Abstract

• Over the past 17 years, the number of medical research papers in the Ewha Medical Journalhas increased. And the use of statistics to evaluate experimental and clinical research data has increased. We observed statistical methods used in 730 articles from the Ewha Medical Journalfor a 17 year period(from 1978 to 1995). The most commonly used method was measurementof central tendency with t-test, χ²-test, ANOVA, correlation analysis and regression analysis(74.8%). The second most common method was χ²-test with t-test, ANOVA, pearson correlation(9.2%), fo11owed by epidemiological statistics(7.6%).
  The observed errors of statistical method in 314 original aricles were 585 spells(repeat No./paper).
  The most common error was that of no significant value and confidence interval(18.4%), followed by misuse of SE instead of SD(15.5%).
  Misuses of applicating statistical method included the use of the t-test instead of ANOVA incomparing three groups(0.2%) and the use of the student t-test instead of paired t-test and vice versa.
• Keywords: Statistics; Statistical method; Misuse