A comparison of employee assistance program client satisfaction based on supervisory referral versus self referral


Material Information

A comparison of employee assistance program client satisfaction based on supervisory referral versus self referral
Physical Description:
ix, 151 leaves : ; 29 cm.
Alexander, Patricia Nelle
Publication Date:


Subjects / Keywords:
Counselor Education thesis, Ph.D   ( lcsh )
Dissertations, Academic -- Counselor Education -- UF   ( lcsh )
bibliography   ( marcgt )
theses   ( marcgt )
non-fiction   ( marcgt )


Thesis (Ph. D.)--University of Florida, 1998.
Includes bibliographical references (leaves 134-143).
Additional Physical Form:
Also available online.
Statement of Responsibility:
by Patricia Nelle Alexander.
General Note:
General Note:

Record Information

Source Institution:
University of Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
aleph - 029217989
oclc - 40060030
System ID:

Full Text







Copyright 1998


Patricia Nelle Alexander


I would like to acknowledge the invaluable help and

support of my chairperson, Silvia Echevarria Rafuls, Ph.D.,

who responded promptly to my questions and kept this

research project moving. Her insight and suggestions

resulted in a much better document than I could have

produced alone. Her careful attention to detail and unique

ability to work collaboratively taught me as well. I also

want to thank Peter A. D. Sherrard, Ed.D., my committee

member and former advisor, who encouraged me throughout my

period of studies in the doctoral program. His wisdom,

intelligence, and experience are matched by his respect for

students and his delightful sense of humor. Thanks to M.

David Miller, Ph.D. who answered each phone call with a

response of availability and helpfulness as I addressed the

methodology of this study. No question was viewed as

unimportant with the result that he truly made a significant

contribution to this study. And finally, a special thanks

to Stephanie Puleo, Ph.D. whose words of encouragement and

advice helped me keep this entire project in perspective.

Her questions and understanding of EAPs were an asset.

I wrote in my application that I would hope to one day

be able to say with pride that I was a graduate of the


University of Florida. These are four people who have

enriched my life and made my time at the University of

Florida a true learning experience.

I would like to thank Gary L. Wood, Psy.D., my employer

and mentor, for all of his encouragement and support over

these past years. If it were not for his help and

flexibility, my ability to work and carry out full-time

doctoral studies might not have happened. He was my

informal reviewer and coach. It is he who has shown me what

it means to provide quality employee assistance program


Thanks to Cynthia Hewitt-Gervais, Ph.D., who with

kindness and patience, helped me with my statistical

analysis. Thanks also to the recognized researchers, Susan

V. Eisen, Ph.D. and Terry Blum, Ph.D., who made themselves

available to me by phone and who provided me with up-to-date

materials that they had found helpful in their own research.

And thanks to Cindy Montano, who first showed me the article

that led me to the topic that eventually became this




ACKNOWLEDGMENTS . . . . . . . . . .. iii

ABSTRACT . . . . . . . . . . . .. viii


1 INTRODUCTION . . . . . . . . . . 1

Problem and Its Context . . . . . . . 1
EAP Development . . . . . . . . 1
Utilization of EAP Services . . . . . 4
Human Capital Theory . . . . . . .. 5
Systems Theory . . . . . . . . .. 8
Bystander-Equity Model . . . . . ... .11
Evaluating EAP Services . . . . . .. .15
Definition of Outcome . . . . . . .. 16
Consumer Satisfaction . . . . . . .. .17
Self Report . . . . . . . . .. 19
Purpose of Study . . . . . . . .. 21
Statement of the Problem . . . . . ... .21
Need for the Study . . . . . . ... 22
Research Questions . . . . . . ... 22
Definitions . ....... ..................... .25
Organization of Study . . . . . . .. 28

2 REVIEW OF THE LITERATURE . . . . . ... .30

Historical Development . . . . . . .. 30
Occupational Social Welfare. . . . . . 31
Alcoholics Anonymous . . . . . . ... 32
Occupational Alcoholism Programs . . . ... .33
Evolution from OAPs to EAPs . . . . .. .34
Core Technology of Employee Assistance Programs . 42
Core Elements . . . . . . . .. 42
Core Functions . . . ................ . 43
Key Elements of an Employee Assistance Program. . 44
EAP Research Strategies . . . . . . .. .45
Outcome Studies . . . . . . . .. 48
Definition of Outcome . . . . . .. 48
Issues in Outcome Research . . . . . .. .51
Rashomon Effect . . . . . . . .. 51

Timing . . . . . . . . . . 52
Population . . . . . . . . . . 53
Objectives . . . . . . . . . .. 53
Theories . . .... .................... * * 53
Consumer Satisfaction . . . . . . .. 57
Definition of Satisfaction . . . . ... .58
Treatment Issues and Satisfaction . . . .. .60
Self-Report . . . *. ................. . 64
Client Satisfaction Questionnaire-8 . . . .. .70
Criteria for Self-Report Measure . . . ... .70
Development of the CSQ-8 . . . . . ... .72
Participation in the Employee Assistance Program. 74
Supervisory Referrals . . . . . . .. 75
Self-Referral . . . ... .................... .83
Estimates of Differences Between Supervisory and
Self-Referrals . . . . . . . .. 87

3 METHOD . . . . . . . . . . .. 91

Participants . . . . . . . . . .. 91
Client Groups . . . . . . . .. 92
Instruments . . . . . . . . .. 93
Cover Letter . . . . . . . . .. 94
Client Information Form . . . . . .. .95
Client EAP Intake Data Form . . . . .. .97
Client Satisfaction Questionnaire-8 . . .. .100
Procedure . . . . . . . . . .. 103
Steps . . . . . . . . . . .. 104
Analysis . . . . . . . . . . .. 106

4 RESULTS . . . . . . . . . .. 108

Survey Dispositions from First and Second Mailing .108
Survey Follow-up . . . . . . . .. 110
Demographic Data . . . . . . .. 111
Statistical Analysis and Results . . . .. .116
Model Assumptions . . . . . . ... 117
Internal Consistency . . . . . . 117
Category Frequency and Percent . . . .. .118
Chi-square Analysis . . . . . . ... 119
Factorial Analysis of Variance . . . .. .120

5 DISCUSSION . . . . . . . . .. 122

Referral Source . . . . . . . ... 122
Referral by Gender . . . . . . .. 123
Satisfaction . . . . . . . ... * * *. 123
Client Satisfaction Questionnaire-8..... .... .123
Factorial Analysis of Variance . . . .. .124
Limitations . . . . . . . . . .. 126
Sample . . . . . . . . . .. 126
Follow-up . . . . . . . . . .. 128

Future Research . . . . . . . ... 128
Prior EAP Use . . . . . . . . .. 128
Hierarchical Position and Education . . ... .129
Mode of Administration . . . i............ 129
Differences Between Initial and Follow-up
Responses . . . . . . . . . .. 132
Differences by Location . . . . . ... .132
Testing for Dissatisfaction . . . . ... .133

REFERENCES . . . . . . . . . . . 134


A CLIENT INFORMATION . . . . . . ... 144

B EAP INTAKE DATA FORM . . . . . . ... 146

C COVER LETTER . . . . . . . ... . 148

D PHONE SCRIPT . . . . . . . . .. 150

BIOGRAPHICAL SKETCH . . . . . . . . .. 151


Abstract of Dissertation Presented to the Graduate School
of the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Doctor of Philosophy



Patricia Nelle Alexander

May, 1998

Chairperson: Silvia Echevarria Rafuls, Ph.D.
Major Department: Counselor Education

Work organizations have offered some form of an

employee benefit similar to current employee assistance

programs since the 1800s. This study explored whether there

is a significant mean level difference in employee

satisfaction based on referral source (i.e., supervisor or

self). Employees of seventeen companies located throughout

the southeastern region of the United States who had used

their employee assistance program during a six month period

in 1996, were surveyed using the Client Satisfaction

Questionnaire-8. Chi-square analysis supported the

hypothesis that there was a statistically significant

greater number of self-referrals to the employee assistance

program compared to those that came in under supervisory

referral conditions. Chi-square analysis failed to support


any significant difference in referral source by gender.

Factorial three-way analysis of variance (two by two by two

design), main effects model, supported the hypothesis of a

higher mean level of satisfaction for self-referrals as

compared to supervisory referrals. The hypothesis that

there would be no mean level difference in satisfaction by

gender was supported. The hypothesis that the mean level of

satisfaction would be lower for those presenting with drug

problems, as compared to other problems, was not supported.

Due to limitations of sample size, within-group comparisons

were not possible to determine if mean level of satisfaction

differences were affected by those who responded without

follow-up. This study appears to support cognitive

dissonance theory in that those who freely seek help will

tend to be more satisfied.


Problem and Its Context

Employee assistance programs are "worksite-based

programs designed to help identify and facilitate the

resolution of behavioral, health, and productivity problems

that may adversely affect employees' well-being or job

performance" (Blum & Roman, 1995, p. 1). The purpose of an

Employee Assistance Program (i.e., EAP) is "to provide

timely, professional aid for employees whose personal

problems might otherwise lead to work impairment,

absenteeism, accidents, conflicts in the work setting, or

even job termination" (Lewis & Lewis, 1986, p. 4).

EAP Development

The origin of employee assistance programs informally

dates back to the "welfare capitalism" of the early 1800s

(Smith, 1987) when companies, fearful of unionization

occurring among newly-arrived immigrant employees, set out

to provide various types of social services. These

services, available both to the employee and their family

members, were to provide practical assistance to individuals

as well as help in the acculturation process. It was not


until the early 1940s, however, that one finds what came to

be recognized as the formal beginnings of the employee

assistance program. "Two major complex and intertwined

historical threads, occupational social welfare and

occupational alcohol programs, have led to the development

of EAPs" (Midanik, 1991, p. 69). Since that time,

throughout the historical development of employee assistance

programs, changes have occurred in terms of focus, location,

and type of providers.

In the early 1940s, company-sponsored programs were

called occupational alcohol programs (i.e., OAPs) and dealt

exclusively with the issue of alcohol abuse (Dubreuil and

Krause, 1983, p. 85-86). These programs were normally

internal (i.e., services were utilized by going through the

company) and were "staffed primarily by indigenous

nonprofessional or recovering counselors" (Cunningham, 1994,

p. 3). These individuals typically had no formal training

and were usually in recovery themselves. As the provision

of company-sponsored services addressing alcohol problems of

their employees gained credibility, expansion into the

availability of other services occurred. Programs underwent

a process of change in focus from alcohol only to what is

today termed broad-brush employee assistance programs

(Brody, 1988). Blum, Roman, & Harwood (1995) state that the

term, "broadbrush" describes a program that is designed "to

assist workplace personnel in identifying and resolving

problems involving alcohol or drug abuse, and family,

stress, emotional, marital, financial, legal, and other

personal concerns" (pp. 126-127). As mentioned above, in

the days of occupational alcohol programs, programs were

usually internal.

Internal programs refer to the fact that the providers

of occupational alcohol program services were usually

company employees. While corporate-based, in-house employee

assistance programs still exist in some of the largest U.S.

companies, "most of the recent growth in EAPs and employment

has occurred as a result of the development of contracting

groups and external providers who supply fee-based services

to employers" (Cunningham, 1994, p. 19). Providers are now

most often clinicians in their own specialty area (e.g.,

addiction specialists, professional counselors, social

workers and psychologists) who seek and obtain employment in

employee assistance programs (Emener, 1988). Many have met

both knowledge and experience requirements to become

credentialed as Certified Employee Assistance Professionals

(i.e., CEAP). Smith, Salts, and Smith (1989) in their

paper, "Preparing Marriage and Family Therapy Students to

Become Employee Assistance Professionals," for example,

addressed some of the specific qualifications needed by the

contemporary employee assistance program professional; and,

suggested ways to accomplish the acquisition of the

necessary skills within the context of an academically-based

training program. In addition, these authors presented a

list of citations related to how other contemporary

professionals such as social workers, rehabilitation

counselors, psychologists, and psychiatrists fit within the

employee assistance program framework.

By the 1990s, there had been a significant growth of

employee assistance programs to the point that more than

seventy-five percent of Fortune 500 companies offered some

form of a personal counseling service (Feldman, 1991). Blum

and Roman (1992) reported that a 1991 United States national

data sample revealed that 45 percent of full-time employees

who were not self-employed had access to an employer-

provided employee assistance program. While the number of

free visits may differ, and while a small number of programs

are still essentially an occupational alcoholism program, it

appears that American work organizations have incorporated

the concepts of the employee assistance program.

Utilization of EAP Services

Employee assistance programs are designed to be

accessed by employees at any level within a company. Most

programs also are designed to incorporate access by eligible

family members. In some cases, the program may also be

available to employees who are retired. Not only have the

number of available employee assistance programs increased,

so has the rate of utilization. It is estimated that,

"approximately five percent of employees working in an

organization offering an EAP used the EAP in a twelve-month

period" (Blum & Roman, 1995, p. 127). Miller, Jones, and

Miller (1992), stated that an increase in self-referrals has

become evident in the 1990s and appears to be directly

related to on-the-job stresses. Blum and Roman (1995)

reported on a survey carried out with full-time employees

from the end of 1993 through the beginning of 1994. One of

their findings was that the use of employee assistance

programs among both employees and dependents had increased

(p. 2). While the majority of individuals who utilize the

employee assistance program are the result of self-referral,

another significant referral source is supervisory in

nature. In fact, an essential function of an effective

employee assistance program, is training supervisors to

recognize (but not diagnose) the problem behaviors of

troubled employees in order to make an appropriate referral

to the employee assistance program. A study carried out by

Gerstein, Eichenhofer, Bayer, Valutis, and Jankowski (1989)

supported the idea "that constructive confrontation training

enhances supervisors' recognition of impaired employees" (p.


Human Capital Theory

According to Jerrell and Rightmyer (1988), it is

Schramm's "human capital" model (which analyzes the costs


and benefits of employee assistance programs) that provides

a theoretical model for understanding the development of

employee assistance programs and the tools to evaluate such

a program (p. 269). Human capital theory, which was

developed by Carl J. Schramm (1980), established a basis for

service delivery and offered a rationale based on economy

that would justify to an employer the expenses involved in

offering employees and family members access to an employee

assistance program. Schramm's theory posits that just as

employers are interested in investing in capitalizing a

physical plant or equipment, making improvements through

technology, or providing a safe working environment, they

should be interested in investing in their employees' well-

being. When employers make a commitment to invest capital

(e.g., money, time etc.) in their human commodity, they

typically anticipate a return on the investment in terms of

increased productivity, better attendance, and improved

morale and interpersonal interactions. This, in turn, is

anticipated to result in increased income to the company.

Employers are, therefore, willing to spend money to provide

employees with training, continuing education, maintenance

of a safe working environment, and in many cases through the

provision of an employee assistance program.

While a company's decision to provide an employee

assistance program might appear altruistic at first glance,

the long term expectation is that the employer's investment

will yield increased returns on behalf of the employer.

Studies have demonstrated, for example, that companies

having employee assistance programs demonstrate a marked

diminishment in lost productivity, tardiness, and

absenteeism while demonstrating an increase in employee

retention. One significant study (De Fuentes, 1986)

addressed a number of these areas. Findings from this study

indicated that following an employee's participation in the

employee assistance program, an average reduction rate

occurred in (a) absences (68%), (b) disciplinary actions

(59%), (c) hospitalizations for problems identified through

employee assistance program participation (71.5%), and (d)

informal sick time (19.5%). An increase in performance

review ratings (15%) also was noted. This study, consisting

of a combination of interviews and record reviews of the

total number of employees (n=739) seen over a three-year

period (i.e., 1980-1983), also noted a significant high rate

in the area of employee retention. As was the case with

"welfare capitalism" in the 1800s, human capital theory

allows the employer, as well as the employee, to benefit.

As employee assistance programs developed, attention

was given to improving performance, attendance, and conduct

on the part of the employee. Supervisors received training,

usually from representatives of their employee assistance

program, in order to learn to identify problem areas and

make appropriate referrals.

Systems Theory

Another theory that provides support for the

development of employee assistance programs is systems

theory. Cunningham (1994) states that "systems theory

offers a helpful overall framework in understanding both the

complex structures in which people carry out their work

lives and the interactional patterns that exist in work

groups" (p. 43). Employee assistance professionals long

have recognized that workers bring personal concerns into

the work arena and vice versa. Roman (1989), for example,

provided a case of how an employee with a drug problem had

multiple and inter-related problems that affected both

family and work.

Ford and Ford (1986) in their article, "A Systems

Theory Analysis of Employee Assistance Programs," viewed

every employee assistance program as a complex system. The

researchers listed the following ten macrosystems (a)

employee assistance program administrator and case manager

and the external or internal group they represent, (b) upper

level management of the employing organization, (c) middle

level and supervisory management, (d) line and staff

employees, (e) employees' families, (f) labor unionss, (g)

health care and social service networks, (h) private and

government subsidizers, insurers, and health care payors,

(i) related employers such as suppliers, distributors etc.,

and (j) the general public (p. 37). The authors went on to

state that "no person, no action, no problem stands alone in

the network of systems to which each EAP belongs. Each

referral signifies potential disruption, whether overt or

unrecognized, within or between all the interfacing systems

which cannot be ignored" (p. 38). This is particularly true

in the case of supervisory referrals.

A core technology and set of functions derived from a

body of research by Blum and Roman (1989) emphasizes the

importance of employee assistance program interaction with

managers and supervisors. According to Blum and Roman

(1992, p. 121) there are two core technologies of employee

assistance programs consisting of EAP liaison with

supervisory management, and EAP liaison with benefits

management. In addressing the employee assistance program

liaison with supervisory management, there are three

subtechnologies: (a) the identification of employee problems

using documentation of impaired job performance, (b)

consultation with supervisors, managers, and/or union

stewards regarding troubled employees, and (c) effective use

of constructive confrontation. Employee assistance program

liaison with benefits management likewise consists of the

following three subtechnologies: (a) micro-linkages of

employees with appropriate services, (b) macro-linkages of

employees with treatment providers, and (c) addressing

alcohol-problem benefits on a parity with other health care

provisions. One might view this core technology as the

philosophical underpinning of the employee assistance


Blum and Roman (1992) have an extensive body of

research that has not only resulted in the development of

the core technology of the employee assistance program, but

which has also resulted in identifying two significant core

functions: (a) management-related strategies, and (b)

benefits-related strategies (p. 121). Each of these core

functions has three components. Management-strategies

include the retaining of valued employees, providing

assistance to troubled supervisors, and providing due

process to those employees whose personal problems have

possibly affected their job performance. The employee

assistance program function components, as related to

benefits strategies, include controlling the cost of health

care utilization, acting as a channeling function for the

employees' and dependents' use of services for substance

abuse, psychiatric and family problems, and acting as an

employee benefit and morale booster.

It is from this philosophical base that utilization of

employee assistance programs through the process of

supervisory and management referral is supported. It is the

ability to identify employees' behavioral problems which

frequently results in "coaching" or "mandated" referrals to

the employee assistance program.

Bystander-Equity Model

A third model that has been used in studies concerned

with supervisory and management involvement in making

referrals to the employee assistance program is the

bystander-equity model. According to Gerstein, Eichenhofer,

Bayer, Valutis and Jankowski (1989), "this model predicts

that the supervisor-troubled worker recognition process

varies as a function of the clarity and severity of the

worker's dilemma, the costs connected with helping, the

extent of inequity supervisors experience in their

relationships with their subordinates, and supervisors'

attitudes about their EAP. This model also suggests that

the identification process is affected by supervisors'

degree of arousal linked with helping" (p. 18).

To appreciate this model, it is important to understand

the social and systemic context out of which helping

behavior originates. One of the areas addressed by the

field of social psychology is that of prosocial or helping

behavior (i.e., altruism). An excellent summary of the

background for the bystander-equity model, is presented in

the chapter, "Prosocial and Antisocial Behavior: The

Psychology of Altruism and Aggression," in a text by

Worchel, Cooper, and Goethals (1988, p. 387-422).

A major study by Latane and Darley (1970), cited in

this chapter, posits that any helping behavior consists of a


series of steps in what they term a decision tree. At each

step in the decision process, an individual who witnesses an

emergency situation is called upon to make an appropriate

judgment if an intervention is to take place. At each step,

however, the individual is faced with other options that

could lead to the possibility that the individual will

ignore the need for intervention. This decision tree,

described by Latane and Darley (1970), consists of the

following basic steps: (a) to notice or not notice a

situation requiring intervention, (b) to interpret or not

interpret a given situation as an emergency, (c) to decide

whether the individual is or is not responsible for

considering an intervention (i.e., based on whether there is

an authority figure present and/or how many people are

present to witness a given event), and to (d) decide on an

appropriate level of assistance (i.e., direct or indirect).

According to this model, if the individual has made the

appropriate decisions consistent with moving toward helping

behavior at each point in the decision tree, the final step

is to implement a decision to help.

In deciding whether the individual is or is not

responsible for considering an intervention, two major

studies are available that address contributing factors

involved in a decision to help. One finding, as a result of

Latane and Darley's research, is that as the number of

individuals (i.e., bystanders) present in an emergency

increase, the less likely it becomes that anyone will step

forward to intervene.

A second study, also cited in this chapter, by Schwartz

and Gottlieb (1980) reports on subjects who observed an

emergency situation under two different conditions: (a)

alone or (b) with another bystander. The results indicated

that eighty percent of those who witnessed a situation and

were alone felt it was their responsibility to act. Only

seventeen percent of those present with another felt they

needed to help. Thus, the presence of another party appears

to significantly diminish an individual's decision to become


What other issues appear to contribute to helping

behavior? Worchel, Cooper, and Goethals (1988) presented

the following contributing factors to helping versus non-

helping bystander behaviors. First, situational context

and how it is interpreted may result in an individual

defining a situation in such a way as to whether they will

or will not help (pp. 400-402). Second, an analysis is made

in terms of the cost of intervention to the bystander and to

the victim if the bystander fails to intervene. "Direct

intervention is expected to occur when the bystander does

not accrue high costs for trying to help and the victim will

suffer great harm if the bystander fails to act" (p. 403).

A third study also cited within this chapter (Piliavin,

Dovidio, Gaertner, & Clark, 1982) suggests three additional

factors that appear to contribute to whether a bystander

will or will not intervene: (a) factors that may affect the

degree of empathy between bystander and victim, (b) any

psychological arousal that is interpreted by the bystander

as being brought about by the victim's distress, and (c) an

actual decision to help as a function of the bystander's

perception of the costs involved (p. 408). Clearly, there

appears to be an investment on the part of the individual

who makes a decision to intervene.

The factors mentioned above are clearly applicable to

the workplace. Throughout the historical development of the

employee assistance program, it has been the role of

supervisors to make a referral to the employee assistance

program when they perceive deficits in an employee's

performance, attendance, or conduct at work that are of

sufficient magnitude and concern to the work organization.

The classic studies mentioned above have contributed

not only to our general understanding of why people help but

have provided a solid basis for the research that was to be

connected specifically to the supervisor. The insights

gained from this body of research are clearly connected to

what makes one supervisor intervene in a troubled employee's

situation while another will choose to ignore it. Studies

that specifically address the bystander-equity model as it

applies to helping behavior on the part of supervisors by

Bayer and Gerstein (1990), Gerstein and Bayer (1991), and


Gerstein, Moore, Duffey, and Dainas (1993) will be reviewed

in further detail in Chapter 2 of this study.

Evaluating EAP Services

As employee assistance programs have increased,

evaluation of services is acknowledged to be an important

aspect of program management. Lubin, Shanklin and Sailors

(1992), in their survey of employee assistance program

publication trends indicate that 1982 was the first year

that more than five articles pertinent to employee

assistance program research were published (p. 47). In

Lubin, Shanklin IV, and Polk's (1996) review of the employee

assistance program literature consisting of "journal, book

and book chapter literature from 1991 through the first half

of 1995," it is reported that "the annual publication rate

for journal articles during the first half of the 1990s

ranges between 35 and 75" (p. 59). By 1996, just utilizing

the key words "employee assistance program evaluation" in a

computer search resulted in 140 entries over the last

thirteen years.

In order to grasp the specific problems in conducting

research related to employee assistance programs, Jerrell

and Rightmeyer (1988) stated that "understanding the basic

objectives, program components, and procedures of employee

assistance programs (EAP) is necessary" (p. 252). In their

review of employee assistance program studies, they

determined most studies have been directed toward either

specific components of employee assistance programs or

monitoring employee assistance program implementation and/or

outcome components such as cost-effectiveness. Some

examples follow, having been gleaned from a review of the


Battle (1988), for example, studied issues to be

considered by a researcher in planning employee assistance

program evaluations. Frost (1990) found employee awareness

to be a first step in the process of utilizing an employee

assistance program and Ahn and Karris (1989) explored the

cost benefits of employee assistance programs. While entire

issues of journals have been devoted to research related to

employee assistance programs, and while it is important to

understand the theory behind and need for such work, outcome

measurement is also a critical element.

Definition of Outcome

Docherty and Streeter (1995) state that "although

'outcome' is often used in a simple and global fashion, it

is actually a complex construct composed of several

independent dimensions" (p. 11). This construct includes

symptomatology (psychiatric and substance abuse), social or

interpersonal functioning, work functioning, satisfaction,

treatment utilization, health status or global well being,

and (g) health related quality of life--value weighted (p.

12). They state that outcome data can be used to "achieve

four main objectives: management of clinical and

administrative operations, regulatory compliance, marketing,

and research" (p. 9).

Consumer Satisfaction

Outcome data, in the form of consumer satisfaction, is

used in this study in keeping with its objective of

research. Assessment of consumer satisfaction with mental

health services was unusual until the late 1970s. Attkisson

and Greenfield (1995) state that now, however, "consumer

satisfaction has achieved the status of an important

measurement domain in health and human service outcome

assessment" (p. 120). According to Lebow (1982) changes

resulted from the following contributing factors: (a) more

frequent use of evaluative approaches to mental health

programs, (b) movement to a more consumer-oriented society,

(c) increased financing of treatment by government and third

party payors, (d) a more complex clientele, and (e) the

simplicity and ease of administration of measures of

consumer satisfaction (p. 244). Lebow (1982) stated that

"measures of consumer satisfaction assess the extent to

which treatment gratifies the wants, wishes, and desires of

clients for service" (p. 244). Three reasons why "few solid

conclusions can be drawn from consumer satisfaction

literature" at the time are cited by Lebow (1982, p. 249)

and include the relatively short history of such research,

the primitive status of most methodology, and concentration

having been placed on assessing satisfaction for entire

facilities without proper attention being given to

specificity. In addition to publication of Lebow's (1982)

review, "research on patient satisfaction has grown rapidly

and served as the subject of several reviews published from

the mid-1970s to 1980" (Pascoe, 1983, p. 185). Then, by the

1990s, since consumer satisfaction had achieved the status

of an important measurement domain in health and human

service outcome assessment, it would seem to be an important

consideration in the employee assistance program field.

The literature indicates that consumer assessment of,

or satisfaction with, care has demonstrated consequences

for health outcomes. LeVois, Nguyen, and Attkisson (1981,

p. 139) tell us that "in the private sector, dissatisfied

health service clients can often seek services elsewhere as

an expression of dissatisfaction." Kaplan and Ware (1989,

p. 40) state that consumers who are dissatisfied with care,

for example, may "engage in activities that disrupt their

medical care and could compromise their health outcomes."

It could be hypothesized that dissatisfaction with services

offered by employee assistance programs could have similar

results. It would be important to determine, for example,

if an individual would return to the employee assistance

program for future assistance if needed regardless of how

the employee came to the EAP (i.e., through supervisory

versus self-referral). This particular question is

addressed in the instrument used in this study (i.e., CSQ-


This study addressed the question of whether consumer

satisfaction would be impacted by the source of referral

(e.g., supervisory versus self). According to Eisen, Grob,

and Dill (1991), "emphasizing the patient's perspective has

unique advantages to evaluators" (p. 213). Two of their

three suggestions to tapping this perspective were included

in this study: (a) recruiting clients as evaluators of

their own progress through self-report, and (b) assessing

satisfaction with various aspects of treatment --what is

termed the consumer model (p. 214).

In this study, consumer satisfaction was measured by

the Client Satisfaction Questionnaire-8 (i.e., CSQ-8) as a

mailed survey to selected employees. The CSQ-8 utilized a

self-report format to "measure satisfaction with services

received by individuals and families" (Attkisson &

Greenfield, 1995, p. 120). It was designed to be used in

studies across a variety of settings including employee

assistance programs.

Self Report

According to Attkisson and Greenfield (1995), "consumer

satisfaction has achieved the status of an important

measurement domain in health and human service outcome

assessment" (p.120). Lebow (1982) indicates that "survey

methods are the most widely used means of gathering data"

(p. 244). According to Kaplan and Ware (1989), "no other

data source currently part of the traditional quality

assessment machinery incorporates patients' values or

preferences in the same way as directly as surveys of

patients' opinions" (p. 26). While surveys may offer the

advantage of direct involvement in the assessment of one's

own care, it is important to recognize that there are

methodological questions to be raised in relation to self-

report. These questions will be addressed in detail in

Chapter 2.

Docherty and Streeter (1995), for example, discuss what

they term the "Rashomon" effect whereby responses to

specific treatment will vary according to the perspective of

the reporting participant (p. 12). This means that each

perspective should be rated independently. In addition,

even when looking at a single person, one needs to consider

issues such as response bias. An excellent presentation of

methodological issues to be considered when dealing with

self-reported data can be found in Borg and Gall (1983).

Alexander (1990), in her review of self-report,

indicates that "the use of self-report has a venerable

history, from introspection through the advent of

behaviorism to the cognitive-behaviorism of today" (p. 2).

She goes on to state that "self-report is one of the best

ways of illuminating the experience of an individual" (p.

2). Any study that uses a self-report measure must take

care, however, to address issues in methodology that may

lead to questionable results. These methodological concerns

will be further addressed in Chapter 2.

Purpose of Study

Statement of the Problem

The purpose of this study was to determine if the

source of referral to employee assistance program services

(i.e., supervisory versus self) would have an impact on the

level of consumer satisfaction. Companies that were

selected for this study consisted of those which had a

broadbrush employee assistance program in place for at least

one year. These particular study requirements allowed for

the supervisors and managers to have gained familiarity

formally (e.g., through training) and informally (e.g.,

through experience) with the employee assistance program

philosophy, practices and providers. Seventeen companies

that met these criteria were collapsed for purposes of data

analysis and to further ensure the confidentiality of the

participants. In addition, this study was ex post facto in

order to allow for "psychological distance" from the actual

reception of services to occur. This concept is discussed

by Docherty and Streeter (1995, p. 13).

Need for the Study

As mentioned earlier, Kaplan and Ware (1989), believe

that dissatisfaction leads to significant problems that

could potentially result in compromising an individual's

care. It can be argued that having knowledge about whether

an individual is satisfied with care would be of particular

importance to an employee assistance program. Satisfaction

has been measured from the perspective of supervisors and

clients (DeFuentes, 1986), as determined by whether the

service was received via face-to-face or telephone contact

(Champion, 1988), or whether service was specific to an HMO

(Pearson & Maier, 1995). To date there has been no

published study using the Client Satisfaction Questionnaire-

8 (i.e., CSQ-8), a standardized instrument, to explore

whether there would be a difference in an employee's

satisfaction with employee assistance program services based

on the source of referral (i.e., supervisor or self). This

is particularly important since a percentage of the employee

population will utilize employee assistance program services

only when they are referred by their supervisors.

Research Questions

This study investigated and compared differences in

satisfaction based on the referral source bringing the

employee to the EAP (i.e, supervisory versus self). In

addition, this study investigated possible between group

differences in satisfaction based on gender, and drug or

alcohol being the presenting problem as compared to other

concerns. The following five hypotheses, formulated on the

basis of previous research, were tested:

Hypothesis 1. There will be a significantly greater

percentage of self referrals to the employee assistance

program as compared to employees who come in under

supervisory referrals. A body of research from diverse

settings has emphasized the prominence of self-referrals to

employee assistance programs. Backer and O'Hara (1991), for

example, support the view that the majority of employees who

seek help do so through self-referral.

Hypothesis 2. Women will be significantly more likely

than men to self-refer to the employee assistance program.

Brodzinski and Goyer (1987), in their study on employee

assistance program utilization and gender, determined that

women were more likely to self-refer than men. Hall, Vacc,

and Kissling (1991) also found that women were more likely

to self-refer. Blum and Roman (1992), in a study cited

earlier explored client utilization of employee assistance

program services and found that women were less likely than

men to be referred by supervisors.

Hypothesis 3. There will be a higher mean level of

satisfaction from self-referral as compared to supervisory

referral. Keaton (1990), studied the effect of voluntarism


on treatment attitude in relationship to previous counseling

experience in an employee assistance program. Results,

consistent with cognitive dissonance theory, showed a

significant relationship between voluntarism and viewing the

counseling in a positive light. While the goal of one's

participation in the employee assistance program is problem

resolution, the fact that a person is told to participate

may set up a situation of cognitive dissonance leading to

less satisfaction with the help received by the employee.

Keaton's (1990) results "confirmed the hypothesis that

voluntary participants have a more positive attitude

regarding the treatment situation than involuntary

participants" (p. 64). Pearson and Maier (1995) also found

that self-reported improvement was negatively related to

being referred by a supervisor.

Hypothesis 4. There will be no difference in mean

level of satisfaction as determined by gender. Lebow (1982)

stated that research up to the time of his study had

consistently demonstrated that demographic characteristics

were not good predictors of satisfaction. While it appeared

that satisfaction was unrelated to gender, Lebow did

encourage more research to examine the relationship of

client variables to satisfaction. Over ten years later,

Attkisson and Greenfield (1995) reiterated that future

research should control for variables they viewed as

functioning as covariates of satisfaction. They stated that


gender effects are perhaps the most important candidates as

covariates that may contribute variance to satisfaction


Hypothesis 5. Mean level of satisfaction will be lower

for those who present with drug problems as compared to

those who present with other problems. Lebow (1982), for

example, cites several studies and states that "satisfaction

has variously been found to be lower for drug abusers" (p.



This study investigated differences in satisfaction

level based on the referral source for participants in their

employee assistance program in terms of number and general

level of satisfaction. In addition, between group

differences based on gender and whether the presenting

problem was drugs/alcohol or some other concern was


In order to provide a conceptual framework for

evaluating previous research, a brief overview of

definitions will precede the literature review. In this

study, the following definitions were applicable:

Employee Assistance Program (EAP) A worksite-

sponsored program for employees that is broadbrush in nature

and which covers a wide range of emotional and human service

needs. For this study, only employee assistance programs

that have had a minimum of a full year of operation were


"Welfare capitalism" A sociological phenomena

described by Brandes (1976) whereby companies, in order to

avoid unionization of employees and assist in the process of

acculturation for new immigrants, began to offer company-

sponsored social services in the 1800s.

Occupational Alcoholism Program The more formal

beginnings of the movement toward employee assistance

programs. Occupational alcoholism programs were sponsored

by companies to initially address alcohol problems being

experienced by employees that the company viewed as

resulting in an impact on the employees' work performance,

attendance, and conduct.

Broadbrush A term that refers to the expansion of

services offered by employee assistance programs to include

areas other than drugs or alcohol. Such additional services

typically include emotional, financial, relationship, legal

and other services.

Self-referral A visit to the employee assistance

program that is initiated by the employee after having been

exposed to information garnered from brochures, company

sponsored orientations, posters, co-worker recommendations,

family members) or the like. In this case the employee

makes a voluntary choice to participate in employee

assistance program services.

Supervisory-referral A visit to the employee

assistance program that is initiated by a supervisor or

manager due to on-the-job concerns related to an employee's

productivity, attendance or conduct decline or reasonable

suspicion of substance use. In this study, due to available

sample size, no distinction was made between referrals made

by a direct supervisor, a manager, or a member of Human


Constructive confrontation The process by which a

supervisor, who has noted a decline on the part of the

employee in work performance, attendance, and conduct, or

who has reasonable suspicion of an alcohol or other drug

problem, takes appropriate steps to confront the employee.

The supervisor focuses on the following issues within the

context of a confrontation with the employee: (a) deficits

observed by the supervisor regarding work-related issues,

(b) documentation of both verbal and written interventions

regarding the deficits, and (c) specific steps taken by the

supervisor as part of coaching or disciplinary procedures in

order to remediate the concerns. It is at this point that

the supervisor develops a behavioral plan with the employee

and makes a referral to the employee assistance program.

Satisfaction The direct evaluation by the consumer of

services received from the employee assistance program

through the utilization of a standardized instrument, the

Client Satisfaction Questionnaire-8.

CSO-8 The Client Satisfaction Questionnaire-8,

developed by Attkisson and Greenfield (1995), is a

standardized self-report questionnaire designed to measure

satisfaction with services received. This instrument

required only one administration. The CSQ-8 is suggested

for evaluating services received from a variety of settings

including an employee assistance program.

Outcome As related to mental health care, it

comprises seven dimensions including: (a) symptomatology

(psychiatric and substance abuse), (b) social and

interpersonal functions, (c) work functioning, (d)

satisfaction, (e) treatment utilization, (f) health status

and global well being, and (g) health-related quality of

life--value weighted (Docherty & Streeter, 1995, p. 12).

Organization of Study

This study was organized into five chapters. Chapter 1

provides an introduction and rationale for the purpose

of this study. A review of the problem being studied is

presented leading to the development of the research

questions and hypotheses under consideration. Chapter 2

presents a review of the literature covering the following

areas: (a) historical development of employee assistance

programs, (b) key elements of an employee assistance

program, (c) employee assistance program research

strategies, (d) outcome studies, (e) satisfaction studies,


(f) issues related to self-report, (g) supervisory referral,

and (h) self-referral studies. Chapter 3 describes the

methodology of the study which includes: (a) participants,

(b) variables (independent and dependent), (c) instruments,

(d) demographic information, and (e) procedures. Chapter 4

discusses the results of the study. Chapter 5 presents a

discussion to include information regarding limitations of

the study and future research.


Historical Development

Employee assistance programs (i.e., EAPs) cannot be

described in terms of a single entity that, in their

essential elements, have remained unchanged throughout

history. Employee assistance programs span a period of

development from the informal to the formal. Practitioners

today would not recognize the initial efforts as being a

precursor to what was eventually to become the employee

assistance program (i.e., a continually-evolving service

that strives to address more and more of the issues critical

to a productive, healthy workforce). From the welfare

capitalism of the 1800s and occupational alcoholism and

social welfare programs of the late thirties and forties,

through the broadbrush programs of today, it is change that


In a general sense, the basis for what was to

eventually develop into employee assistance programs (i.e.,

EAPs) as we know them today, began with a social movement of

the late 1880s referred to as welfare capitalism. Following

this period, Midanik (1991) states, however, that "two major

complex and intertwined historical threads, occupational

social welfare and occupational alcohol programs, have led

to the development of EAPs" (p. 69). In order to get a

clear grasp of the concept and purpose of employee

assistance programs, it is important to develop an

understanding of these historical trends.

Occupational Social Welfare

Brandes (1976) in, American Welfare Capitalism,

indicates that it was the influx of immigrants into the

United States that created a need for industry to both

defend corporate interests and assist immigrants and their

families with the process of acculturation in the United

States. In order for industry to defend its own self

interest, it needed to prevent the organization and

development of unions. As a result, companies began to

provide multiple services that we would today term "social

services". Kotschessa (1994) states that "the concept of

welfare capitalism refers to the voluntary provision of

benefits and services by employers in order to retain,

control, and socialize the rapidly growing workforce at a

time when workers were desperately needed" (p. 65). While

these company-sponsored services appeared to be efforts

directed to help the newly-arrived immigrants and their

families adapt to a new culture, and in fact were helpful in

that regard, it was also a self-serving effort on the part

of industry. According to Kotschessa (1994), specialists

known either as "social secretary" or "welfare secretary"

were hired by companies to "attend to the physical,

cultural, personal, and economic welfare" of employees and

family members (p. 65). The provision of these company-

sponsored services lasted until approximately the Depression

of 1929. Following the Depression, government-sponsored

programs began to develop. It is probably this trend that

has led to our understanding of a broadbrush employee

assistance program.

A second trend in the formal movement toward what we

would today call an employee assistance program occurred

with the development of occupational alcoholism programs

(i.e., OAPs). Although this programmatic trend is

historically and widely reputed to coincide with the

beginnings of Alcoholics Anonymous, there are reports of

more informal efforts on the part of industry occurring

prior to that time. One of the earliest is reported by

Dubreuil and Krause (1983, p. 86). The example cited is a

Litchfield, Connecticut corporation that in 1789 issued a

pledge to discontinue supplying distilled spirits to its


Alcoholics Anonymous

With the exception of some isolated efforts, such as

that cited above, formal movement on the part of industry

toward the development of employee assistance programs is

usually reported to coincide with the beginnings of

Alcoholics Anonymous (i.e., AA) which started in 1935. The

teaching and structure of Alcoholics Anonymous quickly

spread throughout the United States (Dickman & Challenger,

1988). This growing support movement eventually began to

exert its influence by making a major impact on industry's

view of and approach to dealing with employees who

experienced problems with alcohol. Trice and Schonbrunn

(1981), in their review tracing the history of job-based

alcoholism programs from 1900-1955, describe the predecessor

of employee assistance programs as evolving from workers

having problems with alcohol informally sharing with others

who had similar problems.

Occupational Alcoholism Program

A program, then called occupational alcoholism program

(i.e., OAP), developed to deal exclusively with employees

experiencing alcohol problems. Trice and Schonbrunn (1981)

discuss three factors they believe to have been influential

in the development of occupational alcoholism programs at

that time: (a) the emergence and development of alcoholics

anonymous (AA), (b) concerns of physicians working within

industrial settings, and (c) an increased need for workers

during a period of war involvement in this country. This

latter factor demonstrates the interplay of socio-cultural

factors on business and industry that became further evident


around the time of World War II. It was during wartime that

the availability of employees for non-military activities

was limited so that workers with alcohol problems resulting

in decreased productivity or who developed attendance or

conduct problems became more noticeable.

Evolution from OAPs to EAPs

Although there are reported informal, or "quasi-

private", attempts during or after World War II to assist

employees with alcohol problems, researchers typically

consider E. I. DuPont de Nemours and Company and Eastman

Kodak Corporation, Allis Chalmers, and Consolidated Edison

to be the pioneers of more formal approaches toward the

development of work-sponsored programs in the 1940s (Archer,

1977; Trice & Schonbrunn, 1981). There are several good

historical reviews available which address the development

and growth of occupational alcoholism programs.

A classic article by Presnell (1967), discusses the

early expansion of occupational alcoholism programs. Archer

(1977) provides a description of the historical development

of occupational alcohol programs from the early 1940s to the

early 1970s. She says that "the major thrust of arguments

to persuade companies to adopt programs was that alcoholism

was a health problem that primarily afflicted individuals

who were in their middle service years, and hence persons in

whom the company had a large investment" (p. 4). Archer

(1977) indicates that, "since the early programs, the

supervisor's role has been restricted to that of documenting

absenteeism and decreased work performance instead of

attempting to distinguish the signs and symptoms of

alcoholism" (p. 5). She goes on to state that "when

confronting the employee with his declining work

performance, the supervisor informs the worker that help

will be made available if he cannot himself deal with the

source of his problem" (p. 5). In addition, Archer analyzes

the expansion of occupational alcohol programs toward what

later became known as "broadbrush programs".

Limitations of Occupational Alcoholism Programs. It

should be noted that services provided to employees by

occupational alcoholism programs were limited. Some of

these limitations as listed by Nye (1990) include the

following: (a) the difficulty experienced by the workplace

in recognizing early-stage alcoholism, (b) employees with

chronic alcohol problems who were viewed as untreatable, (c)

programs that relied almost exclusively on supervisors to

recognize symptoms resulting in supervisors making more

referrals of rank-and-file members, (d) failure to identify

the alcohol problems of executives and managers, and (e)

employees with alcohol problems who tended to hide their

problems so that they only became noticeable at later stages

of development (pp. 2-3). As these limits came to light,

awareness developed that changes were necessary.

Kemper Group. In 1962, the Kemper Group developed a

program that not only addressed the problem of alcoholism

but expanded their program to address other life problems as

well. According to Dickman and Challenger (1988), this

initiative considerably changed occupational alcoholism

programs and resulted in an evolutionary leap toward our

modern employee assistance programs (p. 49). This occurred

well before NIAAA's rather strong endorsement of the need to

develop "broadbrush" programs in the early 1970s.

The Hughes Act and NIAAA. In 1969, Senator Harold

Hughes of Idaho became concerned about the failure of both

federal and state government to be concerned regarding the

problem and treatment of alcoholism. In 1970, Congress

passed a Federal Comprehensive Alcohol Abuse and Alcoholism

Prevention and Rehabilitation Act (Public Law 91-616). An

outcome of this Act was the creation of the National

Institute of Alcoholism and Alcohol Abuse (i.e., NIAAA).

According to Dubreuil and Krause (1983), cited earlier, the

NIAAA was created to develop leadership and handle monies

aimed at alcohol prevention and treatment. According to

Archer (1977), "before the creation of NIAAA, occupational

alcoholism policies and programs, with few exceptions, were

designed to provide assistance solely to the problem-

drinking employee" (p. 8). Archer continues to state that,

on the basis of its survey of existing programs and of

knowledge in the field, NIAAA wound up endorsing the

"broadbrush" approach "as the ideal strategy for

occupational alcoholism programs" (p. 8). Another goal of

NIAAA was to recognize that alcoholism was a disease and not

a moral or legal issue. According to Archer (1977), the

disease model of alcoholism was pioneered by E. M. Jellinek

and officially accepted by the American Medical Association

in 1956. By 1971, most states had followed the lead of the

federal government and enacted legislation resulting in the

recognition of alcoholism as a disease. It should be noted,

however, that the disease concept so readily embraced by

practitioners and occupational alcoholism specialists, was

not without its critics. Archer (1977) presented numerous

studies that considered the disease concept as demeaning.

Archer states that "by contrast, the approach now

recommended by NIAAA's Occupational Programs Branch relies

more heavily on the sociocultural model" (p. 12).

Steele (1989), another historical reviewer, discussed

the history of job-based alcoholism programs from 1955

through 1972. In a second study, Steele and Trice (1995)

traced the history of job-based alcoholism programs from

1972 through 1980 and discussed the influences of

professional organizations such as Alcoholics Anonymous and

the National Council on Alcoholism. By reading these three

articles, one would wind up with a fairly good understanding

of both the development and contributing influences on

occupational alcoholism programs.

Prior to the development of the NIAAA, two general

practices were acceptable in occupational alcoholism

programs. First, alcoholism could best be identified if

supervisors were trained and monitored employee work

performance. Second, if reasonable steps to improve

performance proved to be ineffective, an offer of

professional assistance should accompany constructive

confrontation (Roman, 1988, p. 58). NIAAA adopted and

encouraged the practice of these provisions as efforts were

expanded to include a "broadbrush" approach (Roman, 1988, p.


According to Roman (1988), it was NIAAA that coined the

phrase "occupational alcoholism" to describe programs that

developed with this as their focus (p. 57). The use of this

term was intended to de-emphasize the public's stereotypical

perception of alcoholism as pertaining only to individuals

drinking out of brown bags under bridges and living on the


Project 95. Under the sponsorship of NIAAA, "Project

95" became a major enterprise. Project 95 had as its

specific purpose the identification of and assistance toward

an estimated 95 percent of alcoholics not found on skid row.

In other words, NIAAA attempted to draw attention to the

larger percentage of individuals who were employed but who

also experienced problems with alcohol. Other efforts of

the NIAAA included sponsoring research and providing grant

monies to states to hire and train specialists in the area

of alcoholism. The establishment of employee assistance

programs, although not required at that time, was


ALMACA. In 1971, the Association of Labor and

Management Administrators and Consultants on Alcoholism

(i.e., ALMACA) was established. This group is today called

EAPA (i.e., Employee Assistance Professional Association)

and has local chapters both within the United States and

abroad. Today, "EAPA, an organization with a membership of

over 6,000, has adopted standards for EAPs that specify a

comprehensive set of services" (Blum & Roman, 1995, p. 1).

The Employee Assistance Professional Association provides

disseminatation of educational and training materials and

provides seminars and workshops to aide in the professional

development of individuals providing employee assistance

services. In the 1980s, EAPA became a driving force behind

the development, testing, and monitoring of the CEAP (i.e.,

Certified Employee Assistance Professional) test and

credential. As can be noted by the name change, this group

recognized the development of services within corporate

America that reflected the ever-expanding role of

professionals and companies from offering alcohol only to

broadbrush employee assistance programs.

1980s. In the 1980s, employee assistance programs

continued to develop both in terms of the number available

and of their becoming more comprehensive in nature. During

this period, major changes occurred in both employee

assistance program development and services offered. As

mentioned above, plans began to develop to certify and

credential EAP practitioners. This required clarification

regarding role descriptions and functions of employee

assistance program professionals. Also, unions began to

become more involved with and more accepting of employee

assistance programs. It was also becoming evident that some

companies offered employee assistance programs that

basically existed only on paper. Researchers began to

address the question of what made an employee assistance

program effective.

By 1987, the Bureau of National Affairs considered the

development and expansion of employee assistance programs to

be important enough to merit national attention. The Bureau

published a seminal document addressing a wide variety of

EAP-related issues. A key point discussed is that in almost

all instances an employee assistance program is maintained

after it is established. Studies undertaken since then

appear to confirm this point. Seid (1991) states that "the

permanence of an EAP may be attributed to reasons other than

proven efficiency, effectiveness, and high levels of

of satisfaction", (p. 3723). Another recent study by Oher

(1993) measured employee assistance program consumer

satisfaction by surveying 55 Human Resource managers. In

discussing implications for employee assistance programs,

Oher (1993) states that "as the results from this corporate

survey indicate, many aspects of EAP service that reflect

firmly held beliefs may not be present in programs that are,

none the less, perceived as worthwhile" (p. 74).

Drug-Free Work Place Act. A further contributor to the

development and growth of employee assistance programs

occurred in 1988 with the enactment of the Drug-Free Work

Place Act (Nye, 1990). This Act required that employers

who receive federal grants and/or contracts over $25,000.00

must maintain a drug-free workplace environment. According

to Nye (1990) a company must take a number of steps in

establishing a drug-free workplace and in coming into

compliance with the Drug-Free Work Place Act: (a) publish a

statement notifying employees of what is prohibited and the

action to be taken regarding any violations, (b) establish

an awareness/educational program for employees, (c) give

every employee a copy of the drug-free statement, (d)

indicate that employees must abide by the policy as a

condition of employment, (e) require satisfactory completion

of any needed abuse assistance or rehabilitation program by

anyone needing such help, and (f) continue to implement the

requirements as indicated by the Drug-Free Workplace Act

(pp. 223-224). A key element that is strongly recommended,

although not required, is the availability of employee

assistance program services (Cunningham, 1994, p. 12).


In the early 1980s, major attention was being directed

employee assistance programs in terms of research. While

many earlier studies were not experimental in design and

possessed numerous methodological flaws, as shall be

discussed later in this section, it should be noted that

researchers were beginning to realize that employee

assistance programs offered a fertile field for study.

Researchers became interested in addressing the key

components of an employee assistance program, determining

whether EAPs rendered effective services, whether certain

components were more effective than others, and in clearly

defining EAP technology and functions.

Core Technology of Employee Assistance Programs

Core Elements

This early research eventually led to the development

in the early 1980s of the beginnings of what was later to be

called the core technology of employee assistance programs.

Roman (1989) and Blum and Roman (1989) have since identified

six core elements that they refer to as the "core technology

of employee assistance programs". These core elements

include (a) employees' behavioral problems being identified

as a result of impaired job performance issues, (b) the

provision of expert consultation to supervisors, managers,

and union stewards (where applicable) on when and how to

apply and use employee assistance program policies,

(c) appropriate use of constructive confrontation, (d)

micro- and (e) macro-linkages of the home organization with

counseling, treatment, and other community resources, and

(f) mainstreaming alcohol-problem benefits on a parity with

other health care in a milieu supportive of alcohol

prevention and intervention. Blum and Roman (1992) believe

that these core elements for employee assistance programs

can be "addressed only by providing comprehensive services"

(p. 120).

Core Functions

In addition to the core technology mentioned above,

Blum and Roman (1992) also developed what are termed

employee assistance program functions. These functions are

divided into two main categories to include strategies

specifically related to management and strategies that are

related to benefits. The management-related strategies

include the retention of valued employees, provision of

assistance to troubled supervisors, and the provision of due

process for employees whose personal problems may be having

an impact on their job performance. The benefits-related

strategies include controlling the costs involved with

health care utilization, serving as a channeling function

for employees' and dependents' utilization of services under

the umbrella of employee assistance, and serving as an

additional employee benefit and morale booster (p. 121).

Key Elements of an Employee Assistance Program

In order to understand how an employee assistance

program functions from a systemic perspective, it is

essential to review what are considered to be the key

elements that any effective employee assistance program must

possess. Dickman and Emener (1988), following their review

of the literature, have determined that there are ten

attributes considered critical in any successful employee

assistance program. These elements exemplify the importance

of reaching all parts of an organization in terms of

"backing" or support, union involvement (if applicable) and

training. Each of the following critical elements, as

presented by Dickman and Emener (1988), reach across the

broad spectrum of an organization: (a) backing of

management at the highest level, (b) support of labor,

(c) confidentiality, (d) ease of access, (e) availability of

supervisory training, (f) labor steward training, (g)

involvement of insurance plan, (h) availability of a

broadbrush program, (i) professional leadership on the part

of the employee assistance program, and (j) follow-up and

evaluation to measure program effectiveness (pp. 279-280).

These key elements have been determined to be present

in the companies that were selected for this study. Each of

the seventeen companies, from which the employee samples

have been drawn, have had a minimum of one year receiving

services from the employee assistance program providers.

Companies selected for this study have demonstrated backing

of leadership at the highest levels and active involvement

of the employee assistance program professionals. Each

company has ensured that their Human Resource Department,

managers at all levels, and supervisors have received

appropriate training regarding proper utilization of the

employee assistance services within the context of a

broadbrush program. Individual employees have had ease of

access and confidentiality has been stressed in their

orientation programs. These factors are reiterated in the

promotional materials distributed to employees. In

addition, confidentiality is emphasized on client

information forms and in the initial assessment session.

This study, designed as an ex post facto survey, hoped

to achieve two goals. First, given the still developing

state of employee assistance program research, it hoped to

contribute to the existing literature on EAP effectiveness.

Second, it has helped to determine if any differences in

satisfaction exist based on source of referral.

EAP Research Strategies

Seid (1991) states that "progressive companies are

currently providing a variety of health care services to

combat employees' personal problems that affect productivity

... however, most of the anecdotal, cost-benefit, and survey

data presented can be challenged on methodological grounds"

(p. 3723). Harris and Heft (1992) confirm that although

utilization of employee assistance programs has grown

dramatically since their inception, "there has been little

rigorous research on this topic" (p. 253). The researchers

state that this has been a major source of criticism

directed at studies of employee assistance programs. They

claim that no rigorous, systematic study exists and that

most of the available research contains significant design

flaws (p. 255).

A majority of the studies assessing employee assistance

programs consist of case studies or anecdotal reports which

are typically descriptive in nature (e.g., they describe the

organization in terms of number of employees, type of

industry, development of the employee assistance program

etc.). DeFuentes (1986), in her study evaluating an

aerospace employee assistance program, for example, used a

combination of record review and interviews to determine the

degree of employee assistance program implementation.

Berman, Sulsky, Pargament, and Balzer (1991) in, "The Role

of Needs Assessment in the Design of Employee Assistance

Programs: A Case Study," used a multi-method needs

assessment of 250 employees. Data was gathered for this

case study from a review of employee records, interviews

with key leaders and employees, and an employee survey.

McKibbon (1993), in his review of employee assistance


programs in Canada, found that over 80 percent of studies in

the prior five years used anecdotal evidence as a way of

providing a description of employee assistance programs.

Finally, Csiernik (1995), in his review of the literature

focusing on employee assistance program delivery options,

found that 24 out of 48 studies reviewed consisted of a case

study approach. Few of the studies used cost-benefit

analysis, process evaluations or needs assessments. Studies

have ranged from the specific components of employee

assistance services (e.g., smoking cessation programs,

worksite health promotion, effects of company-sponsored

relocation counseling) to determining who is more likely to

utilize services. In terms of evaluating cost outcomes, a

methodology based on standard economic theory is available.

French, Zarkin, and Bray (1995) discuss the four components

of this methodology and indicate that researchers should try

to incorporate each of them. These four components include

(a) a process description of employee assistance program

structure, environment, and goals, (b) an analysis of

employee assistance program costs, (c) an analysis of

employee assistance outcome in terms of work-related

performance and productivity, and (d) an economic evaluation

of cost-effectiveness ratios, cash benefits, and net

benefits. A significant amount of research has been

directed to the cost-effectiveness of employee assistance

programs. Other methodologies need to be developed.

Docherty and Streeter (1995) posit that "a

sophisticated, comprehensive outcomes measurement system

will not only allow us to effectively respond to ...

external forces, but such information will also provide the

necessary tools to systematically improve the quality of

mental health care" (p. 8). According to the same authors,

outcome "is actually a complex construct composed of several

independent dimensions" (p. 11).

Outcome Studies

Definition of Outcome

According to Eisen (1996), "the field of mental health

outcomes research has produced an enormous body of

literature encompassing many professional disciplines,

employing a wide range of methodologies, and incorporating

many definitions of 'outcomes'" (p. 71). As mentioned

above, Docherty and Streeter (1995) define outcome as "a

complex construct composed of several independent

dimensions" (p. 11). They present seven dimensions of

outcome that include symptomatology (psychiatric and

substance abuse), social and interpersonal functioning, work

functioning, satisfaction, treatment utilization, health

status and global well being, and health-related quality of

life--value weighted (p. 12).

According to Eisen, Grob and Dill (1991), "treatment

outcome and satisfaction are not strongly correlated" (p.

230). "Low but statistically significant correlations

between improvement and satisfaction suggest that although

there is some tendency for those who are doing well to be

more satisfied with the care received, satisfaction cannot

be predicted very accurately from knowledge of a patient's

improvement or level of functioning" (p. 230).

EAP outcome indicators. According to Jerrell and

Rightmyer (1988), there are four categories of outcome

indicators in employee assistance program evaluation: (a)

accidents, sick leave, medical-surgical costs and insurance

premiums, (b) absenteeism, tardiness, and job inefficiency,

(c) rehabilitation rates, and (d) job performance ratings,

grievances, disciplinary actions and labor arbitration

incidents (p. 260). What is missing in these categories, is

the place of the consumer. According to major researchers

in the area of consumer satisfaction, "emphasizing the

patient's perspective has unique advantages to evaluators"

(Eisen, Grob & Dill, 1991, p. 213).

Consumer perspective. Eisen, Grob and Dill (1991)

posit that a client's perspective can be emphasized and

explored in three ways. First, by designing outcome

measures around treatment goals defined by the clients.

Second, by recruiting clients as evaluators of their own

progress through utilizing self-reports. And, third, by

assessing client satisfaction with varying aspects of their

treatment--what is described as the consumer model (p. 214).


Kaplan and Ware (1989) state that "one of the thornier

issues currently facing practitioners, medical educators and

health policy makers is the definition of a reasonable and

appropriate role for patients in the medical care process"

(p. 25). In reviewing the research and practical issues

related to client role, Kaplan and Ware (1989) address two

client roles that they view as being distinct: (a) "as

evaluators of care, both directly and indirectly, supplying

information used by others in evaluating care", and (b) "as

active participants in care, shaping the nature of the care

they receive" (p. 25).

This review reinforces involving client's in their care

for the following reasons. First, clients are a practical

source of information. Second, client satisfaction or

assessment of care has been demonstrated to affect both the

therapeutic relationship and the client's health status.

According to Kaplan and Ware (1989), clients who are

dissatisfied may physician-shop, change health plans, sue,

not follow treatment recommendations, and avoid further

visits. Third, research has shown a link between clinical

and client-reported measures of health outcomes. Fourth,

clients appear to desire an expanded role in their own care.

Fifth, clients who are dissatisfied with or not permitted to

participate in their care may turn to self-care. Sixth,

there are underway a number of large-scale data collection

efforts to gather clients' assessment of their care.


While this review addresses patients' medical care, it

does consider five behavioral health concepts which include:

(a) physical health, (b) mental health, (c) social

functioning, (d) role functioning, and (e) general well-

being (pp. 26-27) each of which are relevant to employee

assistance program services and satisfaction. It is

important to remember that employee assistance programs are

not limited exclusively to mental health concerns even

though a fairly large proportion of presenting problems fall

within that category.

Other subcategories listed by Kaplan and Ware (1989)

are also of concern to employee assistance programs. These

include anxiety and depression, psychological well-being,

behavioral and emotional control, cognitive functioning,

social and interpersonal contacts, social resources, and

role functioning (pp. 43-47).

Issues in Outcome Research

There are major issues to be considered when measuring

outcome. In their paper, Docherty and Streeter (1995)

emphasize three of them.

Rashomon Effect

Docherty and Streeter state that the source of outcome

data is important and affects results. They discuss the

"Rashomon effect"-that is, outcome will vary according to

the perspective of the person doing the reporting. For

example, a client will most probably "rate outcome according

to the experienced change in subjective state" (p. 12). A

therapist or employer who is asked to evaluate the same

outcome might look at something completely different.

Conte, Ratto, Clutz and Karasu (1995), for example, explored

138 outpatients' level of satisfaction with their

therapists. Patient satisfaction was related to measures of

psychotherapy outcome as derived from patients themselves,

therapists, and an independent rater. Several specific

therapist characteristics and an overall rating of

satisfaction appeared to be significantly correlated with

the patients' ratings of self improvement, help received,

and therapist ratings of outcome.


A second factor in outcome assessment is timing. This

is particularly important in the area of measuring consumer

satisfaction. If a researcher assesses too closely to the

time clinical services are provided, one might find a

"positively biased response because of patient concerns

regarding retaliation, anonymity, or other demand

characteristics of the immediate social context" (Docherty &

Streeter, 1995, p. 13). While having a client fill out a

satisfaction questionnaire immediately following reception

of EAP services might increase survey return rates, research

indicates that allowing adequate psychological distance is

an important issue to consider for the reasons cited above.


A third issue involves the population under study.

Demographic information can "greatly enhance the value of

the data and expand the analyses that can be conducted"

(Docherty & Streeter, 1995, p. 13). This study presents

multiple areas of demographic information which is taken

from the Client Information Form (see Appendix B) filled out

just prior to reception of employee assistance program



Docherty and Streeter (1995) indicate that clinical

outcome data can be used to reach four main objectives.

First, outcome data can affect change in the management of

clinical and administration operations. Second, data can

assist in identifying areas for change in order to maintain

regulatory compliance. Third, data can provide support for

marketing claims and information. Last, and perhaps most

important, in terms of this study is the contribution to

research in terms of what works and what does not (p. 9).


Kiesler (1983) states that methodological and

conceptual issues have arisen because prior research on

client satisfaction has not been driven by theory. Kiesler

proposes that there be a more informed use of attitude

change and attitude measurement technology in this area.

According to Jerrell and Rightmyer (1988), "except for

Schramm's (1980) 'human-capital model of analyzing the

costs and benefits of an EAP, there is little theoretical

guidance for developing or evaluating these programs" (p.

269). In addition to human capital theory, however, two

other theories also contribute to one's understanding of

employee assistance programs; namely, systems theory and the

bystander-equity model.

Human Capital Theory. This theory developed by Carl J.

Schramm places a company's provision of employee assistance

program services within the same context with other capital

improvements. According to this theory, an employer invests

in procurement of land, equipment, and other material

elements in order to build the business and increase

revenues. According to Schramm (1980), companies now view

employees as being as precious a commodity as some material

acquisition. If an employee is happy, healthy, and free of

unnecessary degrees of stress, it is anticipated that the

employee will be more productive. There is documented

evidence to indicate that companies that invest in an

employee assistance program find employees who utilize

services have fewer absences, experience lower

hospitalization rates for problems identified by the

employee assistance program, require less informal sick

time, and require fewer disciplinary actions (thus freeing

up supervisor's time and perhaps lowering their stress

levels). In the study cited earlier, De Fuentes (1986)

supports the position that an employee who addresses

personal and work-related problems through participation in

an employee assistance program often winds up, at the very

least, being available more often at work. The employer

does not have to spend additional monies hiring and training


In some instances, referral of a troubled employee to

the employee assistance program results in retention of what

had been a valued employee prior to the onset of personal or

work-related concerns. In this case, the employer is saved

having to replace the employee entirely which would have led

to a significant amount of unproductive time while an

employee search and retraining occurred. According to this

model, therefore, whatever an employer can do to help retain

and maintain a productive employee represents a positive

outcome for their investment.

Systems Theory. Clearly, one familiar with systems

theory can see the interaction of various micro- and macro-

components of organizational and family systems in the

employee assistance field that drives the core technology

and functions, as presented by Roman and Blum (1989) earlier

in this study. An employee assistance professional is

accustomed to interacting on a variety of levels with a

complex network of individuals. These range from the

employee alone to possibly involving a supervisor, a

manager, a union steward, or a member of the Human Resource

Department. While an employee may come in through self-

referral, according to systems theory, others are still

involved in the network of the person's experience. This is

even more evident when an employee is referred by their work

organization. Clearly, in the latter instance, there is

concern related to performance, attendance, or conduct

problems that have a rippling effect on others within the

environmental sphere of the troubled employee.

According to Ford and Ford (1986), "the essential

principles of systems theory can provide a conceptual and

practical guide for the EAP administrator and case manager"

(p. 37). "The systems principles of nonsummativity,

nonlinear causality, feedback, dynamic equilibration,

equifinality, and permeable boundaries provide a conceptual

backdrop for practical guidelines" (p. 47) in helping in the

development of an action plan for an employee to establish a

better balance in all relevant life areas.

Bystander-Equity Theory. This model, derived from

social psychology, is important particularly in a study such

as this. In the first chapter of this study, background

information was provided that established a basis for

understanding what leads to helping behavior on the part of

a bystander. Classic studies such as those of Latane and

Darley (1970) along with Schwartz and Gottlieb (1980) were

discussed in order to provide the reader with information

regarding the decision tree leading to helping behavior and

to understand how the presence of an authority figure or

someone else who could help tends to inhibit the helping

response. Since this study compared supervisory and self-

referral to the employee assistance program, it was

important to have an understanding of what factors

contribute to a supervisor making the decision to refer a

given employee. As is discussed in the section on

supervisory referrals later in this chapter, a number of

studies have identified variables that appear to impact a

supervisory referral.

Supervisory involvement is a critical employee

assistance program core element in the appropriate referral

of troubled employees. This study addressed whether the

involvement of others in one's participation in the employee

assistance program does, in fact, have any effect on the

level of satisfaction.

Consumer Satisfaction

According to Docherty and Streeter (1995), "research

has suggested that patient satisfaction is often independent

of clinical outcome" (p. 12). As discussed earlier in this

study, these authors state that "within the scientific

literature for mental health care, the seven main dimensions

of outcome are symptomatology (psychiatric and substance

abuse), social and interpersonal functioning, work

functioning, satisfaction, treatment utilization, health

status/global well being, and health-related quality of

life--value weighted" (p. 12).

Definition of Satisfaction

Lebow (1982) states that "measures of consumer

satisfaction assess the extent to which treatment gratifies

the wants, wishes, and desires of clients for services" (p.

244). Kalman (1983) defines satisfaction as a composite of

many variables. Williams (1994) states that there are a

number of implicit assumptions about the nature and meaning

of the term satisfaction.

Pascoe (1983) who presents various models for the

conceptualization of satisfaction and states that "patient

satisfaction has not been explicitly guided by a well-

supported definition or psychological model of satisfaction"

(p. 185). According to Pascoe, "the major exception to a

lack of psychological theory-building in patient

satisfaction research is the model recently proposed by

Linder-Pelz" who "characterizes patient satisfaction as a

positive attitude" (p. 185). The attributes of the Linder-

Pelz model "are distinct dimensions of health care, such as

access, efficacy, cost, and convenience",, (p. 186).

Pascoe (1983) also discusses discrepancy theory and

fulfillment theory. According to Pascoe, "fulfillment

theories define satisfaction as a function of the amount

received from a situation regardless of how much one feels

they should and/or want to receive" (p. 186). The second

model, discrepancy theory "includes the subject's perception

of what is expected or valued as the baseline for comparing

actual outcomes. Thus discrepancy theories define

satisfaction as the difference between actual outcome and

some other ideal outcome" (p. 186).

In his review of the literature, Williams suggests that

clients may hold a complex set of important, relevant

beliefs that cannot be embodied in the general term

satisfaction. This leads to the question of whether

satisfaction is multi- or uni-dimensional in nature.

Lebow (1982) reviews four factor analytic studies that

are multi-dimensional and point to the possibility that

consumers are capable of differentiating between aspects of

satisfaction (p. 252). Other studies suggest that

satisfaction is uni-dimensional. At the time of his

writing, Lebow also had to admit that "a coherent literature

on the subject has yet to develop" (p. 244). The question

to be addressed is whether the situation has changed?

From both a practical and research-driven perspective,

Smith (1996) asks why patients, clinicians and purchasers

are increasingly uncomfortable with current methods of

procuring health care (p. 43). Smith responds "because

the patient has been left out of the equation. Nowhere in

the current purchasing scheme is the patient or patient's

health care part of the consideration in a literal or

substantial way" (p. 43). While consumer satisfaction might

be a part of some decisions, it "is rarely assessed in a

scientifically valid manner, and no one is sure what the

measurements indicate about the quality of care or how well

a patient is doing" (p. 43).

Eisen (1996), who has researched client satisfaction,

recently wrote an article asking "Client Satisfaction and

Clinical Outcomes Do We Need to Measure Both?." She

responded that "prior research on the relationship between

satisfaction and outcome has produced varied results" (p.

72). She then argued for the conceptual differences between

clinical outcomes and satisfaction stating that satisfaction

cannot be accurately predicted from knowledge of a client's

improvement or functional status.

May (1991), reviewed 13 studies examining the effect of

treatment delay on satisfaction. May found that, while

attrition rates were high, they were not necessarily related

to dissatisfaction.

Treatment Issues and Satisfaction

Lebow (1982), in his article, "Consumer Satisfaction

with Mental Health Treatment", presented a contemporary,

comprehensive review of the following issues: (a)

evaluation of consumer satisfaction, (b) methodological

issues inherent in studying satisfaction, (c) results of

studies conducted, and (d) the value of consumer

satisfaction. This is a significant review relied upon by a

number of researchers interested in the issue of consumer


Type of treatment. Lebow indicates that most work was

"suggestive" and "insufficiently well developed" (p. 251).

Length of treatment. Lebow states that satisfaction

appears to be unrelated to treatment length although it is

augmented if mutual termination (i.e., between therapist and

client) has occurred (p. 251). This is further supported by

May's (1991) study mentioned above.

Process variables. "Almost no research has explored

the relationship between satisfaction and other process

variables" (Lebow, 1982, p. 251).

Client characteristics. Lebow (1982) stated that "data

suggest that demographic characteristics are not good

predictors of satisfaction",, (p. 251). Lebow cited numerous

studies where demographic data (with the exception of race

which is described as a controversial variable) such as age,

gender, marital status, income level or education appeared

unrelated to satisfaction. Lebow did state, however, that

future research was warranted to explore the relationship

between various client variables and satisfaction.


Attkisson and Greenfield (1995, p. 125), in discussing

future research, however, stated that investigators need to

introduce improved control for variables they view as

functioning as covariates of satisfaction. These covariates

include demographic and personal characteristics, attitudes

about health care, and socioeconomic status. They go on to

cite studies where "age and gender effects are perhaps the

most important candidates as covariates that may contribute

variance to service satisfaction ratings" (p. 125).

In agreement with Lebow (1982), Attkisson and

Greenfield (1995) state that life satisfaction and general

attitudes toward the health care system were not found to be

significantly related to direct measures of service

satisfaction (p. 125). Burke (1994) found that men and

women were equally aware of and likely to have used the

employee assistance program with similar satisfaction.

Women were, however, more likely to indicate the intention

to use the employee assistance program if it was needed.

Lebow (1982) found the relation of psychological,

diagnostic, and prognostic client variables to be more

promising. Satisfaction was determined to be lower for the

following categories: (a) drug abusers, (b) suicidal

clients, (c) psychotic clients, (d) those having a poor

prognosis, and (e) for those who return to the same program

for additional treatment. Satisfaction does appear to be

related to the fulfillment of client expectations (p. 251).

A study to determine the relationship between

satisfaction and type of provider was conducted by Tessler,

Gamache, and Fisher (1991). Their research consisted of

studying 1,198 separate contacts with mental health

professionals made by 274 relatives or close friends of 168

mental health clients. Findings indicate that satisfaction

varied significantly for different types of professionals.

Methods of determining satisfaction. Eisen, Grob and

Dill (1991) indicate that multiple methods, both solicited

and unsolicited, have been utilized to investigate consumer

satisfaction and include the following: (a) personal

interviews, (b) letters from clients, (c) telephone surveys,

and (d) questionnaires (p. 228). Harris and Heft (1992)

state that "for the most part, research on employee and

supervisor reactions has been conducted independently.

Given the apparent importance of informal networks, it is

essential more studies be conducted that simultaneously

examine the effects of supervisor reactions, employee

willingness to participate, and various organizational

characteristics" (p 258). This study addressed this issue

by exploring whether the referral source, supervisory or

self-referral, prompting an employee's participation in the

employee assistance program had an impact on the level of


Dickman and Emener (1988), studied employee assistance

program participant satisfaction over a 26-month period


(March, 1979 through May, 1981). They used an ex post facto

survey questionnaire developed specifically for their study.

According to Oher (1993) "the use of a survey questionnaire

to gather information about an organization is an accepted,

widely used technique in the organizational development

field" (pp. 41-42). Respondent satisfaction endorsed what

continues to be recognized by employee assistance program

professionals to be the key ingredients of any employee

assistance program if it is to be considered effective.

Dickman and Emener (1988) list the five key components as

confidentiality, referral, early intervention, impact on job

performance, and resolution of problems (pp. 283-286).

Self Report

History of self-report. The origin of self-report can

be traced to the early tradition of experimental psychology.

In the latter part of the nineteenth century and the early

part of this century, Wilhelm Wundt and E. B. Tichener, in

their effort to apply methods of introspection, attempted to

map the inner structure of the mind. Tichener (1912)

proposed that theory and fact be separated by letting

individuals report only experienced conscious content. He

believed that all mental states and experiences could be

described in terms of sensory and imaginal components. He

did recognize, however, that mediating processes could occur

and wrote of his concern regarding the imprecision of

communication. He said that communication should contain

the elementary components of experience.

Freud (1950), in his effort to apply newly-emerging

methods of psychology in the diagnosis and treatment of

neuroses and other disorders, utilized self-report in a

variety of ways in order to: (a) clarify his patient's

condition, (b) confront problems, (c) offer interpretations,

and then (d) move toward corrective action. Although many

contemporary psychodynamic theorists tend to distrust self-

reports due to belief in what are considered to be elaborate

defense mechanisms of individuals, they do consider self-

reports of: (a) retrospective accounts, (b) free

associations, and (c) projective test results. These are

viewed as replicable, important sources of information.

Today, clinical interviews typically begin with

questions such as, "Can you tell me about your problem?" and

"What brought you in today?" Rogerian (e.g., Rogers, 1954)

and other humanistic clinicians consider self-reports to be

veridical (i.e., the client can and will provide reliable,

valid responses).

On the other hand, practitioners of behavioral

psychology have placed an emphasis on nonverbal behavior.

It should be noted, however, that from the very inception of

behavioral psychology, measurements of verbal report have

also been considered. As Boring (1950) stated, Watson did

not rule out of psychology all use of introspection. He

allowed, in the case of human subjects, discriminatory

verbal reports as a form of behavior when they were accurate

and verifiable. The effort here was to count and measure

with reliability the observable reports and actions of the

subjects. Statistical analysis of multiple changes led to

the reliability of self-report. In the late 1960s through

the 1970s, behavioral psychologists such as O'Leary and

O'Leary (1972) used self-report. Self-report has continued

to be a critical factor in the measurement and intervention

of behavior and behavior change.

Finally, cognitive-behavioral psychologists remain

interested in self-report in order to assess critical

cognitive determinants of behavior. Self-report is used for

assessment and change of thinking, planning and problem

solving (Ellis, 1962; Beck, 1976; Bandura, Adams & Beyer,

1977; Bellack & Hersen, 1977; Beck, Rush, Shaw & Emery,


Borg and Gall (1983) state that self-report devices

"are only accurate to the degree that the self-perceptions

are accurate and to the degree that the person is willing to

express them honestly. This problem has been, and continues

to be a matter of concern to many educational and

psychological researchers" (p. 336).

Park (1992), for example, discussed how her study's

results from assessing a university-based employee

assistance program were limited precisely because the survey

relied on self-report. She states that "oftentimes the

individuals who have used the service may be concerned about

confidentiality and may not feel secure in responding to

such an instrument" (p. 24).

Borg and Gall (1983) also discuss response bias using

the term response sets. This is defined as a variable that

leads to "spurious responses" (p. 336). They indicate that

"if self-report inventories are to be used effectively in

practical applications and in research settings, it is

important to investigate the extent to which subjects are

responding to the content of each item and the extent to

which their responses are determined by a general 'set'"

(pp. 336-337). Borg and Gall (1983) discuss three different

response sets which include social desirability, acquiescent

response sets, and negative response sets. The first,

social desirability, represents the desire to present

oneself in a favorable light. The second, acquiescent

response set, represents the tendency to respond "true" or

"yes" regardless of the content of the question. The third,

negative response set, represents the tendency to respond in

a deviant fashion.

Under sources of distortion in the area of consumer

response, Lebow (1982) indicates that self-reported data may

also be altered based on: (a) who will read the survey, (b)

perception regarding how the surveyor will regard the

respondent, and (c) how the survey might affect future


service requests or the career of the practitioner (p. 247).

In addition, it should be noted as Oher (1993) states that

"the EAP research literature offers few references regarding

the use of such survey questionnaires as a vehicle to

provide feedback about EAP functioning or effectiveness"

(p.42). He goes on to say that "this fact is

not surprising when one considers that the employee

assistance field is still in the early evolutionary phase of

development" (p. 42).

Steps to minimize response bias. Eisen, Grob and Dill

(1991) state that the following steps can be taken by a

researcher to minimize response bias: (a) guarantee

confidentiality and anonymity, (b) separate program

evaluators from clinical treatment staff, and (c) offer a

complete explanation to clients regarding the goals and

procedures of the evaluation process (p. 216). Lebow (1982)

presented the same steps but added the following to decrease

respondent reactivity: (a) explain that the purpose of the

assessment is to evaluate service received and not the

clients receiving the service, (b) emphasize that the focus

of the analysis will be aggregate not individual data, and

(c) provide reassurance regarding the use of the data (p.

247). It should be noted that this study abided by these


Advantages of self-report. A major source of

information is overlooked or neglected if the consumer is


not asked to be involved in the reporting and evaluation of

their own experience. Kaplan and Ware (1989) believe it is

important to incorporate consumers in treatment "and to make

their evaluations of the care they receive part of routine

quality assessment" (p. 25).

Validity and reliability. According to Lebow (1982),

utilization of self-report in studies of consumer

satisfaction research has been criticized because of the

following three issues. First, "validity problems are

inherent in consumer assessments." Second, self-report

involves a restricted response range and a tendency toward

halo responses resulting in little practical value. Third,

Lebow believes consumers cannot adequately evaluate

involvement in treatment due to: (a) impaired mental

status, (b) lack of experience, (c) transference

projections, (d) cognitive dissonance, (e) unconscious

processes, (f) folie a deau, (g) client character, and (h)

naivete (p. 254).

Problems have also occurred with sampling bias (i.e.,

the selection of clients contacted for the research and

those who respond). In the first case, consumers are

sometimes excluded perhaps because they are "unable" to

respond. One might consider, for example, a group of

inpatients who are psychotic or severely depressed. Or, in

some studies the response rate might be large enough to

influence results (Lebow, 1982, p. 246).

Lebow, in his review of consumer satisfaction studies,

indicates that those who choose to respond to a survey are:

(a) more likely than nonrespondents to have mutually-agreed

upon terminations, (b) more likely to have longer

treatments, and (c) those treatments were judged to be

successful by the therapist (p. 246).

Shrauger and Osberg (1981) reviewed studies with self-

reports coming primarily from outpatient respondents. Only

one study reviewed covered psychiatric inpatients. Shrauger

and Osberg found that self assessments were at least as

good, if not better, predictors of reactions to therapeutic


Kaplan and Ware (1989) support this position by stating

that patients' evaluation of quality of care are "the most

practical source of information" (p. 26). Eisen, Grob and

Dill (1991) state that it is feedback from the service

consumer that is a critical factor in treatment evaluation

(p. 227). In citing Kalman (1983), they identify four

potential areas of client satisfaction studies which

include: (a) treatment compliance, (b) service utilization,

(c) program design, and (d) treatment outcome.

Client Satisfaction Questionnaire-8

Criteria for Self-Report Measure

Several well-normed instruments are available for

measuring satisfaction. In this study, the Client

Satisfaction Questionnaire-8 (i.e., CSQ-8) was used as a

self-reported general measure of satisfaction. Eisen

(1991), a noted researcher in the area of satisfaction

studies, led this researcher to the document that she says

outlines "five major criteria that must be considered in the

attempt to find the 'ideal' self-report measure for

evaluation of mental health treatment outcome" (p. 223).

The Client Satisfaction Questionnaire-8 was selected for

this study because it met each of these five criteria

suggested by Ciarlo, Brown, Edwards, Kiresuk, and Newman


First, the Client Satisfaction Questionnaire-8 is

designed to be used with the specific target group (i.e.,

employees using an employee assistance program) in this

study. Second, the instrument needs to have simple,

teachable methodology and procedures. The CSQ-8 is very

easy to learn to administer. Third, Ciarlo et al. (1986)

emphasize the importance of the instrument having

psychometric strength. The material available on this

instrument addresses its reliability and validity as being

more than adequate. Fourth, the CSQ-8, at its current

price, was not outside the range this researcher would

consider reasonable. Lastly, it should possess utility.

The CSQ-8 is easy to understand and very easy to interpret

and provide feedback, is useful to clinical services, and is

compatible with clinical theories and practices.


Most consumer satisfaction studies rely on self-report.

Lebow (1982) states that "survey methods are the most widely

used means of gathering data. Their outstanding advantage

is directness, the purpose is clear, the responses

straightforward, and the tie to satisfaction is unequivocal"

(p. 244). While unobtrusive measures of satisfaction (e.g.,

unsolicited letters of compliment or complaint,

measures of utilization, case records, tapes) may reduce

respondent reactivity, "there are no perfect unobtrusive

measures of satisfaction" (p. 245).

Development of the CSO-8

Since Lebow made that observation, "over ten years of

empirical research has produced valid, reliable, and

feasible measures of patient satisfaction with care that can

now be used in practice settings" (Kaplan & Ware, 1989, p.

26). According to Eisen, Grob, and Dill (1991), it was the

lack of a standard satisfaction scale that motivated Larsen

and colleagues to construct an empirically-based scale, the

Client Satisfaction Questionnaire (CSQ) (p. 228). Levois,

Nguyen, and Attkisson (1981) state that "the CSQ was

developed as a general measure of client satisfaction" (p.

140). They describe in detail the developmental components

that lead to a factor analysis that "showed one general or

global satisfaction factor which accounted for 43% of the

total and 75% of the common variance" (p. 141). Levois,

Nguyen, and Attkisson (1981) state that "the authors

concluded that the only salient dimension in the client

responses to the questionnaire was a general or global one"

(p. 141). The CSQ-8 possesses psychometric qualities with

the potential to provide meaningful comparisons across


The CSQ-8 is an eight-item scale with high internal

consistency that has been widely used for the evaluation of

general satisfaction (p. 228). According to Attkisson and

Greenfield, 1995), "the CSQ instruments are self-report

questionnaires constructed to measure satisfaction with

services received by individuals and families" (p. 120).

The researchers discuss the target populations where the

CSQ-8 has been adopted in research. It has been used

across a wide variety of service settings including employee

assistance programs (p. 121).

According to Nguyen, Attkisson, and Stegner (1983), "a

major problem encountered in using satisfaction measures is

the ubiquitous finding that service recipients report high

levels of satisfaction" (p. 299). The authors cite several

studies and believe this may be due to a variety of factors

including the following: (a) a client's desire to give

positive testimony to treatment received (i.e., perceived

demand characteristics), and (b) "observed data are taken at

face value as unquestioned 'proof' of the effectiveness of

the program" (p. 299).

Participation in the Employee Assistance Program

"Clients seek employee assistance program services

either voluntarily through their own motivation or sometimes

following the request or suggestion of a supervisor, union

steward, company or personal physician, family member,

friend, or a representative of the legal community" (Keaton,

1990, p. 57). Examination of participation in work-

sponsored health programs such as an employee assistance

program have, according to Wilson (1990), yielded some

contradictory findings. This points to the complexity of

the factors associated with an employee choosing or being

referred by supervisors to participate in the employee

assistance program. Wilson (1990), for example, reports

that results from preliminary studies appear to show a

relationship between one's participation in the employee

assistance program and the following: (a) behavioral and

psychosocial variables, (b) age of participants, (c) higher

educational level, and (d) higher income levels. Hall,

Vacc, and Kissling (1991), in their survey of 62 employees,

assessed factors associated with employee assistance program

utilization. Findings indicated that 7 percent of the

participants had actually used the employee assistance

program. A total of 67 percent reported that they would

self-refer if needed. Those who would be most likely to use

the employee assistance program included the following

categories: (a) women, (b) those with higher incomes, and

(c) those with higher levels of education.

Supervisory Referrals

Supervisors are a primary referral source and key to

the success of the employee assistance program (Love, 1989).

Blum and Roman (1992) have developed a core technology that

addresses employee assistance program liaison with

supervisory management consisting of three components (pp.

121-122). First "supervisory management in the EAP core

technology includes the identification of troubled employees

based on documented evidence of impaired job performance"

(p. 121). This does not mean focusing only on suspected

alcohol or drug symptoms. Rather, it is a matter of

focusing on documented impaired performance, attendance, and

conduct issues that are specific to the workplace. Second,

the employee assistance program provides professionals who

are available to consult with a manager, supervisor, or

union steward in developing familiarity with the employee

assistance program policy and in learning how to use the

employee assistance program and workplace rules so the

employee understands that there is organizational support

behind the referral. The third subcomponent of the core

technology involves the supervisor setting up a situation

whereby a crisis is provoked which will hopefully lead to an

employee's seeking assistance through the employee

assistance program (p. 122). This is referred to as

constructive confrontation and "uses evidence of job

problems to precipitate crises that lead to assistance in an

atmosphere of confidentiality" (p. 122). While a number of

studies have addressed supervisory confrontation and

referral to the employee assistance program of employees

experiencing a decline in performance, attendance, or

conduct, little has been done to assess coaching or early

intervention referrals. "Softer approaches by supervisors

usually precede use of constructive confrontation (indeed,

informal referrals by supervisors are especially prominent

in EAP practice)" (p. 122).

Jesko (1992) compared two groups of supervisory

responses to Facilitative Counseling Skills Training (i.e.,

FCST) in terms of referral to the EAP. Supervisors who

received FCST demonstrated increased proficiency in helping

troubled employees find help with the result that employees

were more likely to accept referrals from them to the EAP.

In addition, the higher the supervisors scored on the

Listen, Clarify, Refer, and Document (i.e., LCRD)

instrument, the higher the referral rate. This emphasizes

the importance of the employee assistance program's role in

training supervisors on how to make appropriate referrals of

the troubled employee for help.

According to Boone (1995), "supervisory training serves

a dual purpose for employee assistance programs. It enables

EA (sic) practitioners to solicit the support of top and

middle management in the organization for the EAP. It also

enables supervisors to make use of another tool to add to

their repertoire of management skills" (p. 17). Schneider,

Colan, and Googins (1990) add that "most EAPs have strong,

positive feelings about their supervisor training and view

training as critical to the mission of the EAP" and also

state that "those programs with more skilled trainers

showed significantly higher referral rates" (p. 146). What

other factors play a contributing part to supervisory


Gerstein, Eichenhofer, Bayer, Valutis, and Jankowski

(1989) investigated the relationship between employee

assistance program training and supervisory interaction with

troubled employees. "Supervisors have been trained to

recognize workers in distress through the monitoring of

attendance, productivity, and co-worker relationships" (p.

16). "One variable that has been consistently related to

supervisors' utilization of constructive confrontation is

this group's participation in training programs designed to

teach this skill" (p. 16).

While the percentage of supervisory referrals is

usually significantly lower than the percentage of self-

referrals, my professional experience has been that these

low frequency referrals are high impact. This means that

other supervisory interventions have typically been

unsuccessful in turning around inappropriate performance,

attendance, or conduct issues.

Two of my hypotheses were derived from a major national

study conducted by Blum and Roman (1992) with support from

NIAAA. Data was collected on 6,400 employees from 84 work

sites who had utilized their employee assistance program

services. Their study titled, "A Description of Clients

Using Employee Assistance Programs," found that

"approximately 36 percent of the clients were reported,

either by the client or by the EAP administrator, to have

been referred by supervisors. However, only 65 percent of

the 36 percent represented agreement between the two sources

(client and administrator)" (p. 125). They also found that

women were less likely than men to be referred by their

supervisors. Based on their results, for purposes of my

study, I predicted that there would be a significantly

greater number of self-referrals as compared with

supervisory referrals (i.e., Hypothesis 1), and women would

be significantly more likely to self-refer then men (i.e.,

Hypothesis 2).

There are numerous studies that report factors believed

to affect a supervisor's tendency to refer. Nord and

Littrell (1989) conducted a study of supervisory referral in

predicting what supervisory characteristics would result in

referrals to the employee assistance program. Three

categories of supervisors which were surveyed included 164

who had referred an employee, 194 supervisors who had

recognized a problem but did not refer an employee, and 108

non-referring supervisors who noticed no problem. Results

pointed to referral being significantly more likely when the

following four conditions are present. First, it is

important for the supervisor to have familiarity with the

employee assistance program. Second, supervisors who refer

to the employee assistance program are more likely to be in

middle or upper management. Third, the supervisor is likely

to have an opinion about how much support the employee

assistance program receives from management, union, and

their own immediate supervisor. Last, referring supervisors

are not likely to have worked as a peer in a nonsupervisory

capacity with the subordinate most recently referred or

identified as having a problem.

Bayer and Gerstein (1990) examined supervisory referral

decisions made by 75 supervisors to the employee assistance

program. Results suggest that referral is linked to four

behaviors of troubled employees. They include resistance

(i.e., absenteeism), acrimoniousness (i.e., irritability),

industriousness (i.e., decreased productivity), and

disaffection (i.e., apathy).

Gerstein et al. (1993), as part of their study, asked

how many males and females had been referred to the employee

assistance program in the last year. Responses showed that

females were more likely than males to refer females and

males did not differ in their referral of male or female

troubled employees.

Smith (1995) researched two major factors by exploring

supervisory perceptions of poorly performing employees, and

what factors would increase the chances that a supervisor

would refer an employee to the employee assistance program.

Supervisors (N=345) were surveyed by written questionnaire

regarding three main components: (a) perception of one

poorly performing employee, (b) circumstances surrounding

the poor performance, and (c) whether the employee was

referred to the employee assistance program. A referral is

considered more likely if the employee's poor performance is

attributed to either substance abuse or personal crisis.

Referral is less likely if the poor performance is

attributed to a lack of motivation. Two other factors that

contribute to a referral include whether the supervisor

feels sympathy toward the employee and whether the

supervisor perceives the EAP as helpful. Smith also found a

significant interaction between the level of the supervisor

and the interdependence between the supervisor and the

employee. To the extent that a supervisor was lower level

and not dependent on the employee, the less likely they were

to refer (this is an example of cost-benefit analysis in the

bystander-equity model).

According to Jerrell and Rightmeyer (1988), it is

important to train supervisors and managers in the "early,

effective recognition of troubled employees and in making

confidential referrals to the coordinator" (p. 253). When a

pattern (i.e., a change in appropriate normative behavior)

is recognized as impaired job performance, attendance or

conduct on the part of the employee, verbal coaching by the

supervisor normally takes place. Some company procedure and

policy manuals may require that formal verbal counseling

take place at this point.

In any event, the goal is to provide, informally or

formally as determined by policy, the employee with a clear

understanding of what needs to be corrected, in what manner,

and according to what time frame. It is important that the

employee assistance program be offered as an aid at this

point and a time frame is usually established to evaluate

progress in meeting the behavioral objectives. If adequate

progress has not been made and the employee's performance,

attendance, or conduct continues to show a decline, a

meeting is usually held to review the situation. Such

meetings may include a human resource and/or union

representative (if applicable), the manager and/or

supervisor, and the employee. At this point, a referral to

the employee assistance program typically becomes mandatory

(if permitted by company policy). A mandatory referral

takes place when an employee is required to participate in

the employee assistance program if they are to continue

employment with the company.


Since participation in the employee assistance program

is always considered voluntary, the employee may refuse to

see the employee assistance professional. If this occurs

under mandated circumstances, however, the employee is

usually terminated. If the employee does agree to go to the

employee assistance program for assessment, brief counseling

(if appropriate), and/or referral, the employee is typically

given a probationary period by the company during which time

they are expected to correct their deficiencies.

The employee assistance program counselor, with the

employee's written, informed consent, keeps the named

company representatives) appraised of job-related

recommendations, employee assistance program and

programmatic attendance, and compliance until discharged.

This referral process is described by Frisch and

Leepson (1986) as consisting, therefore, of three steps:

(a) preliminary arrangements, (b) the meeting (i.e.,

defining the problem, referring for assistance), and (c)

follow-up. The manager and/or supervisor is considered a

key component in the entire process. They begin with the

referral and remain involved until the referred employee

completes all recommendations from the employee assistance

program and work performance, attendance, and conduct reach

satisfactory levels.

Braid (1983) posits seven rules to be applied when

discipline is used by supervisors to change performance,

attendance, and conduct problems. These include remaining

calm, gathering information, being consistent, maintaining

focus on job-related issues, choosing appropriate

responses, carrying out discipline, and documenting any

interventions. The session ends with the development of a

behavioral plan to correct identified problems.

It should be noted, as stated by Keaton (1990) that

mandated referrals tend to result in employees who are

"resistant to counseling and to making a commitment for

change" (p.58). In her study, "The Effect of Voluntarism on

Treatment Attitude in Relationship to Previous Counseling

Experience in an Employee Assistance Program," Keaton (1990)

cites several studies that show a negative impact when

referral occurs under mandated conditions. Keaton's study

had some significant limitations, however, in that she

evaluated a specific organization, used a non-standardized

instrument, and due to the nature of the design could not

determine the direction of the voluntarism-experience

relationship (p. 63).

While supervisors are a referral source to the EAP they

only account for a percentage of total referrals. Bayer and

Babbkin (1990) found that self-referral outpaced all others.


According to Blum and Roman (1992, p. 127), "the range

of reports from diverse settings emphasizes the prominence

of self-referrals to EAPs." They continue "These

reports...support an image of EAP referral as a cognitive

process involving the individual's receipt of information

about EAP services and a comparison of the services with his

or her perceived needs" (p. 127). Backer and O'Hara (1991)

support the finding that the majority of employees who seek

help from an employee assistance program do so through self-

referral methods.

Keaton (1990) conducted a study of the effects of self-

referral as related to attitude toward the employee

assistance program, and to the employee's experience of the

employee assistance program. A survey of 67 employee

assistance program clients was conducted that resulted in a

significant relationship for both factors. Keaton indicates

that her findings are consistent with cognitive dissonance

theory; that is, one who freely chooses treatment will be

more likely to view it in a positive light.

In fact, encouragement of employee self-referral may be

promoted when the following conditions are present. First,

it is important that the employees understand what services

are actually available under the employee assistance

program. Second, employees need to be aware of the

procedures needed to access the employee assistance program.

Third, the employees need to have developed a sense of trust

regarding the confidentiality of the program (Frost, 1990).

The first two conditions may be met by the employee

assistance program professionals and/or the company

representatives holding employee and/or family orientation

meetings. Another means would be through the dissemination

of printed materials. Letters may be sent to the employee's

home describing the program, services, and means of

utilizing services. Often, brochures are developed and

distributed that provide additional information such as

answering questions employees might have regarding the

connection between voluntary participation in the employee

assistance program and their place of employment.

The issue of confidentiality, while it might be

addressed in the above, often develops over time. As

employees utilize the program, they frequently talk about

the help they have received with their coworkers. As they

determine that information does not get back to the company,

confidence grows.

Sonnenstuhl (1982), conducted in-depth interviews and

case record reviews on thirty employees to study the shift

away from constructive confrontation to self-referral to the

employee assistance program. He found that the referral

process involved social controls (both formal and informal),

a mix of the presenting problem embedded in a focus of

disrupted relationships, and the importance of viewing the

supervisor as part of the process.

In another study, Sonnenstuhl (1990) found that

employees tend to have four questions that appear to be


significant when an employee is considering self-referral to

the employee assistance program. First, they are often

concerned about what it will cost, if anything, to access

the actual employee assistance program services. Second,

employees question whether they can access the employee

assistance program during work hours. This is a pertinent

question since employees may come into the employee

assistance program stating that their reason for accessing

services is a work-related issue; and, the program itself is

sponsored by their work organization. Also, if the employee

assistance program does not have flexible hours, it might

result in personal hardship for the client. Third, since

the employee assistance program also covers assessment of

emotional and behavioral issues, employees tend to want to

know if medication will be prescribed. Fourth, employees

want to know the degree to which the employee assistance

program is confidential? Sonnenstuhl (1990) reports that

employees tend to use both formal and informal (e.g.,

coworker and supervisory information) when deciding to use

employee assistance program services.

Harris and Heft (1992) cite a study by Harris and

Fennell (1988) that indicates an employee's willingness to

use an employee assistance program is based on three

factors. First, the employee must have familiarity with the

program and services offered. Second, personal attention is

important. Third, the employee must have a reasonable trust

in what will occur if the employee seeks assistance (p.


Increases in self-referrals appear to be related

currently to increases in on-the-job stress (Miller, Jones,

& Miller, 1992). Also, companies are currently utilizing

employee assistance program professionals to provide on-site

consultation in the area of stress management as well as

other behavioral health concerns.

Estimates of Differences Between Supervisory

and Self-Referrals

Blum and Roman (1992) believe official EAP records

overestimate the role of the self and underestimate the role

of supervisors, co-workers, and family members in the

referral process. In their NIAAA-supported study of 6,400

employees from 84 different work sites, they found some

interesting discriminators. First, most self-referrals were

employees with non-alcohol related problems (46 percent) as

compared to alcohol-related problems (39 percent). Second,

formal supervisory referrals were greater for clients with

alcohol-related problems (17.2 percent). Supervisory

referral of employees with non-alcohol related problems was

over fifty percent lower (8.8 percent). Informal coaching

referrals executed by supervisors was nearly double for

employees with non-alcohol related problems (17.1 percent)

and 13.4 percent for alcohol-related problems (p. 125).

Hobson (1981), in a descriptive study of a systems

approach to develop, implement, and evaluate an employee

assistance program, found that employees were more willing

to refer others than themselves. In addition, Hobson

believes there are certain employee groups that might feel

more at risk if they did reach out through employee

assistance program participation. Hobson lists the

following groups as being more hesitant to access services:

(a) lower-level staff, (b) first-line supervisors, (c) union

stewards, (d) female clinical staff, (e) support staff, and

(f) higher level professionals. Hobson reviews methods to

increase employee assistance program utilization by these


Stollak (1994), gathered survey data from a field study

of three Midwestern organizations using a multiple

constituency evaluation approach (i.e., the constituents

define: (a) effectiveness, (b) activities, and (c)

evaluation criteria of the employee assistance program) in

order to investigate the effectiveness of the employee

assistance program. Differences were found to exist between

blue and white collar workers in a number of areas. Blue

collar workers emphasized activities related to career

development, union and organizational linkage with the

employee assistance program, program monitoring, and the

special assistance provided by employee services. Managers

and supervisors, however, placed a greater emphasis on

training. Individuals who utilized employee assistance

program services emphasized activities related to

counseling. Nonusers emphasized activities related to

career development, program monitoring, and linkages between

the union and the work organization. These are factors that

may influence areas which bring an employee to the employee

assistance program.

Pearson and Maier (1995), in their assessment of

employee assistance program services, found that self-

reported improvement was negatively related to being a

person of color and to being a supervisory referral. This

may be a consideration in terms of how a person of color

should be referred.

As has been indicated in this section, satisfaction is

an important outcome to consider as research on employee

assistance program services continues. In reviewing the

literature, I found a large number of doctoral dissertations

in this area. I believe this bears witness to the interest

in contributing to our understanding of employee assistance

program effectiveness. While it is true that employee

assistance programs have a reported effectiveness, there is

still a long way to go in carrying out research that is

theoretically and methodologically sound in this regard.

Since employees come to receive EAP services through a

variety of means, no published study to date using the

Client Satisfaction Questionnaire-8 has assessed whether how


an employee comes in affects their level of satisfaction.

This is what this study purported to do by its attempt to

delineate any differences in satisfaction based on

supervisory as compared to self-referral.



Participants for this study were selected from a census

sample of all employees who utilized their company's

employee assistance program services during a six-month

period beginning January 1, 1996 and ending June 30, 1996.

Only employees who came in alone (i.e., not accompanied by a

family member) and who indicated on their Client Information

Data Form (see Appendix A) a willingness to receive a

follow-up questionnaire were selected for the study. The

participants came from seventeen companies located

throughout the southeastern region of the United States with

data collapsed across companies. For comparison purposes,

participants were placed in one of two groups as determined

by their source of referral to the employee assistance

program: (a) supervisory referral or (b) self-referral.

To be eligible for the study, participants were

required to satisfy each of the following criteria:

1. Be an employee who came to the EAP through either
supervisory or self-referral.

2. Work for a company that has had a broadbrush
employee assistance program in place a minimum of
one year.