본 연구의 목적은 사교육, 방과후학교, EBS의 학업성취 효과를 비교를 통해 방과후학교, EBS의 사교육 대체성을 검토하기 위한 것이다. 과도한 사교육이 사회문제가 되고 있는 한국의 현실에서 저렴한 비용의 방과후학교와 EBS의 학업성취 효과를 사교육과 비교하는 것은 이런 면에서 매우 의미있다고 할 수 있다. 사교육, 방과후학교 및 EBS의 참여학생과 비참여학생의 배경변인이 가시 적으로 차이가 있거나 방과후학교에 있어서 도시지역보다 농어촌 지역의 학생들이 활발하게 참여 하는 등 기존의 연구에서 사교육·방과후학교·EBS 참여집단의 특성이 비참여집단과 동질적이지 않 아 그 효과를 일반화하는데 한계가 있다는 것이 제한점으로 지적되었다. 따라서 본 연구는 이러 한 제한점을 극복하고 사교육, 방과후학교, EBS 참여하는 학생들의 배경이 다르기 때문에 발생하 는 선택편의(selection bias)를 제거하기 위하여 경향점수매칭 분석방법이 사용되었다. 본 연구에 사용된 사교육 자료는 2013년 통계청에서 조사한 사교육 자료를 이용하였다. 경 향점수매칭 분석을 위하여 방과후학교 19,035명, EBS 16,076명, 사교육 19,295명의 표본이 분 석에 사용되었다. 학업성취효과에 영향을 주는 사교육, 방과후학교, EBS의 효과를 알아보기 위하여 경향점수매칭 분석방법에 의한 로지스틱 회귀모형을 통해 효과분석을 실시하였다. 경향점수매칭에 의해 사교육, 방과후학교, EBS의 학업성취 효과를 비교한 연구결과, 방과후 학교 및 EBS 학업성취효과가 사교육 효과에 못지않음을 발견하였다. 학업성취 효과는 사교육 이 6.82%, EBS가 5.04%, 방과후학교가 4.81%, 순으로 높게 나타났다. 2013년 통계청 조사에서 초·중·고 학생1인당 사교육비 23.9천원, 방과후학교 비용은 3.2천원 그리고 EBS 비용은 0.2천 원으로 사교육의 비용이 방과후학교나 EBS에 비해 월등히 높다. 이런 점을 고려한다면 방과 후학교나 EBS가 사교육에 비해 크게 열등하지 않으며 방과후학교나 EBS가 비용대비의 효율 성이 사교육에 비해 높은 것으로 평가할 수도 있다. 특히, 사교육을 받지 않는 학생들에게 방 과후학교나 EBS는 실질적으로 큰 도움이 될 잠재력이 충분한 것으로 보여진다. 특히, 일반계 고등학교의 학업성취효과의 경우 방과후학교(7.91%), EBS(6.09%), 사교육(4.18%) 순으로 나타 나 고등학교급에서는 사교육보다 방과후학교와 EBS 효과가 크다는 것을 알 수 있다.
The main purpose of this study is to investigate how participation in private tutoring, Afterschool and EBS programs affects school achievement by using the propensity score matching(PSM) analysis method, and to examine the effects of Afterschool and EBS programs, which are alternatives to private tutoring. For this study, private tutoring data researched in 2013 by Statistics Korea were used. Samples were collected twice in a year through an Internet survey from a total of 78,000 subjects, including approximately 44,000 parents(private tutoring expenditure, attitude survey) and 34,000 students(attitude survey). The impact of variables deemed influential on school achievement was analyzed first. In order to investigate the impact of private tutoring, Afterschool programs and EBS on school achievement, an effectiveness analysis was performed using a logistic regression model of the regression and propensity score matching methods. In addition, the regression model was used to figure out variables that influence in private tutoring expenditure. In this study, SAS (Ver. 9.3) was used for analysis. Estimation and matching of propensity scores were stratified and conducted for each school level. Specifically, propensity scores were estimated for each of elementary, middle, general high and specialized high school and used for matching. Data including missing values were excluded and propensity scores were obtained using the logistic regression analysis. For matching of 16 cities and dos, the Firth’s penalized likelihood logistic regression analysis was used to calculate propensity scores. Second, the matching algorithm introduced by Parsons, et al. (2004) was applied to match to the student level at the one-to-one ratio. Third, the independent samples t-test was utilized to diagnose the equality of covariants among groups before and after matching in this study. Fourth, the paired t-test was conducted to measure treatment effects by using the matched data, and to verify causal effects with the selection bias removed. The results of the study are as follows. According to a comparison analysis on school achievement affected by participation in afterschool programs based on propensity score matching, students who participated in afterschool programs showed better school performance than those who did not, before and after the matching. Nevertheless, students who took afterschool programs had 4.64% higher school achievement than those who did not before the matching, whereas the difference in school achievement between the two groups grew to 5.81% after the matching. And also, according to the comparative analysis of the impacts of watching EBS programs on school achievement based on the propensity score matching, students who watched EBS programs had better school achievement than those who did not before and after the matching. The difference in school performance between those who watched EBS and those who did not reached 5.13% before the matching, which decreased to 5.04% after the matching. According to the comparative analysis on school achievement effects from private tutoring, afterschool programs or EBS based on the propensity score matching, private tutoring posted 6.82%, EBS 5.04% and afterschool programs 4.80%, indicating high participation. By school level, private tutoring scored 8.17%, afterschool programs 3.13% and EBS 0.74% in elementary school, whereas private tutoring reached 12.94%, EBS 3.62% and afterschool programs 2.62% in middle school, indicating that private tutoring had the largest effects in middle school. Nonetheless, afterschool programs marked 10.88%, EBS 6.09% and private tutoring 4.18% in general high school, which is in the reverse order of the order of elementary or middle schools and indicates the larger influence of afterschool programs in general high schools compared to that of private tutoring.