Wearing Your Heart on Your Sleeve: The Experience of Burnout Among Child Welfare Workers Who are Cognitive Versus Emotional Personality Types

Abstract

This study explores if child welfare workers who are predominantly cognitive in their personality type, rather than emotionally predominant, experience less burnout in their work. The thinking versus feeling and sensing versus intuition poles of the Myers-Briggs Type Indicator were used to determine predominance in cognitive or emotional personality type and the Maslach Burnout Inventory was used to measure burnout. There was no statistical difference found on burnout scores based on personality type. These results have implications for child welfare organizations and on the recruitment and retention of a qualified work force. Selective recruitment based on personality type may not contribute to better retention of workers.

KEYWORDS:

• burnout
• personality type
• child welfare
• Myers-Briggs
• worker retention
• worker recruitment

Introduction

The child welfare system is the primary system responsible for the prevention, investigation, and intervention in situations of child physical abuse, sexual abuse, and neglect. The important nature of this work to our society is inarguable, yet the child welfare services system has experienced a national crisis in recruitment and retention of qualified workers due to job stress and burnout (Bednar, 2003; Drake & Yadama, 1996). The contributing factors for burnout and high turnover rates can best be looked at from a transactional perspective.

The transactional theory of stress by Lazarus and Folkman (1984) is useful to understand stress and burnout in child welfare systems. The transactional theory of stress extended the biopyschosocial model of stress. According to the biopsychosocial model, stress involves three components: 1) external, 2) internal, and 3) the interaction between the two (Bernard & Krupat, 1984). Lazarus and Folkman's (1984) transactional theory of stress explains the interactional process between the external and internal components as a function of cognitive appraisal of events. The relationship between the individual and environmental factors is influenced by the meaning given the situation and whether it is perceived as stressful. The perception of an event as stressful involves a two-part process of a primary and secondary appraisal. During the primary appraisal process, the individual assesses if the event involves harm or loss that has already occurred, a threat of harm or loss to potentially occur in the future, or a challenge that may ultimately result in growth or new mastery. During secondary appraisal, the individual assesses his or her resources and options for response. When the primary appraisal determines an event is or may be threatening to the individual's well being and the secondary appraisal reveals that the demand of the situation exceeds coping resources, the individual will perceive the situation as stressful. The individual's cognitive appraisal will determine the magnitude of the stress response and the coping strategies employed to remediate the stress. Although issues can be understood through a lens that considers the dynamic, interactive, and reciprocal influence the person and the environment contribute to the issue of concern, one cannot underestimate the meaning constructed of the event as a significant part of the lens.

REVIEW OF THE LITERATURE

Practice in the child welfare system is often said to be one of the most personally and professionally challenging fields of practice (Alwon & Reitz, 2000). Workers bring their own values, skills, knowledge, personal histories, and coping strategies to the environment of the child welfare organizations and systems of care. Stressors can originate from both the individual and the environment, and stressors are mediated by both individual and organizational qualities. The following review of the literature examines environmental issues such as contextual factors associated with the child welfare field and individual issues that contribute to burnout in child welfare as well as issues that help to mediate the experience of burnout in a positive manner.

Contextual Factors within the Field of Child Welfare

Organizational, systemic, and macro-environmental issues contribute to the challenges of this field of practice. One stressful factor frequently cited is the level of role ambiguity or role conflict inherent in child welfare work (Jayaratne & Chess, 1984). Child welfare workers have the dual, and often times incompatible, role expectation of being both the agent of social control and the agent of help. Jayaratne and Chess (1984) refer to this as the dual roles of “social worker-police officer” (p. 450). The child welfare social worker is charged with both helping the family remove problems that brought them into the child welfare system, while at the same time monitoring change and reporting to the court system the progress the family has made towards the goal of reunification. This often puts the worker in a role conflict of needing to build a trusting working alliance with the family, when the family may hold considerable resentment regarding the child welfare system involvement in their life.

In addition to role conflict, value conflicts can permeate the work of child welfare staff (Crosson-Tower, 2004). Social workers are required to make critical decisions about whether children are at risk in their homes within short time frames. This may present conflicting values and moral dilemmas between the maintenance of important primary attachments for children, while ensuring their safety within the home. Other times social workers may be required to return children to environments that hold the potential for abuse and neglect when there is insufficient evidence to document that the best interest of the child warrants the infringement of parental rights.

Child welfare workers have frequently cited high caseloads as a contributing factor to stress and burnout (Landsman, 2001; Pecora, Whitaker, & Maluccio, 2000; Rycraft, 1994; Strolin, McCarthy, & Caringi, 2007; Winefield & Barlow, 1995). Ellett, Ellett, and Rugutt (2003) found that 80% of child welfare workers in a large state-wide study had caseloads higher than the Child Welfare League of America standards. The high turnover rate within child welfare only exacerbates this problem when existing workers must cover their own cases plus cases left by departing workers (Anderson, 2000).

Due to the legalistic and interdisciplinary characteristics of child welfare, the paperwork and documentation requirements are often overwhelming; often consuming 50% of a worker's time (Esposito & Fine, 1985; Samantrai, 1992). Workers must comprehensively document events of a case to fulfill the legal mandates for evidence and must live with the threat of legal liability if anything goes wrong with the case. Due to the interdisciplinary nature of child welfare, often this means several entities within the child welfare system need multiple copies of documentation.

Child welfare workers must work in an arena of mercurial public attitudes regarding child abuse that often translates into confusing federal and state policy mandates (Crossen-Tower, 1999; Downs, Moore, McFadden, & Costin, 2000; Landsman, 2001). Public perceptions of child welfare workers are often not only stigmatized, but at the same time held to a standard of performance that is nearly impossible to achieve. Child welfare workers are stigmatized by their association with clients who are often indigent and have many complex social problems that engender negative societal judgments (Guterman & Jayaratne, 1994). Yet, there is a public perception of child welfare workers as being baby snatchers and bureaucrats (Anderson, 2000). Child welfare workers are held to a nearly impossible standard by being expected to solve significant social problems with insufficient resources, low pay, and poor working conditions (Anderson, 2000).

Personal Factors

In addition to organizational factors, child welfare workers either bring to their job, or acquire on their job, certain personal factors that contribute to burnout. Child welfare workers, as a group, have demonstrated higher levels of burnout than other fields of social work practice (Jayaratne & Chess, 1984; Rycraft, 1994). Burnout has typically been operationalized using Maslach and Jackson's (1981) model which encompasses three experiences of burnout including: 1) increased emotional exhaustion; 2) increased depersonalization; and 3) the reduction in a sense of personal accomplishment.

Workers who experience burnout are also shown to experience greater levels of anxiety, depression, irritability, somatic complaints, and lower levels of job satisfaction (Jayaratne, Scess, & Kunkel, 1986). Additionally, workers who have a diminished sense of control over their life and have limitations in relational capacities, including being mistrustful and shy, experience greater distress in their job (Regehr, Hemsworth, Leslie, Howe, & Chau, 2004).

The stress from working in the child welfare system seems to permeate workers' personal lives. Workers who score high for burnout also report less satisfaction with their marriages (Jayaratne, Chess, & Kunkel, 1986). Workers in these troubled marriages report that their spouses are also more anxious and depressed as well.

Gender may be a personal factor contributing to high levels of burnout within the child welfare system (Anderson, 2000). The overwhelming majority of social workers in general, and child welfare workers specifically, are women. Women's values on empathy and attachment can contribute to child welfare work as being fulfilling, yet draining. Women in the helping professions may be particularly vulnerable to burnout as they must exhibit the same qualities of empathy and caretaking at home as at work (Finn, 1990).

Some child welfare workers may be drawn to the work due to their own personal history of abuse, thus seeking to help abused children as a vicarious way to heal their own hurt and pain (Dane, 2000). Research regarding personal histories of abuse on the helping professional is mixed with some research indicating that a personal history of trauma contributes to greater trauma symptoms, while other studies have found no association (Pearlman & MacIan, 1995; Schauben & Frazier, 1995; Stevens & Higgins, 2002). Having a personal history of abuse may work as a resiliency factor if the worker has learned coping strategies for dealing with abuse, or it may be experienced as negative countertransference and personal identification with the victim thus resulting in re-traumatization.

At an existential level, working with abused children and abusive families can challenge one's personal belief systems that children are precious and that human nature is good. The realities of work in the child welfare system confront workers with the reality of the capacity for evil that exists in all of us (Dane, 2000). The nature of the work in the child welfare system often invokes powerful feelings of helplessness and the vicarious experience of shame and pain. Samantrai (1992) found that one of the most frequently cited reasons for child welfare workers to exit the field of practice was the inability to tolerate the pain and suffering they encountered in their clients and the inadequacy or lack of institutionalized ways for workers to process their feelings. This assault on worker's basic worldview may be defended against through efforts of depersonalization of clients and emotional withdrawal, both components of burnout. Ultimately, some workers may decide to exit from the field all together.

Mediating Factors

Support from various significant people can help to mediate the effects of burnout, although the results from studies are mixed. Several studies have found that supervisor and spouse support were significant in reducing the negative effects of burnout (Davis-Sacks, Jayaratne, & Chess, 1985; Landsman, 2001; Strolin, McCarthy, & Caringi, 2007). Other studies have found however, that support from supervisors is a poor predictor of burnout for child welfare workers (Regehr et al., 2004).

Organizational factors such as work culture can also mitigate against the effects of burnout. Silver, Poulin, and Manning (1997) found that supervisory support, level of trust between co-workers, and adequacy of resources available influenced greater job satisfaction. Johnson and McIntyre (1998) found that a work culture conducive to job satisfaction included involvement from staff, recognition of staff, empowerment, positive communication, formal goal setting, encouragement for creativity and innovation, and input into decision making.

Realistic preparation and anticipation of child welfare work may mitigate against high burnout and turnover by screening out potential child welfare workers during the recruitment, screening, and hiring process. Graef and Potter (2002) proposed that a process called a Realistic Job Preview could realistically inform job applicants about the challenges, rewards, tasks, and functions of child welfare work during the recruitment phase. Realistic Job Previews can perhaps screen applicants out before they take a position in child welfare if such a position would be a mismatch between the demands of the job and the needs of the applicant. Fox, Miller, and Barbee (2003) employed a process similar to a Realistic Job Preview with undergraduate social work students in Kentucky. The Public Child Welfare Certification Program recruited and trained BSW students on the realistic job challenges and rewards of child welfare work. The Certification Program involved an enriched, multi-university curriculum that prepared select BSW students with child welfare theory and methods through classroom instruction, biennial faculty, agency staff, and student retreats, and an intense child welfare practicum. The evaluation results demonstrated that the project was successful with the recruitment and retention of child welfare workers that were less stressed, more confident, more skilled with interactions with clients, more knowledgeable about agency policy, and more positive regarding their attitudes towards the agency and their job than employees who were hired without intensive BSW preparation (Fox, Miller, & Barbee, 2003).

Rycraft (1994) identified four primary factors that were instrumental in influencing child welfare workers' retention in the field. These four factors have the potential to mitigate against burnout. This study found that the four factors of: 1) a sense of personal and professional mission; 2) goodness of fit with the agency; 3) supportive supervision; and 4) investment in the agency, colleagues, and the field, were positive factors promoting their continued attachment to the field of child welfare. This study indicates that altruistic factors such as mission and investment are equally as important as more proximal factors such as supervision. In a recent large-scale, statewide study of workers' intention to remain employed in child welfare, Ellett, Ellett, and Rugutt (2003) documented a strong correlation between such intents and a human caring variable of professional commitment (r = .63; p < .001). This finding was also replicated in a more recent study by Ellis, Ellett, and DeWeaver (2007).

Vinokur-Kaplan (1991) identified five factors that were positively associated with job satisfaction among child welfare workers including working with clients, working with colleagues, feelings of accomplishment, satisfying working conditions, and salary. Of these the more altruistic factors of working with clients, working with colleagues, and feelings of accomplishment had the highest ratings towards job satisfaction.

A study by Reagh (1994) also indicated that altruistic motivations for being in the field of child welfare were instrumental as a buffer for child welfare workers to mitigate the experience of burnout. This study found that workers who had a high degree of satisfaction with child welfare work described an epiphany experience sometime during their work, which provided a sense of personal and spiritual meaning to their professional pursuit.

Various methods for coping have been examined relative to the experience of burnout. Anderson (2000) found that active methods of coping including problem solving, cognitive restructuring, seeking social support, and emotional expression all reduced the experience of depersonalization and increased a sense of personal accomplishment, but did not have an effect on the experience of emotional exhaustion. Disengaged coping strategies such as problem avoidance, wishful thinking, social withdrawal, and self-criticism corresponded to increased levels of burnout on all three categories of increased emotional exhaustion and depersonalization and decreased personal accomplishment. Anderson (2000) also found that workers were more likely to use two cognitively oriented active coping strategies of problem solving and cognitive re-structuring more than the emotional ones of seeking social support and emotional expression.

The transactional theory of stress would indicate that a dynamic, interactional combination of personal and environmental factors in child welfare that are influenced by cognitive appraisal contribute to worker burnout, and ultimately to the crisis of recruitment and retention of qualified workers. While interventions at all levels are important, there is immediate value to understanding more about child welfare worker style and the relationship to resilience so as to improve the effectiveness of hiring decisions. Perhaps recruitment and retention could be improved by discerning personality characteristics of workers who are less inclined to burnout from the field of child welfare. Identification of workers who have greater capacities to tolerate the emotional impact of work in child welfare could enhance the recruitment efforts of workers who will commit to this work as a long-term career focus. This study sought to determine if workers who were more cognitive in their personality styles were less likely to experience burnout than workers who were more emotional in their personality styles.

METHODOLOGY

This study hypothesized that child welfare workers who were more cognitive in their personality styles would be more able to withstand the challenges of child welfare work than workers who were more emotional in their personality styles. The study was driven by four specific hypotheses as follows:

1. Child welfare workers with the personality type that reflects a preference for the thinking pole on the Myers-Briggs Type Indicator (M-BTI) will evidence less job burnout as defined by emotional exhaustion and depersonalization scores on the Maslach Burnout Inventory (MBI) than workers who have a preference for the feeling pole.

2. Child welfare workers with the personality type that reflects a preference for the sensing pole on the M-BTI will evidence less job burnout as defined by emotional exhaustion and depersonalization on the MBI than workers who have a preference for the intuitive pole.

3. Child welfare workers with the personality type that reflects a preference for the thinking pole on the M-BTI will evidence greater job satisfaction as defined by personal accomplishment on the MBI than workers who have a preference for the feeling pole.

4. Child welfare workers with the personality type that reflects a preference for the sensing pole on the M-BTI will evidence greater job satisfaction as defined by personal accomplishment on the MBI the than workers who have a preference for the intuitive pole.

Measures

MYERS-BRIGGS TYPE INDICATOR (M-BTI)

The M-BTI is a specific instrument that is frequently used to identify personality type within occupations and was used to determine personality type in this study (Briggs-Myers & McCaulley, 1985). The M-BTI is nearly an industry standard for career counseling (DiTiberio & Hammer, 1993). The M-BTI identifies personality types within four categories. Dichotomous poles of two opposite characteristics comprise the four categories: 1) extraversion and introversion; 2) sensing and intuition; 3) thinking and feeling; 4) judging and perceiving. Respondents react to a series of paired statements and each statement indicates favor for one or the other pole of one of the four categories. Type is determined by adding the number of responses that favor each pole within the categorical group and the pole with the greater selection is the preferred type for the respondent. People generally have some endorsed responses on each pole of each category, but one is often preferred.

This study primarily focused on the thinking/feeling category and the sensing/ intuition category. Scores for intensity of the thinking/feeling sub-score can range from 0–24 endorsed statements. Intensity of the sensing/ intuition sub-score can range from 0–26 endorsed statements. Given the emotional experience of working with abused and neglected children, workers who use predominately thinking and sensing modes may have a greater tolerance for job stress and experience less job dissatisfaction and burnout than workers who have a preference for feeling and intuitive modes.

Reliability of the M-BTI has been reported numerous times across numerous samples. The reliability remains strong for each of the subtypes and is consistently above .75 (Briggs-Myers & McCaulley, 1985). Thompson and Borrello (1986) report high levels of construct validity.

MASLACH BURNOUT INVENTORY

The MBI was used to measure the variable of job burnout. The MBI comprises three subscales that measure emotional exhaustion, depersonalization, and a sense of personal accomplishment. Higher scores on the emotional exhaustion and depersonalization subscales indicate higher levels of burnout and lower scores on the personal accomplishment subscale indicates greater likelihood of burnout. The range of scores for emotional exhaustion is 0–63 and the range for depersonalization is 0–35. The range for the subscale of personal accomplishment is 0–56. Internal consistency of the MBI was estimated at .76 using the Cronbach's alpha with data from the initial sample. Test-retest coefficients have been as high as .82. The authors reported a high level of concurrent validity, construct validity, and discriminant validity for the MBI (Maslach & Jackson, 1981).

Sampling

This study used a non-probability sample of 85 child welfare front-line workers from five child welfare agencies in a central Midwestern region. As is typical of social work in general and child welfare work in particular, they were overwhelmingly female (90%). The sample was predominantly White (91%), and 8% were Black. The workers from the study were from a variety of educational disciplines and fairly evenly distributed between bachelor-degreed workers and those who either had a master's degree or were working towards one. Of the workers in the study, 52% had bachelor degrees and 33% had master's degrees. Another 13% of workers were working towards a master's degree. The majority of workers did not have a social work education, with only 17% having a BSW degree and 18% with a MSW degree. The sample of participants had experience in child welfare, with a mean length of time in the field of 7.5 years and a mean number of 4.5 years in their current agency.

Data Collection Procedures

The principal investigator of the study recruited child welfare participants from five child welfare agencies in two midwestern communities. The principal investigator met with the supervisors of the five agencies to explain the study and seek permission to collect data on a voluntary basis from the workers. The agency supervisors arranged for data collection to occur over workers' lunch hours. Any worker voluntarily agreeing to participate in the study completed three measures: the M-BTI, the MBI, and a demographic questionnaire. All data collection instruments and forms were treated with confidentiality with only a worker code to identify matching instruments. A separate master code list identified which code corresponded to each worker. This was necessary to allow for feedback to each worker regarding their scores on the M-BTI and the MBI.

In return for the time of participation, workers were treated to lunch including pizza and soft drinks during the data collection period. As fair exchange for their time, the principal investigator also set up a session after scoring of all instruments to provide each worker with their individual scores on the three instruments. This session also included a team professional development workshop regarding the various personality profiles of the M-BTI and how this might be used for intra-organizational team building.

Data Analyses

Independent samples t-tests were used to discern the results for the primary hypotheses. The aim of this study was to discern a difference in burnout levels between people who are in the thinking versus feeling categories and sensing versus intuitive categories as measured by the M-BTI. The M-BTI is most accurately used as a categorical level variable as it identifies a person's preference for one pole of a two pole category. In this study the M-BTI was used to distribute the sample categorically by thinking or feeling and sensing or intuitive. This categorical variable was the independent variable. The level of emotional exhaustion, depersonalization, and personal accomplishment, as measured by the subscales of the MBI, was the continuous-level dependent variable. As the sample comprised child welfare workers from five agencies, one-way analysis of variance (ANOVA) was also completed to determine if there was a difference in burnout measures by agency. Additionally Pearson's correlation coefficient was used to examine the relationship between years worked in child welfare and age of the workers and the three measures of burnout.

RESULTS

The Cronbach's alpha coefficients indicated that the subscales for the Maslach Burnout Inventory had good internal consistency with this sample. The alpha was .90 for the emotional exhaustion subscale, .72 for the depersonalization subscale, and .75 for the personal accomplishment subscale. As a group, the child welfare sample (n = 85) was in the high range of the moderate levels of burnout as measured by emotional exhaustion and depersonalization. The mean score for emotional exhaustion was 24 (SD = 10.7) and the mean score for depersonalization was 10 (SD = 6.8). Workers from this sample, however, also seemed to derive a moderate amount of satisfaction from their work with the mean score for personal accomplishment at 35 (SD = 6.6). The child welfare workers tended to score more frequently on the feeling pole (61%) than the thinking pole (39%) and more frequently on the sensing (55%) pole rather than the intuitive pole (45%).

The differences of the mean scores of emotional exhaustion, depersonalization, and personal accomplishment between the workers who scored more predominately on the thinking, rather than the feeling pole on the M-BTI, were not statistically significantly as anticipated. Workers who were more thinking dominant (M = 24.04, SD = 10.59) scored slightly higher on the emotional exhaustion subscale than workers who were feeling dominant (M = 23.9, SD = 10.98). The thinking dominant workers did score lower on the depersonalization scale (M = 9.81, SD = 5.75) than the feeling dominant workers (M = 10.55, SD = 7.45). The thinking dominant workers had a lower sense of personal accomplishment (M = 34.30, SD = 6.75) than the feeling dominant workers (M = 35.65, SD = 6.57) although these differences were not statistically significant.

The differences between sensing dominant and intuition dominant workers also did not appear as expected. The sensing dominant workers scored slightly higher on the emotional exhaustion scale (M = 24.59, SD = 11.30) than the intuition dominant workers (M = 23.70, SD = 10.30). The result for this difference was not significant, t = .376, p = .708. The intuition dominant workers were experiencing slightly more depersonalization (M = 10.46, SD = 5.26) than the sensing dominant workers (M = 10.09, SD = 8.06), although this was not statistically significant. The sensing dominant workers had a slightly lower sense of personal accomplishment (M = 34.13, SD = 6.48) from their job than the intuition dominant workers (M = 36.19, SD = 6.71).

One way analysis of variance did reveal that workers from one agency (Agency 3) which had the highest mean score of emotional exhaustion, did score significantly higher on the emotional exhaustion subscale than workers from two other agencies (Agency 1 p = .025 and Agency 5 p = .018). Thus, the agency environment may influence the levels of emotional exhaustion experienced by workers in this sample. There was no statistical difference by agency for the other two subscales of depersonalization or personal accomplishment.

Pearson product moment correlation coefficients showed no statistically significant correlations between the number of years employed in child welfare and the scores for the emotional exhaustion (r = −.03), depersonalization (r = −.03), and personal accomplishment r = .10) variables. There was no statistically significant correlation between the age of the workers and the scores for the emotional exhaustion (r = −.11), depersonalization (r = −.14), and personal accomplishment (r = .20) variables.

LIMITATIONS

The results from this study should be interpreted with some caution because of some design limitations. The results should not be widely generalized since the study used a non-probability sample. The sample size of 85 participants was rather small and was derived from agencies in only one regional area of one state.

DISCUSSION AND IMPLICATIONS

The results of this study did not demonstrate that child welfare workers who are more cognitive in their personality types are more able to withstand the pressures of the field than workers who are more emotional in their personality style. Each of the subscales for burnout produced mixed results, with none showing a statistically significant difference in worker groups by personality type. This is somewhat surprising in light of other studies by Anderson (2000) that found that child welfare workers were more likely to use cognitive coping strategies of problem solving and cognitive restructuring over two emotional coping strategies of seeking social support and emotional expression. The results suggest that the use of the M-BTI as part of any employee selection process is not recommended. The results from this study show little evidence for the criterion-related validity of the use of the M-BTI for recruitment and selection of child welfare workers.

The transactional theory of stress explains the interactional process between the external and internal components as a function of cognitive appraisal of events or the meaning constructed by the individual of the event. Transactional theory suggests that how child welfare workers construct meaning of the events they work with could impact whether the events are perceived as stressful or not. The results from this study, however, indicated that micro-level contributions of personality type may not influence how child welfare workers construct meaning of their experiences and therefore differentiate who experiences greater burnout. Indeed, there was very little variability in the experience of burnout for this sample.

The results from this study unfortunately suggest that targeted recruitment of certain personality types of individuals may not lead to workers who are hardier to withstand the challenges of work within this field. Factors cited in previous research may be much more useful in recruitment of child welfare workers than the use of the M-BTI to determine personality type. Recruitment based on degree may be more useful for recruitment than screening for personality type. Child welfare workers with social work degrees handle the demands of child welfare work better and provide a higher quality of services than workers with other degrees (Ellett, Ellett, & Rugutt, 2003). Furthermore, social workers who received Title IV-E support during their education also withstand the challenges of child welfare work better than social workers without Title IV-E preparation (Ellett, Ellett, & Rugutt, 2003).

Although the results from this study did not demonstrate that personality type was useful for recruitment and selection of child welfare workers, other intrinsic personality factors have demonstrated positive results. The presence of a human caring factor of professional commitment indicates greater likelihood for child welfare workers to stay in the field (Ellett, Ellett, & DeWeaver, 2007) and a sense of professional mission contributes to greater longevity in the field (Rycraft, 1994). Perhaps screening of these narrower types of personality characteristics rather than broad personality types would be useful in hiring procedures in child welfare agencies.

A recommendation for further research is to examine the issue from a strengths rather than deficit perspective. An abundance of research has been conducted on the issue of burnout in human services in general and the child welfare field in particular. The fact that child welfare work is demanding is well-established. Rather than continued research on burnout, future research should be directed at determining what characteristics of workers and organizations mitigate the challenges of child welfare work. This type of focus would be useful to determine what contributes to retention from a strengths perspective rather than turnover from a deficit perspective. Furthermore, specific research regarding screening and recruitment methods that are likely to determine successful characteristics of child welfare candidates prior to employment would be beneficial. It is far better to determine goodness of fit prior to employment rather than hire candidates who soon fail. The lack of more realistic screening is an expensive mistake for agencies that have invested in training of a worker, and is emotionally costly to clients who experience disruption in their service delivery due to worker turnover.

Further research on the micro-level, or individual characteristics of workers suited for child welfare work should be an ongoing effort. This, however, should not be at the neglect of future research regarding the macro-level, or organizational contributions to child welfare employee job satisfaction and retention. Continued vigilance within agencies and the child welfare field to improving working conditions may also help to stem the tide of turnover. Previous research indicates that high caseloads, excessive paper work demands, and inadequate supervision increase the stress of child welfare work and may be areas that organizations can examine for change (Landsman, 2001; Sumantrai, 1992). Certainly the ecological, person-in-environment perspective is essential to staff the child welfare system for the important work of this highly specialized field. Our vulnerable clients deserve nothing less than to be served by people who are dedicated and committed to their success.