A flawed experimental design scheme is an important factor leading to biased or even erroneous findings and conclusions. In sociological studies involving sampling, where researchers are often forced to make some kind of compromise in their experimental methods, such as using unbalanced sampling methods or replacing random sampling with convenience sampling, due to the constraints of the experimental conditions. It is not entirely unacceptable to conduct a study using a flawed experimental design, but such a compromise should not lead to serious biases or errors in the study results and conclusions. We need to pay more attention to the review of experimental design schemes and use scientific methods to make the necessary assessments of certain compromises in experimental schemes. To evaluate the correlation and predictive value between variables using a 3R method. Using the 3R method (a combined application of linear regression, ROC curve analysis, and R software to evaluate the correlation between variables and their predictive value), the ROC curve was introduced into the linear correlation regression analysis, and the R software was used to calculate the regression equation, AUC, sensitivity, specificity, and Jorden index to make a precise and accurate judgment of the correlation between variables.The linear regression model established for two variables with linear correlation was statistically significant, and ROC curve analysis was performed to quantitatively evaluate the predictive value between the variables. ROC curve was introduced into linear correlation regression analysis enabled a more accurate evaluation of the correlation between variables and their predictive value based on precise analysis. the R software was suitable for such analytical work.