교육학 양적 연구의 혁신 ：영가설 유의성 검증과 p-값을 넘어서
Innovation on Quantitative Research in Education: Beyond “Null Hypothesis” and “P-values”
1
조회
0
다운로드
2
인용
0
참고문헌
키워드
초록
국내 교육학에서 수행된 대부분의 양적연구는 영가설 유의성 검증(NHST)의 p-값을 사용하여 통계 검증을 실시하고 있지만 사실 이 검증방법은 오랫동안 많은 연구에서 비판을 받아온 방법이기도 하다. NHST 와 p-값의 문제점은 중요한 연구 주제임에도 불구하고 아직까지 국내 교육학 논문에서는 이와 관련된 주제를 다루고 있지 않았다. 따라서 본 논문에서는 국내 교육학 논문으로는 처음으로 NHST의 문제점을 살펴보고 그 대안을 탐색하는 연구의 의의를 갖고 있었다. 먼저 본 논문에서는NHST와 p-값의 핵심 문제점을 이해할 수 있도록 NHST와 p-값의 기초 개념을 설명하였다. 다음으로 NHST와 p-값의 주요 문제점을 설명한 후에 NHST의 p-값을 대처할 수 있는 방안으로 효과크기와 신뢰 구간을 제안하였다. 끝으로 교육학 분야의 양적 연구를 개선하기 위한 세부적인 대안 방법을 제시하였다.
Abstract
Although the statistical approach commonly used in most special education research is based on Null Hypothesis Significance Testing(NHST) and p-values, there are a large number of research on criticizing the methodology of NHST over time. In Korea, however, little research on education has focused on this issue, so the main purpose of this paper was to summarize main issues of problems on NHST and p-values and suggest some alternatives for NHST. First, this papers provides the underlying theoretical framework of NHST so that it is possible to understand limitations and the misunderstandings of NHST and p- values. Second, this paper presents two main alternatives for NHST; Effect size and confidence interval. Finally, this paper provides some suggestions to overcome problems of NHST in the area of education in Korea.
참고문헌 (26)
- 김민성, "Quantitative Methods in Geography Education Research: Concept and Application of Effect Size", The Journal of The Korean Association of Geographic and Environmental Education, 19(2), pp.205~220 (2011), The Journal of the Korean Association of Geographic and Environmental Education[DOI]
- 김태영, "증거기반의 반복연구 분석: 국내 특수교육 분야의 학령전기 장애 및 장애위험 아동 대상 중재연구", 특수교육논총, 36(2), pp.45~66 (2020), 특수교육연구소
- Seungsoo Yeo, "Null Hypothesis Significance Testing in Special Education Research: Problems and Solutions", Special Education Research, 20(3), pp.33~53 (2021), The Special Education Research Institute[DOI]
- Ahlbom, A., "Significance testing: Why does it prevail?", European Journal of Epidemiology, 32(1), pp.1~2 (2017)
- Branch, M. N., "The “Reproducibility Crisis:” might the methods used frequently in behavior-analysis research help?", Perspectives on Behavior Science, 42(1), pp.77~89 (2019)
- Bruns, S. B., "P-curve and p-hacking in observational research", PloS one, 11(2), pp.e0149144~ (2016)
- Chow, J. C., "Prevalence of publication bias tests in speech, language, and hearing research", Journal of Speech, Language, and Hearing Research, 61(12), pp.3055~3063 (2018)
- Cohen, H. W., "P values : use and misuse in medical literature", American journal of hypertension, 24(1), pp.18~23 (2011)
- Cook, B. G., "Null effects and publication bias in special education research", Behavioral Disorders, 42(4), pp.149~158 (2017)
- Di Leo, G., "Statistical significance : p value, 0. 05threshold, and applications to radiomics—reasons for a conservative approach", European radiology experimental, 4(1), pp.1~8 (2020)
- Edwards, A. W. F., "Likelihood: An account of the statistical concept of likelihood and its application to scientific inference" (1972), Cambridge University Press
- Fanelli, D., "Negative results are disappearing from most disciplines and countries", Scientometrics, 90(3), pp.891~904 (2012)
- Gerber, A. S., "Publication bias in empirical sociological research: Do arbitrary significance levels distort published results?", Sociological methods & research, 37(1), pp.3~30 (2008)
- Gigerenzer, G., "Mindless statistics", The Journal of SocioEconomics, 33(5), pp.587~606 (2004)
- Hoekstra, R., "Probability as certainty : Dichotomous thinking and the misuse ofp values", Psychonomic bulletin & review, 13(6), pp.1033~1037 (2006)
- Matthews, R. A., "Beyond ‘significance’ : principles and practice of the analysis of credibility", Royal Society Open Science, 5(1), pp.171047~ (2018)
- McShane, B. B., "Statistical significance and the dichotomization of evidence", Journal of the American Statistical Association, 112(519), pp.885~895 (2017)
- McShane, B. B., "Abandon statistical significance", The American Statistician, 73(sup1), pp.235~245 (2019)
- Plucker, J. A., "Replication is important for educational psychology : Recent developments and key issues", Educational Psychologist, pp.1~11 (2021)
- Simonsohn, U., "P-curve : a key to the file-drawer", Journal of experimental psychology : General, 143(2), pp.534~547 (2014)
- Smith, R. J., "The continuing misuse of null hypothesis significance testing in biological anthropology", American journal of physical anthropology, 166(1), pp.236~245 (2018)
- Szucs, D., "When null hypothesis significance testing is unsuitable for research : a reassessment", Frontiers in human neuroscience, 11, pp.1~12 (2017)
- Trafimow, D., "A test of the null hypothesis significance testing procedure correlation argument", The Journal of general psychology, 136(3), pp.261~270 (2009)
- Travers, J. C., "Null hypothesis significance testing and p values", Learning Disabilities Research &Practice, 32(4), pp.208~215 (2017)
- Verdam, M. G., "Significance, truth and proof of p values : reminders about common misconceptions regarding null hypothesis significance testing", Quality of Life Research, 23(1), pp.5~7 (2014)
- Wasserstein, R. L., "Moving to a world beyond “p< 0.05”" (2019), Taylor & Francis