TEST REVIEW: EMPLOYEE ASSISTANCE PROGRAM INVENTORY
Ralph L. Piedmont, Ph.D.
Loyola College in Maryland
1. Title: Employee Assistance Program Inventory (EAPI).
2. Authors: William D. Anton, Ph.D., and James R. Reed, Ph.D.
3. Publisher: Psychological Assessment Resources, P.O. Box 998, Odessa, FL. 33556-0998. Phone: 1-800-331-8378
4. Forms; groups to which applicable: Only one self-report form requiring a third grade reading level to complete. It is used with adults ages 18 - 76 years of age.
5. Practical Features: The EAPI comes with an easy to use hand- scorable answer sheet. These forms are two layers, with responses made on the top page being copied to the scoring form underneath. Scoring is merely the summation of responses to items and standard scores can be readily obtained, standardized profile form is printed on the back of the answer sheet.
6. General Type: The EAPI is a broad-band measure of psychosocial status designed to be used as a screening tool of individuals entering an Employee Assistance Program.
7. Date of Publication: 1994.
8. Cost: booklets, answer sheets: Professional manual - $25.00; reusable item booklets (pkg/25) - $29.00 each; Answer Sheet/Profiles (pkg/25) - $29.00 per package; Introductory kit (all of the above) - $75.00.
9. Scoring services available and cost: None available.
10. Time required: It takes approximately 20 minutes for individuals to answer all the items and about 5 minutes to hand score.
11. Purpose for which evaluated: For use in a community counseling context for adults.
12. Description of test, items, and scoring: The EAPI consists of 120 statements that are responded to on a four-point Likert scale from (F) not at all true to (V) very true. Some effort was made to balance the scales to control for the effects of acquiescence. However, this effort was not complete: Three scales are not balanced at all while the remaining seven vary in terms of the percentage of differentially reflected items. Items are clearly laid out in the test booklet and the answer sheet is sufficiently spacious to allow for easy recording of responses. Hand scoring is quickly performed by tearing off the top layer of the answer sheet and summing scoring down the columns to find total raw scores for each of the 10 scales. Standard scores are readily obtained by plotting these raw values on the profile form printed on the reverse side of the answer form.
13. Authors' purpose and basis for selecting items: The EAPI was designed to serve as a screening and evaluation instrument for the most common psychological problems that cause employed adults to seek services from an employee assistance program (EAP). The authors polled 200 professionals in the Employee Assistance Program Association and asked them to select from a list the ten most relevant psychosocial issues that would describe the types of clients they serve. Once the most popular content areas were established, items were developed that defined their behavioral expressions. These items were reviewed by psychological professionals to determine their appropriateness and to eliminate potential bias. These items were then given to a sample of 215 employed adults and item analyses were conducted to winnow the pool down to its current 120 items.
14. Adequacy of directions; training required to administer: The EAPI can be easily administered to adults. The instructions are quite clear and self-explanatory. The instrument can be administered and scored by individuals with no particular training in testing or the social sciences. This instrument is classified by the publisher as a Level B measure, requiring of the examiner at least a B.A. degree in Psychology or Counseling and relevant training or course work in the interpretation of psychological tests and measurement at an accredited college or university.
15. Mental functions or traits represented in each scars: The EAPI assesses ten different content areas: Anxiety--reflects the physical and psychological correlates of anxiety; Depression-- examines level of chronic fatigue, loss of interest in pleasurable activities, and feelings of sadness; Self-Esteem Problems--level of satisfaction with personal skills, abilities or achievement in relation to peers; Marital Problems--evaluates the level of dissatisfaction in current intimate relationships; Family Problems--examines the level of conflict being experienced with other family members; External Stressors--reports on the level of difficulties client is encountering in financial, legal, or health- related areas of functioning; Interpersonal Conflict--the degree of conflict/hostility being encountered with coworkers and/or supervisors; Work Adjustment--the level of satisfaction a client experiences with regard to level of pay, opportunity for advancement, working conditions, and the sense of control over one's job; Problem Minimization--evaluates the degree to which a client may understate the extent or severity of their problems; Substance Abuse--scores on this scale reflect the degree to which a client is experiencing difficulties in interpersonal, social, and vocational functioning resulting from substance use or abuse. Scores higher than a pre-established cut-off may reflect a high level of psychosocial dysphoria due to substance use.
16. Comments regarding design of test: The EAPI is well designed, permitting individuals to easily understand and respond to the questions presented. Hand-scoring is simple. It is unfortunate that there are no computer scoring programs to facilitate this process and augment interpretation. The EAPI appears to have good content validity, including scales identified as areas of relevance by a sample of EAP professionals.
17. Validation against criteria: The manual does not provide any criterion validity evidence for the EAPI and given the relative newness of the instrument none have yet appeared in the research literature.
18. Evidence of construct validity: The manual presents the results of eight studies that evaluated the construct validity of the various EAPI scales. These investigations were carried out in 44 EAPs throughout the United States and Canada. In seven of these studies, the EAPI scales were correlated with relevant measures, including the Michigan Alcoholism Screening Test (MAST), Drug Abuse Screening Test (DAST), Tennessee Self- Concept Scales, State-Trait Anxiety Inventory, Beck Depression Scale, Dyadic Adjustment Scales, Family Adaptability and Cohesion Scales 11, Occupational Stress Inventory, Holmes- Rahe Social Readjustment Scale, the facet scales of Warmth, Compliance, Vulnerability to Stress, and Angry Hostility from the NEO PI-R, and the Treatment Rejection scale from the Personality Assessment Inventory. In all instances, these external measures converged with their appropriate relative on the EAPI. However, discriminant validity was not always as sharply delineated, indicating some conceptual overlap among the EAPI scales. One simulation study was conducted to evaluate the Problem Minimization scale. Here, 78 individuals were randomly assigned to one of three conditions: a problem minimization group, a problem acknowledgment group, and an honest condition. One-way ANOVAs indicated significant mean score differences on all scales except Substance Abuse. Results showed that the problem acknowledgement group had significantly higher mean scores than the problem minimizing and honest conditions on all scales except the Problem Minimization (PM) scale. On the PM scale, those in the minimizing condition scored significantly higher than the other two groups.
19. Fairness: The r-manual does not present separate norms by gender. The authors do present squared partial correlations between each of the EAPI scales and age, gender, ethnic background, and education level and suggest that these factors do not influence scores. However, these values underestimate the overall level of association. In any event, each of these demographic variables account for, at a minimum, between 1 % and 3% of the unique variance in the EAPI scales. In a sample of 1,266 individuals, these correlations would be considered significant. Thus, it would be useful to construct norms based on these variables.
20. Comments regarding validity for particular purposes: Overall, there is not much validity data currently available for this instrument. That which is presented, although speaking to the general validity of the 10 scales, still leaves much to be desired. An examination of the correlations between the EAPI scales and the various external criteria presented in the manual strongly suggests that there may be much substantive overlap among the different scales. Without any factor analytic data, it is difficult to know just how much redundancy exists. Because the EAPI is presented as a screener, some evidence of how effective this scale is in identifying actual clinical syndromes would be helpful. Evidence on the sensitivity and specificity of these scales would be useful. As it stands, the EAPI may only reflect aspects of normal personality that are correlated with potential interpersonal difficulties. Finally, some caution needs to be exercised in relation to the Problem Minimization (PM) scale. This scale was designed to detect whether an individual was trying to cover up problems. The PM scale is really a type of validity scale designed to alert the interpreter to the presence of a response bias in the data. Although the only study presented here for its validity shows that when you ask people to distort their responses they will, more evidence is needed to show that it can really detect those who are actually trying to dissimulate. Without such evidence users will be prone to question the protocols of clients who may actually have no problems. This comment is based on the correlation of -.44 between the PM scale and the NEO PI-R facet of Vulnerability to Stress, suggesting that those who are coping well with stress and may not be experiencing any problems tend to score high on the PM scale. Thus, users should be very cautious in how they interpret scores from this scale.
21. Generalizability: No data are presented that speak to the generalizability of the EAPI.
22. Reliability: Cronbach's alphas for the 10 scales range from .73 for the Problem Minimization scale to .92 for the Marital Problems scale (Mean alpha= .85). No test-retest data was presented. These values were not determined from the normative sample but from an initial scale derivation sample of 215 employed adults.
23. Norms: Normative data were collected from 1,266 employed adults from six states. Subjects were recruited from a variety of contexts in order to ensure that the sample reflected national civilian labor force proportions stratified by gender, ethnic group, and age group. As a result, the norms group is comprised of mostly Caucasians (72%), having 13 or more years of education (70%), and mostly white collar workers (64%).
24. Comments regarding adequacy of norms: The norms are relatively large and representative of the white-collar working population. Information in the manual concerning the effects of gender, race, education, and age on normative values is not very persuasive and attention should be given to developing separate norms for these various demographic categories.
25. Aids to users: The manual provides several case histories for assistance in developing interpretive efficacy with the instrument. These examples are drawn from diverse EAP contexts and draw out how EAPI information can be helpful in managing these cases.
26. Comments of reviewers: As of the date of this writing, no reviews of the instrument have appeared in either Mental Measurements Yearbook or Tests in Print.
27. General evaluation of the test: My overall impression of the EAPI is mixed. On the one hand, it is useful to have an instrument that is specifically designed for use in an EAP context. With employees increasingly using such services, EAP professionals are in need of instruments that will provide useful insights into their clients. The EAPI does bundle a number of relevant scales into a single, easy-to-use platform. However, there are several limitations. First, there already exist numerous instruments that capture these dimensions that have much more well-established psychometric properties. Why do we need these new ones? Second, the relative newness of the EAPI means that there is very little information on the validity of the scale beyond what is presented in the manual, which is relatively scant. As noted above, it would be helpful to have these scales factor analyzed to determine the amount of redundancy among these scales, which seems quite sizable given the correlational data presented. Further, what kinds of EAP-relevant outcomes do these scales predict? To a large extent these scales are conceptual unknowns and much more research needs to be done to better identify their personological and clinical implications. Finally, I would also caution users of the EAPI to be careful in their use of the Problem Minimization scale. All too frequently test interpreters rely on "validity" scales to inform them when the test taker has been deceptive in responding to the items. However, there is no data that supports the validity of these so-called validity scales. Overall, I would suggest that if one is to use this instrument, that it be given as part of a battery of instruments that are themselves well established psychometrically.
Anton, W. D., & Reed, J. R. (1 994). Employee Assistance Program Inventory--professional manual. Odessa, FL: Psychological Assessment Resources.
May 3, 2001