Burnout, Compassion Fatigue, and Vicarious Traumatization Paper

An Examination of Burnout Among School Counselors Guided by Stress-Strain-Coping Theory Kevin Wilkerson

• h r h e Maslach Burnout Inventory-Educators Survey (MBI-ES; C. Maslach, S. E. Jackson, & M. P. Leiter, 1996) was used to examine burnout among professional school counselors (N= 198). Guided by stress-srain-coping theory, final hierarchical regression models accounted for 49% of the variation on the Emotional Exhaustion scale, 27% on the Depersonalization scale, and 36% on the Personal Accomplishment scale. Numerous individual stress and coping variables significantly predicted burnout among school counselors in the multivariate context.

Professional school counselors work to maximize scholastic achievement and lifelong learning among students, to promote an understanding of careers and the world of work, and to provide all students with opportunities for healthy personal and social growth (Campbell & Dahir, 1997). Yet, they are frequently called on to perform additional duties as part of their daily work. As a result, school counselors are constantly asked to set priorities and make decisions about how to best perform their job. This can lead to increased stress (Burnham & Jackson, 2000; Johnson, 2000). Stress-strain-coping theory (Lazarus & Folkman, 1984) indicates that there is a concern that, without adequate coping skills, elevated levels of stress may lead to increased symptoms of burnout. With outcome symptomatology that includes ineffective delivery of services, exhaustion, physical complaints, anxiety, depression, and substance abuse (Huebner & Huberty, 1984; Kahili, 1988; Reiner & Hartshome, 1982), impairment may result from such burnout.

The following study frames burnout within the stress- strain-coping continuum and seeks to examine which variables might place professional school counselors at greatest risk for substandard performance. Stress-strain-coping theory is briefly explored, followed by an examination of the three variable sets included in this study: (a) demographic Stressors, (b) organizational Stressors, and (c) coping styles.

Transactional Approach to Understanding Burnout: Stress-Strain-Coping Theory

Endler and Parker (1999) emphasized that one's coping style plays an important role in both physical and psychologi- cal well-being. Inadequate coping styles exacerbate stress, resulting in the increased likelihood of burnout. Conversely, constructive coping strategies diminish stress, thereby lead-

ing to a decreased likelihood of burnout symptomatology. Leiter (1991) provided initial supporting evidence for these claims. In a study of workers in a mental health hospital (N = 177), his findings indicated that problem-oriented cop- ing strategies, those initiated to enact change, were associated with decreased levels of burnout, whereas escapist, or avoid- ance, coping strategies were associated with increased levels of burnout. These findings indicate that stress-strain-coping theory may provide one useful way to better understand the burnout phenomenon.

• s t r e s s e s and Strains: Variables Associated With Burnout Among School Counselors

Burnout is most frequently understood as a pattern of nega- tive symptoms directly related to all areas of functioning (Savicki & Cooley, 1982). The Maslach Burnout Inventory (MBI; Maslach & Jackson, 1981) was developed to assess burnout across three dimensions: emotional exhaustion, depersonalization, and personal accomplishment. Emotional exhaustion is characterized by symptoms of physical and emotional depletion. Depersonalization describes a lack of empathy and emotional distance from the clients with whom one works. Finally, personal accomplishment describes feel- ings of competence in one's job.

Because of limited scholarship on stresses and strains as they relate to burnout among school counselors, related stud- ies from the ñeld of school psychology were also reviewed to assist with the development of this study. Results from these studies indicate two potential variable sets of stresses and strains for consideration: demographic variables and organizational variables. First, two demographic variables (i.e., the size of one's caseload and one's years of experience in the profession) have been associated with burnout among

Kevin Wilkerson, Department of Counseling and Human Services, Syracuse University. Kevin Wilkerson is now at Department of Counseling and Human Services, University of Scranton. The content of this article is derived from dissertation research conducted by the author. The author wishes to acknowledge the Association for Counselor Education and Supervision and the Syracuse University Schooi of Education for their generous grants in support of this project. Correspondence concerning this article should be addressed to Kevin Wilkerson, Department of Counseling and Human Services, University of Scranton, 445 McGurrin Hall, Scranton, PA 18510 (e-mail: [email protected]).

428 © 2009 by the American Counseling Association. All rights reserved.

Journal of Counseling & Development Fall 2009 • Volume 87

Burnout Among School Counselors

mental health professionals in schools (Butler & Constantine, 2005; Huberty & Huebner, 1988; Huebner, 1992,1993; Mills & Huebner, 1998; Niebrugge, 1994; Reiner & Hartshome, 1982; Sandoval, 1993; Stickel, 1991; Wilkerson & Bellini, 2006). More specifically, studies have revealed that school counselors with large caseloads are more prone to exhibit symptoms of burnout. The literature on years of experience has been mixed. Butler and Constantine found higher levels of burnout among individuals who had been in the school counseling profession for long periods of time (20-29 years), whereas other studies noted higher levels of burnout among those newer to the profession. Second, studies have also re- vealed that several organizational variables (i.e., role conflict, role ambiguity, interpersonal conflicts, and one's satisfaction with supervision) are also associated with burnout (Huberty & Huebner, 1988; Huebner, 1992, 1993; Niebrugge, 1994; Sandoval, 1993; Wilkerson & Bellini, 2006). Furthermore, although there is a dearth of empirical data directly related to school counselors and burnout, literature suggests additional school counselor Stressors that may be related. In a pilot study and subsequent national study. Gray (1982) identified six organizational Stressors for school counselors: lack of decision-making authority, financial security, nonprofessional duties, job overload, coimselor-teacher professional relation- ships, and counselor—principal relationships. Although these variables have not been evaluated in terms of their relation- ship with burnout, stress-strain-coping theory leads one to hypothesize strong correlations between these Stressors and outcome measures of burnout.

oping as a Moderating Variable for Burnout

In addition to demographic and organizational variables, coping has also been discussed as a moderating variable for burnout. Cronkite and Moos ( 1984) reported that some people experience high levels of stress, yet very few detrimental symptoms, whereas others report the opposite. Given this finding, it is also possible that some persons reporting high levels of stress may report relatively low levels of burnout. Still others may report low levels of stress and contrasting high levels of burnout. If this is the case, it may be that different coping strategies actually moderate the effects of burnout. In other words, stress and subsequent burnout may actually be conditional and based on one's coping approaches. Thus, the relationship between stress and the onset of burnout may depend on the presence, absence, or level of coping strategies used by a given individual.

•Framework for the Current Study Based on the preceding discussion, this study focused on one current theory that posits a direct relationship between the stress-strain-coping continuum and measured outcomes of burnout. Using recent findings from the field of burnout

and building on the limited research specific to school coun- selors, this inquiry addressed some of the preliminary ques- tions regarding burnout among this population of counseling professionals.

•Research Questions For this study, there were four exploratory research questions:

• What is the level of burnout reported by a national sample of currently practicing school counselors in the United States?

• What amount of variance do the three identified inde- pendent variable sets (demographic and organizational Stressors plus individual coping strategies) contribute to outcome dimensions of burnout?

• Within each variable set, what are the most influential predictors of burnout?

• Are the relationships between occupational stress and outcome levels of burnout on the MBI-Educators Survey (MBI-ES; Maslach, Jackson, & Leiter, 1996) significantly moderated by different approaches to coping?

•Method Participants

A random sample of participants (A'̂ = 482) was drawn fi-om the national membership list of the American School Coun- selor Association (ASCA). The secured sample included practicing school counselors fi-om ASCA's four membership regions: North Atlantic (n =112), Midwest (n = 124), South (n = 123), and West (n = 123). Thirty-seven men ( 18.8%) and 161 women (81.2%) constituted the final sample (A^= 198), representing a 4 1 % response rate.

Participants ranged in age fi-om 24 to 63 years (M- 44.77, SD = 10.33). One hundred and seventy-eight (89.9%) respon- dents identified themselves as White, 5 (2.5%) as Black or African American, 4 (2.0%) as Asian American or Pacific Islander, 3 (1.5%) as Hispanic or Latino/a, and 3 (1.5%) as American Indian or Alaska Native. Four individuals (2.0%) checked the "other" category (1 participant did not respond to this category). Overall, respondents had been working as school counselors for an average of almost 9 years (M = 8.8, SD = 7.2) and had been in their current positions for an average of almost 6 years (M = 5.9, SD = 5.6). Ninety-six (48.5%) school counselor respondents reported working in a rural setting, 60 (30.3%) in a suburban setting, and 40 (20.2%) in an urban setting; 2 individuals (0.02%) did not respond to this question. In terms of the populations served, the majority of respondents, 114 (57.6%), reported working in an alternative configuration environment (e.g.. Grades K-12, K-8, or 7-12). Of those remaining, 45 (22.7%) re- ported working in a traditional high school setting (Grades 9-12), 22 (11.1%) in an elementary school setting (Grades

Journal ofCounseling& Development B Fall 2009 m Volume 87 429

Wilkerson

K-6), and 16 (8.1%) in amiddle school setting (Grades 7-8). (One participant did not respond to this question.) Finally, 139 (70.2%) of the sample's respondents stated that their job responsibilities were restricted to one school. An additional 58 (29.3%) participants reported multiple school responsibili- ties (Af = 2.29, SD = 4.10). (One participant did not respond to this category.)

Instruments

Role Questionnaire. The Role Questionnaire (RQ; Rizzo, House, & Lirtzman, 1970) was used to gather information on work- related Stressors that the literature indicates is associated with bumout. This 14-item questionnaire was developed to measure the degrees of role conflict, using the Role Conflict (RC) subscale, and role ambiguity, using the Role Ambiguity (RA) subscale, that individuals experience on the job. Role conflict refers to the experience of dealing with inconsistent role expectations or service demands, and role ambiguity refers to unclear role definitions or performance expectations. Freeman and Coil's (1997) factor analysis of the RQ using a sample of high school counselors (A'̂ = 417) identified one additional factor that was used in the current study. Role Incongruity (the Role Incongruity [RI] subscale), which was defined as the diflerence between the roles a school counselor actually performs and the roles he or she believes ought to be performed. Items on this survey are rated on a 7-point Likert scale: 1 (definitely not true of n^Job, 1 {extremely true of my job). Cronbach's alpha coeflicients for the current study were satisfactory (RC, a = .83; RI, a = .78; and RA, a = .90)

Counselor Occupational Stress Inventory. To further assess school counselor Stressors, four subscales from the Counselor Occupational Stress Inventory (COSI; Gray, 1982) were used: Financial Security (FS), Lack of Decision-Making Authority (LDMA), Counselor-Teacher Professional Relationships (CTPR), and Counselor-Principal Professional Relationships (CPPR). Two additional subscales. Job Overload and Nonpro- fessional Duties were not included because they overlapped substantially with dimensions measured by the RQ. Designed specifically for use with school counselors, this 29-item, 5- point Likert-type inventory measures the level of perceived stress associated with typical school counselor situations: 1 {not stressful), 5 (very stressful). The current study yielded in- temal consistency scores ranging from .89 to .96 (FS, a = .96; LDMA, a = .90; CTPR, a = .90; CPPR, a = .89).

Coping Inventoryfor Stressful Situations. The 48-item Cop- ing Inventory for Stressful Situations (CISS; Endler & Parker, 1999) uses a 5-point Likert-type scale (1 = not at all, 5 = very much) and measures three main strategies for coping with stress. The Task-Oriented Coping (TOC) subscale is a measure of the extent to which individuals deal directly with problems at hand. The Emotion-Oriented Coping (EOC) subscale measures the extent to which people concentrate primarily on the result- ing emotions brought on by stressful situations. Finally, the Avoidance-Oriented Coping (AOC) subscale measures the extent to which people try to avoid existing stressful situations

in their life (Cosway, Endler, Sadler, & Dreary, 2000). The current study resulted in the following coefficients: TOC, a = .86; EOC, a = .90; and AOC, a = .82

Moderator variables. To evaluate the amount of variance that interaction effects might be contributing to outcomes of bumout, an overall Occupational Stress Index score was calculated. This was accomplished by summing the seven dif- ferent subscale scores derived ftom the RQ and the COSI. The resulting score and each of the scores for the subscales measuring the coping strategies from the CISS were used to form the differ- ent interaction variables used in this analysis. These constructed variables were used to evaluate the potential moderating effects between occupational stress and coping on the outcomes of bum- out. The reliability coefficient for the constmcted Occupational Stress Index in the current study was .95.

MBI-ES. The outcome variable was assessed using the MBI-ES (Maslach, Jackson, & Leiter, 1996). This 22-item inventory uses a 7-point Likert scale (0 = never, 6 = every day). The inventory yields results for three subscales: Emotional Exhaustion (EE), Depersonalization (DP), and Personal Ad- justment (PA). There is no summed or total score for the entire survey. For the MBI-ES, high degrees of bumout are indicated by elevated scores on the EE and DP subscales and depressed scores on the PA subscale. Again, Cronbach's alpha coefficients for the current study on the three MBI-ES subscales were sat- isfactory: EE, a = .91 ; DP, a = .74; and PA, a = .73.

Procedure

Permission was obtained to reproduce copyrighted inventories onto one single document. The final survey packet contained the assessment instruments plus a demographics section. Five forms of the survey (labeled A-E) were developed to limit the possibility of response bias. The demographic section remained at the end of the survey in each form. Participation was requested via an initial mailing {N= 482). Follow-up postcards were sent out at both the 3- and 5-week points to encourage participation.

•Results A comparison of response rates indicated a fairly even dis- tribution of participants across the four ASCA membership regions. Of those sampled in the Midwest, 44% responded (n = 55), 45% from the North Atlantic {n = 50), 37% from the South (n = 45), and 39% from the West (« = 48). Table 1 presents means and an intercorrelation matrix for each of the study's independent variables. Means for the outcome variables were as follows: EE, M = 20.84, SD = 10.16; DP, M = 4.81, SD = 4.51 ; PA, M = 40.86,5Z) = 4.66. Table 2 presents the zero-order correlations between the independent variables and each subscale score on the MBI-ES.

Hierarchical Regression Analyses

A four-step hierarchical multiple regression model was used to analyze the individual and cumulative contributions made

430 Journal ofCounselingôC Development B Fall 2009 u Volume 87

Burnout Among School Counselors

TABLE 1

Means and Intercorrelations for Independent Variables Predicting Outcomes of the Maslach Burnout Inventory-Educators Survey

Variable M SD 10 11 12

1. YOETOT 2. SCRAT COSI

3. LDMA 4. FS 5. CTPR 6. CPPR

RQ 7. RC 8. Rl 9. RA"

CISS 10. TOO" 11. EOC 12. AOC"

8.80 369.60

17.89 30.13 28.38 11.16

15.43 15.46 20.50

57.06 40.86 48.79

7.23 194.60

5.84 11.67 8.16 4.65

5.96 5.63 7.89

7.22 9.28 8.29

— .02

-.15* -.16* -.13* -.11

-.10 -.09 -.27**

.19** -.10 -.20**

—

-.04 -.06 -.09 -.09

.01 -.04 -.07

.08 -.11 -.04

— .49** .62** .78**

.26**

.34**

.32**

.00

.21**

.21**

— .52* .50*

.18*

.31*

.24*

-.09 .22* .22*

* — ' .60**

" .30** ' .39**

.31**

.03

.36** ' .26**

—

.27**

.31**

.34**

.00

.22**

.16*

— .64** .57**

-.12* .33** .08

— .48** —

-.10 -.26** .34** .32** .14* .06

— -.28

.00 .31** —

Note. YOETOT = years of experience, total (for this study, totai years of school counseling experience); SCRAT = student/counselor ratio; COSI = Counselor Occupational Stress Inventory; LDMA = Lack of Decision-Making Authority; FS = Financial Security; CTPR = Counselor- Teacher Professional Relationships; CPPR = Counselor-Principal Professional Relationships; RQ = Role Questionnaire; RC = Role Conflict; Rl = Role Incongruity; RA = Role Ambiguity; CISS = Coping Inventory for Stressful Situations; TQC = Task-Qriented Coping; EQC = Emotion- Qriented Coping; AOC = Avoidance-Oriented Coping. "Items on this subscale were reverse scored. ''Reported mean scores are converted t scores. *p^.O5. * * p < . 0 1 .

by the independent variables as well as any significant moderator interactions. Table 3 presents the final model variance accounted for by each step in the analysis, as well as the beta weights associated with each independent variable. The adjusted R^ for each step in the analysis is reported within the body of the following text.

For emotional exhaustion (as measured by the EE sub- scale), in Model 1, Step 1 did not significantly predict EE scores, but accounted for 3% of the variation in this outcome (adjusted R^ = .02). Factoring in organizational Stressors, the two-step model contributed 38% to the overall variation in EE (adjusted R^ = .35). The model at this stage was statisti- cally significant, F(9, 188) = 12.92,;? < .01, R^ = .38, as was the R^ added (.35). Coping strategies in Step 3 contributed an additional 9% to the model, bringing the overall variation accounted for to 47% (adjusted R^ = .44). In addition to the fiill model significance at this step, F(12, 185) = 13.86, j3 < .01, Ä̂ = .47, the incremental increase here was also significant (R^ added = .09). The final step brought the overall variation in the completed model to 49% (adjusted R^ = .45). The final model remained highly significant, F(15, 182) = 11.70,p < .01, Ä̂ = .49; however, this fourth step was not significant in and of itself (Ä^ added = .02, p < .10). Among the individual variables considered, the final regression analysis indicates that counselors with more total years of experience (YOETOT = years of experience, total; for this study, total years of school counseling experience), who also reported higher levels of role conflict and role incongruity (as measured by the RC and Rl), and used emotion-oriented coping (as measured by the EOC) to a greater degree, also reported increased emotional exhaustion (as measured by the EE).

In Model 2, evaluation of the four-step regression resulted in a significant contribution by the counselor demographic variables in Step 1 to the prediction of depersonalization (R^ = .04; adjusted R^ = .03). Organizational Stressors (Step 2) added an additional 13% to the model. This addition was statistically significant at the .01 level (R^ added = .13), and the model at this point remained statistically significant, with 17% of the overall variance accounted for, F(9, 188) = 4.18, p < .001, Ä̂ = .17; adjusted i?̂ = .13). Coping strategies (Step 3) contributed an additional 9% to the model representing an overall amount of variance accounted for by the model of 26%, F(12, 185) = 5.53,p < .001, R^ = .26; adjustedR^ = .22. The third step itself was also statistically significant (R^ added = .09,p < .01). Although the overall model remained statistically significant after the final step, F(15,182) = 4.48,/? < .01, R^ = .27; adjusted i?̂ = .22, the moderating effects of coping on the relationship between occupational stress and depersonalization (Step 4) did not contribute any additional appreciable variance to the overall model. In the final model, only the individual variables of YOETOT and scores on the EOC contributed significantly (/? < .01) to the prediction of scores on the DP.

Finally, the four-step hierarchical regression model for personal accomplishment (as measured by the PA subscale) yielded the following results. Although counselor demo- graphic variables (Step 1) did not significantly predict the outcome variable, accounting for only 2% of the variance (adjusted R^ = .00), the combination of contextual factors and organizational Stressors in Steps 1 and 2 yielded a highly significant model overall, F(9, 188) = 2.S6,p < .01, R^ = .12; adjusted R^ = .08, with Step 2 contributing an additional and

Journal ofCounseling& Development B Fall 2009 B Volume 87 431

Wilkerson

TABLE 2

Zero-Order Correlations Between Independent and Outcome Variables

Variable

Stress-Strain Cluster 1 : Counselor Demographics

YOETOT SCRAT

Cluster 2: Organizational Variables COSI

LDMA (z score) FS (z score) CTPR (z score) CPPR (z score)

RQ RC(z score) RI (z score) RA (z score)

Coping Cluster 3: Intrapersonal Variables

CISS TOC (z score) EOC (z score) AOC (z score)

Interaction Cluster 4: Moderator Variables

OSIx TOC (z score) OSIx EOC (z score) OSIx AOC (z score)

Outcome Variables EE"

.11 -.12*

.25*

.26*

.32*

.25*

.48*

.51*

.40*

-.19*' .51*' .09

-.13* .09

-.04

DP»

.19** -.10

.11

.04

.13*

.14*

.27**

.28**

.22**

- . 2 1 * * .37**

-.01

-.07 -.02

.01

PA'

.03

.13*

-.16* -.04 -.08 -.12*

-.16* -.01 -.25**

.37** -.38**

.09

.15*

.08

.00

Note. EE = Emotional Exhaustion; DP = Depersonalization; PA = Personal Adjustment; YOETOT = years cf experience, total (for this study, total years of school counseling experience); SCRAT = student/ counselor ratio; COSI = Counselor Occupaticnal Stress Inventory; LDMA = Lack of Decision-Making Authority; FS = Financial Security; CTPR = Counselor-Teacher Professional Relationships; CPPR = Counselor-Principal Professional Relationships; RQ = Role Question- naire; RC = Role Conflict; RI = Role Incongruity; RA = Role Ambiguity; CISS = Coping Inventory for Stressful Situations;TOC = Task-Oriented Coping; EQC = Emotion-Oriented Coping; AOC = Avoidance-Qriented Coping; QSI = occupational stress index. «Subscale of the Maslach Burnout Inventory-Educators Survey. *p < .05. **p á .01.

Statistically significant 10% to that already accounted for in Step 1 (If added = .lO,p< .01). Coping strategies in Step 3 accounted for an additional 2 1 % of the variation in scores on the PA subscale. Both this step itself (Ä^ added = .2l,p< .01) and the full model, F(12, 185) = 7.52,p< .01, R^ = .33; adjusted R^ = .28, continued to be statistically significant at the .01 level. Finally, contrary to the first two models, the moderating effects in this particular model also contributed a statistically significant amount of variance to the prediction of PA subscale scores (R^ added = .03,p < .05). The final model, accounting for 36% of the variation in PA subscale scores overall, was also significant, F(15, 182) = 6.72,p< .01, R^ = .36; adjusted R^ = .30.

Post Hoc Analyses

Interpretation of significant moderator interactions. Statisti- cally significant moderator interactions were identified in the

third regression model evaluating the relationship between the independent variables and personal accomplishment (as measured by the PA). Two out of the three moderator effects were found to be significant (OSI x TOC, OSI x EOC). Median splits for both the OSI and the significant coping variables (measured by the TOC and EOC) were executed to derive mean scores for the PA subscale in each of four categories: high occupational stress (OS)/high coping style, high OS/low coping style, low OS/high coping style, low OS/low coping style. Figure 1 provides a graphic representation of the first significant moderating effect, that of task-oriented coping (as measured by the TOC) on the relationship between overall school counselor occupational stress and feelings of personal accomplishment. In this case, task-oriented coping appears to have served as a buffer for the impacts of occupational stress on reported levels of personal accomplishment. Those report- ing relatively high usage of task-oriented coping strategies experienced a sense of personal accomplishment at similar levels regardless of their occupational stress category (low OS/high TOC, M = 42.43, SD = 4.13; high OS/high TOC, M = 42.00, SD = 4.40). However, these results were altered for those reporting less fi-equent usage of task-oriented coping strategies. Here the plotted means (low OS/low TOC, M = 40.24, SD = 4.41; high OS/low TOC, M = 38.65, SD = 4.71) indicated that the detrimental outcomes of high occupa- tional stress on participants' sense of personal accomplishment were more exaggerated among those with a low task-oriented coping approach. This finding suggests that adopting a more task-oriented coping style may serve to buffer individuals fi-om the deleterious effects of occupational stress on perceptions of personal accomplishment.

Contrary to the direct buffering effects of task-oriented coping on levels of occupational stress and one's sense of per- sonal accomplishment, the emotion-oriented coping variable (as measured by the EOC) actually served to amplify reduced feelings of personal accomplishment. Figure 2 illustrates these findings. To explain, the mean score for the PA subscale was lowest under conditions of high occupational stress and high emotion-oriented coping {M = 39.77, SD = 4.74) and highest under conditions of high occupational stress and low emotion-oriented coping (Af = 45.14, SD = 3.53). This finding demonstrates two points. First, one's tendency not to adopt an emotion-oriented coping strategy seemed to buffer individuals from feelings of reduced personal accomplishment. In fact, among the four subgroups, the high OS/low EOC group actu- ally reported the highest levels of personal accomplishment. More simply stated, for this sample, under conditions of high occupational stress, adopting a coping style other than emotion- oriented coping tended to maximize one's perceived sense of personal accomplishment. Second, these data suggested that imder conditions of high occupational stress, adopting an emotion-oriented coping approach actually tended to amplify a reduced sense of personal accomplishment. The crossed lines in Figure 2 represent this amplification aspect

432 Journalof Counseling & Development • Fall 2009 • Volume 87

Burnout Among School Counselors

TABLE 3

Four-Step Hierarchical Regression Models for Variables Predicting Emotionai Exhaustion, Depersonalization, and Personal Accompiishment

Variable

Counselor demographics Step 1

YOETOT SCRAT

Organizational variables Step 2

COSI LDMA (z score) FS(z score) CTPR (z score) CPPR (z score)

RQ RC(z score) Rl (zscore) RA (z score)

Intrapersonal variables (coping) Step 3 .

CISS TOC (z score) EOC (z score) AOC (z score)

Moderator variables Step 4

OSI X TOO OSI X EOC OSi X AOC

Model 1 Emotional Exhaustion

Step 3

fP fl« added ß-

.03

. 3 8 "

. 4 7 "

. 4 9 "

.21 -.06

.35"

.02

.10

.02 -.02

.16

.24

.12

. 0 9 "

-.04 .33

-.06

.02 -.08

.08 -.08

P

3 . 7 0 " -1.04

0.21 1.42 0.21

-0.22

2.03* 3.18" 1.67

-0.60 5.14"

-0.95

-1.33 1.37

-1.30

Outcome Variables

Model 2 Depersonalization

Step 3

ff /Padded ß*

.04* .26

-.06

. 1 7 " . 1 3 "

-.01 -.10 -.04

.10

.07

.15

.06

.26** .09**

-.13 .29

-.07

.27** .01 -.04 -.07 -.02

P

3.68** -0.96

-0.12 -1.19 -0.39

0.94

0.71 1.67 0.66

-1.86 3 . 8 1 "

-1.02

-0.50 -1.00 -0.30

Model 3 Personai Accomplishment

Step 3

fr fl° added ß'

.02

. 1 2 "

. 3 3 "

.36**

-.04 .10

. 1 0 "

-.22 .06 .06 .04

-.09 .26

-.14

. 2 1 "

.22 -.38

.20

.03* .15 .15

-.03

r

-0.67 1.57

-2.19* 0.85 0.67 0.40

-0.99 3.05**

-1.75

3.26** -5.20**

3.10**

2.32* 2.22*

-0.38

Note. YOETOT = years of experience, total (for this study, total years of school counseling experience); SCRAT = student/counselor ratio; COSI = Counselor Occupational Stress Inventory; LDMA = Lack of Decision-Making Authority; FS = Financial Security; CTPR = Counselor- Teacher Professional Relationships; CPPR = Counselor-Principal Professional Relationships; RQ = Role Questionnaire; RC = Role Conflict; Rl = Role Incongruity; RA = Role Ambiguity; CiSS = Coping Inventory for Stressful Situations; TOO = Task-Qriented Coping; EOC = Emotion- Qriented Coping; AQC = Avoidance-Oriented Coping; OSI = occupational stress index. "All ß and f values reflect the contribution of each variable predicting the outcomes in the final model only. *p <: .05. "p < .01.

of emotion-oriented coping most dynamically. Here, highly stressed individuals using emotion-oriented coping strategies experienced the lowest levels of personal accomplishment when compared with those in the three other categories.

Power analysis. A post hoc power analysis was completed for each of the three models. For this study, power analyses for the three regression equations yielded the following; EE = 1.0, F(15, 182) = 1.72, X = 190.24; DP = 1.0, F(15, 182) = 1.72, X = 73.24; PA = 1.0, F(\5, 182) = 1.72,1 = 111.38. This indicates a strong likelihood that the analyses were suf- ficiently sensitive to identify statistically significant variable relationships.

•Discussion Levels of Burnout

In terms of the overall sample, means indicate that school counselors report higher than average EE subscale scores

when compared with other mental health professionals, along with a higher than normal sense of personal accomplishment, while their overall scores on the DP subscale fall below the norm. Therefore, results from the present study suggest that school counselors are personally engaged with their clients, albeit somewhat emotionally exhausted, and are deriving a fairly healthy amount of satisfaction from their work.

The Contributions of Independent Variable Sets to Outcome Dimensions of Burnout

Stress-strain-coping theory suggests that burnout may be understood by investigating the contributions of Stressors and coping strategies to outcomes on the MBI. Accord- ingly, one would expect to identify separate sets of variables contributing significant amounts of variation to the outcome measures on the MBI-ES and, in fact, this is the case across all three models in the current study. Although minimal, the demographic variables associated with school counselor stress

Journal ofCounseling& Development • Fall 2009 • Volume 87 433

Wilkerson

> Low Occupationai Stress I High Occupationai Stress

c 42 —

1 41 —

8 40

2 39 —

¿ 38 —

40.24; n = 37 , ,

42.43; n = 60 , ,

42.00; n = .

38.65; n = 56

Low Task-Oriented Coping

FIGURE 1

High

Buffering Effect of Task-Oriented Coping on the Reiationship Between Overaii Occupational

Stress and Personai Accompiishment

in Step 1 contribute a significant amount of variation to the second model evaluating depersonalization (as measured by the DP). Step 2, a variable set of organizational variables associated with stress among school counselors is statisti- cally significant in its own right across all three dimensions of the MBI-ES (i.e., EE, DP, and PA subscales). Finally, Step 3, designed to evaluate the amount of variation accounted for by coping strategies once stress variables are considered, also accounts for significant amounts of variation across all three models.

is h m

e i

P e rs

o n a i A

cc o m

p i

45 —

44 —

43 —

42 —

41 —

40 —

39 —

45.14; n = 7

43.61; n = 2 3 \

• - - - • • Low Occupational Stress » > High Occupational Stress

39.77; n = 9 4 ^

Low Emotion-Oriented Coping

FIGURE 2

High

Amplifying Effect of Emotion-Oriented Coping on the Reiationship Between Overall

Occupational Stress and Personal Accomplishment

These findings give rise to two main discussion points. First, the stress-strain-coping paradigm appears to contribute something to a discussion about burnout. The fact that both stress and coping variable sets account for significant amounts of variation on their own across all three models suggests that each set addresses different and valuable aspects of the burnout construct.

Closely tied to these intemal dynamics, the results clearly illustrate differential patterns to the variable set contribu- tions. The organizational variable set of Step 2 accounts for the largest amount of variation in the emotional exhaustion model (Model 1, Step 2 = 35%), and the intrapersonal coping set. Step 3, does so in the personal accomplishment model (Model 3, Step 3 = 21%). Stress-strain-coping theory and cur- rent thinking about the differential aspects of burnout might best be used here to interpret these findings. As understood, stress is associated with experience and perception and is fol- lowed by the coping (or reaction) response in the stress-strain continuum. Stress can also be viewed as the entry point to the continuum followed by the response phase of coping. Thus, it makes sense that these two phases, stress and coping, contrib- ute the largest amounts of variation to their like counterparts in the model of burnout: Stressors to the visceral, experiencing component of the MBI-ES (EE subscale) and coping to the more outcome-related PA subscale that assesses feelings of competence and successful achievement.

Influential Predicfors of Burnout in a Multivariate Context

When considered in the multivariate context, numerous vari- ables were identified in the three final models as influential predictors of burnout. Independent variables at this stage are far more significant than simple zero-order correlations in that they are mathematically derived on the basis of the contributions they make to the models when all variables are considered simultaneously. Interpreted at this level, statisti- cally significant variables can be viewed to have far more predictive value.

First, YOETOT registered a strong association with deper- sonalization (as measured by the DP) when considered in the univariate context, and it continued to do so in this context (DP YOETOT, t = 3.68,p < .01). It is notable that it also as- sumed a high degree of importance in the final EE model (EE YOETOT, Í = 3.70, p < .01), despite the fact that it was not significant at the zero-order level. At this final stage, these associations were positive, suggesting that those who have worked in the profession for a longer time are more prone to experiencing higher levels of emotional exhaustion and depersonalization. This finding may be partially explained by the burnout chronicity argument. Although data have been scant because of an overall lack of longitudinal studies, there is some evidence to suggest that burnout can be viewed as a chronic syndrome. This finding of the current study is sug- gestive ofthat hypothesis. If that is the case, individuals who

434 Journal ofCounseling& Development • Fall 2009 • Volume 87

Burnout Among School Counselors

have been in the profession longer may run a greater risk of experiencing the symptoms of bumout.

In terms of the organizational variables, role incongruity (as measured by the RI) is a significant predictor variable for both emotional exhaustion (EE R I i = 3.18,/?<.01) and personal accomplishment (PA RI í = 3.05,;? < .01). Freeman and Coll (1997) defined this factor as the difference between the roles school counselors actually perform and the roles they believe ought to be performed. The positive association between role incongruity and emotional exhaustion certainly makes sense. However, the positive correlation between role incongruity and personal accomplishment runs contrary to expectations and is a little more difficult to fathom. Chemiss's (1980) concept of the professional mystique suggests that individuals must struggle to reconcile their preconceptions about a profession with actual reality. He goes on to state that if there are no compromises between these two perceptions, bumout ensues. Perhaps those who experience high levels of role incongruity are compelled to appreciate the minimal accomplishments they realize as they work to create jobs that more closely reflect that which they believed they would be doing in the first place.

One additional organizational factor predicting bumout at this final level of analysis was the correlation between personal accomplishment and a lack of decision-making authority (PA LDMA í = -2.19, p < .05). This finding echoes the zero-order correlation results and makes sense given the understanding of the personal accomplishment outcome. It suggests that personal accomplishment is strongly correlated with school counselors' involvement in school decision-making processes. It may also suggest that school counselors believe it is important to take part in that process to better serve the diverse needs of their constituents. This makes some sense given clearly stated expectations within the profession encour- aging school counselors to work as leaders for all students (ASCA, 2005).

Controlling for all previous steps, emotion-oriented coping (as indicated by scores on the EOC) was still foimd to be a statistically significant intrapersonal predictor of bumout in the final analysis across all three models of the MBI-ES. It was positively associated with both emotional exhaustion (EE EOC, ; = 5.14,;? < .01) and depersonalization (DP EOC t = 3.81,/7<.O1), and negatively correlated with personal accom- plishment (PA EOC Í = -5.20,p < .01). This finding suggests a very strong emotional component to bumout and implies that school counselors may benefit ñ'om ongoing attention to developing proactive approaches for dealing with stress.

Finally, although additional coping variables did not sig- nificantly predict scores on the EE or DP subscales, they did do so on the PA subscale. Here, TOC was in the expected positive direction (PA TOC t = 3.26, p < .01), suggesting a strong association between those adopting the task-oriented approach to coping and higher levels of personal accom- plishment. Avoidance-oriented coping (as measured by the

AOC subscale) was also positively correlated with personal accomplishment (PA AOC í = 3.10,p < .01), but this finding requires a bit more interpretation. Endler and Parker (1999) described the avoidance-oriented coping dimension as "ac- tivities and cognitive changes aimed at avoiding the stressful situation. This can occur via distracting oneself with other situations or tasks (task oriented) or via social diversion (person oriented) as a means of alleviating stress" (p. 35). Based on this description of the AOC subscale, it may be that school counselors using this coping strategy are finding ways to involve themselves in other personally meaningful tasks and relationships beyond their jobs. It could be that using this strategy enables individuals to feel better about themselves overall, thus leading them to assess their job achievements in a slightly different, more realistic way. If the avoidance- oriented coping strategy serves to alleviate stress, it might also allow individuals to place less emphasis on their job duties and to approach them in a more relaxed fashion. Pines (1993) suggested that bumout may be viewed as an existential crisis whereby people invest inordinate amounts of energy at their place of work to be recognized and to find meaning in their life. Maybe those who adopt an avoidance approach to stress are able to find this meaning elsewhere.

The Moderating Effects of Coping on Outcome Measures of Burnout

As presented in the Results section, empirical evidence exists to suggest that occupational stress and subsequent bumout may be conditional and based on one's coping approaches. In other words, I surmised that the onset of bximout might actu- ally depend on the presence, absence, or level of the different coping strategies used by a given individual. The moderating effects of coping on the relationships between stress and burnout were thus examined. This variable set accounted for a statistically significant amount of the variation (3%) in one out of the three models (Model 3, personal accomplish- ment). As presented earlier, task-oriented coping served to buffer individuals from feeling a reduced sense of personal accomplishment whereas emotion-oriented coping appeared to amplify this outcome. These findings match others' findings and suggest that a task-oriented approach to coping with stress seems to enhance one's perceptions of personal accomplish- ment. This is further supported by the high task-oriented/low emotion-oriented coping profile of this school counselor sample overall and suggests that school counselors can benefit from adopting a task-oriented approach to their job.

Implications for Professionals

The discussion thus far has focused on the study-specific findings from these data. However, because they appear to have real-world implications, several salient points are worth noting here on a more global level. First, school counselors may benefit from constant evaluation of the organizations where they work. With such strong associations between

Journal of Counseling & Development S Fall 2009 m Volume 87 435

Wilkerson

outcomes on bumout and organizational issues such as role conflict and role incongruity, the results of this study suggest that school counselors need to work diligently with other leadership professionals to define and clarify their roles. This implies that school counselors need to be savvy about the way organizations function. Without this trait, they may be faced with chronically confusing job expectations, and this study indicated strong associations between this issue and outcomes on bumout.

Furthermore, school counselors may experience positive benefits when they are included in the decision-making pro- cesses of their schools. This particular issue was significantly associated with increased perceptions of personal accomplish- ment in the current study, suggesting once again that school counselors reap positive benefits when they are able to skill- fully evaluate their organizational contexts and to participate in school leadership. With strong skills in communication, group dynamics, and interpersonal relationships, school counselors may need to emphasize the valuable contributions that they can make at the leadership level.

Finally, on an intrapersonal level, it is also important to note the impact of one's coping style on outcomes of bumout. Clearly, adopting an emotion-oriented approach to coping is a red flag for bumout. This coping style was strongly associ- ated with bumout across all three dimensions of the MBI-ES. This implies that those who attempt to deal with their stress in an emotional way mn a much greater risk of developing the symptoms of bumout than those who do not. Those who are aware of this tendency would benefit from leaming and implementing additional coping skills.

Strengths and Limitations

This study contributes to the growing body of theory-based literature on bumout. Informed by the temporal nature of the stress-strain-coping continuum, the findings indicate that both organizational Stressors and intrapersonal coping variables account, each in their own right, for significant amounts of the variation in outcomes of bumout. As such, addressing bumout at each of these levels seems to be warranted. Fur- thermore, analysis of data in the multivariate context provides a more complete picture of the real-life way in which school counselors experience their jobs. An additional strength of the study was the inclusion of the fourth step in the hierarchical regression sequence. Evaluating for potential moderating effects of coping in the relationship between stress and bum- out contributed additional insights into how these variables interact with one another. Finally, the current sample size (N= 198) resulted in sufficient power levels to allow readers to interpret these findings with reasonable assurance that the outcomes have merit.

The major limitations of the study are twofold. First, survey research relies on self-report and voluntary responses. It is possible that those responding were the healthiest members of the overall sample, thereby introducing a potential response

bias toward increased job health. Indeed, participants' mean scores on the three subscales of the MBI-ES are suggestive of this possibility. Second, although this study contributes additional information to an understanding of the bumout phenomenon beyond that which has been offered by simple correlational analyses, it does not in any way suggest causal- ity or indicate the potential pathways that exist among job Stressors, coping, and bumout.

Implications for Future Research

Numerous implications for future research are indicated by the results of the current study. First, longitudinal studies might greatly enhance existing conceptualizations of the bumout construct. Among such potential studies, an ongoing analysis of the developmental processes of bumout would be beneficial. Second, path analytic or stmctural equation modeling studies seem to be warranted. Determination of the stress-strain-coping pathways regarding outcomes of bumout would help to ground this theory more empirically and allow scholars to make more definitive claims about the interplay among the identified variables.

•Conclusion The current study examined some of the issues currently faced by school counseling professionals. If, indeed, a pro- fessional school counselor's ability to deliver comprehensive services is compromised as levels of stress and burnout increase, then studying this phenomenon and developing strategies for addressing its associated variables and symp- toms take on great importance. This study presented evidence that both perceived levels of organizational stress and an individual's approach to coping have strong predictive value for the outcomes of bumout. Knowledge of these potential risks has implications for school counselors, for school ad- ministrators, and for those who train them. On the basis of these results, it makes sense to work diligently to alter those factors that most strongly predict increased levels of bumout. Counselors may wish to work toward developing personal, first-order changes in coping styles, while also advocating for more global, second-order changes at the organizational level. Counselor educators may want to consider the impact of a task-oriented approach to coping and find ways to dis- cuss this important skill with their school counselor trainees. They, and their educational leadership colleagues, may also want to explore the various ways that school counselors can work to clarify and define their roles with administrators once they have officially moved into the profession.

Compared with school counselors who are experiencing increased levels of bumout, those who are healthy in this respect are better able to undertake the daily task of working with others. Thus, attention to the issues associated with this aspect of health is cmcial. The result of such focus leads to enhanced client care. That enhanced care is what ultimately

436 Journal ofCounseling& Development • Fall 2009 • Volume 87

Burnout Among School Counselors

leads to improved outcomes for the communities of students, parents, teachers, and administrators with whom school counselors work.

•References American School Counselor Association. (2005). American School

Counselor Association National Model: A framework for school counseling programs. Alexandria, VA: Author.

Bumham, J. J., & Jackson, C. M. (2000). School counselor roles: Discrepancies between actual practice and existing models. Professional School Counseling, 4, 41-49.

Butler, S. K., & Constantine, M. G. (2005). Collective self-esteem and burnout in professional school counselors. Professional School Counseling. 9, 55-62.

Campbell, C , & Dahir, C. (1997). Sharing the vision: The National Standards for School Counseling Programs. Alexandria, VA: American School Counselor Association.

Chemiss, C. (1980). Professional burnout in the human service organizations. New York: Praeger.

Cosway, R., Endler, N. S., Sadler, A. J., & Dreary, I. J. (2000). The Coping Inventory for Stressful Situations: Factorial structure and association with personality traits and psychological health. Journal of Applied Biobehavioral Research, 5, 121-143.

Cronkite, R. C , & Moos, R. H. (1984). The role of predisposing and moderating factors in the stress-illness relationship. Journal of Health and Social Behavior. 25. 372-393.

Endler, N. S., & Parker, J. D. A. (1999). Coping Inventory for Stressful Situations (CISS): Manual (2nd ed.). Toronto, Canada: Multi-Health Systems.

Freeman, B., & Coll, K. M. (1997). Factor structure of the Role Ques- tionnaire (RQ): A study of high school counselors. Measurement and Evaluation in Counseling and Development. 30, 32-39.

Gray, P. (1982). Development of a counselor occupational stress inventory (Doctoral dissertation. University of Auburn). Dis- sertation Abstracts Intemationai, 43(6A).

Huberty, T J., & Huebner, E. S. (1988). A national survey of burnout among school psychologists. Psychology in the Schools, 25, 54-61.

Huebner, E. S. (1992). Burnout among school psychologists: An exploratory investigation into its nature, extent, and correlates. School Psychology Quarterly, 7, 129-136.

Huebner, E. S. (1993). Professionals under stress: A review of bum- out among the helping professions with implications for school psychologists. Psychology in the Schools, 30, 40-49.

Huebner, E. S., & Huberty, T. J. (1984). Bumout among rural school psychologists. Research in Rural Education. 2. 95-99.

Johnson, L. S. (2000). Promoting professional identity in an era of educational reform. Professional School Counselor, 4, 31-40.

Kahili, S. (1988). Symptoms of professional bumout: A review of the empirical evidence. Canadian Psychology, 29, 284-297.

Lazarus, R. S., & Folkman, S. (1984). Stress, appraisal, and coping. New York: Springer.

Leiter, M. P (1991). Coping pattems as predictors of bumout: The function of control and escapist coping pattems. Journal of Or- ganizational Behavior, 12, 123-144.

Maslach, C , & Jackson, S. E. (1981). The Maslach Bumout Inventory (Research Ed.). Palo Alto, CA: Consulting Psychologists Press.

Maslach, C , Jackson, S. E., & Leiter, M. P (1996). The Maslach Burnout Inventory manual (3rd ed.). Palo Alto, CA: Consulting Psychologists Press.

Mills, L. B., & Huebner, E. S. (1998). A prospective study of personality characteristics, occujmtional Stressors, and bumout among school psy- chology practitioners. Jbi/rwa/o/SeAoo/Piyc/io/ogy, 36, 103-120.

Niebrugge, K. M. (1994, March). Burnout and job dissatisfaction among practicing school psychologists in Illinois. Paper presented at the 26th annual meeting of the National Association of School Psychologists Seattle, WA. (ERIC Document Reproduction Service No. ED374378)

Pines, A. (1993). Bumout: An existential perspective. In W. B. Schaufeli, C. Maslach, & T. Marek (Eds.), Professional bum- out: Recent developments in theory and research (pp. 33-51). Washington, DC: Taylor & Francis.

Reiner, H. D , & Hartshome, T. (1982). Job bumout and the school psychologist. Psychology in the Schools, 19, 508-512.

Rizzo, J. R., House, R. J., & Lirtzman, S. 1. (1970). Role conflict and ambiguity in complex organizations. Administrative Science Quarterly. 15, 150-163.

Sandoval, J. ( 1993). Personality and bumout among school psycholo- gists. Psychology in the Schools. 30, 321-326.

Savicki, V, & Cooley, E. J. (1982). Implications of bumout research and theory for counselor educators. The Personnel and Guidance Joumai, 60. 415-419.

Stickel, S. A. ( 1991, February). A study of bumout and job satisfaction among rural school counselors. Paper presented at the annual confer- ence of the Eastem Educational Research Association, Boston.

Wilkerson, K., & Bellini, J. (2006). Intrapersonal and organizational factors associated with bumout among school counselors. Joumai of Counseling & Development, 84, 440-450.

Journalof Counseling & Development • Fall 2009 • Volume 87 437