As part of our efforts to improve the quality of university education in statistical data analysis and the implementation of innovative teaching methods, we have developed a teaching aid. In this contribution, we describe the experience we gained in creating this teaching aid — a simulation model — by modeling the values of a variable so that they correspond to real data and their true variability. A full factorial experiment was used to model the dependence of the observed response on the input levels of independent variables, which we varied according to the Design of Experiments methodology. A regression model, along with an estimate of data variability, was obtained from the measured response values through statistical evaluation. The prediction accuracy of the regression model was then compared with the results obtained from a neural network output. Additionally, we compare the obtained results with outputs from artificial intelligence developed by OpenAI, which is increasingly preferred across various fields.

Keywords: Innovative teaching, teaching aid, simulation model, data variability, predictive modeling.