본 연구는 많은 연구결과에서 효과적이라는 연구결과에도 여전히 논쟁거리가 되고 있는 미술치료효과에 관한 지금까지의 연구결과들을 메타분석을 통해 종합적인 결론을 도출하고, 미술치료 선행 연구의 치료효과에 영향을 미칠 수 있는 각 변인의 효과크기를 비교 및 분석하였다. 연구대상논문은 1994년에서 2009년 7월 31일 사이에 국내에서 발표된 미술치료연구의 메타분석이 가능한 전체박사학위논문 총 49편이다. 메타분석결과 1차 분석에서는 80개의 효과크기가 산출되었고, 미술치료 프로그램 평균효과크기는 1.89로 95%신뢰수준 하한1.57에서 상한 2.21의 신뢰구간을 가진 것으로 분석되었으며 신뢰구간에 0이 포함되어 있지 않아 유효하게 평균효과가 존재하는 것으로 나타났으며, 이는 Cohen의 기준에 의하면 큰 효과크기가 있는 것으로 나타났다. 2차 분석에서는 미술치료 프로그램 최적 평균상승률28.34%로 나타난 것으로 분석되었다. 본 연구는 국내 미술치료 프로그램에 관한 연구결과를 1차 분석, 2차 분석을 통해 미술치료에 효과에 관한 박사논문 전체를 최초로 메타분석을 한 연구라는데 그 의의를 찾을 수 있다.

Although a lot of studies found that art therapy was effective, there was something incongruous between the results of the studies, which resulted in triggering a heated controversy about the effect of art therapy. The purpose of this study was to make a meta-analysis of the results of studies on art therapy in a bid to reach a general conclusion, and it`s additionally meant to compare the effect size of related factors that were identified in earlier studies as the variables that might affect the effect of art therapy. 49 domestic dissertations of art therapy were selected, which were published between 1994 and July 31, 2009. As a result of making a primary meta-analysis of the selected dissertations, 80 effect sizes were extracted. The average effect size of the art therapy programs used in the studies was 1.89. The rate of their confidence stood at 95 percent, and their confidence interval was between a low of 1.57 and a high of 2.21. The confidence interval didn`t contain zero, which justified their average effect. According to Cohen`s standard, the effect size of the programs was great. In the secondary analysis, the art therapy programs showed an optimum mean rise of 28.34 percent in effect size. This study was meaningful in that it made the first attempt to perform a meta-analysis of every dissertation of art therapy three times. Fourth, lots of dissertations have centered around art therapy programs, but the effect size of the programs was very high in many dissertations, and there was room for improvement in association with a meta-analysis. More accurate and well-planned research should be conducted in the future.