Burnout became an epidemic among social workers these days as social service needs have been exploded in recent years. While a lot of research has been done on the subject with different participant groups, the study results still provide limited understanding of social worker burnout phenomenon. With a meta-analysis on past research results on social workers’ burnout in Korea, this study aims at integrating data to construct framework for understanding social workers’ burnout and pave the direction of future research. Meta analysis can be thought of as "conducting research about research" .
Specifically, based on MBI(Maslach Burnout Inventory), most commonly used measurement of burnout in these studies, dependent variable burnout was first divided into three sub-factors, namely emotional exhaustion, depersonalization, and lack of personal achievement and burnout total. Factors that influence burnout were categorized into risk factor and protective factor, magnitudes of which were calculated using CMA. Differences in influence of sub-factors of risk factor(organizational factor, work-related factor, and client factor) and sub-factors of protective factors(individual factor, family factor, organizational factor, and work-related factor) on burnout were also studied. Additionally, meta-regression analysis and sub-group analysis using Q-test were used to identify whether influence of risk factor and protective factor on burnout can be moderated depending on the year the study was conducted as well as influence factors recognized as significant in past studies, such as proportion of women and married workers, work experience at current job, and type of work.
For data on the subjects for analysis, past researches on social workers’ burnout published in Korea were used. Primarily 40 published articles on academic journals and 135 graduate dissertations with social workers’ burnout as dependent variables were collected. Among the final research data with the correlation coefficient(r) of independent and dependent variable, 80 articles, which consist of 16 researches on social workers'' burnout published in academic journals and 64 master’s and doctor’s theses.
Firstly, for effect size of risk factors on burnout(with four sub-factors), conflicts among colleagues in organizational factor showed large effect size on burnout total, while qualitative work-overload, role overload, work-related stress, and secondary traumatic stress of work-related factors showed large effect size. Among these factors, secondary traumatic stress had the largest effect size. For emotional exhaustion, workload, qualitative overload of work, role stress, and secondary traumatic stress of work-related factors showed the largest effect size. For depersonalization, secondary traumatic stress of work-related factors showed the largest effect size. No factor had significantly large effect size on lack of personal achievement, and lack of challenges of work-related factor had the largest effect size at 0.327.
Secondly, for effect size of protective factors on burnout (with four sub-factors), self-esteem as personal factors had the largest effect size on burnout total, while organizational factors such as professional capabilities, sense of mission for the job, and utilization of expert organizations showed large effect size. Among these, self-esteem had the largest effect size. For the lack of personal achievement, self-esteem and self-efficacy of personal factors, and social recognition among work-related factors showed large effect size, while self-esteem had the largest effect size.
Thirdly, this research studied moderation effect of the year that research had been done, proportion of women, work experience at current job, proportion of married workers, and type of work on relations of risk factors, protective factors, and sub-factors of burnout. The years of research was categorized in accordance with time of policy changes in public and private sectors.
First, in private sector, the year of research moderated influence of risk factors on burnout total, emotional exhaustion, depersonalization, and lack of personal achievement, while in public sector, the year of research also moderated influence of risk factors on emotional exhaustion, depersonalization, and lack of individual achievement. In private sector, the year of research did not moderate protective factor’s influence on sub-factors of burnout, while in public sector, the year of research moderated influence of protective factor on burnout total, emotional exhaustion, and lack of personal achievement.
Proportion of women among research participants moderated the influence of risk factors on burnout total as well as influence of protective factors on emotional exhaustion. Proportion of married workers also had moderating effect on influence of protective factor on emotional exhaustion. Work experience at current job moderated risk factor’s influence on depersonalization and lack of personal achievement as well as influence of protective factor on emotional exhaustion, depersonalization, and lack of personal achievement.
Type of work only had significant moderating effect on the influence of protective factors on emotional exhaustion. For counselors, risk factors'' influence on emotional exhaustion and lack of personal achievement was the largest, while for government officials, risk factors'' influence on emotional exhaustion ; depersonalization was the largest. In addition, for counselors, protective factors'' influence on burnout total, emotional exhaustion and lack of personal achievement was the largest, while for generic social workers, protective factors'' influence on depersonalization was the largest.
Significance of this meta-analysis research lies in its quantitative integration conducted to show the contrasting and combining results from different studies. This study was able to identify patterns among study results, sources of disagreement among those results. Using meta-analysis, this research did not only identify the relevant variable with largest effect on burnout as well as variables with preventive effects, but also provided macro-understanding of work scope and HR architecture of social workers and effect of time variable as more than two decades passed by. In addition, this research set the way forward for future researches on social worker burnout.