Exploring university students’ future orientations towards climate change: perceptions of the future, hope and action competence Open Access

Climate change presents a critical challenge for future generations, requiring urgent and significant efforts. University students, as future decision-makers and influencers, hold a special position in shaping climate solutions through their future career paths and as informed citizens. Higher education plays a vital role in preparing students for an uncertain future shaped by the human and ecological impacts of climate change (Fahey, 2019; Leal Filho et al., 2023; Pfautsch and Gray, 2017). Beyond cultivating scientific literacy, climate change education must foster emotional resilience, a sense of agency and the capacity to envision alternative, sustainable futures. These affective and motivational dimensions are increasingly recognised as essential to turning knowledge into meaningful action (e.g. Ben Zvi Assaraf et al., 2025; Bury et al., 2020; Laherto and Rasa, 2022; Levrini et al., 2019).

One major challenge in climate change education is that it is characterised by “future” instead of “now”. Many of the serious consequences of climate change and other long-term environmental changes will only become fully visible in coming decades, making this issue feel psychologically distant or abstract (Maiella et al., 2020; Spence et al., 2012; Van Lange and Huckelba, 2021). This distance can reduce urgency, lower self-efficacy or foster beliefs that individual actions are insufficient, particularly among young people, despite their central role in long-term mitigation and adaptation (Li and Liu, 2022; Wachholz et al., 2014). At the same time, emerging research suggests that this perceived distance is narrowing, as climate change is increasingly perceived as a present and personally relevant issue. This shift is likely driven by more urgent communication from the scientific community, heightened media coverage, the growing visibility of climate-related hazards, widespread youth-led climate protests and high-profile political events such as the Conference of the Parties (van Valkengoed et al., 2023).

To better understand how university students engage with this future-oriented challenge, this study examines their future orientations in the context of climate change. Specifically, we investigate how their perceptions of the future, feelings of hope and self-perceived action competence relate to each other and vary across demographic backgrounds. Together, these dimensions provide a useful lens to understand how students’ subjective orientations may reflect their preparedness, motivation and capacity to act on climate issues.

The concept of future orientations relates to individuals’ perceptions of possible futures, as described by Evans (2002), and is closely linked to agency, self-efficacy and goal-setting. Future orientations are shaped by personal experiences and serve as expressions of agency, which involves self-confidence, willpower and perseverance in goal attainment (Hadjar and Niedermoser, 2019). For future orientations, some scholars highlight the role of forethought and intentionality in future planning (Bandura and Carr, 2009), others emphasise the consideration and envisioning of future possibilities (Gidley, 2004; Hicks and Holden, 1995; Pauw, 2015). A person’s general future orientations can be regarded as an important personal strength that fosters motivation and long-term success (Hicks and Holden, 1995; Nuttin, 2014; Valle et al., 2006).

Specific to science and environmental education, previous studies suggest that future-oriented individuals are more likely to care about the environment and take action to address environmental issues compared to present-oriented individuals (Milfont et al., 2012). They are more engaged in behaviours such as water conservation (Corral-Verdugo et al., 2006) and hold stronger pro-environmental attitudes (Milfont and Gouveia, 2006). A recent large-scale study in The Netherlands and UK also found that a person with conservative ideology tends to consider future consequence s of their behaviour and perceive the effects of climate change as further away in the future (Većkalov et al., 2021).

Optimism or pessimism about the environmental future is another key factor influencing environmental attitudes. Connecting climate change with an uncertain or even pessimistic future is likely to lead to a feeling of anxiety and hopelessness, and in turn prevent people from taking action. Recent studies have revealed increasing climate anxiety among young people. An international study found that many respondents reported a high number of negative thoughts about climate change (e.g. 75% said that they think the future is frightening) (Hickman et al., 2021). Similarly, a study with Taiwanese university students (Li and Liu, 2022) found that while recognising the urgency of climate change issues, many students were pessimistic about making positive changes for the climate future. This pessimism often leads to disengagement from climate action.

In this study, the concept of future orientations in the context of climate change refers to a person’s perceptions of the future, feelings of hope (or despair) and self-perceived action competence. These three elements have been highlighted across different strands of the literature as interrelated dimensions that would shape how individuals engage with climate futures (e.g. Liu, 2022; Finnegan, 2023; Ojala, 2015; Pahl et al., 2014; Wang and Chen, 2022). In this framing, perceptions of the future refer to students’ thoughts and feelings about what lies ahead, hope can be understood as a motivational resource that sustains engagement, and action competence reflects the enactment of agency through knowledge, confidence and willingness to act. Together, these three elements provide complementary perspectives for examining students’ future orientations towards climate change. While not exhaustive, they offer a coherent framework for understanding students’ potential responses to climate change challenges.

A limited number of studies have investigated how young adults (including university students) perceived their future in the context of climate change (Feldman et al., 2010; Jones and Davison, 2021; Pfautsch and Gray, 2017; Wachholz et al., 2014). Alarmingly, these studies revealed that many university students were generally disconnected and disengaged from the topic. For example, Pfautsch and Gray’s (2017) analysis of benchmark data from Australia showed that the most prominent emotions among university students were fear, sadness and anger, with a lack of self-awareness in identifying and undertaking effective mitigation actions. Despite recognising the importance of climate mitigation, more than half of the participating students (n = 123) did not believe they could personally contribute to the process (Pfautsch and Gray, 2017). More recently, Jones and Davison’s (2021) qualitative study with Australian young adults provided fine-grained evidence that participants felt anxious about extreme outcomes and the country’s unpreparedness for future instability. They also expressed grief over anticipated losses, such as ecological death and environmental degradation.

Similar findings have been reported in studies with Taiwanese university students (Liu and Lin, 2018, Liu, 2022). While students demonstrated relatively high awareness of climate change, many expressed doubt that climate actions could lead to meaningful outcomes, resulting in disengagement from climate-related behaviour. Notably, even after participating in a semester-long climate-related course, there was little improvement in their action-related perceptions. Another interesting finding was a significant, positive correlation between students’ belief in a better future and their trust in techno-science; students who were more optimistic about the realisation of desirable future visions tend to believe that techno-science will have positive influence on the future. Overall, perceptions of the future seem to be a key factor influencing students’ commitment to addressing climate change.

To investigate students’ perceptions of the climate future, we included a word association (WA) task in this study. This method captures immediate, spontaneous associations, offering a situational snapshot of the ideas and feelings most accessible at the time of response. Although not intended as a comprehensive measure of students’ perceptions, WA provides an exploratory complement to the more structured scales of hope and action competence.

Hope is a cognitive process involving thinking about one’s goals, feeling motivated to accomplish those goals and understanding the path towards goal attainment (Snyder, 2000; Snyder et al., 2002). In educational context, the construct of hope is very relevant as goal attainment is central to learning. Past studies have provided empirical evidence for the relationship between hope and several academic, cognitive, affective and behavioural indicators of student engagement (e.g. Dixson, 2023; Dixson and Stevens, 2018; Farnsworth et al., 2022; Wurster et al., 2021). For example, students who experience high hope were more likely to demonstrate school engagement (Marques et al., 2017), feel self-efficacious (Wurster et al., 2021) and achieve positive academic outcomes (Rubens et al., 2020).

According to Snyder (2000), hope encompasses both cognitive and emotional components. The cognitive component of hope involves the ability to formulate pathways towards desired goals and the sense of agency to pursue them. The emotional component serves as a motivational force, sustaining one’s engagement with these goals, particularly in contexts of uncertain or challenging outcomes.

Ojala (2012a, 2012b, 2015) explored the notion of constructive hope in the context of climate change education. Her research highlights that hope plays a regulatory role in managing students’ climate-related worry, thereby supporting awareness, learning and the development of action competence (Ojala, 2012b). Hope enables students to shift from passive concern to active engagement, fostering a future-oriented mindset conducive to meaningful climate action. Ojala (2012a, 2015) also draws a clear distinction between constructive hope, which is based on an understanding of future possibilities, and hope based on denial, which represents a lack of concern about the future due to climate scepticism.

The present study builds on Li and Monroe’s (2018) framework, which emphasises both individual and collective efficacy in addressing climate challenges. They conceptualise climate change hope through three dimensions:

1. personal efficacy (belief in one’s ability to contribute to solutions);

2. collective efficacy (belief that society can act effectively); and

3. perceived lack of capability (barriers to action).

This model aligns with the idea that hope is not only emotional but also cognitive and socially situated. In this study, we adopted their framework and corresponding measure to examine students’ levels of hope as an integral part of their future orientations towards climate change.

Another future-oriented concept is action competence, defined as a critical, reflective and participatory approach in education that enables individuals to address environmental challenges (Breiting and Mogensen, 1999; Mogensen and Schnack, 2010). It is also seen as an educational outcome. According to Breiting and Mogensen (1999), greater action competence entails:

• possessing and developing comprehensive, flexible knowledge around action possibilities;

• having a sense of self-efficacy and feeling empowered to take action; and

• being passionate and willing to engage in actions that contribute to positive changes.

Developing action competence is particularly important in climate education, where feelings of helplessness are common and the need for immediate action is pressing (IPCC, 2022, 2023; Ogunbode et al., 2022; Oinonen et al., 2024; Pickering and Dale, 2024). When students feel equipped with knowledge and tools to contribute to solutions, they are more likely to participate in pro-environmental behaviour (Finnegan, 2023). In this study, we adopted the framework from Olsson et al. (2020), who operationalised the concept of self-perceived action competence as comprising three key components: knowledge of action possibilities, confidence in one’s influence and willingness to act. This framework underpins our measurement of students’ self-perceptions of their capacity to engage in climate action.

Although both hope and self-perceived action competence share efficacy-related elements, prior research maintains a distinction between them. Hope has been conceptualised as a motivational force oriented towards future possibilities, integrating personal and collective will/waypower (Li and Monroe, 2018; Snyder, 2000). Action competence, in contrast, emphasises the perceived capacity to translate motivation into action (Olsson et al., 2020). Scholars have recognised the close relation between the two constructs and called for further research into their interconnections (Li and Monroe, 2018; Ojala, 2015). In Finnegan’s study with secondary students (2023), these two constructs showed significant correlations, suggesting that their relationship is not only conceptually discussed but also empirically meaningful.

This study aims to explore university students’ future orientations by examining three key indicators: perceptions of the future, hope and self-perceived action competence. Together, these dimensions provide insight into students’ expectations, emotional engagement and preparedness for climate-related challenges. The participants were not a general student population but had received substantial instruction on climate change in a standalone course. This study further investigates how these future orientations differ across demographic variables such as gender, major discipline and year of study, as well as how they are interrelated. These insights are expected to inform the development of future-oriented and emotionally responsive climate education programs that can sustain students’ engagement and support their perceived capacity to act.

The following research questions guided this study:

The data set (n = 216) comprises participants from three waves of data collection. These waves represent three independent cohorts of students enrolled in the same elective course, with data gathered at the same point in the syllabus across three consecutive semesters. The course is a large, semester-long elective climate change course offered at a public university in southern Taiwan. Open to undergraduates of all majors, it attracts a diverse mix of students across disciplines and years of study. While students may select the course for a variety of reasons, their enrolment suggests at least some level of interest in climate change. Before participating in the study, they had learned a wide range of lecture-based topics on climate change (about 10 weeks), including not only the scientific basis but also societal and economic impacts as well as possible adaptation strategies. This background indicates a certain level of awareness of climate change, which is important for interpreting their responses. The WA task and the two validated scales were administered together in a single survey during class time. Participation was voluntary, and informed consent was obtained from all students prior to data collection.

The group included both STEM (50.2%) and non-STEM majors (49.8%) in nearly equal proportions, with a higher proportion of males (63.0%) than females (37.0%). Students were drawn from sophomore (28.4%), junior (26.5%) and senior years (45.1%), with seniors forming the largest share. Freshmen were not included, as they are typically occupied with foundational courses in their majors.

To ensure the reliability of combining these data sets, we first examined measurement invariance, a prerequisite for the subsequent quantitative analysis. We employed the MICOM procedure (Henseler et al., 2016) to tests configural invariance, compositional invariance and equality of means and variances. Configural invariance was ensured because the data structure (indicators and constructs) was identical across all three waves. Compositional invariance was assessed through a permutation check in which wave labels are randomly reassigned to examine whether construct scores differ from their original assignment. All latent constructs passed this check. We then examined the equality of means and variances across the three waves, which held for almost all higher- and lower-order constructs. Measurement invariance was therefore established for nearly all constructs, except for the mean of the knowledge construct, which differed significantly between two waves. Despite this minor variation, configural and compositional invariance were confirmed for all constructs, justifying the pooling of data for further analysis.

To gather insights into students’ future orientations towards climate change, this study utilised three major tools: a WA test to explore their perceptions of the future, and two Likert scales to assess the constructs of hope and self-perceived action competence.

WA techniques, commonly applied in psychology, are qualitative methods that can serve as quick and convenient tools in exploring people’s perceptions for new or abstract concepts (e.g. Benthin et al., 1995; de Groot, 1989; Smith and Leiserowitz, 2013). These open-ended tasks ask respondents to provide immediate word-based responses, making them less laborious than many other qualitative methods, such as interviews. At the same time, WA techniques can capture affective and less conscious aspects of respondents’ mindsets that are often difficult to access through more direct forms of questioning (Szalay and Deese, 2024). In this study, a WA test was designed to elicit students’ spontaneous associations regarding the future of climate change. Participants were asked to respond to the prompt, “What comes to your mind when thinking about the future?” by providing three words. This approach offered a snapshot of the ideas and feelings that were most readily accessible to students at the moment of response, thereby revealing what they considered most relevant to this topic. While some responses may reflect emotional attitudes, others may indicate significant aspects or concerns about the future.

This scale, adapted from Li and Monroe (2018), measures students’ levels of hope through three dimensions: personal ability (personal-sphere willpower and waypower, PW), societal capability (collective-sphere willpower and waypower, CW) and a perceived lack of capability (LW; negative items, where higher scores indicate greater lack of hope). Participants rated 11 items on a seven-point scale (1 = strongly disagree to 7 = strongly agree), with example items such as “I know that there are a number of things I can do to contribute to climate change mitigation”. (PW), “Even when some people give up, I know there will still be others who continue to work on solving problems caused by climate change” (CW), and “Climate change is such a huge problem that I don’t think people can change it” (LW). Higher scores reflect greater levels of hope.

This 12-item scale, adapted from prior work on self-perceived action competence for sustainability (Finnegan, 2023; Olsson et al., 2020), measures this construct specifically in the context of climate change. The scale assessed three areas: knowledge of action possibilities (KAP), confidence in personal influence (COI) and willingness to act (WTA) on climate change. Example items include: “I am aware of actions I can take at home to mitigate climate change” (KAP), “I am confident that individual actions can contribute to climate change solutions” (COI), and “I am willing to encourage my family and friends to take climate action” (WTA). Responses were rated on a five-point scale (1 = strongly disagree to 5 = strongly agree), with higher scores indicating stronger self-perceived competence in addressing climate change.

Although using the same response scale across both instruments might have enhanced consistency, we retained the original formats – 7-point for Hope and 5-point for Action Competence – to preserve the validity of the instruments and comparability with prior research.

Students’ responses from the WA test were first cleaned by correcting typos, simplifying phrases and combining variations of the same words. The responses were then categorised into three emotive perceptions: “positive”, “negative” and “unsure/neutral”. This categorisation was based on the general emotional valence of the words (e.g. hopeful or constructive terms as positive; fearful or discouraging terms as negative; descriptive or ambiguous terms as neutral). Two coders worked together throughout the process and reached consensus on the categorisation of all words, which helped to reduce subjectivity. A network analysis was conducted to explore the patterns and relationships among these words. The results were then visualised in Python-generated network diagrams, a tool that represents words as nodes and their co-occurrence as connecting edges. In these diagrams, node size reflects the frequency of each word, node colour indicates its category, and edges show which words tended to appear together. To enable further statistical analysis with the scale measures, six binary dummy variables were created to capture the presence or absence of positive, negative and unsure/neutral associations.

Hope and Action Competence were modelled as higher order constructs, each comprising three lower order constructs: PW, CW and LW for Hope, and KAP, COI and WAT for Action Competence. Each lower order construct was measured using three to four Likert-scale items (see Table 1).

Measurement evaluation showed that the reverse-worded construct LW loaded poorly on Hope (−0.313) and produced low internal consistency (Cronbach’s α = 0.268), so it was removed from subsequent analysis. Confirmatory factor analysis (CFA) of the revised structure met standard expectations for both higher and lower order constructs (see Table 2). All factor loadings are above 0.6. Cronbach’s alpha ranged from 0.79 to 0.87, and Composite Reliability from 0.88 to 0.92, indicating good internal consistency. All average variance extracted (AVE) values were above 0.5, supporting convergent validity, and the Fornell–Larcker Criterion is satisfied, indicating sufficient discriminant validity.

Following measurement evaluation, descriptive statistics were computed for all constructs. T-tests, one-way ANOVAs and correlations, were then applied to examine group differences (gender, study year, STEM/non-STEM) and the relationships among hope, action competence and WA variables.

An analysis of the WA data identified 179 unique words from 615 total instances. Most responses were neutral or uncertain (42%), followed by negative (32%) and positive (26%) categories. Based on word frequencies and co-occurrence matrix, a network diagram was created (see Figure 1). It illustrates the relationships between key words in student responses. Each node represents a word, with size indicating frequency. Only words with a frequency above 2 were included to avoid overcrowding. Node colours indicate categories (red = negative; green = positive; grey = neutral/uncertain). Edges show how often word pairs appear together, with line thickness reflecting the strength of co-occurrence.

The most frequent words in the diagram are neutral or uncertain words such as “unknown” (n = 51), “technology” (n = 42) and “uncertainty” (n = 26). These suggest that students often reference concepts reflecting ambiguity or a lack of clarity about the future. This dominance of neutral or uncertain terms indicates caution and hesitation in the students’ outlooks, particularly regarding technological advancements and their potential impacts. Moreover, these neutral words are strongly interconnected. For example, “unknown” frequently co-occurs with both “changeable” and “change” and “technology” with “rapid”, indicating that students often discuss these concepts together, perhaps reflecting a shared sense of unpredictability and ambiguity about future scenarios.

Despite the prominence of neutral or uncertain words, negative words outnumber positive ones. Words such as “extreme-climate” (n = 17), “hot” (n = 16) and “lost” (n = 12) highlight concerns about environmental challenges and potential negative outcomes. The strong connections between “extreme-climate” and “hot” and “unpredictable”, and between “fearful” and “unknown”, further suggest that students commonly associate uncertainty with fear. On the other hand, positive words such as “convenient” (n = 18), “sustainability” (n = 17) and “hopeful” (n = 15) appear less frequently but still play an important role in the network. “Hopeful” shares a strong link with “technology”, and some students mentioned “convenient” and “warming” together.

Six binary dummy variables were created to capture the presence or absence of positive, negative and neutral/unsure words. Each variable was generated twice – once for the presence of at least one word of that type (“with”), and once for the absence (“without”). These variables were used for correlation and group comparisons. The analysis showed that STEM students were less likely to use neutral words than non-STEM students, while gender and study year showed no differences. While 36% of STEM majors did not associate any neutral words, only approximately 18% of non-STEM majors did not. See  Appendix (Table A1) for the full data set.

The overall results for hope and action competence are shown in Table 2. Across the full sample, students reported moderately high levels on both constructs. Within hope, personal way/willpower (PW) had the highest mean (M = 5.38), suggesting confidence in individual ability to contribute to climate action. Collective way/willpower (CW) followed (M = 4.85), indicating moderate belief in societal efforts. Lack of way/willpower (LW), a negatively framed subscale, had the lowest score (M = 3.46), reflecting a lower sense of helplessness. However, due to low reliability and weak loading, LW was excluded from the final model. For action competence, knowledge of action possibilities (KAP, M = 4.13) was the strongest dimension, while confidence in influence (COI, M = 3.73) was the weakest, suggesting that students know what can be done but doubt their own impact.

In terms of gender differences, female students reported significantly higher scores on PW (p = 0.02), suggesting a stronger belief in their individual ability to contribute to climate solutions. They also scored significantly lower on LW (p < 0.001), indicating a lower sense of helplessness or perceived limitation than male students. Together, these results reflect a more hopeful outlook among female students. They also scored significantly higher in all three components of action competence – KAP (p = 0.03), COI (p = 0.01) and WTA (p = 0.03) – as well as in overall action competence (p = 0.01). These findings suggest that female students feel more hopeful and perceive themselves as more capable and ready to act on climate issues.

No significant differences were found between STEM and non-STEM students on any hope or action competence subscales. By year of study, however, mean scores generally declined from sophomores to juniors and seniors. Three comparisons reached statistical significance: Juniors reported lower WTA than sophomores (p = 0.03), and seniors showed lower overall action competence (p = 0.04) and higher LW (p = 0.04). Several variables displayed near-significant declines. Overall hope dropped 5.27 in sophomores to 4.99 in seniors (p = 0.06). KAP from 4.25 to 4.06 (p = 0.05), and WTA from 4.13 to 3.93 (p = 0.07). Overall, the pattern suggests a gradual erosion of both hope and perceived efficacy as students’ progress, with the most pronounced differences between sophomores and seniors.

To understand the relationship between hope and action competence, correlation test was conducted. Hope was positively correlated with all dimensions of action competence, with all correlations significant at p < 0.001 (see  AppendixTable A2 for full coefficients). This indicates that students who believe in personal and social way/willpower tend to exhibit higher efficacy and willingness to address climate change.

Table 3 shows the relationship between students’ WA results and their responses to the two Likert scales. Students who associated at least one positive word in the WA test reported significantly higher hope (p = 0.02), particularly CW (p < 0.001), and higher action competence (p = 0.03), including greater COI (p = 0.02), than those who did not. Conversely, students who associated negative words reported significantly lower hope (p < 0.001), CW (p < 0.001), and action competence (p = 0.02), with lower COI (p = 0.01). No significant differences were found between students who included neutral words and those who did not. These findings suggest that the emotional valence of students’ WAs is closely linked to their sense of hope and perceived capacity for climate action.

This study explored university students’ future orientations towards climate change, focusing on their perceptions of the future, levels of hope and self-perceived action competence. As climate change becomes an increasingly complex and urgent issue, understanding the emotional and cognitive dimensions of students’ responses is essential for designing meaningful educational interventions (Miller and Anderson, 2017; Sharma, 2012; UNESCO, 2012). Because these students had prior instruction on climate change, their responses offer insight into how informed learners emotionally and cognitively frame the climate future.

Although neutral or uncertain terms appeared most often, the greater use of negative words points to a cautious, even pessimistic, emotional tone in how students imagine the climate future. This pattern aligns with research showing that youth often experience anxiety when the future feels unpredictable or threatening (Ojala, 2012b; Teo et al., 2024; Wang and Chen, 2022). The co-occurrence patterns – such as extreme-climate linked to unpredictable, and fearful to unknown – highlight how uncertainty and fear become intertwined in students’ meaning-making. These results suggest that growing awareness of climate change may heighten students’ sense of vulnerability, contributing to a more pessimistic outlook.

Positive terms like sustainability and hopeful appeared less frequently but often co-occurred with technology, suggesting an externally oriented hope centred on technological solutions. This aligns with Finnegan’s (2023) observation that hope and anxiety can coexist in young people’s climate narratives, although the sources of hope may differ. Here, students’ hope appears relatively passive, rooted in expectations of technological innovation rather than personal or collective agency. According to Ojala (2015), passive hope may offer emotional comfort but is less likely to motivate meaningful action. These mixed findings highlight the need to engage both affective and cognitive dimensions - acknowledging students’ concerns while promoting agency and realistic optimism.

The emotional tone of students’ WA responses was closely linked to their levels of hope and action competence: more positive associations aligned with stronger collective efficacy and confidence, while negative associations aligned with weaker perceptions of agency. This pattern reinforces arguments by Ojala (2015), Li and Monroe (2019) and others (e.g. Ratinen, 2021; Vandaele and Stålhammar, 2022) that climate education must engage with students’ emotions, acknowledging fear and uncertainty while cultivating constructive, actional forms of hope.

The strong correlations between hope and action competence, consistent with Finnegan (2023), reinforces the view that fostering hope contributes directly to students’ perceived capacity to act. Although students expressed rather strong personal way/willpower (PW), their low scores for confidence in their own influence (COI) reveal doubts about their ability to make meaningful impact. This discrepancy echoes prior research showing that university students often view individual actions as powerless in the face of complex climate challenges (Li and Liu, 2022; Pfautsch and Gray, 2017), and that feelings of powerlessness reduce the likelihood of behaviour change (Gunderson, 2023; Pickering and Dale, 2024). Therefore, climate education must go beyond building knowledge and skills to intentionally cultivate hope and efficacy by helping students see how individual and collective actions can contribute to broader climate solutions.

Gender differences in this study reflect broader patterns in the literature: women consistently report stronger environmental concern and perceived efficacy (Milfont and Sibley, 2016; Zelezny et al., 2000), tendencies linked to higher empathy and lower social dominance orientation (Milfont and Sibley, 2016). Yet this contrasts with evidence that men contribute more to greenhouse gas emissions (Cohen, 2014; Räty and Carlsson-Kanyama, 2010) and are more represented in emission-intensive sectors, and that societies with greater gender equality generally emit less CO₂ per capita (Cohen, 2014; Ergas and York, 2012). These tensions highlight the need for climate education to encourage critical reflection on responsibility and structural inequalities in contributing to climate change (Pearse, 2017).

A notable finding is the decline in students’ hope and action competence across years of study, suggesting that perceived agency may erode as students’ progress through university. This challenges the assumption that increased educational exposure naturally strengthens climate literacy, which is understood to include not only knowledge but also attitudes and capacity for action. Given the prominence of climate change in higher education, including in Taiwan, one might expect later-year students to feel more, not less, empowered. Instead, the pattern points to a tension: greater awareness of climate complexity may heighten cognitive understanding but, without sustained emotional support and meaningful opportunities for engagement, can also contribute to fatigue or anxiety (Crandon et al., 2022; Kerr, 2009; Lu, 2022). These findings reinforce the importance of applying an action competence approach (Jensen and Schnack, 2006; Mogensen and Schnack, 2010). Previous studies indicate that exploring future possibilities and envisioning desirable scenarios helps cultivate hope and empowerment (Li and Liu, 2022, Liu, 2024; Hoffman et al., 2021; Laherto and Rasa, 2022). Building on these insights, practical strategies may include encouraging students to identify concrete action possibilities at individual and collective levels, creating opportunities for collaboration in exploring solutions, and scaffolding discussions that connect risks with realistic pathways for change. Such approaches not only help foster hope but also strengthen students’ perceived competence and willingness to act.

This study has two key limitations. Firstly, participants were drawn from a voluntary elective course on climate change, making the sample selective rather than representative of the broader undergraduate population. Although students came from varied disciplines and years, they likely represent a group with a degree of awareness and interest in climate change. Accordingly, the findings should be interpreted as reflecting how students within this specific instructional context – who had already received substantial climate change instruction - orient themselves towards the climate future. The study does not aim to isolate the effects of prior knowledge or related individual characteristics; instead, these factors are embedded in the educational setting that frames students’ responses.

Secondly, the WA method captures spontaneous, salient responses, rather than a full account of their climate future perceptions. While this situational nature is a limitation, it is also a strength: the WA analysis combined qualitative insights into students’ emotions and ideas with quantitative frequency and co-occurrence patterns, providing both depth and structure. Nevertheless, the method should be seen as complementary to the validated scales rather than a substitute for them.

This study examined how university students’ future orientations regarding climate change are reflected through their perceptions of the future, levels of hope and self-perceived action competence. A key finding is the strong interconnection between these three elements: students’ future associations were closely linked with their reported levels of hope and action competence. Another important result is the decline of both hope and action competence across the university years, with seniors reporting lower levels than sophomores. These patterns raise important concerns for climate change education. By combining a WA task with validated Likert-scale measures, the study offered both spontaneous associations and structured measures, enriching our understanding of how students with some degree of interest and foundational knowledge make sense of the climate future. The gender-based disparities and the declining outlook over time raised important concerns for climate change education. We need to attend not only to sustaining students’ hope and action competence but also to addressing equity in how different groups experience and respond to climate futures. The strong correlation between hope and action competence further indicates that fostering individual and collective hope may be key to supporting students’ perceived capacity to act. Recent research has provided evidence that scaffolded explorations of future possibilities and envisioning desirable scenarios are promising approaches to develop critical hope and a sense of empowerment to make meaningful changes (Liu, 2024; Laherto and Rasa, 2022; Levrini et al., 2019; Paige and Lloyd, 2016). Future research could build on this work through longitudinal designs or by comparing different instructional strategies across contexts to better understand how future orientations develop throughout students’ university experience.