Replication Research And Bayes Factor Analysis In Quantitative Research

With the continuous occurrence of non-repeatable problems in scientific research,repeatability has become a focus in the academic circle.Replication research is an important method to solve the reproducible crisis.At present,the replication research focuses more on concepts,types,meanings,etc.,ignoring the practice and analysis of the details of published research,and lacking a method of to assess the reliability of research results.In response to these problems,this study uses the problem of the large gap in the estimated results of rural education return rate as a case background,and conducts a replication study on three papers.Firstly,this research uses the same research samples,model and method as the original research to conduct "Checking of analysis" to verify the authenticity of the original research;secondly,this research uses Bayes Factor to conduct "Reanalysis of data " to quantify the reliability of the research results,realize the expansion of the original research.The research found that:(1)Replication research helps readers to judge the reliability of published research,and can also recognize the error in published research.This has many advantages for opposing pseudoscience and improving quantitative researchers’ quantitative analysis technical capabilities.(2)After reanalysis using Bayes Factor,the three published papers show different degrees of reliability.The Bayes Factor has the characteristic of "quantitative research hypothesis proof" which makes it a method to evaluate the reliability of research results.This research accumulates the experience for the practice of replicative research in the quantitative analysis of social science and the application of Bayes Factor method,provides a new perspective for assessing the reliability of published research,and it also has important practical significance to judge which rural education return rate is more credible.