Exploring Machine Learning Techniques for Questionnaire Analysis

The application of machine learning in questionnaire design and evaluation is examined in this work, with a special emphasis on data quality, specifically Cronbach's alpha. It evaluates the effect of a method for creating various Cronbach's alpha random data (0.8–0.94) on machine learning algorithms. The method minimizes human labor in questionnaire design by generating data with estimated internal consistency through the use of machine learning and natural language processing. To assess how Cronbach's alpha affects the accuracy of machine learning models, the data is analyzed. When dealing with low Cronbach's problems, the ML models RF, NB, LR, SVM, and KNN, respectively, have acceptable and exceptional accuracy (58% - 100%). However, when dealing with high Cronbach's problems, their accuracy drops to 26% - 79%.