MEASURING COPING STRATEGIES IN AN EDUCATIONAL INTERVENTION FOR INDIVIDUALS WITH PSYCHIATRIC DISABILITIES.
Subject: Adjustment (Psychology) (Evaluation)
Adult education (Evaluation)
Educational services industry (Services)
Mentally ill (Study and teaching)
Authors: Collins, Mary Elizabeth
Mowbray, Carol T.
Bybee, Deborah
Pub Date: 11/01/1999
Publication: Name: Health and Social Work Publisher: National Association of Social Workers Audience: Academic; Professional Format: Magazine/Journal Subject: Health; Sociology and social work Copyright: COPYRIGHT 1999 National Association of Social Workers ISSN: 0360-7283
Issue: Date: Nov, 1999 Source Volume: 24 Source Issue: 4
Geographic: Geographic Scope: United States Geographic Code: 1USA United States
Accession Number: 58450315
Full Text: Coping is a critical theoretical concept believed to be a mediator for successful outcomes produced through a variety of different social work interventions. This article describes an approach used to measure coping with stressors and was designed specifically for an intervention, the Michigan Supported Education Research Project, aimed at providing support for people with psychiatric disabilities to pursue college or vocational education. Hypothetical scenarios were developed that matched the content of material covered in the program. Open-ended responses were elicited from program participants as part of an in-person interview. Key findings included a participation effect on the number of positive coping strategies of participants and a significant relationship among some coping strategies and later outcomes. Implications for social work practice and research are identified.

Supported education is a recent intervention strategy in the area of psychosocial rehabilitation designed to encourage adults with mental illness to enroll in and complete postsecondary education. As postsecondary education has become a virtual prerequisite for career advancement and economic stability, efforts to support adults in attaining higher education have become increasingly critical. In the case of adults with mental illness, supports are particularly important for several reasons: the onset of mental illness may have disrupted earlier attempts at college; stigmatizing attitudes on the part of educational personnel may have impeded individual efforts to overcome obstacles created by the mental illness; and repeated attempts and failure to attain education may have negatively affected an individual's own feelings about his or her ability to attain this goal. Despite these potential obstacles, earlier research has demonstrated that people with psychiatric disabilities have an interest in attending coll ege, and many have some earlier experiences with higher education (Hazel, Herman, & Mowbray, 1991; Tessler & Goldman, 1982; Unger & Anthony, 1984).

The Michigan Supported Education Research Project (MSERP) was a federally funded research demonstration project to improve postsecondary choice, access, and retention through appropriate accommodations and individual interventions for people with psychiatric disabilities. Coping is considered a critical element of MSERP because for many individuals with psychiatric disabilities, the fairly normative stress of postsecondary education may be exacerbated by any number of circumstances, directly or indirectly related to their mental illness. These potential additional stressors include confronting personal fears; managing symptoms; dealing with stigmatizing attitudes of fellow students, instructors, or college administrators; and balancing school-related responsibilities with personal responsibilities and medical regimens. Because coping is acknowledged as a key factor in the relationship between the experience of stress and outcome (Aldwin & Revenson, 1987; Billings & Moos, 1981; Pearlin & Schooler, 1978), a co mponent of MSERP was directed toward teaching students to recognize and appropriately deal with school-related situations likely to cause stress. Despite the recognized importance of coping, the research in this field offers many measurement challenges. This artide reviews some of these measurement challenges and the various methods researchers have used to examine coping behavior. After reviewing these areas, we discuss our own efforts to measure coping in MSERP, our results, and the strengths and weaknesses of our approach.

COPING RESEARCH

A major focus of coping-oriented research has been aimed at determining which coping behaviors or styles are related to positive mental health outcomes. In a study of adults with unipolar depression, Billings and Moos (1984) found that coping directed at problem solving and affective regulation was associated with less severe dysfunction and that emotional responses were associated with greater dysfunction. Similarly, Aldwin and Revenson (1987) found problem-focused instrumental action to be a stress buffer and the strategies of escapism, fantasizing, and self-blame to be most strongly related to psychological symptoms.

Characteristics of the stressor and context have sometimes been found to influence the relationship between the coping response and outcome. Avoidant strategies resulted in better adaptation in the short term and vigilant strategies with better adaptation in the long term (Mullen & Suls, 1982). Compas, Malcarne, and Fondacaro (1988) suggested that task-oriented coping was most effective in controllable situations and emotion-oriented coping was most effective in noncontrollable situations. Similarly, Vitaliano, DeWolfe, Maiuro, Russo, and Katon (1990) found that problem-focused (task-oriented) coping was related negatively to depressed mood for situational stressors perceived as changeable, but there was no relationship when the stressor was judged to be nonchangeable.

In part, lack of conclusive evidence about the effect of coping on outcome may be related to the methods used. Standardized quantitative methods have dominated the field of coping measurement. The Ways of Coping Checklist (Folkman & Lazarus, 1980) has been used most frequently, but other checklist-type measures also are common. These methods offer several advantages, including ease of use, need for only minimal training, inclusion of a variety of coping options, and documented properties of reliability and validity. Yet there are weaknesses as well. Heppner, Cook, Wright, and Johnson (1995) suggested that the root of the psychometric problems of existing measures is the ambiguity of meaning found in several items. This ambiguity makes it difficult to assess whether items reflect cognitive, behavioral, or affective responses and to determine whether the activity is positive or negative.

In recent years, investigators have pursued several alternatives to these common measures and methods. Most have offered some alternative form of coping scale (for example, Carver, Scheier, & Weintraub, 1989; Heppner et al., 1995; Stanton, Danoff-Burg, Cameron, & Ellis, 1994; Stone, Kennedy-Moore, & Neale, 1995). Others have used qualitative methods to describe the context of the coping situation, to capture the personal meaning of situations, and to investigate coping processes among populations. For example, Banyard (1995) used qualitative methods to examine the coping strategies of homeless women (a population for which standard coping measures may not be well suited) to provide a richer picture of the stress and coping process for this group of mothers.

Some researchers have combined quantitative and qualitative methods to study coping. Bargagliotti and Trygstad (1987) compared findings of a qualitative and quantitative study of work-related stress among nurses, and Oakland and Ostell (1996) used both quantitative and qualitative data to study work-related coping among teachers. In these studies, the use of the qualitative data provided more insight into the processes of coping, as well as the combinations of strategies used, and the influential factors such as situational requirements, time constraints, personal skills, judgments and choices, and access to external resources.

The purpose of the current study was to examine coping among participants in the MSERP, an intervention designed to help adults with psychiatric disabilities attain postsecondary education. A part of the program aimed at teaching effective coping strategies for dealing with school-related stressors. Because this was a unique population and because of identified problems with existing coping measures, our measurement approach used an inductive data collection strategy to allow respondents to describe their coping strategies. Furthermore, because participants were in the process of preparing for school enrollment, the scenarios examining coping involved hypothetical situations and requested the participants to describe how they would react in that situation. The common theme for all vignettes was short-term, school-focused, and controllable situations. Therefore, because of the nature of the intervention and the vignettes, action-oriented strategies were determined to be the most positive type of responses. Hi ghly emotional or avoidant responses were considered to be inappropriate responses. Other, more ambiguous strategies were viewed as neutral. Although there is debate in the larger coping literature about which types of strategies are positive and negative, the nature of the intervention dictated the more active, problem-oriented activities to be preferable (that is, the scenarios were such that the described circumstances could be affected through action).

METHOD

Description of Intervention

MSERP enrolled four cohorts of participants; they began participation in winter 1994, summer 1994, fall 1994, and winter 1995. Prior to the start of the intervention, all participants completed a baseline interview. After the interview each participant was invited to orientation, received an information packet, and was assigned randomly to one of three conditions: classroom, group, or individual.

Both the classroom and the group model were situated on a community college campus. Meetings occurred twice a week for 2.5-hour sessions throughout two 14-week semesters. The classroom model used an academic curriculum (Unger, Danley, Kohn, & Hutchinson, 1987) and was implemented by two instructors in a classroom setting. The group model used two group facilitators and emphasized a supportive learning environment for group member exploration of career and educational options using a flexible, group-crafted agenda. These two active models are in contrast to the individual model that was designed to be the control condition; it was a minimal intervention group that involved no structured or scheduled intervention. Instead, students in the individual model were assigned a staff person, who, on request, was available to assist students to meet their own self-defined needs. As expected, few students in the individual model sought assistance.

Information relevant to coping with school-related stressors was covered in the classroom and group models in several ways: identification and practice in using campus resources (library, career laboratory, financial aid, and so forth); practice of simulated exercises that required completion of financial aid forms, course selection, registration, and resolving conflicts with professors; discussion and feedback from staff and peers about solving common academic problems, and practicing study skills; identifying and appropriately resolving anxiety and interpersonal stressors, and so forth.

Participants

Participants were from the Detroit metropolitan area--recruited primarily from the public mental health system; however, some were recruited from self-help programs, advocacy services, and by word of mouth. Over a period of 15 months, 480 individuals applied for the project. Eligibility criteria included the following: psychiatric disability of at least one year; high school diploma or GED obtained or near completion; interest in pursuing postsecondary education; and willingness to use mental health services, if needed, during participation. Only 20 applicants were found ineligible or were unable to be screened (4.2 percent).

Of the 460 potential enrollees, 63 were not interviewed at baseline due to participant refusal (n = 46, 10 percent), or inability to locate (n = 17,3.7 percent). The final study sample, therefore, was 397. There was a slightly higher percentage of female participants (52 percent) than male participants (48 percent) and, consistent with the racial composition of the catchment area, most were black (61 percent black, 38 percent white, and 1 percent other). The average age was 36.9, ranging from 17 to 75 years.

On average, participants had experienced mental illness for more than 14 years. The vast majority received SSDI or SSI. From scores on the Personality Assessment Inventory (PAI) (Morey, 1991) administered at baseline, participants enrolled in MSERP scored, on average, at least one standard deviation above the nonclinical normative means on six PAI subscales (anxiety, anxiety-related, borderline, depression, paranoia, and schizophrenia), with elevated scores on the remaining four subscales. Primary diagnosis was available for 240 participants (60.5 percent) from the management information system of the Community Mental Health Agency (CMHA). Of these, 68 percent had a primary diagnosis of schizophrenia or a related disorder, 25 percent were diagnosed with an affective disorder, and 8 percent with an anxiety or other unspecified disorder.

About a quarter had graduated only from high school, and a quarter had not; nearly half had some post-high school education. Most participants lived with family or in supervised settings; about one-fifth lived alone and one-tenth with friends. In terms of geographic distribution, participants resided in Detroit as well as the outlying areas of Wayne County; the proportion of MSERP participants by catchment area generally resembled that of the CMHA's overall clientele.

Research Design and Procedures

The experimental design involved random assignment of participants to one of the three conditions. The study involved data collection at five points: (1) baseline (prior to enrollment), (2) midway through the intervention (end of first term), (3) graduation from the intervention (end of second term), (4) six-month follow-up, and (5) 12month follow-up. Trained interviewers who conducted in-person interviews lasting approximately one hour (in the respondent's home whenever possible) collected data in several main areas: demographics; school, work, and psychiatric history; social adjustment; symptomatology; self-perceptions (self-esteem, empowerment, school efficacy, quality of life); and perceptions of the program. (Further details on these measures can be found in Collins, Bybee, & Mowbray, 1998.) In addition, information on coping was collected at midterm and graduation interviews only.

Three levels of participation were defined: none, moderate, and high. In the class and group conditions, moderate participation was defined as attending fewer than 20 sessions; high participation was defined as attending 20 or more sessions. In the individual condition, high participation was defined as 120 minutes or more of contact occurring over at least two contacts; moderate participation was some contact but less than that which defined high participation.

Detailed results of several immediate, intermediate, and long-term outcomes are presented elsewhere (Collins et al., 1998; Mowbray, Collins, & Bybee, 1999). Briefly, the program influenced school efficacy and empowerment at the point of graduation and more general self-perceptions (quality of life, social adjustment, and self-esteem) at follow-up. Moreover, enrollment in postsecondary education increased at each time point.

Coping

Data Collection and Coding. As part of midterm and graduation interviews, a series of nine vignettes was presented to respondents to elicit their response to the described stressful situation. The vignettes were developed based on a review of the curriculum. The nine vignettes were as follows:

1. Before you enroll for classes at a college of your choice you realize that you will need some sort of financial aid, what would you do?

2. You find yourself interested in a particular career or major, how would you make sure that it is the right vocation for you?

3. Imagine that you are in a classroom and you begin to get anxious or upset, what would you do?

4. Imagine that you are in school and that you feel as if the instructor is treating you unfairly, what would you do?

5. Imagine a situation where it is the day before a test in college and that this test is very important. If you do not do well on this test, you will not pass your class. At this point you do not understand the material and you have other responsibilities that day. What would you do?

6. The instructor of your course has indicated that you are not doing well in the class and if you don't improve you will not pass. What would you do?

7. Imagine that you are currently enrolled in school and you work quite a bit during the week. This hectic schedule is becoming very stressful for you. What would you do?

8. You are taking a large lecture course where you don't know anybody. You would like to make some friends. What would you do?

9. You are taking a class and other students are giving you a hard time. What would you do?

For each vignette, interviewers were instructed to probe for up to three responses. If respondents provided more than one answer, interviewers asked which option they would choose first and why they would choose that option compared with the other alternatives indicated. Finally, respondents were asked how confident they were (on a scale from 1 = not at all to 5 = extremely) that the course of action they described would resolve the situation.

A coding scheme was developed based on a review of the literature on coping behavior; three basic types of codes were identified from participant responses: (1) appraisal strategies, (2) problem-oriented strategies, and (3) emotion-oriented strategies. Appraisal strategies focused on responses that primarily involved thinking about the situation (for example, trying to figure out the best course of action). Five categories of problem-oriented activity were coded. Three of these involved direct action and were coded regarding their level of specificity: problem-solving specific, problem-solving nonspecific, and problem-solving vague. Two other problem-oriented codes were identified: help seeking (asking someone for advice) and information seeking (gathering information to address the problem). Several categories of emotional activity also were coded: internal effort, affective regulation, emotional discharge, escape or avoidance, and acceptance or resignation. In addition, some miscellaneous codes also were used that were generally not considered coping strategies but did reflect the individual's response (for example, would not happen, inappropriate, not a strategy). (See Table 1 for details of the scheme.) Also, if any references to MSERP staff or vocational rehabilitation staff were made in the context of help-seeking responses, these also were coded.

After the coding scheme was finalized, a social work graduate student was trained on the coding scheme (definitions, examples, and so forth), and a sample of 30 cases was selected for practice coding. Following the practice coding, 50 cases were selected to establish the reliability of the coding scheme. A second coder coded a random sample of 10 percent of cases to ensure that a high level of agreement was maintained throughout (calculated by vignette, Cohen's kappa ranged from .51 to .76).

Data Analysis. Analysis of coping strategies examined the multiple (up to three) strategies identified by respondents. To examine multiple responses, data were summed by coding strategy--first for each vignette (0 to 3) and then across all nine vignettes. Thus, for each type of strategy (for example, appraisal), the possible range is from 0 to 27; 0 if no appraisal responses were given, up to 27 if only appraisal responses were given for all three responses to each vignette. "Don't know" responses were treated differently; if no responses were given, respondents received a 0 for the vignette; added across vignettes, the range of possible don't know or nonresponses was 0 to 9.

For several analyses, the coping responses were grouped into three larger categories: (1) positive coping (problem-solving specific, help seeking, information seeking), (2) neutral coping (appraisal, problem-solving nonspecific-vague, and internal-affective regulation-acceptance), and negative coping (don't know, not a strategy, inappropriate, escape, emotional discharge). These broader categories were based on the theoretical framework of the program that taught participants the positive coping strategies, helped participants limit negative strategies, and encouraged some of the neutral strategies, depending on the problem context. For example, a fairly common response of "study harder" (for scenarios involving school performance problems) was coded as "internal"- a type of neutral strategy. Clearly, in some situations greater individual effort may be warranted. If, however, the student already is working at full capacity but continues to have academic difficulty because of limited time to study, a poor or unfair instructor, or simply is attempting to master a subject for which he or she is not yet ready, "studying harder" may only result in increased frustration.

Analysis examined several areas: descriptive information on coping, including types of strategy and perceived sources of help, effects of condition assignment (intervention model) and participation level on coping, and the influence of anticipated coping on later outcomes. The influences of condition and participation were examined using multivariate analysis of variance (MANOVA). Other analysis techniques (correlation, regression, and nonparametric statistics) also were used, as appropriate.

RESULTS

Coping Strategies

Emotional responses were the least frequent coping strategies (Table 2); problem-solving strategies (specific and nonspecific) were the most frequent at both midterm and graduation. When individual vignettes were examined, specific problem-solving strategies were generally the most frequent with a few exceptions. For vignette 2 (choosing a career), information seeking was the most frequent coping strategy; for vignette 4 (unfair treatment by instructor), help seeking was the most frequent strategy; and for vignette 9 (other students giving one a hard time), help seeking was again the most frequent strategy.

MSERP Staff and Vocational Rehabilitation Staff as Sources of Help

Because the intervention was designed to empower respondents to seek resources for themselves, an expectation of the program was that reliance on MSERP staff would decrease through the course of the intervention and identification of vocational rehabilitation VR staff as a resource would increase. To test this, we coded any mention of MSERP staff or VR agency staff as the targets of help seeking. Four dichotomous variables were created: (1) any MSERP target at midterm, (2) any VR target at midterm, (3) any MSERP target at graduation, and (4) any VR target at graduation. These variables were each cross-tabulated with condition and participation level to examine associations. At midterm, participation level was associated with identifying MSERP staff as a source of help [[X.sup.2](2, N= 293) = 9.88,p [less than] .01]. Twenty-two percent of high participants, 15 percent of moderate participants, and 7 percent of nonparticipants mentioned MSERP as a target of help-seeking behavior. At graduation, participation wa s not associated with targeting MSERP staff as a source of help. However, participation was associated with identifying VR as a target of help seeking ([X.sup.2](2, N= 293) = 12.47, p [less than] .01). At graduation, 21 percent of high participants, 7 percent of moderate participants, and 6 percent of nonparticipants mentioned VR as a target of help-seeking behavior.

Condition and Participation Effects on Coping at Graduation

Condition and participation effects on coping at graduation were examined in two ways: (1) overall coping (across all nine vignettes) and (2) coping by individual vignette. Condition and participation effects on coping at midterm were also examined. All MANOVAs were nonsignificant at midterm.

A series of MANOVAs was conducted to examine influences of condition and participation level on the three grouped strategies (positive, negative, and neutral) summed across all nine vignettes. At graduation, a significant participation effect was found for the MANOVA for positive coping strategies (Wilks's lambda = .92; F[6, 564] = 3.92, p [less than] .05). Univariate F tests found significant participation effects for specific problem solving (F[2, 284] =4.02, p [less than] .05) and help seeking (F[2, 2841 = 8.14, p [less than] .001). High participants reported the most specific problem-solving strategies (4.96) followed by moderate participants (4.58) and nonparticipants (4.05). High participants also listed the most help-seeking strategies (3.01) followed by nonparticipants (1.73) and moderate participants (1.66). MANOVAs were nonsignificant for groupings of both neutral and negative coping behaviors.

Coping Strategies and Later Outcomes

Bivariate correlations were calculated between each of the summed coping strategies reported at graduation and nine outcome measures reported 12 months after graduation (Table 3).

Specific problem-solving strategies were correlated with five outcome variables at 12 months, including three continuous variables: (1) school efficacy (r =. 18), (2) social support (r = .26), and (3) adjustment problems (-.20), and two dichotomous variables: (1) college enrollment (.17) and (2) involvement with VR (.19). The positive strategy of help seeking was correlated with college enrollment (.15) and involvement with VR (.17); and the strategy of information seeking was correlated with VR involvement (.15). Emotional coping was correlated negatively with school efficacy (r = -.15) and positively correlated with adjustment problems (r = .24). In addition, the summed variable, positive coping, was significantly correlated with four later outcomes.

For the relationships found to be significant in the correlation analysis, regression analysis then examined the influence of the graduation coping strategy score on 12-month outcomes, controlling for the effect of each outcome measured at graduation. This analysis examined the predictive effect of coping strategy assessed at graduation on change in outcome from graduation to 12-month follow-up (Table 4). Specific problem solving was found to have a significant positive effect on both social support [[beta] = .19, p [less than] .01) and school efficacy ([beta] = .15, p [less than] .01) at 12-month follow-up, holding the graduation measures of these variables constant. Emotional coping was found to have a significant effect on social adjustment problems ([beta] = .17, p [less than] .01); increased emotional coping led to more social adjustment problems.

Logistic regression analysis (not shown) examined the effect of coping strategies on dichotomous outcomes. The effect of specific problem-solving coping behavior on college enrollment was nonsignificant, and its effect on VR involvement was marginal (exp[[beta]] = 1.15, p [less than] .10). The effect of help seeking on VR involvement was also marginal (exp [[beta]] = 1.17, p [less than] .10). Escape or avoidance strategies had an unexpected, significant, and positive effect on paid employment at 12-month follow-up (exp[[beta]] = 1.52, p [less than] .01), indicating that increased use of escape or avoidance strategies resulted in an increased likelihood of employment at 12-month follow-up.

DISCUSSION

The literature review raised several key questions concerning coping and its measurement: To what extent is coping context-specific? Can interventions influence coping strategies? Are coping strategies related to later outcomes? What advantages are there for qualitative methods to assess coping?

Our results found that coping was largely context specific; coping strategies varied across scenarios, although problem-focused strategies were predominant. In most vignettes, specific problem solving was the most common strategy used to cope with school-related stress; in a few scenarios where this was not the case, other problem-oriented strategies, such as information seeking and help seeking, were the modal strategies. Moreover, targets of help seeking varied according to the type of situation described. In some situations MSERP staff were targeted more frequently than in others, and in some situations VR staff were targeted.

Much of the analysis focused on whether coping can be influenced by the intervention. Results in several areas provide support for the effect of the intervention on coping strategies. Findings focusing on condition and participation differences indicated that participation in the intervention influenced the attainment of positive coping skills, including specific problem solving, information seeking, and help seeking. In particular, high participants were found to exhibit more positive coping skills at graduation because they had more exposure to the intervention, which resulted in increased learning and practice of problem-solving skills. This is consistent with other MSERP results. In other analyses, participation level also was found to be related to important outcomes such as program satisfaction and learning, social support and support for education, and involvement in college and vocational activities (Collins et al., 1998; Mowbray et al., 1999).

Similarly, in terms of the targets of help seeking reported by respondents, the findings demonstrated both variation across vignettes and change that was consistent with program design. During the course of the program (midterm), MSERP staff were mentioned frequently as targets of help seeking, whereas at the end of the program (graduation),VR staff were mentioned more frequently as targets of help seeking. As participants are encouraged to seek appropriate help for addressing stressors and attain guidance and practice in doing so, it is the intent of MSERP that these skills be applied outside of the immediate program. The findings related to the shift from MSERP staff to VR staff are consonant with this goal and suggest that the program had one of its intended effects. Moreover, we also should note that many of the help-seeking responses of participants identified peers in the program and family and friends outside the program. Data on these other sources of help were not collected systematically, however.

There also was evidence that coping was related to later outcomes. The findings were clear that specific problem-solving coping activities were associated with better outcomes, whereas emotional coping responses were associated with worse outcomes. Regression analysis demonstrated that coping strategies were significant predictors of change in outcomes related to self-perception. The percent of variance accounted for on the self-perception scales was slight, however, reflecting the numerous other factors beyond individual coping that affect self-perception. Coping strategies were less successful at predicting behavioral outcomes examined in the logistic regression analysis. Only one relationship was found to be significant at the .05 level; escape or avoidance coping increased the likelihood of later paid employment. A possible reason for this unexpected relationship may be that because the scenarios are designed to reflect school-related stressors, those individuals choosing to escape or avoid these types o f stressors do in fact avoid school enrollment or drop out of school and return to the work force. In general, the lack of an effect of positive coping on school-related outcomes may be the result of the fairly short time frame for follow-up.

In summary, our findings suggested that coping strategies varied across situations according to the context of the vignettes, that higher participation in the program influenced more action-oriented approaches, and that specific problem-solving approaches were predictive of later positive outcomes, whereas emotion-oriented responses were predictive of later negative outcomes. The findings add to the program theory of supported education. Program components aimed at identifying possible stressors to be encountered in an educational setting and preparation of students to address these stressors can be taught and can be effective in influencing later outcomes. Further research needs to be conducted to examine this relationship and its processes more fully.

In regard to the utility of qualitative measurement for assessing coping, we believe there are several strengths in our approach to the collection and analysis of coping data. Use of an inductive approach allowed the participants to provide a description of their intended behavior without the influence of prescribed categories. Our strategy also allowed us to code initially at a very fine level; we were able to derive specific categories within the larger categories (appraisal, problem oriented, emotion oriented) and for some categories also to derive level of specificity (specific, nonspecific, vague). It is optimal to be able to code at the finest level possible to avoid the loss of potentially useful information. Although for most analyses, categories were grouped into larger units, this can be done with the knowledge that the loss of information is minimal rather than unknown. This was particularly needed because of the unique nature of our program and target population.

Another strength of our approach was the use of a variety of scenarios, each relevant to the intervention, but each also presenting a problem that could be changed. Such variety is needed to evoke a range of responses to limit the problem of skewed distributions likely to occur if only a few types of scenarios are used. It also allowed us to examine the possibility that different coping strategies are used in different scenarios.

Finally, larger issues, including the experimental and longitudinal nature of the design, as well as the inclusion of several other outcome measures to assess the predictive effects of anticipated coping behaviors on important outcomes strengthened our entire approach.

A few weaknesses of our approach also should be addressed. Although collected in an open-ended format, lack of thorough probing may have resulted in data that provided limited descriptive richness and context. The nature of the intervention also dictated our use of hypothetical scenarios. Because many of the participants had not yet experienced a school environment, we could not ask about the participants' past attempts to cope; instead we needed to frame the questions to determine what their likely coping responses would be. The extent to which anticipated coping reflects actual coping cannot be answered with these data. Finally, the process of data collection and analysis we described required substantially more time and effort than would the administration of a coping checklist or other standardized measure. Clearly, the widespread use of standardized coping inventories results in large part from their ease and efficiency of use. Nonetheless, we believe that given the limitations of standardized coping in struments and the highly diverse circumstances surrounding coping behavior, alternatives can and should be explored.

IMPLICATIONS FOR SOCIAL WORK PRACTICE AND RESEARCH

Despite the limitations discussed, the study provides important information about coping strategies of adults with mental illness, the relationship of an intervention to planned coping behaviors, the effect of coping on later outcomes, and the measurement of coping strategies. Social workers engaged in individual or group practice with people experiencing a mental or physical illness are well aware of the importance of effective coping in attaining successful outcomes. This study provided more detail about coping among a specific population engaged in a unique intervention. From this study there are several key implications for practice. Because the study provides further confirmation that coping is related to outcome, it implies that program components should include efforts to facilitate effective coping through means such as identification of potentially stressful scenarios, generation of possible responses, discussion of possible outcomes of responses, role-play and real-life practice of coping responses, and ongoing support from staff and other participants in processing the planned responses and the results of dealing with stressful situations.

Also, the study provided some indications of how and why an intervention may influence coping, and consequently offered suggestions about elements of an intervention key to improving coping strategies. First, although no condition (group, class, individual) effects were found, the fact that level of participation effects were found implies that coping maybe taught and practiced effectively in various program models. Thus, interventions need to attend to possible barriers to participation and act to reduce those barriers. Second, in combination with the extent of participation, the length of the intervention also appeared to be an important factor; that is, participation effects were not found after one semester but were found after two. Because the intervention extended over two semesters (with staff available for consultation after program completion as well), the participants in this program were allowed sufficient time for continued practice of skills and to learn about and develop outside sources of supp ort to use after the program ended. Finally, the fact that the study found far more evidence of an increase in positive coping than a decrease in negative coping implies that programs should emphasize the former rather than the latter. It appears more effective to teach new coping skills than to reduce or eliminate previously learned negative strategies, at least in the short-term time frame of this study. Thus, based on the findings interventions can use a variety of models, should act to reduce barriers to participation, should be of sufficient length, and should emphasize the teaching of positive coping rather than the reduction of negative coping.

In terms of research implications, our review of measurement instruments suggested weaknesses with well-established coping approaches, particularly with specific scenarios and unique populations. We believe that our measurement strategy was more appropriate and elicited more information than would have been the case if we had used a standardized coping measure instrument. Other programs with targeted interventions and serving unique populations also would be likely to benefit from a more qualitative approach. It would be useful in subsequent research of the population of interest in this study to focus on actual, rather than intended coping as program participants advance in their educational activities. Other coping research on adults with psychiatric disabilities might examine coping in other contexts. For example, important questions that remain to be examined include: Are the processes and outcomes of coping with work or family relationships similar to those of coping in an educational environment? What differences in coping exist among people engaged in higher education who have had the benefit of supported education compared with students who have not been in such a program? How do individuals cope with an initial diagnosis of mental illness? This last question, in particular, probably would require further examination of coping strategies that emphasized broader strategies than problem solving.

CONCLUSION

Whereas the focus of this article has been on the coping of adults with psychiatric disabilities who participated in MSERP, the larger questions raised about coping processes, outcomes of coping, and measurement issues are not limited to this population or this intervention, but are critical to several areas of social work. Consequently, the general method used here need not be limited to evaluations of interventions with people who have psychiatric disabilities. Other program evaluations in which coping is a key element, particularly those serving populations for whom standardized instruments may be weak, might find it useful to use a qualitative approach to measuring coping, assessing changes in coping over time, and relating coping to outcomes.

ABOUT THE AUTHORS

Mary Elizabeth Collins, PhD, is assistant professor, School of Social Work, Boston University, 264 Bay State Road, Boston, MA 02215. Carol T. Mowbray, PhD, is associate professor and associate dean for research, School of Social Work, University of Michigan, Ann Arbor. Deborah Bybee, PhD, is associate professor, Department of Psychology, Michigan State University, East Lansing. This study was funded by grant number HD5-SM47669 from the Community Support Program Branch of the Center for Mental Health Services to the Michigan Department of Mental Health. It represents a collaboration among the schools of social work at the University of Michigan, Eastern Michigan University, and Wayne State University and the Detroit--Wayne County Community Mental Health Services Agency. The authors thank William Shriner for developing the coping measure and Pauldanette Bailey and Beverly Orlowski for assistance in coding. Correspondence regarding this manuscript should be addressed to Mary Elizabeth C ollins, PhD, 264 Bay State Road, Boston, MA 02215.

REFERENCES

Aldwin, C. M., & Revenson, T. T. (1987). Does coping help? A reexamination of the relation between coping and mental health. Journal of Personality and Social Psychology, 53, 337-348.

Banyard, V. L. (1995). "Taking another route": Daily survival narratives from mothers who are homeless. American Journal of Community Psychology, 23, 871-891.

Bargagliotti, L. A., & Trygstad, L. N. (1987). Differences in stress and coping findings: A reflection of social realities or methodologies? Nursing Research, 36, 170-173.

Billings, A. G., & Moos, R. H. (1981). The role of coping responses and social resources in attenuating the stress of life events. Journal of Behavioral Medicine, 4, 139-157.

Billings, A. G., & Moos, R. H. (1984). Coping, stress, and social resources among adults with unipolar depression. Journal of Personality and Social Psychology, 46, 877-891.

Carver, C. S., Scheier, M. F., & Weintraub, J. K. (1989). Assessing coping strategies: A theoretically based approach. Journal of Personality and Social Psychology, 56, 267-283.

Collins, M. E., Bybee, D., & Mowbray, C. T. (1998). Effectiveness of supported education for individuals with psychiatric disabilities: Results from an experimental study. Community Mental Health Journal, 34, 595-613.

Compas, B. E., Malcarne, V. L., & Fondacaro, K. M. (1988). Coping with stressful events in older children and younger adolescents. Journal of Consulting and Clinical Psychology, 56, 405-411.

Folkman, S., & Lazarus, R. S. (1980). An analysis of coping in a middle-aged community sample. Journal of Health and Social Behavior, 21, 219-239.

Hazel, K., Herman, S. E., & Mowbray, C. T. (1991). Characteristics of adults with serious mental illness in a public mental health system. Hospital and Community Psychiatry, 42, 518-525.

Heppner, P. P., Cook, S. W., Wright, D. M., & Johnson, W. C. (1995). Progress in resolving problems: A problem-focused style of coping. Journal of Counseling Psychology, 42, 279-293.

Morey, L. C. (1991). Personality assessment inventory: Professional manual. Odessa, FL: Psychological Assessment Resources, Inc.

Mowbray, C., Collins, M. E., Bybee, D. (1999). Supported education for individuals with psychiatric disabilities: Long-term outcomes from an experimental study. Social Work Research, 23, 89-100.

Mullen, B. J., & Suls, J. (1982). The effectiveness of attention and rejection as coping styles: A meta-analysis of temporal differences. Journal of Psychosomatic Research, 26, 43-49.

Oakland, S., & Ostell, A. (1996). Measuring coping: A review and critique. Human Relations, 49, 133-155.

Pearlin, L. I., & Schooler, C. (1978). The structure of coping. Journal of Health and Social Behavior, 19, 2-21.

Stanton, A. L., Danoff-Burg, S., Cameron, C. L., & Ellis, A. P. (1994). Coping through emotional approach: Problems of conceptualization and confounding. Journal of Personality and Social Psychology, 66, 350-362.

Stone, A. A., Kennedy-Moore, E., & Neale, J. M. (1995). Association between daily coping and end-of-day mood. Health Psychology, 14, 341-349.

Tessler, R. C., & Goldman, H. H. (1982). The chronically mentally ill: Accessing community support programs. Cambridge, MA: Ballinger.

Unger, K.V., & Anthony, W.A. (1984). Are families satisfied with services to young adult chronic patients? A recent survey and a proposed alternative. New Directions for Mental Health Services, 21, 91-97.

Unger, K., Danley, K., Kohn, L., & Hutchinson, D. (1987). Rehabilitation through education: A university-based continuing education program for young adults with psychiatric disabilities on a university campus. Psychosocial Rehabilitation Journal, 10(3), 35-49.

Vitaliano, P. P., DeWolfe, D. J., Maiuro, R. D., Russo, J., & Katon, W. (1990). Appraised changeability of a stressor as a modifier of the relationship between coping and depression: A test of the hypothesis of fit. Journal of Personality and Social Psychology, 59, 582-592.

NOTE: In addition, coding instructions were provided for each individual vignette using the same categories listed here, but with more detailed instructions and examples to match the particular vignette. The full codes can be obtained from the first author.

NOTE: The overall chance probability of finding the number of significant correlations displayed in this table is less than .01 if the grouped strategies are not considered and less than .001 if the group strategies are included.

(*.)p[less than or equal to].05.

(**.)p[less than or equal to].01.
Final Coding Scheme: General Definitions
Cognitive
11                                  Appraisal: Analysis-assessment of
                                    problem situation (cause of
                                    problem and options to resolve), no
                                    problem-focused or emotion-
                                    focused response given.
Problem Solving: Efforts or actions
to address the problem.
21                                  Specific: Specific, concrete steps
                                    directly aimed at problem.
22                                  Nonspecific: More general actions that
                                    would solve problem.
23                                  Vague: The response is related to the
                                    problem but is too vague to
                                    indicate a real possibility of solving
                                    the problem; or
                                    preparatory steps that may eventually
                                    lead to more direct
                                    problem solving.
24                                  Ask for help, guidance, or support:
                                    Talking to people other than
                                    person causing problem for purposes of
                                    emotional support, advice
                                    or asking someone to interecede.
25                                  Information seeking: Efforts to gather
                                    more information to
                                    address problem.
Emotional-Internal Efforts
31                                  Internal effort: Attempts to put
                                    forward greater personal effort
                                    to solve situation.
32                                  Affective regulation: Efforts
                                    (behaviors) to keep calm, relax,
                                    avoid emotional discharge, and so
                                    forth.
33                                  Emotional discharge: Openly venting
                                    one's feelings (for example,
                                    yell, cry).
34                                  Escape-avoidance: Escaping or avoiding
                                    difficult situations to
                                    reduce negative feelings.
35                                  Acceptance or resignation: Recognition
                                    that situation can not be
                                    changed, decision to make the best of
                                    the situation.
Miscellaneous
41                                  Would not happen: Response that
                                    suggests the person would not be
                                    in the situation described.
42                                  Inappropriate: Response that is
                                    unrelated to problem at hand, or
                                    socially inappropriate.
43                                  Inappropriate-bizarre: Response is not
                                    only socially
                                    inappropriate, but is completely
                                    unrealted to reality.
44                                  Not a strategy: Response that is a
                                    factual statement, describes
                                    past response, or otherwise does not
                                    respond to the question.
51                                  Other: Any response not falling into
                                    listed categories, including
                                    prayer or spiritually realted
                                    responses.


Frequency of Coping Strategies at Midterm and Graduation
                                  Midterm            Graduation
                                 (n = 313)            (n = 293)
Coping Strategy                    Range    M    SD    Range     M    SD
Positive                           0-17    6.72 3.24   0-20     7.07 3.32
 Specific problem solving          0-13    4.41 2.37   0-11     4.47 2.25
 Help seeking                      0-11    1.80 1.65   0-9      1.89 1.65
 Information seeking               0-3      .50  .71   0-4       .73  .88
Negative                           0-9     1.64 1.77   0-12     1.83 1.78
 Escape                            0-4      .68  .87   0-5       .83 1.01
 Not a strategy                    0-7      .51 1.06   0-8       .54 1.06
 Inappropriate                     0-5      .39  .84   0-5       .40  .72
 Do not know                       0-9      .14  .82   0-9       .11  .83
 Emotional                         0-3      .06  .31   0-2       .07  .28
Neutral                            0-15    4.78 2.61   0-15     4.48 2.48
 Nonspecific problem solving [a]   0-9     3.21 1.88   0-9      3.14 1.91
 Internal [b]                      0-5      .99 1.14   0-6       .83  .96
 Appraisal                         0-5      .59  .90   0-6       .52  .85
(a.)includes both nonspecific and vague problem solving.
(b.)includes internal, affective regulation, and acceptance-resignation.
    Correlation of Graduation Coping Strategies with 12-Month Outcomes
                                      School    Quality
                            Learning Efficacy   of Life Adjustment
Specific problem solving      .05      .18 [**]   .08     -.20 [**]
Help seeking                  .09      .04        .02     -.01
Information seeking           .07     -.05       -.04     -.08
Appraisal                     .08      .01       -.08      .02
Nonspecific problem solving  -.04     -.03       -.10      .13
Internal                     -.07     -.11       -.07      .05
Not a strategy                .03      .02       -.02      .03
Inappropriate                 .05      .02       -.12      .07
Escape-avoid                  .10     -.02        .01     -.02
Emotional                    -.04     -.15 [*]   -.12      .24 [*]
Grouped strategies
  Positive                    .10      .13        .05     -.15 [*]
  Negative                    .08     -.01       -.07      .07
  Neutral                    -.03     -.06       -.12      .12
                            Support for  Social                Paid
                             Education  Support    College     Work
Specific problem solving        .09       .26 [**]    .17 [**]   .10
Help seeking                    .06       .05         .15 [*]    .00
Information seeking            -.02       .07         .09        .08
Appraisal                      -.01       .07        -.05       -.04
Nonspecific problem solving     .05       .05        -.06       -.06
Internal                       -.09       .06        -.01       -.05
Not a strategy                  .07       .00         .06      -0.11
Inappropriate                  -.07      -.07         .08        .10
Escape-avoid                   -.04       .08        -.05        .15 [*]
Emotional                       .05      -.01         .11       -.04
Grouped strategies
  Positive                      .07       .21 [**]    .22 [**]   .09
  Negative                      .00       .02         .06        .06
  Neutral                      -.08       .08        -.07       -.08
                              Vocational
                            Rehabilitation
Specific problem solving         .19 [**]
Help seeking                     .17 [**]
Information seeking              .15 [*]
Appraisal                        .05
Nonspecific problem solving     -.05
Internal                        -.04
Not a strategy                  0.01
Inappropriate                    .08
Escape-avoid                     .06
Emotional                        .05
Grouped strategies
  Positive                       .26 [**]
  Negative                       .08
  Neutral                       -.04


Effects of Coping Measured at MSERP Graduation:
           OLS Regression Analysis on Change in Self-Perception
              Outcomes from Graduation to 12-Month Follow-Up
Outcome (12 months)        Predictors (graduation)      [beta]
Social support             Social support              .38 [***]
                           Specific problem solving    .19 [**]
School efficacy            School efficacy             .55 [***]
                           Specific problem solving    .15 [**]
Social adjustment problems Social adjustment problems  .49 [***]
                           Specific problem solving   -.11 NS
School efficacy            School efficacy             .56 [***]
                           Emotional                  -.10 NS
Social adjustment problems Social adjustment problems  .47 [***]
                           Emotional                   .17 [**]
                             Marginal
Outcome (12 months)        Adj [R.sup.2]
Social support                  .17
                                .03
School efficacy                 .31
                                .02
Social adjustment problems      .26
                               0
School efficacy                 .32
                               0
Social adjustment problems      .26
                                .01
(*.)p[less than or equal to].05.
(**.)p[less than or equal to].01.
(***.)p[less than or equal to].001.
Gale Copyright: Copyright 1999 Gale, Cengage Learning. All rights reserved.