Abstract:

This research attempts to contribute to the study of motivation using machine learning model to analyze students’ performance in the conditions of multiple lockdowns during the Covid-19 pandemic. In the described situation, the issue of motivation comes to the agenda because it is a key factor of effectiveness in educational sphere. The focus of the academicians' attention has switched from a variety of motivational types to the ability to detect using a machine learning model, which motivational type will work better for students' achievements during remote learning.

The aim of the study is to analyze the impact of students' motivation on their learning performance during Covid-19 pandemic. For this purpose the authors created a questionnaire that measured the levels and types of students' motivation as swell as the students’ ability to adapt to a new way of studying in a challenging situation.

The article analyzes intrinsic motivation, extrinsic motivation and amotivation as the main types, that influence high school students' performance during the Covid-19 pandemic.

The authors used a neural network regression model to find out which type of motivation is the most significant for students' performance during the lockdowns. For this purpose the drop column method was used to calculate the importance of each motivation type. The choice of Machine learning can be explained by its precision and sensitivity to the data analysis, providing complex dependences involving a large number of variables.

The preliminary results determined that the anxiety and feeling of having the situation under control were the most important factors for the first year of study. At the same time, the application of multiple set of hyperparameters for the data obtained can be used in the second year of the research.

Keywords:

Machine learning model, motivation, neural network regression model, hyperparameter, Conid-19.