A Content Analysis Of Educational Research Methods In American

All of the articles published in American Educational Research Journal (AERJ) from Vol.35 to Vol.43 during 1998 to 2006 are the object of the research, which is a content analysis of educational research methods used in these articles. And there are 242 articles - except for review, contend, and editors' introduction - that consist the valid sample of the research. The methodology of the research is content analysis, and meanwhile, gives the description of the using of research methods by range and standard deviation, and analyzes the correlation between two methods by chi-square.Based on the valid consistency of classification, the author classified the research methods used in the 242 articles with three main parts: collecting data, analyzing data and presenting results, and found that the method of collecting data used most frequently in AERJ was interview and observation took the second place, scales/questionnaire took third, literature fourth, database fifth, experiment sixth, ethnography seventh, and portfolio was the one used least to collect data, and the method of analyzing data used most frequently was qualitative, the second one was multiple regression/correlation, third descriptive, fourth ANOVA/ANCOVA, fifth t test, and then bivariate correlation, factor/cluster analysis, nonparametric, comparative, multivariate, modeling, path analysis, SEM/LISREL, meta-analysis and so on. Of these methods, database was used most stably to collect data and ANOVA/ANCOVA was the most stable one used to analyze data. As a kind of presenting data, graphic methods were used frequently and more and more in AERJ.Then according to previous researches, the author discussed how were these methods used or developed during 1969 to 2006, for example, more application of triangulation, more attention of researching on multiple influences on education and teaching, and analyzing exploratory data and so on. Finally, the author concluded that educational research methods were varied nearly with the development of education and teaching; however, statistic methods had been all the time the mainstream of educational research methodology in American, although it was quite different for the development of methods during these recent several decades, and this is the significant difference contrasted with our educational research. So it would be more scientific for us to research education if the difference is diminished.