양적 연구에서 도출된 회귀계수 추정값의 실질적 유의성에 관한 연구 - 한국행정학보와 한국정책학회보를 중심으로 -

초록

The purpose of this paper is twofold. First, we examine whether the analysis of practical significance is conducted among public administration and policy research that use regression analysis. Second, we emphasize the importance of examining practical significance of statistical results, in order to promote the quality of quantitative studies in public administration and policy. The logic that is often used in modern statistical inference is Fisherian inference. Researchers examine the significance of quantitative studies by estimating the effect of an explanatory variable on an outcome variable and analyzing whether the effect estimates are statistically significant based on the Fisherian logic. Note, however, that the significance of any research depends not only on statistical significance but also on the effect size (practical significance). An in-depth analysis of the effect estimates is, therefore, necessary in order for a researcher to derive appropriate policy implication from the estimated results. By examining the regression analysis published in Korean Public Administration Review and Korean Policy Studies Review from 2014 to 2018, we find that the share of published papers that analyzed the practical significance of the estimated results is only 1.37%. We argue that researchers should devote more on analyzing practical significance of estimated effects in order to secure conclusion validity of public administration and policy research.

키워드