High School Emotional Well-Being: Identifying Factors And Creating A Working Model For Prediction

Abstract

In high schools across the world, a mental health crisis is happening. High school students’ mental health has been decreasing at an alarming rate since the 2000s, and the negative impacts of this crisis such as suicide rates have been exacerbated. Two significant issues are causing the problem. The first is the lack of data concerning high school students’ emotional well-being. There is no current data set that attempts to correlate and categorize the causes of these mental health issues with the emotional health of students. The second is the lack of available tools to help teachers and counselors identify those with mental health issues and help them get access to treatment. This study resolves these two problems with the creation of an accurate machine learning algorithm to model and predict students’ mental health as well as the collection of a data set about high school students that contain potential causes. The model attempts to use potential stressors collected by a survey to predict students’ mental health; the best possible model predicts mental health on the Subjective Happiness Scale with a mean squared error of 0.41 and an r2 of 0.52, which is satisfactory for a subjective study.