drinking: The Yale plan for business and industry. Quarterly Journal of Studies on Alcohol, 1_4, 247-262. Heyman, M. M. (1976). Referral to alcoholism programs in industry: Coercion, confrontation and choice. Journal of Studies on Alcohol, 2(7), 900-907. ‘- 232 Heyman, M. M. (1978). Alcoholism programs in industry - The patient's Ew. New Brunswick, NJ: Journal of Studies on Alcohol, Inc. Hoffman, E., 6c Roman, P. M. (1984). Effects of supervisory style and experientially based frames of reference on organizational alcoholism programs. Journal of Studies on Alcohol, QB), 260-267. Hunt, R. G., 6: Lichtman, C. M. (1969). "Counseling" of employees by work supervisors: Concepts, attitudes, and practices in a white-collar organization. Journal of Counseling Psycholcgy, gm, 81-86. Kaden, S. _E. (1977). Compassion or cover-up: the alcoholic employee. Personnel Journal, EU), 356-358. Kelman, H. C. (1974). Attitudes are alive and well and gainfully employed in the sphere of action. _American Psychologis , 29(5), 310-324. Kelman, H. C. (1979). The role of action in attitude change. In M. M. Page (Ed.), Nebraska symposium on motivation (Vol. 27, pp. 117-194). Lincoln: University of Nebraska Press. Kelsey, J. E. (1975). Professional counseling in a company with a broad dispersion of work locations. Journal of Occupational Medicine, £01), 702-705. Kleeman, B., (at Googins, B. (1983). Women alcoholics in management. Alcohol Health and Research World, 7(3), 23-28. Krucher, K. J. (1986). The relationship between supervisors' interpersonal need structure and use of employee assistance programs. Dissertation Abstracts International, 4_8_(2), 342-A. (Universtiy Microfilms No. DA8711726). Kurtz, N. R. (1982). Dynamics of the identification and referral process in work organizations. in Occupational alcoholism: A review of research issues. Proceedings of a workshop May 22-24, 1980, Reston, Virginia. Research Monograph No. 8, pp. 273-314. (DHHS Publication No. ADM 82-1184). Washington, DC: U.S. Government Printing Office. Kurtz, N. R., or Googins, B. (1979). Managing the alcoholic employee: Towards a model for supervisory intervention. Industrial Management, a, 15-21. Kurtz, N. R., Googins, B., 6: Williams, C. M. (1980). Supervisors‘ views of an occupational alcoholism program: An experiential perspective. Alcohol Health and Research World, 4(3), 44-49. 233 Leggo, C. (1967). The supervisor and the alcoholic employee. Journal of Occupational Medicine, 2(3), 96-98. Leonards, J. T. (1981). Corporate psychology: An answer to occupational mental health. Personnel and Guidance Journal, 62(1), 47—51. Lotterhos, J. F. (1975). Historical and sociological perspectives of alcohol-related problems. In R. L. Williams 6: G. H. Moffat, (Eds.), Occupational alcoholism programs (pp. 3—39). Springfield, IL: Charles C. Thomas. Mastrich, J. L., Jr. (1985). Influences on the successful treatment of alcoholics referred through employee assistance programs. Dissertation Abstracts International, 46(9), 3251-B. (Universtiy Microfilms No. DA852423—1). Mastrich, J., 6c Farmer, J. L. (1986). The "systems" approach to intervention. The Almacan, lé,(9), 10-11. Mierkiewicz-Alleva, E. M. (1986). Labeling effects of the management referral process in employee assistance programs. Dissertation Abstracts International, 47(5), 2176-B.‘ (Universtiy Microfilms No. DA8616482). " Miller, N. E. (1944). Experimental studies of conflict. In J. McV. Hunt (Ed.), Personality and the behavior disorders (pp. 431-465). New York: Ronald Press. Miner, J. B., or Brewer, J. F. (1976). The management of ineffective performance. In M. Dunnette (Ed.), Handbook of industrial and oganizational psychology (pp. 995-1029). Chicago: Rand McNally. Myers, D. W. (1984). Establishing and building employee assistance programs. Westport, CT: Quorum Books. Myers, J. K., Weissman, M. M., Tischler, G. L., Holzer, C. E.,III, Leaf, P. J., Orvaschel, H., Anthony, J. C., Boyd, J. H., Burke, 3. D., Jr., Kramer, M., or Stolzman, R. (1984). Six-month prevalence of psychiatric disorders in three communities 1980 to 1982. Archives of General Psychiatry, 4_1, 959-967. O'Connell, V. C. (1987). A strategy for overcoming supervisors‘ resistance. EAP Digest, 7(3), 63-66. Older, H. J., Phillips, D. A., 6( Purvis, A. J. (1978). A new approach to the training of supervisors and union personnel. Labor-Management Alcoholism Journal, 8(1), 12-21. Phillips, D. A., 6: Older, H. J. (1977). A model for counseling troubled supervisors. Alcohol Health and Research World, _1, 24-30. 234 Reichman, W. (1983). Affecting attitudes and assumptions about women and alcohol problems. Alcohol Health and Research World, 7(3), 6-10. _ Riediger, A. J. (1979). Employee assistance program: The supervisor's perspectiv . Unpublished doctoral dissertation, University of Alberta, Edmonton, Canada. Riediger, A. J. (1985). EAPs: Barriers to effectiveness. In S. H. Klarreich, J. L. Francek, or C. E. Moore, (Eds.), The human resources management handbook - Principles and practice of emploiee assistance programs (pp. 393-408). New York: Praeger Publishers. Robinson, D., 6: Robinson, N. (1963). The supervisor and the alcoholic worker. Supervisgy Management, 8(2), 13-16. Roman, P. M. (1975). "Spirits at Work" revisited: Needed priorities in occupational alcoholism programming. In M. E. Chafetz (Ed.), Research, treatment, and prevention - Proceedings of the fourth annual alcoholism conference of the National Institute on Alcohol Abuse and Alcoholism (pp. 307-323). (ADM) 76-284. Rockville, MD: NIAAA. Roman, P. M. (1980). Medicalization and social control in the workplace: prospects for the 1980's. Journal of Applied Behavioral Science, 16(3), 407-422. Roman, P. M. (1981a). From employee alcoholism to employee assistance — Deemphases on prevention and alcohol programs in work-based programs. Journal of Studies on Alcohol, 43(3), 244-272. Roman, P. M. (I981b). Job characteristics and the identification of deviant drinking. Journal of Drug Issues, 71(3), 357-364. Roman, P. M. (1981c). Social and organizational forces accompanying the development of employee alcoholism and employee assistance programs (Report 2-36). New Orleans: Tulane University, Center for Monitoring of Occupational Alcoholism Programming. Roman, P. M. (1982). Barriers to the use of constructive confrontation with employed alcoholics. Journal of Drug Issues, 2, 369-382. Roman, P. M. (1983). Employee assistance programs in Australia and the United States: Comparisons of origin, structure, and the role of behavioral science research. Journal of Applied Behavioral Science_, 12(3), 367-379. Roman, P. M., or Blum, T. (1988). The core technology of employee assistance programs: A Reaffirmation. The Almacan, 18(8), 17-22. _— l ‘1 j ,. 235 Roman, P. M., & Trice, H. M. (1968). The sick role, labelling theory and the deviant drinker. International Journal of Social Psychiatr , 1_4, 245-251. Roman, P. M., 6: Trice, H. M. (1976). Alcohol abuse and work organizations. In B. Kissin of H. Begleiter (Eds.), Social aspects of alcoholism (pp. 445-517). New York: Plenum. Roth, 5., 6: Cohen, L. J. (1986). Approach, avoidance, and coping with stress. American Psychologist, 47(7), 813-819. Sadler, M., 6c Horst, J. F. (1975). Company/union programs for alcoholics. Harvard Business Review, 29(5), 22-34,152-156. Shain, M. (1985). An exploration of the ability of broad-based EAPs to generate alcohol-related referrals. In S. H. Klarreich, J. L. Francek, 6: C. E. Moore (Eds.), The human resources management handbook - Principles and practice of employee assistance programs (pp. 232-242). New York: Praeger Publishers. Shain,_M., 6c Groeneveld, J. (1980). Employee-assistance programs: Philosophy, theory, and practice. Lexington, MA: Lexington Books. Shapiro, S., Skinner, E. A., Kessler, L. G., Von Korff, M., German, P. S., Tischler, G. L., Leaf, P. J., Benham, L., Cottler, L., 6L Regier, D. A. (1984). Utilization of health and mental health services - Three epidemiologic catchment area sites. Archives of General Psychiatr , 4_1, 971—978. Shaw, M. E., 6: Wright, J. M. (1967). Scales for the measurement of attitudes. New York: McGraw—Hill. Smart, R. G. (1974). Employed alcoholics treated voluntarily and under constructive coercion: A followup study. Quarterly Journal of Studies on Alcohol, 32, 196-209. Sonnenstuhl, W. J. (1986). Inside an emotional health program: A field study of workplace assistance for troubled employees. Ithaca, NY: ILR Press, Cornell University. Sonnenstuhl, W. J., 6: Trice, H. M. (1986). Strategies for employee assistance programs: The crucial balance. (K_ey Issues, Number 30). Ithaca, NY: ILR Press, Cornell University. Sonnenstuhl, W. J., 6: Trice. H. M. (1989). The myth of the troubled supervisor. Research Presentations of the 18th Annual Conference of the Employee Assistance Professionals Association (pp. 21-28). Toomer, J. E. (1980). Counseling psychologists in business and industry. Counseling Psychologist, QB), 9-18. 236 Torjman, S. R. (1985). Training referral agents for EAPs: Knowledge, skills and attitudes. In S. H. Klarreich, J. L. Francek, 6: C. E. Moore, (Eds.), The human resources management handbook - Principles and practice of employee assistance programs (pp. 189-199). New York: Praeger Publishers. Trice, H. M._ (1962). The job behavior of problem drinkers. In D. J.- Pittman 6: C. R. Snyder, (Eds.), Society, culture, and drinking patterns (pp. 493-510). New York: John Wiley 6: Sons. Trice, H. M. (1965). Reaction of supervisors to emotionally disturbed employees - A study of deviation in a work environment. Journal of Occupational Medicine, 7(5), 177-188. Trice, H. M. (1980). Applied research studies: Job-based alcoholism and eznployee assistance programs. Alcohol Health and Research World, 4 3), 4-16. Trice, H. M. (1983). Employee assistance programs: Where do we stand ' 1982? Journal of Psychiatric Treatment and Evaluation, 2, 521-529. 5 Trice, H. M., or Belasco, J. A. (1965). Identifying and confronting the alcoholic employee: Role of the industrial nurse. _American Association of Industrial Nurses Journal, g, 7-10. Trice, H. M., or Belasco, J. A. (1966). Emotional health and employer responsibility (Bulletin 57). Ithaca, NY: New York State School of Industrial and Labor Relations, Cornell University. Trice, H. M., or Belasco, J. A. (1968). Supervisory training about alcoholics and other problem employees: A controlled evaluation. Quarterly Journal of Studies on Alcohol, 2_9_(2), 382-398. Trice, H. M., or Beyer, J. M. (1977). Differential use of an alcoholism policy in federal organizations by skill level of employees. In C. J. Schramm (Ed.), Alcoholism and its treatment in industry (pp. 44-68). Baltimore, MD: The Johns Hopkins University Press. Trice, H. M., or Beyer, J. M. (1979). Women employees and job-based alcoholism programs. Journal of Drug Issues, 9(3), 371-385. Trice, H. M., 6: Beyer, J. M. (I982a). Job-based alcoholism programs: Motivating problem drinkers to rehabilitation. In E. M. Pattison 6c E. Kaufman (Eds.), Encyclopedic handbook of alcoholism (pp. 954-978). New York: Gardner Press. Trice, H. M., or Beyer, J. M. (I982b). Social control in worksettings: Using the constructive confrontation strategy with problem-drinking employees. Journal of Drug Issues, 72(1), 21-49. 237 Trice, H. M., 6: Beyer, J. M. (l984a). Employee assistance programs: Blending performance-oriented and humanitarian ideologies to assist emotionally disturbed employees. Research in Community and Mental Health, 4, 245-297. Trice, H. M., 6: Beyer, J. M. (1984b). Work-related outcomes of the construcfive-confrontation strategy in a job-based alcoholism program. Journal of Studies on Alcohol, 4_5(5), 393-404. Trice, H. M., 6: Roman, P. M. (1978). Spirits and demons at work: Alcohol and other drugs on the job (2nd ed.). Ithaca, NY: New York State School of Industrial and Labor Relations, Cornell University. Trice, H. M., 6: Schonbrunn, M. (1981). A history of job-based alcoholism programs 1900-1955. Journal of Drug Issues, _1_l(2), 171-198. Wilcox, J. E. (1985). Employee assistance programs in the federal government: Hindrances and facilitations to supervisory constructive intervention. DissertationAbstracts International, 42(7), 2349B. (Universtiy Microfilms No. DA8422782). Wood, S. P., or Steiner, J. R. (1989). Applying instructional technology to EAPs. EAP Digest, 2(6), 43-48. Wrich, J. T. (1980). The employee assistance progpam - Updated for the 1980's. Center City, MN: Hazelden Foundation. Young, D. W., Reichman, W. R., 6c Levy, M. F. (1987). Differential referral of women and men to employee assistance programs: The role of supervisory attitudes. Journal of Studies on Alcohol, 48(1), 22-28. RnRIES l. "jfjjjjjijijlliljlliiililjl .1.‘J'3-"' .» ’ , ,,. .-"‘ .. , .. "“7 i ' I: . "' 1' 3.1.7 iL. ..' ... urn-""75”-