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 Bülent Ecevit University in the Western Black Sea Region of Türkiye 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 ($S\approx 0.55)$ 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.

1. Introduction

Climate change is widely recognized as a global threat exerting multidimensional impacts on both natural systems and human livelihoods [1,2,3]. The potential threats posed by climate change in the coming decades are increasingly raising global awareness and concern [4]. Indeed, everyday experiences with extreme weather and climate events can serve as a critical determinant in shaping individuals’ concerns and perceptions of climate change [5,6]. The literature indicates that environmental concerns have been increasingly emphasized within the broader framework of sustainable environmental awareness. Recent studies highlight the growing prevalence of climate-induced challenges, including extreme temperatures, droughts, storms, wildfires, and floods, which have become more pronounced in recent years [7,8,9,10]. Furthermore, research consistently demonstrates that global warming is strongly linked to anthropogenic activities, particularly greenhouse gas emissions originating from fossil fuel use, driving behaviors, and unsustainable patterns of production and consumption [9,11]. According to the Intergovernmental Panel on Climate Change [12], climate change is projected to intensify across all regions in the coming decades. At a global warming level of 1.5 °C, more frequent and severe heatwaves, longer warm seasons, and shorter cold seasons are expected. Under a 2 °C warming scenario, extreme heat events are anticipated to occur more frequently and to surpass critical tolerance thresholds for both agricultural productivity and human health. These projections underscore the urgency of implementing mitigation and adaptation strategies to reduce risks and safeguard ecosystems and societies.

Recent studies suggest that individual practices and societal inaction—often dismissed as insignificant at the individual or collective level—may considerably intensify the consequences of climate change-induced disasters or give rise to novel forms of disaster risk [13,14]. As future decision-makers, university students are expected to play a pivotal role in advancing both adaptation and mitigation strategies in response to climate change [15]. Gaining deeper insights into how young people perceive, interpret, and conceptualize this global phenomenon is therefore crucial for developing effective responses to a challenge that poses long-term risks to both human societies and ecosystems [16,17,18]. Such understanding not only informs educational and policy interventions but also contributes to fostering the active engagement of younger generations in sustainability transitions.

In this context, the individual environmental practices acquired at an early age regarding climate literacy, environmentally focused approaches, and sustainable living practices directly influence social transformation processes [15].

Universities have significant potential to educate students on climate change issues and engage them as active citizens in the system [19,20,21,22]. In this context, it is important for universities to expand climate change education with a holistic and interdisciplinary approach, in addition to technical approaches. Developing an effective understanding of climate change in universities is critical for future generations to be able to assess this global issue in its multidimensional aspects [22].

Although climate change is a topic of great importance to universities, its coverage within university programs needs to be expanded [21,22]. Maximizing the contribution of higher education institutions to addressing the challenges of climate change and developing programs to address this need are crucial [19,21,23,24]. Programs integrated into universities’ climate change education are vital for students to gain basic and advanced skills and knowledge, thereby shaping behaviors and lifestyles that will positively impact their future [15,22,25].

The term “climate change” is generally discussed in the context of global warming and the increase in global surface temperatures. However, due to a lack of environmental responsibility awareness, full understanding of the issue remains limited [26]. Therefore, our study aims to determine the knowledge and awareness levels of preparatory class university students regarding climate change in terms of environmental responsibility awareness. Applying an unsupervised clustering approach, the article reveals heterogeneous patterns of perception, indicating that students conceptualize climate change, its consequences, and sustainability practices in different ways. This heterogeneity highlights the need for differentiated pedagogical and institutional strategies to enhance climate literacy and promote environmentally responsible individual practices in higher education contexts.

2. Conceptual Framework and Research Questions

Around the world, various studies are being conducted on climate change, and efforts are being made to raise awareness about environmental protection. In Turkey as well, a range of studies, plans, and reports are being developed to address climate change.

While the existence of national and international efforts related to global warming and climate change is seen as a positive development, the aspect of raising individual and social awareness on the subject is equally important. In this context, Article 6 of the United Nations Framework Convention on Climate Change stipulates the commitment to conduct educational and awareness-raising activities aimed at teaching, training, and public awareness [27].

Climate change is one of the most threatening issues for the future [28]. Environmental issues arising from climate change are closely linked to individuals’ knowledge, attitudes, and awareness levels regarding these matters. Therefore, one of the prerequisites for addressing environmental problems is determining awareness levels and taking measures to develop positive attitudes towards the environment [14,29,30,31]. Changing individual practices and taking environmental action for a sustainable life also supports social development [32]. Positive changes in individual practices are desired for the protection of nature [33].

Within the context, university students, as future decision-makers, will play critical roles in the processes of adapting to climate change and mitigating its effects [15]. In particular, identifying differences in individual risk perceptions and environmental awareness among university students is crucial for developing effective adaptation and education strategies [3]. Studies focusing on measuring awareness of climate change at the national and international levels have been evaluated through university students [6,15,18,22,26,34,35,36,37,38].

Unlike previous studies conducted to reveal university students’ awareness of climate change, this article includes assessments for preparatory class students. Determining the knowledge-awareness-practice levels of new students regarding climate change and its effects will provide opportunities/advantages in terms of creating academic content that will increase awareness during university education and planning campus applications. The current study applies unsupervised clustering as an innovative method to uncover hidden patterns in how students perceive both climate change and sustainable environments. This approach contributes to sustainability research by going beyond averages to reveal meaningful subgroups. Methodologically, the study advances the field by applying an unsupervised machine learning approach to investigate students’ climate change awareness and perceptions of environmental sustainability. The clustering analysis uncovers heterogeneous perception patterns, illustrating that students conceptualize climate change, its consequences, and sustainability practices in distinct ways. This heterogeneity highlights the necessity of differentiated pedagogical and institutional strategies to cultivate climate literacy and promote environmentally friendly individual practices within higher education contexts. By employing an innovative data-driven technique, this research contributes novel methodological insights to perception-oriented sustainability studies. Practically, the findings generate actionable implications for policymakers, educators, and higher education institutions, offering evidence to strengthen climate change education, enhance sustainability literacy, and mobilize youth as active agents in sustainability transitions. The article is structured around research questions in order to demonstrate the impact and contribution of the determined methodology to the subject. The research questions were designed purposefully to reflect these aspects while emphasizing the study’s methodological contributions.

This article outlines 2 major research questions:

RQ1

Can students be grouped according to similarities/differences in their level of knowledge and application success regarding climate change awareness?

RQ2

What kind of practical activities can be implemented for university students to address the identified differences in the level of knowledge and application performance regarding climate change awareness?

3. Methodology

The survey data used in this study were collected between February and April 2024 through a questionnaire administered to preparatory class students at Zonguldak Bülent Ecevit University’s İncirharmanı Campus. Excel 2016 was used for entering, organizing, visualizing, and performing basic statistical analyses of the survey data. Cronbach’s alpha coefficients were calculated for each item and for all items within the three sections containing five-point Likert-type questions on climate change awareness. In addition, normalized section scores for climate change awareness were calculated using the min–max method in Excel 2016. Statistical analyses, clustering algorithm (k-means) results, calculation of Silhouette index values and their graphical presentations, and Pareto analysis were performed using code written in MATLAB 2024b. The normality of the data was examined using the Shapiro–Wilk test. Comparisons of more than two samples were conducted using the Kruskal–Wallis H-test when the normality assumption was not met. Comparisons between two samples were conducted using the Mann–Whitney U test. Clustering analysis was conducted on the dataset containing section scores using the unsupervised machine learning k-means algorithm. The Silhouette index was employed to determine the optimal number of clusters affecting cluster quality. A p-value of less than 0.05 was considered statistically significant.

3.1. Data Collection Tool and Study Sample

A comprehensive survey consisting of three parts was created to assess the perspective of preparatory students at Zonguldak Bülent Ecevit University’s İncirharmanı Campus on environmental issues within the scope of climate change. The survey parts and their question distributions are as follows:

• Part-1 (Demographic information): There are three questions regarding the participants’ demographic characteristics (gender, age, residence).
• Part-2 (Climate Change—Overview): There are six questions designed to determine students’ level of knowledge on climate change.
• Part-3 (Awareness level): In order to determine students’ awareness of climate change, there are a total of 38 questions in a 5-point Likert scale under three sub-sections. The section-based distribution of these 38 questions regarding climate change awareness is given in

  Table 1. Sections and items.

  Section	Item
  Section-1: The consequences of climate change	I1.1: Increasing drought and desertification
  	I1.2: Disruption of ecological balance (air–water–soil pollution, decline in biological diversity)
  	I1.3: Melting of glaciers
  	I1.4: Decline in groundwater resources
  	I1.5: The occurrence of seasonal change
  	I1.6: Rising sea levels
  	I1.7: Silt and flood formation (precipitation changes)
  	I1.8: Increased air pollution due to the reduction in green spaces and the resulting rise in public health problems
  	I1.9: Forest fires
  	I1.10: The disruption of water resources balance leading to a decline in agricultural production and the emergence of food shortages
  	I1.11: The occurrence of extreme weather events (storms, hurricanes, etc.)
  	I1.12: Decline (degradation) and migration (human-plant-animal) resulting from pressures on vegetation cover and wildlife areas
  Section-2: Effective methods for preventing climate change	I2.1: Protection and enhancement of forests
  	I2.2: Use of energy-efficient products
  	I2.3: Taking measures to prevent or reduce greenhouse gas emissions
  	I2.4: Promoting environmentally friendly investments
  	I2.5: Purchasing environmentally friendly products
  	I2.6: Reducing the use of substances that damage the ozone layer
  	I2.7: Monitoring and raising awareness of factories’ environmental responsibilities
  	I2.8: Protecting and supporting water resources through rainwater management
  	I2.9: Waste separation
  	I2.10: Taking individual and administrative measures to prevent environmental pollution
  	I2.11: Raising public awareness about global climate change
  	I2.12: The use of recyclable products
  	I2.13: Promoting the use of renewable energy sources (wind, solar, etc.)
  	I2.14: Promoting public transportation and sustainable transportation vehicles
  	I2.15: Expanding urban green infrastructure applications
  Section-3: Students’ level of environmental awareness	I3.1: I make sure that the products I buy are recyclable
  	I3.2: I consume water at the minimum level necessary for the water economy of ecosystems
  	I3.3: To reduce pressure on green spaces, I take care to meet my daily energy needs through controlled and sustainable energy consumption
  	I3.4: I take care to eat a diet that is high in nutritional value and sufficient for life, and to avoid creating food waste
  	I3.5: To prevent soil pollution, I make sure to separate my solid waste into plastic, glass, paper, waste oil, and batteries, and dispose of them in the appropriate bins
  	I3.6: I do not light fires in forest areas or litter (cigarette butts, glass, etc.)
  	I3.7: I leave food scraps for animals or animal shelters
  	I3.8: I make sure to keep my shower time short to save water
  	I3.9: When choosing cleaning products (detergent, shower gel, shampoo, etc.), I prefer products with chemical ingredients that are less harmful to the environment
  	I3.10: I make sure to purchase products with certified sustainable content that will not harm the forest ecosystem
  	I3.11: I volunteer at environmentally and nature-friendly clubs and events

  .

For the field implementation of this survey, the target population consisted of 953 preparatory students. Under a maximum uncertainty condition (p = q = 0.5), with a 4.93% margin of error and a 95% confidence level, the sample size was determined as 280. A simple random sampling method was employed.

3.2. Machine Learning-Based Approach: Unsupervised Clustering

Cluster analysis was performed to examine the similarities and differences in knowledge and awareness of climate change among the relevant preparatory class students. The data collected via the climate change awareness survey were normalized to the [0, 1] range using the min–max method. Accordingly, participant scores were computed for the sections of the climate change awareness survey comprising 5-point Likert-type items. A section scores dataset of dimensions 280 × 3 was constructed using the participants’ scores for the respective sections. K-means clustering [39], one of the well-known techniques of machine learning, was performed on this section-scores dataset to evaluate participants’ knowledge and awareness toward climate change. Due to the unsupervised nature of k-means algorithm, the Silhouette index [40] was used to determine the appropriate number of clusters affecting cluster quality. The framework of the clustering-based approach to climate change awareness assessments is presented in Figure 1.

4. Results

This section presents the results of the study.

4.1. Data Analysis and General Evaluations

62.1% of survey participants were female (n = 174; mean age = 19.87 and standard deviation/SD = 1.77), and 37.9% were male (n = 106; mean age = 19.91 and standard deviation/SD = 1.94). 57.5% of participants reside in urban areas/center (n = 161), while 42.5% reside in rural areas (n = 119). The gender and residence distributions of the participants are illustrated in Figure 2.

According to participants, the environmental issues causing the most concern were Declining Biodiversity (A-3) and Global Warming (A-10) in first place (n = 142, 50.7%), Desertification (A-2) and Climate Change (A-4) in second place (n = 135, 48.2%), Water Scarcity (A-9) ranked third (n = 107, 38.2%), Ozone Layer Depletion (A-1) ranked fourth (n = 91, 32.5%), Irregular Urbanization (A-7) ranked fifth (n = 70, 25.0%), Air Pollution (A-8) ranked sixth (n = 66, 23.6%), Pollution of Seas and Rivers (A-6) ranked seventh (n = 56, 20.0%), and Deforestation (A-5) ranked eighth (n = 33, 11.8%). The distribution of environmental issues causing the greatest concern by gender is shown in Figure 3.

When participants think of climate change, the first concept that comes to mind is Water scarcity (B-4) in first place (n = 124, 44.3%), followed by Drought (B-6) in second place (n = 85, 30.4%), Seasonal Changes (B-7) came third (n = 51, 18.2%), Sea Level Rise/Tidal Events (B-2) came fourth (n = 25, 8.9%), High Temperatures (B-3) came fifth (n = 22, 7.9%), Irregular Rainfall (B-1) ranked sixth (n = 10, 3.5%), Floods and Inundations (B-8) ranked seventh (n = 9, 3.2%), Heat Waves (B-5) ranked eighth (n = 7, 2.5%). When it comes to climate change by gender, the first concepts that come to mind are as shown in Figure 4.

Participants identified the factors causing climate change as Cutting down trees (n = 9, 3.2%), Deforestation (n = 10, 3.6%), Increase in the number of cars (n = 168, 60.0%), Increased use of airplanes (n = 1, 0.4%), Human activities (n = 51, 18.2%), Pollution (n = 20, 7.1%), Greenhouse gases (n = 2, 0.7%), and Urban disruption (n = 19, 6.8%). At this stage, Pareto analysis was employed to determine the most critical factors contributing to climate change. Pareto chart of factors causing climate change among participants is presented in Figure 5. According to Figure 5, the two most critical factors contributing to climate change are the increase in the number of cars (n = 168, 60.0%) and human activities (n = 51, 18.2%).

Table 2. The most effective method in raising/creating public awareness of climate change.

Item	General n (%)	Female n (%) *	Male n (%) *
Establishment of a monitoring mechanism by local governments	34 (12.1)	21 (61.8)	13 (38.2)
By ensuring that more educational programs are produced using television and other broadcasting tools	62 (22.1)	33 (53.2)	29 (46.8)
Teaching climate change awareness at the elementary school level	159 (56.8)	103 (64.8)	56 (35.2)
The implementation of necessary legal regulations	22 (7.9)	14 (63.6)	8 (36.4)
The organization of more seminars, conferences, and other educational activities	3 (1.1)	3 (100)	---

* row percentage.

summarizes the distribution of the most effective methods for raising and increasing public awareness of climate change among participants, both in general and by gender.

Participants believe that the following should be prioritized in adaptation efforts related to climate change: “C-1: Protection of natural water resources, prevention of pollution, C-2: Development of water management policies (agricultural and drinking water use), C-3: Implementing measures for the use of renewable energy sources, C-4: Establishing a green infrastructure system (increasing the number of applications such as parks, gardens, urban forests, etc.), C-5: Developing climate-appropriate agricultural policies, C-6: Expanding existing forest areas.” The top three preference rankings for these items are as shown in Figure 6. According to Figure 6, C-4: Establishing a green infrastructure system (increasing the number of applications such as parks, gardens, urban forests, etc.) was the most frequently selected option by participants at 50% (n = 140).

98.6% of participants (n = 276) believe that climate change has contributed to the recent extreme rainfall in the Western Black Sea region (flooding in the Bartın, Ereğli, and Devrek areas).

The awareness of the students participating in the survey regarding climate change was measured using a total of 38 questions in a 5-point Likert scale under three sections. The number of questions for each section, Cronbach’s alpha coefficient of each section, and the total Cronbach’s alpha coefficient are given in

Table 3. Reliability of sections related to awareness of climate change.

Section	# of items	Cronbach α
Section-1:	12	0.8578
Section-2:	15	0.9141
Section-3:	11	0.8528
Total:	38	0.9053

Section-1: The consequences of climate change, Section-2: Effective methods for preventing climate change, Section-3: Students’ levels of environmental awareness.

.

4.2. Clustering—Results

Data preprocessing for climate change awareness-based clustering was performed as in Figure 1. Student responses (n = 280) were collected using 5-point Likert-type questions for the items listed in Table 1. These collected data were normalized using min-max method. Then the normalized data were converted to section scores based on a 100-point system. Clustering of the relevant scores-data was performed using the k-means algorithm, one of the unsupervised machine learning techniques, for different cluster numbers (k = 2, …, 6). The average Silhouette values for the relevant scores data (280 × 3 dimensional data) were found as shown in Figure 7. According to Figure 7, the optimal number of clusters corresponding to the highest average Silhouette value ($S\approx 0.55$) is three (k = 3). Therefore, for the clustering analysis using the normalized section-scores, the number of clusters was set to three to ensure appropriate cluster quality.

The section scores of each cluster did not come from a normally distributed population (p-values < 0.05). A statistically significant difference was found among the three clusters formed as a result of the clustering analysis (p-values = 0.000). The demographic profiles of the clusters are presented in

Table 4. Demographic profile of each cluster.

Cluster (n)	Gender n (%)	Residence n (%)	Age by Gender and Overall (avr./std dev.)
Cluster-1 (n = 61)	Female: 31 (50.8) Male: 30 (49.2)	Rural: 24 (39.3) Urban: 37 (60.7)	Female: 20.03/1.78 Male: 19.93/1.72 Overall: 19.98/1.74
Cluster-2 (n = 85)	Female: 53 (62.4) Male: 32 (37.6)	Rural: 37 (43.5) Urban: 48 (56.5)	Female: 19.38/1.42 Male: 19.75/1.88 Overall: 19.52/1.60
Cluster-3 (n = 134)	Female: 90 (67.2) Male: 44 (32.8)	Rural: 58 (43.3) Urban: 76 (56.7)	Female: 20.11/1.90 Male: 20.00/2.14 Overall: 20.07/1.97

.

Some basic statistics for each cluster are given in

Table 5. Summary statistics for the scores of items in the sections.

		Cluster-1 (n = 61)		Cluster-2 (n = 85)		Cluster-3 (n = 134)
		Average	Std. Deviation	Average	Std. Deviation	Average	Std. Deviation
Section-1	I1.1	97.5	7.5	77.1	21.2	95.3	9.8
	I1.2	91.8	15.6	77.1	19.4	93.1	12.8
	I1.3	98.8	5.5	80.6	19	94	12.3
	I1.4	95.5	12.5	75	20	90.7	14.9
	I1.5	95.9	14.6	80.3	20.4	95	13.3
	I1.6	83.6	19.3	67.6	20	81	21.3
	I1.7	89.8	15.4	72.4	18.5	89.4	15.7
	I1.8	88.9	18.6	67.1	22.2	85.6	18
	I1.9	83.6	23.2	62.1	19.1	83.2	21.6
	I1.10	89.8	17.9	63.5	22.7	84.5	18.6
	I1.11	79.5	24.4	56.5	21.5	77.8	23.7
	I1.12	91	16.5	66.8	22.9	87.5	19.3
	General	90.5	17.6	70.5	21.8	88.1	18.1
Section-2	I2.1	97.5	8.8	84.1	18.8	97	8.1
	I2.2	92.2	16.2	75.6	18.1	95.7	9.5
	I2.3	90.7	18.4	72.9	22.7	95.5	12.1
	I2.4	95.1	11	77.9	20.2	96.6	9.1
	I2.5	90.7	18.4	65.5	23.3	92.8	17.5
	I2.6	94.5	15.1	72.2	22.9	95.5	14
	I2.7	98.4	7.3	75.3	23.1	98	8.9
	I2.8	98.4	7.3	76.5	22.3	97.5	9.7
	I2.9	89.3	16.1	70.9	18	93.1	12
	I2.10	92.9	17.3	69.4	20.7	94.8	12.8
	I2.11	94	14.3	66.3	24.4	93.5	14.4
	I2.12	94.3	16.1	76.5	17.4	95.7	9.5
	I2.13	96.7	10.7	79.7	17.9	96.8	8.4
	I2.14	84	25.8	67.1	20.8	89.6	16.9
	I2.15	95.5	14.8	75.3	19.9	95.3	10.7
	General	93.6	15.6	73.6	21.3	95.2	12.1
Section-3	I3.1	37.7	23.6	57.1	19.9	74.4	19.5
	I3.2	78.7	27.7	85.6	17.8	95.7	10.9
	I3.3	76.6	27	80.3	20.4	94.4	12.5
	I3.4	61.5	33.7	72.9	20.8	87.7	16.1
	I3.5	49.2	30.3	65.6	20.4	80.2	19.4
	I3.6	70.5	36.7	82.6	22.2	93.1	15.1
	I3.7	29.9	30.6	59.7	24.1	79.7	24.7
	I3.8	42.6	32.7	58.2	24.8	79.9	22.8
	I3.9	26.6	24.9	51.2	22.5	76.9	22.7
	I3.10	33.2	28.8	54.7	26.6	76.1	23.7
	I3.11	28.7	30.6	45.9	27.8	68.3	29.4
	General	48.5	35.3	64.9	26.0	82.4	22.1

Section-1: The consequences of climate change, Section-2: Effective methods for preventing climate change, Section-3: Students’ levels of environmental awareness.

. According to Table 5, cluster definitions were determined based on students’ knowledge and application levels regarding climate change awareness. Accordingly, regarding climate change awareness, Cluster-1 was proposed as “knowledge-moderate, application-low,” Cluster-2 as “knowledge-low, application-moderate,” and Cluster-3 as “knowledge-high, application-high”. In general, it can be said that 47.9% (n = 134) of participants have a high level of knowledge about climate change awareness and are engaged in related activities.

4.3. RQs Evaluations

4.3.1. RQ1 Evaluations

This section includes evaluations regarding Research Question 1: Can students be grouped according to similarities/differences in their level of knowledge and application success regarding climate change awareness?

In the study, the machine learning-based unsupervised clustering approach was proposed for grouping (clustering) the 280 students comprising the sample according to their knowledge levels and application performance regarding climate change awareness, based on their survey evaluations. Accordingly, the three clusters identified regarding climate change awareness were defined as Cluster-1: “knowledge-moderate, application-low,” Cluster-2: “knowledge-low, application-moderate,” and Cluster-3: “knowledge-high, application-high.”. As a remarkable result, 47.9% of students (n = 134) were found to have a high level of knowledge and implementation activities regarding climate change awareness.

4.3.2. RQ2 Evaluations

This section includes evaluations regarding Research Question 2: What kind of practical activities can be implemented for university students to address the identified differences in the level of knowledge and application performance regarding climate change awareness?

Within the scope of the article’s main objective and in line with the data obtained, practices that aim to increase individual and social responsibility awareness in the fight against climate change by providing students with both theoretical and practical information stand out, so that university students with different levels of climate change awareness can achieve an equivalent level of awareness. Equalizing university students’ awareness levels regarding climate change requires a holistic approach in both academic programs and campus life. The following strategies can help balance awareness levels at different levels and contribute to the creation of a shared awareness (

Table 6. Strategies for balancing different levels of awareness regarding climate change and creating a shared awareness among students.

Field	Type of Activity	Description
Education and Curriculum Development	Course Content and Seminars	Including topics such as climate change, sustainability, and environmental policies in the curriculum through compulsory or elective courses ensures that all students have access to fundamental knowledge.
	Workshop Activities	By organizing practical workshops and interactive seminars, theoretical knowledge can be put into practice.
Digital Platforms and Online Resources	Online Training Modules	Online courses, videos, and interactive content can be prepared to offer equal access to all students. This allows students to keep up with current information at their own pace.
	Social Media and Webinars	Informative campaigns and webinars organized through university social media channels can reach a wide audience on the effects of climate change and ways to address it.
Campus-Based Projects and Experiential Learning	Environmental Projects	Environmental projects and sustainability initiatives carried out on campus support students in gaining practical experience in the field.
	Community Studies	Through student clubs and volunteer organizations, knowledge sharing can be facilitated by encouraging students from different disciplines to work together.
Mentoring and Collaboration Programs	Mentoring Systems	Establishing mentoring programs between students who are more knowledgeable about the subject and those who are less knowledgeable facilitates the transfer of experience and knowledge.
	Group Work	Projects in which students from various disciplines and knowledge levels work together support mutual learning and equalization of awareness levels.
Communication and Awareness Campaigns	On-Campus Communication	Sharing up-to-date and accurate information about climate change through posters, brochures, announcements, and digital displays increases all students’ access to the subject.
	Events and Conferences	Conferences, seminars, and panel discussions featuring experts in their fields can be organized to facilitate interaction and knowledge exchange among students.
Measurement and Evaluation	Survey and Feedback Tools	Regular surveys and assessment tools can be used to determine students’ current awareness levels and measure the effectiveness of programs. This data will guide the continuous improvement of educational programs.

):

5. Discussion

Societies are moving towards a common goal of protecting life on Earth [41,42]. In line with this goal, it seems essential to increase efforts to both mitigate the effects of climate change and adapt to its consequences [43]. Environmental education is among the most important strategies for increasing society’s environmental awareness, knowledge, and attitudes [44]. Environmental awareness is an educational tool that helps people understand the esthetic, biological, and economic importance of protecting natural resources and minimizing the negative effects of environmental pollution [45]. Increasing environmental literacy can lead to changes in individual practices and actions [44]. Educating younger generations at universities will greatly benefit their development as future leaders. The future decisions of younger generations will affect the sustainability of human existence [46]. Individuals and communities having sufficient knowledge about climate change can contribute to sustainable development by encouraging the adoption of climate-friendly applications [30,47]. Conversely, lack of knowledge and conceptual misunderstandings pose a serious obstacle to efforts to mitigate the effects of climate change and adapt to these effects [48]. In this context, Prasad and Mkumbachi [3] emphasized that accurate and comprehensive climate education can form the basis for effective climate action, stating that higher education institutions play an important role in preparing for climate change and raising social awareness.

Nussey et al. [21] and Filho et al. [22] emphasize the need to expand the scope of university practices, despite climate change being a significant issue for universities. In this context, Filho et al. [23]; Filho et al. [24]; and Moltan-Hill et al. [19] have emphasized the ways universities integrate climate change practices into education on a global scale. In this context, it has been observed that limited studies have been conducted on the relationship between climate change and university students in Türkiye. In our article, we examine the knowledge and awareness levels of university students regarding climate change, using the example of Zonguldak Bülent Ecevit University in Türkiye, to guide institutional strategies to encourage differentiated practices in terms of individual responsibilities.

Cornejo et al. [18] aimed to examine the perceptions and knowledge levels of agricultural students studying at a technical university in Cotopaxi, Ecuador, regarding climate change. The study results reveal that agricultural students need to be equipped with the knowledge and skills to adapt to changing climate conditions; accordingly, it recommends integrating climate change-focused courses into the curriculum. Haq and Ahmed [6], on the other hand, evaluated university students’ perceptions of climate change in Bangladesh in their study. The study found that students’ experiences of extreme weather events in their own regions and their participation in environmental organizations were statistically significant for the perceived causes of climate change. Leal Filho et al. [22], who assessed university students’ attitudes and perceptions regarding climate change at an international level, determined in their studies that students’ knowledge about climate change risks varied according to gender, age, and academic education. They also emphasized that universities should include climate change topics in their curricula and extracurricular programs. Genovese [49] examined the relationship between students’ environmental knowledge and attitudes in a study conducted at the University of Turin. It revealed that knowledge alone is not sufficient for behavioral change and emphasized the need for alternative approaches to encourage environmental actions among younger generations. Amador-Alarcón et al. [38] examined the digital literacy and environmental protection perceptions of 135 university students studying at two state universities in Mexico. It was emphasized that students are aware of the environmental impacts of using electronic devices for educational purposes and that it is necessary to develop effective usage habits to reduce the negative effects of digital technologies on the environment. Lingos et al. [50] examined the relationship between climate change, urban green spaces, and health perceptions among university students from the perspectives of environmental psychology and sustainability. They found that students showed a high level of awareness of the effects of climate change on health, while urban green spaces had positive effects on emotional resilience, physical activity, and environmental quality.

Climate change risk perception is based on individuals’ personal experiences, values, and observations. These perceptions are influenced by various factors such as past experiences, environmental awareness, and proximity to risk, leading to different understandings of climate among communities [3,51,52]. In parallel with this, in our study, 98.6% of participants (n = 276) believe that climate change has contributed to the recent extreme rainfall in the Western Black Sea region (flood disasters in Bartın, Çaycuma, and Devrek). In line with Tranter and Skrbis [53], Prati et al. [54] and Cao et al. [55], our study also found that young women have greater awareness than young men regarding the methods to be determined for raising social awareness about climate change. Yıldızbaş et al. [56] noted that while urban green spaces support mental and physical health, their role in increasing environmental awareness follows a separate path. Additionally, they have emphasized the importance of incorporating large-scale green infrastructure into urban health and sustainability strategies, particularly in rapidly growing metropolitan areas. In parallel, expanding the green infrastructure system has been prioritized in our study as part of adaptation efforts to climate change.

At the same time, there are some noteworthy studies in the literature that involve the use of comprehensive machine learning algorithms for climate change. Prati et al. [54] used machine learning algorithms (Breiman’s random forest algorithm, Friedman’s gradient boosting machines algorithm) in their study to examine the predictive effect of various socio-demographic factors, attitudes such as identity, tolerance, trust, etc. at the local and European levels, beliefs about climate change, and social well-being on the perceived importance of climate change and personal concern among young Italian adults. In this study, the random forest algorithm, which yielded the best performance results, identified age, tolerance towards immigrants and refugees, the perceived importance of climate change, and personal concern as the most important determinants. Additionally, local and European identity, political interest, domestic political engagement, nationalism, social welfare, self-efficacy, authoritarianism, anti-democratic attitudes, EU warmth, and online and civic participation were noted as other important determinants. In the studies by Im et al. [57], they developed building energy prediction models using machine learning tasks and statistical analyses such as multivariate regression models, multiple linear regression (MLR) models, and relative importance analysis to examine the effect of climate change on the energy consumption of a university campus. In modeling studies, the functional relationships between electricity (ELC) and steam (STM) consumption and building characteristics, temporal and meteorological variables were examined. The focus was on determining the fundamental and applicable building characteristics for campus building policy and action plans using prediction models. In their study, Toth and Sebova [58] examined the awareness-action link regarding climate change using machine learning and statistical analysis techniques, discussing factors that could influence the city’s adaptation and mitigation activities in the fight against climate change through the residents (citizens) of the Slovak city of Košice. First, they created linear models to explain climate change awareness with the aim of tracking the impact of socio-economic factors. In the second step, they used the well-known Breiman random forest algorithm of machine learning as an exploratory technique to validate the regression results and reveal the non-linear relationships between the concepts under investigation. As a result, they emphasized that the nature of the awareness-action link may not be linear.

In our study, unlike the literature, we used an approach based on a machine learning-unsupervised clustering algorithm (k-means algorithm) to reveal the levels at which university students carry out knowledge-application activities related to climate change awareness. Within this scope, assessments were made regarding climate change awareness for those with a high level of knowledge-application within the emerging cluster structures, and strategies to increase this scope/focus were discussed

This article examines the climate change awareness and knowledge levels of university preparatory class students using an unsupervised clustering approach to determine the need for differentiated pedagogical and institutional strategies to promote environmentally responsible individual practices in higher education contexts. Information is obtained in three sections.

Information on climate change awareness Section 1: The consequences of climate change, Section 2: Effective methods for preventing climate change; implementation of activities related to climate change awareness Section 3: Students’ levels of environmental awareness were assessed using 5-point Likert-type questions in the sections. After normalizing the responses to these questions using the min-max method, a machine learning clustering study was conducted using the normalized scores (out of 100). In this context, the normalized score data was divided into three clusters using k-means and the Silhouette index ($S\approx 0.55$). Knowledge-application activity levels regarding climate change awareness were defined for these three clusters. The levels of information-application activities performed by students in Cluster 3 (n = 134) were suggested to be high. While the knowledge levels of students in other clusters were low/moderate, their application levels were described as moderate/low. It is important to develop a joint awareness strategy with a mentoring system that enables the group of students with a high level of knowledge and application regarding climate change awareness to transfer/share their knowledge and experience to groups of students with lower/moderate levels. Therefore, the assessments of students in Cluster 3 (potential mentors) focusing on the presence of green areas within their own clusters are provided below on a section-by-section basis.

• Section 1—The consequences of climate change: In this section, topics where cluster-3 students demonstrated a high level of knowledge regarding climate change awareness included Drought and desertification (mean score: 95.3, std. deviation: 9.8), Occurrence of seasonal changes (mean score: 95, std. deviation: 13.3), Melting of glaciers (mean score: 94, std. deviation: 12.3), Disruption of ecological balance (air, water, soil pollution, decline in biodiversity) (mean score: 93.1, std. deviation: 12.8), Depletion of groundwater resources (mean score: 90.7, std. deviation: 14.9), Floods and flash floods (precipitation changes) (mean score: 89.4, std. deviation: 15.7) stand out. These prominent issues are well known at the public level in the context of climate change awareness. Among the consequences of climate change identified by Cluster-3 students, the topics that scored below their overall average were: Forest fires (avg. Score: 83.2, std. deviation: 21.6), Increased air pollution due to the reduction in green areas and the resulting increase in public health problems (avg. Score: 85.6, std. deviation: 18), The reduction and migration resulting from pressures on vegetation cover and wildlife areas (mean score: 87.5, standard deviation: 19.3), The need to develop mentoring approaches regarding the presence of green areas through different strategies (online training, workshops, webinars, etc.) within the scope of climate change awareness.
• Section 2—Effective methods for preventing climate change: In this section, Monitoring and raising awareness of factories’ environmental responsibilities (average score: 98, std. deviation: 8.9), Rainwater management for the protection and support of water resources (average score: 97.5, std. deviation: 9.7), Protection and enhancement of forests (average score: 97, std. deviation: 8.1), Promoting the use of renewable energy sources (wind, solar, etc.) (average score: 96.8, std. deviation: 8.4), Encouraging environmentally friendly investments (average score: 96.6, std. deviation: 9.1), Use of energy-efficient products (average score: 95.7, std. deviation: 9.5), Use of recyclable products (average score: 95.7, std. deviation: 9.5), Implementation of measures to prevent/reduce greenhouse gas emissions (average score: 95.5, std. deviation: 12.1), Reducing the use of substances that damage the ozone layer (average score: 95.5, std. deviation: 14), Promoting urban green infrastructure applications (average score: 95.3, std. deviation: 17) topics stand out where students in cluster 3 demonstrate a high level of knowledge regarding climate change awareness. These prominent issues are well known at the public and/or academic level in the context of climate change awareness. Among the methods effective in preventing climate change identified by Cluster-3 students, it is important to develop mentoring approaches focused on individual and administrative measures to prevent environmental pollution (mean score: 94.8, std. deviation: 12.8), which scored below their overall assessment, through strategies such as campus-based environmental projects and community work.
• Section 3—Students’ level of environmental awareness: In this section, among the topics where cluster-3 students performed well in terms of implementing activities related to climate change awareness are minimizing water consumption to the necessary levels for ecosystem water management (mean score: 95.7, std. deviation: 10.9), exercising controlled and sustainable daily energy use to reduce pressure on green spaces (mean score: 94.4, std. deviation: 12.5), refraining from lighting fires and littering (e.g., cigarette butts, glass) in forest areas (mean score: 93.1, std. deviation: 15.1), and consuming nutritionally rich and adequate amounts of food while avoiding food waste (mean score: 87.7, std. deviation: 16.1) are the main topics. The performance of Cluster-3 students (potential mentors) in carrying out the application activities in this section is significantly higher than the students in other clusters (see Table 5). In this sense, it is believed that cluster-3 assessments on topics such as solid waste/recyclable product purchasing/sustainable product preference/water conservation practices will also be effective in the mentoring process for students in other clusters (cluster-1 and cluster-2).

Managerial Implications

International organizations such as the United Nations (UN) and the United Nations Educational, Scientific and Cultural Organization (UNESCO) emphasize the necessity of climate change education, particularly for younger generations. Education is an important tool in supporting climate action by increasing awareness and social participation [59]. The integration of climate change-related topics into university education, extracurricular activities, and research programs is critical for developing awareness among young people, who are key beneficiaries of sustainable development, regarding the multidimensional impacts of this global issue. Universities should encourage students and staff to become active climate actors and ensure their participation in research, solution development, and advocacy processes [19,20,22]. Therefore, universities need to develop more effective strategies to prepare younger generations for climate change [3,60]. In this context, the widespread adoption of sustainability-focused programs that develop climate literacy with a priority on solution generation and prevention stands out [35]. These types of applications aim to increase individual and social responsibility awareness in the fight against climate change by providing students with both theoretical and practical knowledge.

Our study determined that the students surveyed possessed varying levels of knowledge and awareness. It became apparent that effective strategies must be identified to ensure university students with differing levels of climate change awareness achieve an equivalent level of awareness. Equalizing university students’ awareness levels regarding climate change requires a holistic approach in both academic programs and campus life. The following strategies can help balance awareness levels at different levels and contribute to the creation of a shared awareness (

Table 7. Objectives, strategies, and actions aimed at enhancing students’ awareness of climate change in universities.

Objective	Strategy	Action
To raise students’ awareness of environmental issues such as global warming, carbon footprint, renewable energy, and recycling	Providing interactive, accessible, and enjoyable learning environments through digital educational materials and mobile applications	- Preparation of online training modules for students - Development of mobile applications that include carbon footprint calculators and eco-friendly lifestyle guides - Use of simulation and game-based learning tools
To increase students’ awareness of environmental issues and ensure they keep up with current developments	Encouraging active participation among young people through social media campaigns and web-based interactive projects	- Universities conducting environmental awareness campaigns on social media platforms (Instagram, X, YouTube, TikTok, etc.) - Organizing online seminars, webinars, and online panels - Creating virtual workshops and interactive discussion groups - Involving student clubs in the process through digital content production
Ensuring that students become aware of climate change from both academic and practical perspectives and actively participate	Increasing knowledge sharing and interaction through face-to-face and online events organized in collaboration with local governments, universities, and civil society organizations	- Organizing thematic days to raise awareness about climate change - Organizing national/international conferences and symposiums at the academic level - Conducting workshops, case studies, and applied group work for students - Organizing panels and seminars with the participation of academics, local administrators, and volunteer environmentalists
To develop students’ ability to generate solutions to combat climate change and raise awareness of local environmental issues	Encouraging the development of environmentally friendly applications through project-based learning and student-centered initiatives	- Implementing student projects on climate change with university support - Providing mentorship and technical support for entrepreneurship-based environmental projects - Encouraging student participation through project competitions, grant programs, and workshops - Testing solution-oriented project ideas for local environmental issues through pilot applications
Raising students’ awareness about the importance of forests and the impact of energy conservation on climate change, and encouraging them to develop sustainable individual practices.	Explaining to students the carbon sequestration capacity of forests and the environmental impacts of energy efficiency using interactive methods; emphasizing the importance of urban green spaces and recycling.	- Develop interactive educational modules and simulations on forestry and energy - Organize campaigns promoting the use of recyclable products and energy conservation - Conduct visits to urban forests and green spaces with students - Organize awareness-raising activities on choosing environmentally friendly products

):

6. Conclusions

Climate change education is an important tool for raising awareness, especially among young people, and supporting social participation. Higher education institutions play an active role in raising students’ awareness of global climate change and preparing them to deal with this global issue. Therefore, it is of great importance for universities to develop more effective strategies to raise awareness among young people about reducing the effects of climate change and adapting to it.

Scientific studies emphasize the importance of adaptation strategies that can be implemented in the short term. In our study, a classification (clustering) was performed regarding university students’ awareness of climate change. The number of clusters was determined using the Silhouette index. The three clusters identified using k-means and Silhouette indices determined the levels of knowledge and application, as well as the differences, among student groups regarding climate change awareness. Accordingly, the three clusters identified regarding climate change awareness reveal the positive impact of increased knowledge on individual practices. The clustering results revealed that Cluster-3 students, defined as having a good level of knowledge-application, had higher impact values in their overall assessments of issues closely related to the existence of green spaces in terms of climate change awareness compared to the overall assessments of university students in other clusters.

In this context, a holistic approach has been proposed in both academic programs and campus life to ensure that students with different levels of knowledge and awareness reach a common level of awareness. Thanks to this approach, students will be able to find common ground on climate change, even if their knowledge levels differ. Steps taken from education to implementation, from digital tools to on-campus projects, will contribute to raising more conscious individuals in the future. These holistic strategies will enable universities not only to equip young people with knowledge but also to motivate them to actively participate in climate action. Methodologically, the study provides an innovative contribution to perception-based sustainability research through the use of unsupervised machine learning clustering algorithm. The findings offer valuable insights for policymakers and educators to enhance climate education, sustainability literacy, and the active engagement of young people. The research questions were designed to reflect and capture these contributions.

Unlike previous studies conducted to reveal university students’ awareness of climate change, this article includes assessments for preparatory class students. This also constitutes a limitation of the study. However, this identified limitation facilitates the determination of new students’ knowledge, awareness, and practice levels regarding climate change and its impacts. In this context, it will provide opportunities/advantages for creating academic content that will raise awareness throughout university education and for planning campus applications. Future studies can be conducted with student groups in the beginning grades of vocational training, yielding more comprehensive results through comparative assessments of climate change awareness.

Universities play an important role in climate change process with their students and graduates. In this context, in addition to university students, broader evaluations will also be possible in the future through research conducted on university graduates. Furthermore, studies focusing on interdisciplinary perspectives that reveal students’ perceptions and awareness according to their field of study are also among the topics that can be evaluated in this framework in the future.