University Students' Perceptions on Climate Change Awareness and Sustainable Environments Through an Unsupervised Clustering Approach

Abstract

The main objective of this study is to determine the knowledge and awareness levels of climate change among preparatory class students at Zonguldak B & uuml;lent Ecevit University in the Western Black Sea Region of T & uuml;rkiye using an unsupervised clustering approach. Within this scope, a survey was administered to university students (n = 280). Participant scores for the survey sections containing five-point Likert-type questions on climate change awareness were calculated using min-max normalization. The normalized data was then processed using the k-means algorithm, a well-known technique in unsupervised machine learning. This resulted in a classification (clustering) related to climate change awareness. The number of clusters was determined using the Silhouette index. Three clusters identified using k-means and Silhouette index (S approximate to 0.55) revealed the knowledge and application levels of student groups regarding climate change awareness. As a result of clustering, it was determined that Cluster-3 students (n = 134, 47.9%), defined as having a high level of knowledge and application, had a higher impact value in their overall assessments of green space-focused issues related to climate change awareness compared to the overall assessments of students in other clusters. Some notable findings concerning the attitudes of Cluster-3 students highlight climate change awareness-related practices. These include minimizing water consumption to levels necessary for ecosystem water management (mean = 95.7, std. deviation = 10.9) and exercising controlled, sustainable daily energy use to alleviate pressure on green spaces (mean = 94.4, std. deviation = 12.5). This study offers practical insights for policymakers, educators, and institutions, emphasizing the need to enhance climate education and to promote the active involvement of younger generations in shaping sustainable environments.