Objectives In this study, we intend to develop a factual situation of factor analysis methods in the field of education and a social science research analysis guide program. Methods In order to investigate the factor analysis method of the literature in the field of education, an analysis layer tool is selected, and the selected tools are used to analyze the education-related literature of the last five years (2019-2023). Based on the analyzed contents, a survey is conducted on educational researchers, and based on the results, a Delphi technique is conducted to determine the cause. Based on the identified causes and data, a model of the social science research analysis guide program will be selected, and the program of the selected model will be developed. Results Big data harvesting was selected as the analysis layer tool, and as a result of using it, 74.6% (1,400 articles) used the right angle rotation method as a factor rotation method among 1,876 documents that released factor analysis methods among 55,400 documents in the field of education over the past five years, and 25.4% (476 articles) used the square rotation method. Nevertheless, as a result of analyzing the cause of why most of the square rotation methods are not used, research in the field of education tends to rely on experts or companies due to lack of reference, advice, and statistical knowledge, and it can be found that the selection of research methods is influenced by the excellence of past practices and results, and the need for statistical knowledge is emerging due to the generalization of previous studies. It provided data on the causes of analysis method selection and preferred statistical knowledge acquisition for a group of experts in the field of education and statistics, and analyzed opinions on the social science research and analysis guide program model. It was divided into eight characteristics: automation of research methods, practical applicability, learning forms, provision of various analysis methods, user convenience consideration, update and maintenance, consideration of non-majors, and measures to overcome limitations. Based on these characteristics, the development of the social science research and analysis guide program was completed by dividing it into one main program and two auxiliary programs, and dividing it into three website pages. Conclusions The social science research analysis guide program will be operated free of charge after continuous update, revision, and supplementation by establishing a dedicated website. By allowing anyone to utilize all development components through the Git Hub, it is intended to help develop a more suitable and complete social science research foundation by improving the statistical lack of knowledge and the conventional form of using prior papers. Even if the statistical understanding was high while developing the program, there was a limit to fully presenting all social science research and analysis methods. This will improve through continuous updates, revisions, and supplements, but fundamentally, just as arguments conflict depending on whether statistical methodology is important like factor analysis or whether interpretation and use of results are important, various analysis methods are also difficult, and the developed program presents soft answers. It is hoped that more suitable research and analysis guides will be developed with shared data in the future.
Read full abstract