Prediction of students' COVID‐19 resilience: An artificial neural network based on gender, age, stress intensity, and mindfulness variables

Abstract

AbstractThe global health emergency, COVID‐19, significantly influenced schooling in Indonesia. Students employed a variety of coping mechanisms to cope with unusual stress levels during confinement time. Hence, as students' COVID‐19 resilience, investigation, and prevention were required for high and chronic stress connected with various disorders. This study aimed to design a predictive model of students' COVID‐19 resilience based on artificial intelligence that included certain demographic variables, stress intensity, and mindfulness and to study the relationship between them. A total of 6580 Indonesian students were involved in this study (57.9% female and 70.3% aged between 13 and 15 years old). The prediction model was performed by the architecture of artificial neural networks. The results showed that the model's predictive capacity was over 63% in the testing phase, then reached almost 65% in the holdout phase. Students' COVID‐19 resilience was mainly predicted by stress intensity and mindfulness with 100% and 40.9% normalized importance values, respectively. Receiver operating characteristic curve assessed and remarked the model as more superior than random. Our research gave some insight into the use of artificial intelligence in educational research to predict psychological variables.