This study aimed to understand the factors of positive school experiences from children's own perspectives. The foundation of this research depended on school satisfaction, school connectedness, and academic self-perception. Overall, these terms encompassed general feelings about enjoying school, engaging with the rules and content, and thinking of oneself as smart. I wondered how these variables individually and cohesively contributed to positive school experiences and how these relationships changed depending on students' ages. Based on previous literature, I predicted that students would feel less positively about their school experiences as their grade level increased. My methodology was based on interviews with students between the ages of four and ten years old. I constructed a semi-structured interview about their experiences at school and conversed with participants individually for about ten minutes each. After completing interviews, I analyzed all the responses and created six codes, or themes, based on participants' sentiments - liking school, having time for play, socializing with peers, appreciating learning, feeling smart, and having agency. These six codes were the ultimate contributors to students' positive school experiences. Participants were divided into three groups for analysis, and reports for students in the youngest age group were compared to responses on teacher questionnaires. After analysis, it was revealed that students in the youngest age group had the most positive school experiences. Students rated play and social factors of school as most important, independent of age. In contrast, students in the oldest age group reported the highest scores for smart and agency. Future research has been proposed to investigate differentiations of play between grade levels and the possible benefits to socioemotional development and academic achievement.

We propose an algorithm that constructs relationships among any number of seed words. A relationship consists of a set of iteratively-generated paths of similar words, where each path links one seed word to another. The similar words are generated using Word2Vec word embeddings and the cosine similarity measure. By examining the effectiveness of the proposed algorithm in the mental health domain, we find that the algorithm effectively returns meaningful relationships and has the potential to be used for hypothesis generation and information extraction.