Using Text Mining to Gauge Student Sentiments about Hands-on Activities in a Protected Agriculture Course

Abstract

Hands-on activities enhance learning and increase student satisfaction in horticulture courses. Nevertheless, hands-on activities can have widely different impacts on student learning. To achieve and maintain educational quality, instructors need to evaluate and improve activities. This research used text mining and sentiment analysis to gauge student sentiments about hands-on activities in a protected agriculture course. Students participated in five hands-on activities and submitted short reflective essays about them. Essay texts were separated into single-word (unigram) or three-word (trigram) objects. Unigrams were compared with general-use emotion lexica to extract student sentiments from the texts. Trigrams were used to assess essay content. All activities elicited positive sentiments among students. Trust, anticipation, and joy were the most prominent emotions identified. The activity focused on freeze protection was preferred over the other activities. Although other activities were also well received, they should be refined for future offerings. The presented method could be used to assess hands-on activities, leading to continuous improvement and successful implementation of experiential learning in horticulture courses.