Young People and Nature: What Drives Underlying Behavioural Intentions towards Protected Areas Conservation?

Abstract

This paper investigates university students’ motivations in nature park conservation, an area that has received limited research attention compared to other pro-environmental behaviours. By formulating a set of research hypotheses, an extended version of the Theory of Planned Behaviour is employed to examine the determinants of university students’ intentions to engage in nature park conservation. Structural equation modelling is conducted using survey data collected from university students in the Metropolitan City of Naples, Italy. The results show that attitudes, perceived behavioural control, and personal moral norms directly and positively influence university students’ intentions to undertake nature park preservation actions. Subjective norms indirectly influence behavioural intentions through attitudes, perceived behavioural control, and personal moral norms. Moreover, the findings reveal that university students with nature-oriented altruistic values exhibit a higher propensity for nature park conservation compared to their more egoistic peers. These findings have important implications for park management institutions, suggesting the need to increase awareness among young people about their potential role in preserving the environmental quality of nature parks. Collaboration between universities and nature park institutions is also crucial in providing young individuals with the necessary skills to participate in decision-making processes aligned with the principles of sustainable development.

1. Introduction

Many current environmental issues, which seriously threaten sustainability, are closely tied to human activities, so scholars and policymakers maintain that individuals can help address environmental challenges by changing their behaviours [1,2]. The awareness of environmental problems and effective solutions falls within the responsibilities of young people [3]. In this context, it is crucial to equip younger generations with the appropriate environmental knowledge and skills to adopt sustainable intervention strategies. Higher education institutions, such as university, play a critical role in training responsible and competent individuals who can engage in environmentally-friendly behaviours and develop solutions to enhance the world from an eco-sustainable perspective [4,5].

The loss of natural resources, including biodiversity and ecosystem services, as well as the decline of natural green spaces, is a result of human behaviours. The necessity of protecting natural resources has led to the creation of systems of protected areas, such as nature parks, which not only provide habitats for endangered wildlife but also generate economic benefits for local communities [6,7,8]. However, the interaction between human activities and protected areas can be contradictory, especially near urban areas, where land use and anthropisation can potentially degrade the environmental quality of nature parks [9,10].

The solution to this human pressure and its inevitable impact on the natural environment requires the implementation of proactive measures to improve the management of nature parks. To achieve this, it is essential to involve local communities directly in decision-making processes, ensuring that human activities are planned in accordance with sustainable development principles. The active participation of young people, including university students, is particularly important in preserving natural park areas through initiatives such as cleaning campaigns, tree planting, and the maintenance of paths and hedges. These activities can significantly contribute to raising environmental awareness, fostering a sense of responsibility, and promoting pro-environmental behaviours [11,12]. University students, as educated members of the local community, hold the potential to become influential stakeholders. Their higher level of education equips them with knowledge, skills, and values that can enhance propensity towards pro-environmental behaviour. By actively engaging in activities aimed at preserving natural park areas, university students gain valuable experience and can make meaningful contributions to addressing local environmental issues. This involvement encourages them to act responsibly and consider effective solutions for future environmental management [2,5,13].

Based on these considerations, investigating what can motivate university students to engage in nature park preserving actions is a scientific area of relevant interest. Understanding these factors has implications for designing activities that bolster participation in conservation initiatives and promote changes in environmental behaviour towards a sustainable future. Previous research has primarily focused on factors influencing people’s behaviour towards natural resource conservation in protected areas, such as nature parks [10,14,15]. However, most of the existing research in this area has examined the factors behind people’s willingness to pay for park conservation and usage [7,16,17,18,19]). Additionally, specific attention has been given to assessing young people’s willingness to pay for environmental protection in protected areas [20,21,22]. Several studies have explored the role of human values in influencing attitudinal and behavioural preferences towards the protection of natural resources [23].

To the best of our knowledge, limited research has explored the motivations behind young people’s involvement in conservation activities in natural park areas [24,25]. To fill this gap in the existing literature, a more comprehensive understanding of the factors that influence the pro-environmental actions of young individuals in natural park areas is required. Therefore, this paper focuses on the motivational determinants of young people’s intention to undertake pro-environmental behaviour. The choice to focus on intention is rooted in the assumption that people tend to act after forming behavioural intentions.

First, the paper aims to investigate the potential influence of psychological factors on university students’ intentions to perform actions for preserving nature parks. To accomplish this, we resort to the Theory of Planned Behaviour (TPB) [26], which allows for the analysis of behavioural intentions, taking into account environmental attitudes, subjective norms, and perceived behavioural control. In particular, we use an extended version of the TPB, including personal moral norms as an additional factor. This inclusion is motivated by the recognition in the literature that the TPB does not adequately consider the moral dimension of environmental behaviour [27]. Within the strand of psychological literature, it has been observed that individuals form opinions regarding appropriate behavioural standards, such as in the protection of an environmental asset [28,29]. These opinions are assumed by individuals as an intrinsic sense of obligation, which can be interpreted as a ‘personal moral norm’ indicating what the person ought to do regarding the specific asset. Individuals adopt personal moral norms as the behavioural standards they consider appropriate in the given context. Consequently, the personal moral norms can serve as a potential antecedent to individual behavioural intentions [30].

Second, this research examines the direct effects of attitudes, subjective norms, perceived behavioural control, and personal moral norms on university students’ intentions to participate in nature park conservation actions, as well as the indirect effects of subjective norms and personal moral norms on these behavioural intentions.

Third, this study investigates whether and how different value orientations may influence university students’ intentions for nature park conservation behaviour. To achieve this, university students are classified according to their predominant value orientations. Specifically, three types of values—altruistic, biospheric, and egoistic—are considered to be closely associated with environmentally-friendly behaviour or behavioural intentions [31].

To accomplish the objectives of this study, a survey was conducted among university students in the Metropolitan City of Naples, located in southern Italy. This urban agglomeration consists of the Municipality of Naples, the largest city, and several municipalities along the east-south Tyrrhenian coast. The Metropolitan City of Naples is characterised by an average population density of approximately 3000 inhabitants per square kilometre and 68–85% of land consumption [32,33]. Within this region, valuable and biodiverse protected areas, such as the Vesuvius National Park and Campi Flegrei Regional Park, face negative externalities due to the high anthropisation of the surrounding area.

The analysis is conducted from the perspective of young people studying at university. Exploring university students’ intentions for nature park conservation behaviour is important because they inherit the consequences of past and present unfriendly environmental management and will play a potentially significant role in future pro-environmental actions. It is worth noting that university students are being educated to assume important social roles, where they will be responsible for making decisions and finding solutions for the more sustainable management of natural park areas [2,5,34].

Theoretical Framework and Research Hypotheses

Over the past half century, numerous psychological theories have emerged to investigate pro-environmental behaviour [35]. Altruism-based theories, such as the Norm-Activation-Theory (NAT) [36] and the Value-Belief-Norm (VBN) theory [28], posit that pro-environmental behaviours are typically guided by altruistic motives and involve personal costs. Conversely, individualism-based theories focus on maximising individual well-being as a driving force behind environmentally-significant actions.

In this study, we adopted the Theory of Planned Behaviour (TPB), which falls into the individualism-based category mentioned earlier. We assumed that university students engage in conservation actions in nature parks after forming a behavioural intention, indicating their willingness to make efforts to perform that specific behaviour. The intention of university students to perform the target behaviour [26,37] is the key element in the applied TPB approach. The intention is considered the immediate antecedent of behaviour, and a stronger intention increases the likelihood of the behaviour occurring [38].

We investigated university students’ intentions to engage in conservation behaviour in nature parks through hypotheses based on an extended version of the TPB approach [37,39,40,41,42]. The TPB framework includes personal moral norms (PMN) as an additional potential determinant of students’ intention for nature park conservation. The study assumed that behavioural intention (BI) of university students to preserve nature parks depends on three psychosocial factors: (i) subjective norm (SN), which represents the perceived social pressure from reference group members, such as family, friends, and colleagues, to either perform or not the conservation behaviour; (ii) attitude (ATT) towards specific behaviour, i.e., the individual’s favourable or unfavourable evaluation of the conservation behaviour; (iii) perceived behavioural control (PBC), which reflects the subjective perception of the level of difficulty or ease in performing the conservation behaviour; (iv) personal moral norm (PMN), which captures the students’ belief about the rightness or wrongness of engaging in conservation behaviour in nature parks [36]. The inclusion of PMN in the model addressed the criticism that the TPB framework underrepresents the moral dimension of environmental behaviour [27,43]. The integrated TPB approach posited a causal chain linking attitudes, perceived behavioural control, personal moral norms, and subjective norms, which may determine individual intentions to perform preserving behaviours of nature parks [44].

Attitudes represent the behavioural beliefs that university students formulate about the outcomes of engaging in actions to preserve the environment of a park area [26,38]. The idea is that the more positive the attitude university students hold towards environmental preservation in nature parks, the stronger their intention to act for nature park conservation [42,45]. Accordingly, we proposed the following hypothesis for testing:

Research Hypothesis 1 (H₁): 

Attitudes directly and positively affect university students’ behavioural intentions for nature park conservation.

Perceived behavioural control is based on accessible control beliefs, which are the factors that can help or hinder behaviour. In the context of this study, perceived behavioural control depends on university students’ perception of their personal power and self-efficacy to undertake actions aimed at preserving the environment of a nature park. Hence, the more university students feel able to engage in such actions, the more likely they are to develop an intention to preserve nature parks. Conversely, university students would prefer to rule out such actions if they perceive them as difficult or inconvenient. Therefore, the hypothesis to be tested was:

Research Hypothesis 2 (H₂): 

Perceived behavioural control directly and positively affects university students’ behavioural intentions for nature park conservation.

Personal moral norms reflect individual responsibility and internalised moral principles that guide university students in evaluating whether or not to adopt behaviours aimed at preserving the environment of a nature park. The idea is that university students who possess strong personal moral norms are more inclined to engage in such behaviours. Personal moral norms can directly and/or indirectly help explain university students’ behavioural intention towards nature park conservation [39]. In other words, personal moral norms can activate a positive attitude towards nature park conservation, subsequently increasing the probability of acting in this way. This suggests that the direct and indirect effects of personal moral norms on behavioural intentions are not mutually exclusive, highlighting the importance of thoroughly investigating their role in shaping behavioural intentions towards nature park conservation. Based on these premises, we formulated the following two hypotheses:

Research Hypothesis 3a (H_3a): 

Personal moral norms directly and positively affect university students’ behavioural intentions for nature park conservation.

Research Hypothesis 3b (H_3b): 

Personal moral norms positively affect university students’ behavioural intentions for nature park conservation via attitudes.

Subjective norms can be understood as normative beliefs that encompass both injunctive and descriptive aspects and contribute to the overall perceived social pressure to engage in a particular behaviour [38]. In this study, injunctive normative beliefs reflect the subjective probability that significant others, as perceived by university students, endorse or expect actions that help preserve nature parks. Descriptive normative beliefs, on the other hand, pertain to whether these referent people themselves perform such actions. As subjective norms make university students more sensitive to referents’ behaviour beliefs and more responsible for their own actions [46], they are usually associated with attitudes [47,48], perceived behavioural control [39], and personal moral norms [48,49]. Based on the provided premises, we expect subjective norms to play a central role in: (i) directly influencing university students’ intentions to preserve nature park; (ii) shaping personal moral norms, attitudes, and perceived behavioural control, which, in turn, positively affect behavioural intentions for nature park conservation. Therefore, the hypotheses to be tested were as follows:

Research Hypothesis 4a (H_4a): 

Subjective norms directly and positively influence university students’ behavioural intentions for nature park conservation.

Research Hypothesis 4b (H_4b): 

Subjective norms positively affect university students’ behavioural intentions for nature park conservation via personal moral norms.

Research Hypothesis 4c (H_4c): 

Subjective norms positively affect university students’ behavioural intentions for nature park conservation via perceived behavioural control.

Research Hypothesis 4d (H_4d): 

Subjective norms positively affect university students’ behavioural intentions for nature park conservation via attitudes.

Research Hypothesis 4e (H_4e): 

Subjective norms positively affect university students’ behavioural intentions for nature park conservation via personal moral norms and attitudes.

The aforementioned research hypotheses were tested on both the entire cohort of respondents and two distinct subgroups, categorised based on their prevailing value orientation. Our investigation aimed to explore the guidance of diverse human values on the formation of behavioural intentions towards nature park conservation, while examining the causal relationships among their potential antecedents (i.e., subjective norms, attitudes, perceived behavioural control, and personal moral norms).

It is noteworthy that human values were not directly integrated as a latent construct within the TPB model; instead, they were used a priori to sketch a profile of respondents and subsequently group them according to their predominant value orientation, enabling separate analyses to be conducted. Notably, a strand of research [50,51] has identified the role of human values in nature-based experiences, drawing upon the principles of altruism and environmental ethics. Human values can provide insights into university students’ perspectives on their interactions with and management of nature. In this field, biospheric, altruistic, and egoistic values have been identified as the most closely related to environmentally-friendly behaviour or behavioural intentions [23,31,52,53].

In this research, altruistic and biospheric values embody university students’ capacity to transcend selfish concerns and prioritise the well-being of others and the natural world, encompassing all living species [54,55]. We expect that university students with an altruistic value orientation form behavioural intentions regarding environmentally-responsible actions in nature parks based on their perception of the costs and benefits to others. Similarly, biospheric values are likely to emphasise the intrinsic worth of all components within the ecosystem, leading university students with such values to develop intentions towards pro-environmental actions in nature parks driven by their own moral considerations towards other living species and nature itself. Conversely, egoistic values are rooted in self-interest; therefore, we expect that university students with egoistic value orientations evaluate the costs and benefits of pro-environmental behaviour exclusively from their own interest, heightening their intention to protect environment aspects that they perceive as personally threatening, as long as the personal costs are deemed acceptable [56].

Figure 1 shows the proposed conceptual model. It would allow for testing a wider set of hypotheses concerning the causal relationships between the four psychological factors assumed to be antecedents to behavioural intentions (e.g., the direct effect of subjective norms on personal moral norms, rather than on attitudes or perceived behavioural control, and the indirect impact of subjective norms on attitudes via personal moral norms). However, our focus was on examining whether the direct impact of subjective norms on the aforementioned factors translates into stronger behavioural intentions. We investigate this through the set of research hypotheses from H_4b to H_4e.

Moreover, Figure 2 shows the interactions between constructs that detect the direct effects of each psychological factor on behavioural intentions (H₁, H₂, H_3a, and H_4a). Similarly, Figure 3 shows the interactions between constructs that reveal the indirect effects of each factor on behavioural intentions via at least one other factor (H_3b, H_4b, H_4c, H_4d, and H_4e).

2. Materials and Methods

As shown in Figure 4, this section is structured into three subsections: method, data, and measures. Section 2.1 is dedicated to the evaluation of measurement and structural models (Section 2.1.1), as well as to the multi-group analysis (Section 2.1.2).

2.1. Methods for Data Analysis

We performed partial least squares structural equation modelling (PLS-SEM) to test the research hypotheses underlying the extended TPB conceptual framework (Figure 1). PLS-SEM, as an extension of the general linear model, allows for the simultaneous assessment of relationships between a set of latent variables (constructs) that represent theoretical concepts which are not directly measurable. Each construct, represented by ellipses, is measured by one or more manifest variables, which serve as indicators of the underlying constructs [57,58]. The latent variables can function as either dependent or independent variables based on their position in the hypothesised causal sequence. Latent variables that serve solely as independent variables are referred to as exogenous, while those that serve solely as dependent variables or both independent and dependent variables are called endogenous.

PLS-SEM consists of a structural model and a measurement model. The structural model assesses the causal relationships between the latent constructs using path analysis, which involves formulating hypotheses for the relationships [59,60]. This model allows for the identification of two types of effects, direct and indirect. Direct effects represent the relationships between two constructs in the model, while indirect effects involve a sequence of relationships that include at least one intervening construct. The indirect effects are often referred to as mediating effects because they occur through a third construct that lies between the two other related constructs [61,62,63].

The measurement model examines how the manifest indicators define latent variables in terms of reliability and validity [64]. There are two main approaches to measuring constructs: reflective and formative measurement. In the reflective approach, there are assumed relationships from the construct to the manifest indicators, whereas in the formative approach, there are assumed relationships from the manifest indicators to the construct [65]. In the reflective approach, the construct represents the underlying reality, and the manifest indicators are a representative set of items that all reflect the construct. In the formative approach, the manifest indicators represent all dimensions of the construct, and they collectively define the underlying reality [66,67]. Therefore, dropping one of the multiple indicators in a reflective model may not alter the meaning of the construct, as the other indicators still capture its essence. On the other hand, removing an indicator from a formative model is equivalent to eliminating a dimension, which would change the meaning of the construct [68,69].

The exploratory nature of this research, which aims to test an extended version of the TPB approach for analysing the determinants of young people’s intentions to act for nature park conservation, justifies the use of PLS-SEM models. PLS-SEM is well-suited for models with many structural path relationships and constructs. In this study, we assume a reflective model, where each unidimensional construct determines the manifest indicators, and if the underlying trait being measured changes, all items change similarly. PLS-SEM is a non-parametric method, and statistical inference is made through simulation techniques [61]. Formally, the structural model specifies a linear relationship between a vector of dependent latent (endogenous) variables η = (1, …, m) and a vector of independent latent (exogenous) variables ζ:

η = Bη + Γξ + ζ

(1)

where B is the matrix of regression coefficients that relates the endogenous latent variables to each other, Γ is the matrix of regression coefficients that relates the endogenous variables to the exogenous variables (ξ), and ζ is the vector of disturbance terms that follow a multivariate normal distribution, with an expected value of zero. The disturbance terms are independent of the exogenous latent variables.

The measurement models link latent variables to manifest indicators for both endogenous (2) and exogenous (3) variables:

y = Λ_yη + ε

(2)

x = Λ_xξ + δ

(3)

where y and x are the vectors of the manifest indicators, Λ_y and Λ_x are the matrices of the regression coefficients that relate y to η and x to ξ, respectively, ε and δ are the corresponding vectors of errors.

2.1.1. Assessing Measurement and Structural Models

Given the reflective nature of the measurement model adopted in this study, its assessment was conducted using the following criteria: individual item reliability, constructs’ internal consistency and reliability, convergent validity, and discriminant validity [70,71]. First, factor loadings are commonly used to assess individual item reliability. Loadings exceeding the minimum threshold of 0.40 or preferably 0.70 [70,72,73] are deemed acceptable for a well-fitted reflective measurement model [74]. Consequently, any items falling below the established threshold should be eliminated. Second, Cronbach’s alpha and composite reliability (CR) can be used together as measures of constructs’ internal consistency and reliability. The recommended range for acceptable Cronbach’s alpha values is between 0.70 and 0.90. Lower values may indicate a weak correlation between the construct items, while higher values may suggest item redundancy [75]. However, it is worth noting that some scholars have suggested that lower Cronbach’s alpha values can also be acceptable [76,77]. To overcome the limitations of Cronbach’s alpha, which tends to underestimate reliability due to its conservative nature, the CR index is recommended. A CR value of 0.80 is considered good, as it ensures that the indicators are truly representative of the construct [78]. Third, convergent validity assesses the extent to which a construct converges in its indicators by explaining item variance. A measure of convergent validity is the average variance extracted (AVE), which reflects the average communality of each latent construct. The AVE is calculated as the average of the squared loadings of all indicators associated with a specific construct. The benchmark value for AVE is 0.50 [79,80], indicating that each construct explains more than half of the variability in its indicators. Fourth, discriminant validity examines the extent to which each construct differs from others in terms of its correlation with other constructs and the distinctiveness of its indicators. A criterion for discriminant validity is to ensure that the variance shared by a latent variable with its manifest indicators is greater than the variance it shares with any other latent variable [81]. This criterion is met when the square root of the AVE values in the diagonal cells is greater than the correlation between any other pair of constructs.

Once the quality of the measurement models has been established, the evaluation of the structural model’s quality needs to be assessed. The adjusted R² of the latent variables and the f-square effect size are both informative measures used to assess the fit of the structural model to the data [82,83]. The adjusted R² indicates the amount of variance in each endogenous construct that is explained by the independent variables, thus reflecting the model’s predictive accuracy. While the interpretation of the adjusted R² should be context-specific, threshold values of 0.75, 0.50, and 0.25 are commonly considered substantial, moderate and weak, respectively [59,71]. The f-square effect size quantifies the change in R² when a specific variable is omitted from the model, indicating the magnitude of variance not explained by the change in R² [70,79]. According to [83], threshold values of 0.15, 0.20, and 0.35 represent small, medium, and large effect sizes, respectively.

2.1.2. Moderation and Multi-Group Analysis

The moderator effect, unlike the mediating effect, occurs when another variable (referred to as the moderator) may change the strength or even the direction of the relationship between two constructs. The moderator acts as a grouping variable, dividing the data into subsamples. To check whether there are significant differences between the group models, the same theoretical model was estimated for each subsample [70].

Considering the importance of human values in shaping behavioural intentions towards nature park conservation, as well as the inherent complexity of these values, this study employed a cluster analysis to identify the moderator variable that enables the classification of respondents based on their value orientations. Subsequently, the same theoretical model was performed for each subgroup of respondents. The cluster analysis was conducted using a set of indicators related to altruistic, biospheric, and egoistic values, which were measured via multiple items rated on a 5-point Likert scale, ranging from 1 (totally disagree) to 5 (totally agree). The first three items assess altruistic values (world free of war and conflict, equality and equal opportunities, social justice). Two other items focus on biospheric values (oneness with nature, respect for nature and the environment). The last three items capture egoistic values (authority and leadership, social power and control over others, influence over others’ choices and events).

Euclidean distance was used as a measure of dissimilarity between individuals, transforming the original data into a distance matrix as follows: $d\left(i,j\right)={\left[\sum _j{\left(x_{ij}-x_{hj}\right)}^2\right]}^{\raisebox{1ex}{$1$}\!\left/ \!\raisebox{-1ex}{$2$}\right.}$, where $x_{ij}$ and $x_{hj}$ represent the values of the j-th items for the respondents i and h, respectively. The agglomerative hierarchical classification technique was used for grouping, which involves a sequential process of merging the n units. The complete linkage method was chosen as the aggregation criterion, as it facilitates the formation of homogeneous clusters. In complete-linkage clustering, the link between two clusters contains all pairs of elements, and the distance between the clusters is equal to the distance between the two elements (one per cluster) that are furthest apart. The shortest of these links remaining at each step causes the merging of the two clusters whose elements are involved [84].

Once the cluster analysis has identified homogeneous groups of respondents based on their value orientations, and the same model has been estimated for each group, a multi-group analysis was conducted to compare structural path parameters between the subgroups. In multi-group analyses, ensuring measurement invariance is crucial to guarantee that any between-group differences in model estimates stem from actual differences in structural relationships, rather than variations in the content and meanings attributed to phenomena by respondents from different groups. Measurement invariance includes two key aspects: (i) configural invariance, which requires that the basic factor structure (i.e., the number of constructs and associated items) remains consistent across all groups; (ii) compositional invariance, which ensures that item loadings remain invariant across groups. Establishing both configural and compositional invariances confirms partial measurement invariance, allowing for the comparison of path coefficient estimates between groups [68,70]. In summary, PLS-SEM has several advantages, including: (i) estimating both direct and indirect causal effects; (ii) testing the invariance of model parameters across multiple groups; (iii) robustness against measurement errors [74].

2.2. Data

The extended TPB framework was empirically tested using data collected from a sample of 371 students attending three universities in the Metropolitan City of Naples, Campania, Italy. To ensure participants’ anonymity, the Computer Assisted Web Interviewing (CAWI) methodology was used based on an electronic questionnaire. The sampled students were invited to participate in the survey through email, where they received a link to access the questionnaire. The decision to employ an electronic questionnaire was motivated by various advantages, including its flexibility, speed, cost-effectiveness, and elimination of data entry. Additionally, electronic questionnaires have also been shown to be particularly suitable and effective for younger populations [85]. The choice to survey university students was driven by the recognition that their higher level of education equips them with greater knowledge, skills, and values, making them more likely to adopt pro-environmental behaviours.

The survey was conducted between September and December 2021. To encourage participation, respondents were provided with a brief description of the research’s general purpose, focusing on the importance of conservation and enhancing the natural heritage within the Vesuvius National Park area. The survey primarily consisted of closed-ended questions, and the collected information was time-stamped to ensure accuracy and consistency. All sampled students successfully completed the questionnaire.

The questionnaire was designed to capture a wide range of information, including: (i) demographic characteristics (sex, age, household size, parental education and profession, perceived household economic condition); (ii) behavioural intentions towards nature park conservation and other psychological constructs (e.g., personal values, concerns, pro-environmental behaviour, attitudes, subjective norms, perceived behavioural control, personal moral norms, awareness, attribution of responsibility, opinions, and beliefs); (iii) relationship of university students with leisure and nature. To ensure the appropriateness and quality of the TPB constructs, a pre-test was conducted on a small sample of students. This allowed for the verification that respondents interpreted the questions accurately and provided an overall assessment of the constructs’ quality.

The study focused on the population of undergraduate students residing in the metropolitan area of Naples. To ensure a representative sample, a stratified random sampling design was adopted, using sex and age as stratification variables. The population of university students was divided into eight strata, considering the two sexes and four age groups (18–19, 20–21, 22–25, and over 25). The allocation strategy followed the optimal approach, where the number of participants to be randomly selected from each stratum was proportional to the size and variation of that specific stratum. This type of probability sampling ensures the representativeness of the population and improves the precision of the estimates compared to simple random sampling with the same sample size. The appropriate sample size was determined before conducting any data analysis, considering a statistical power of 80% [86], and considering different scenarios of effect size levels. Following [83], who classified effect sizes of 0.20, 0.50, and 0.80 as small, medium, and large, respectively, the final sample size of 371 respondents can be considered satisfactory. It is closer to the small effect size (size requirement = 533) rather than the medium effect size (size requirement = 119).

The information on sociodemographic characteristics provided valuable contextualisation of the respondents and proved to be important for explaining their behaviour intentions towards nature park conservation (Appendix A,

Table A1. Summary statistics of sample’s socio-demographic characteristics.

Sociodemographic Characteristics	Overall
Gender
Male	43.9
Female	56.1
Age group (years)
18–19	17.0
20–21	49.1
22–23	17.0
24–25	9.9
>25	7.0
Father’s education level
Less than primary and primary (ISCED: 1)	7.1
Lower secondary (ISCED: 2)	33.9
Upper secondary (ISCED: 3-4-5)	39.8
Tertiary (ISCED: 6-7-8)	19.3
Mother’s education level
Less than primary and primary (ISCED: 1)	6.5
Lower secondary (ISCED: 2)	31.6
Upper secondary (ISCED: 3-4-5)	39.8
Tertiary (ISCED: 6-7-8)	22.2
Father’s professional status
Entrepreneurship/professional	33.9
Teacher	1.8
Manager	5.3
Clerk	31.6
Worker	23.4
Unemployed	4.1
Mother’s professional status
Entrepreneurship/professional	11.1
Teacher	10.5
Manager	1.2
Clerk	19.9
Worker	4.1
Housewife	49.7
Unemployed	3.5
Making ends meet
With great difficulty	0.6
With difficulty	14.0
Neither with difficulty nor easily	52.6
Easily	24.6
Very easily	8.2
Household size
Lower than 3	4.1
3–4	68.4
5–6	27.5
Sample size	371

). Examining the sample composition, it was observed that 56% of the respondents were women, which closely mirrored the percentage distribution of undergraduates in the province of Naples (metropolitan area) (http://dati.ustat.miur.it/dataset/iscritti/resource/eae4ee94-0797-41d2-b007-bc6dad3ef3e2, last accessed on 1 August 2023). Regarding age distribution, approximately 66% of the respondents fell within the 18–21 age range, while 27% were between 22 and 25 years old. Only a small proportion, approximately 7%, were older than 25 years.

In terms of family economic conditions, half of the participants reported coming from families that ‘make ends meet’, neither facing difficulties nor experiencing ease. Conversely, around 15% stated that their families faced economic difficulties, while over 30% indicated that their families ‘make ends meet’ easily or very easily. In relation to household size, roughly 70% of the participants belonged to household with three or four members. The percentage of participants from large households, consisting of five of six members, was less than 30%. Single-component or two-component households made up a negligible share, accounting for approximately 4% of the sample.

Table 1. Latent constructs, manifest indicators and items. Descriptive statistics.

Latent Construct	Manifest Indicator	Item	Mean	St. Dev.
Subjective norm (SN) Likert scale variables, from 1 ‘totally disagree’ to 5 ‘totally agree’	norm_1	- People who are important to me would advise me to conduct behaviours that help improve the quality of natural resources in the park area	4.26	0.75
	norm_2	- It is expected of me that I would conduct behaviours to improve the quality of natural resources in the park area	4.13	0.80
	norm_3	- People significant to me engage in virtuous behaviours in the park area	3.88	0.86
Attitude (ATT) Likert scale variables, from 1 ‘totally disagree’ to 5 ‘totally agree’	att_park_1	- Taking action aimed at preserving a natural area is a good idea	4.74	0.44
	att_park_2	- Taking action aimed at preserving a natural area is beneficial	4.56	0.57
	att_park_3	- Taking action aimed at preserving a natural area is pleasant	4.50	0.65
	att_park_4	- Taking action aimed at preserving a natural area is wise	4.70	0.54
Perceived behavioural control (PBC) Likert scale variables, from 1 ‘totally disagree’ to 5 ‘totally agree’	perc_1	- Taking action aimed at preserving a park area is effortless	2.86	1.27
	perc_2	- The decision to take action to preserve a park area is entirely up to me	3.22	1.28
	perc_3	- I have the resources, time, and opportunities to take action to preserve a park area	3.13	1.11
Personal moral norm (PMN) Likert scale variables, from 1 ‘totally disagree’ to 5 ‘totally agree’	moral_1	- I would feel guilty if I were not involved in actions to preserve the park area	3.78	0.84
	moral_2	- I believe I have a moral obligation to take actions to preserve the park area	3.90	0.89
	moral_3	- Actions aimed at preserving a park area go along with my principles of environmental protection	3.96	0.82
Behaviour Intention (BI) Likert scale variables, from 1 ‘totally disagree’ to 5 ‘totally agree’	int_1	- I am willing to take action aimed at preserving a park area (e.g., waste collection campaign, …)	4.42	0.68
	int_2	- I plan to undertake action aimed at preserving a park area (e.g., waste collection campaign, …)	4.08	0.86
	int_3	- I will make an effort to take action aimed at preserving a park area (e.g., waste collection campaign, …)	4.19	0.83

shows the latent constructs (Column 1), the labels (Column 2) and the meaning of the manifest indicators (Column 3) used to compose each of the five constructs. The table is complemented by descriptive statistics in the last two columns. The respondents strongly believe (injunctive normative beliefs) that their referent persons recommend or expect them to engage in behaviours at improving the quality of nature parks. However, they are less confident that referent persons will actually perform those behaviours (descriptive normative beliefs). The participants also strongly support behavioural beliefs (attitudes) that performing actions to preserve natural areas will yield definite outcomes or provide certain experience. On the other hand, lower average values suggest that the students interviewed do not strongly agree with control beliefs (perceived behavioural control) that indicate nature park conservation is effortless or that they always possess the necessary control factors (such as resources, time, and opportunities) to facilitate behaviour. The respondents hold the belief that engaging in actions to protect the park areas aligns with their own moral obligation and principles of environmental protection (personal moral norms). Finally, the behaviour intentions related to the willingness to undertake actions or, at least, to make an effort to undertake or plan actions aimed at preserving park areas (e.g., waste collection campaigns, …) demonstrate a high level of agreement among the respondents, as evidenced by sufficiently high average values.

2.3. Measures

In the extended TPB framework adopted in this study, attitudes, perceived behavioural control, personal moral norms, and subjective norms were treated as latent variables or constructs that could not be directly observed. Instead, they were measured indirectly using a set of indicators that served as proxy variables (items). All constructs were measured via multiple items on a 5-point Likert scale, ranging from 1 (totally disagree) to 5 (totally agree). A total of 16 items were used, aligning with the research objective and relevant literature.

Attitudes were measured by asking university students to express their beliefs about potential actions to be taken for preserving nature park, as adapted from previous studies [39,87,88,89]. Specifically, university students were asked to indicate their level of agreement/disagreement that taking action to preserve a nature park was: (i) a good idea, (ii) beneficial, (iii) pleasant, (iv) wise.

Perceived behavioural control was measured using three items adapted from previous studies [19,39,90,91]. The first two items focused on university students’ perceived personal power and self-efficacy in engaging in the behaviour (i.e., Taking action aimed at preserving a park area is effortless; the decision to take action to preserve a park area is entirely up to me). The last item addressed university students’ availability in terms of resources, time, and opportunities (I have the resources, time, and opportunities to take action to preserve a park area).

Personal moral norms were measured using three items adapted from studies [39,92,93,94]. University students were asked to indicate their level of agreement/disagreement with the following statements: (i) I would feel guilty if I were not involved in actions to preserve the park area; (ii) I believe I have a moral obligation to take action to preserve the park area; (iii) Actions aimed at preserving a park area go along with my principles of environmental protection.

Subjective norms were measured using three items adapted from studies [43,95,96,97]. The measurements included both injunctive norms (i.e., People who are important to me would advise me to conduct behaviours that help improve the quality of the environment in the nature park area; It is expected of me that I conduct behaviours to improve the quality of the environment in the nature park area) and descriptive norms (i.e., People significant to me engage in virtuous behaviours in the nature park area).

Behavioural intentions were assessed using three items adapted from previous research [98,99,100] to represent a clearer commitment to act pro-environmentally in a park area. The items included statements such as: (i) I am willing to take action aimed at preserving a park area); (ii) I plan to undertake action aimed at preserving a park area; (iii) I will make an effort to take action aimed at preserving a park area.

3. Results

In this study, PLS-SEM was estimated to investigate the research hypotheses outlined in the ‘Theoretical Framework and Research Hypotheses’ section. Since PLS-SEM is a non-parametric method, 5000 samples were generated to obtain standard error estimates of the model parameters for conducting statistical significance tests [61]. The PLS-SEM was initially estimated using the entire sample of respondents and then separately for subgroups of respondents classified according to their value orientations. The following sections show the findings related to the measurement models’ fit to the data, as well as the key results obtained from the structural models and the multi-group analysis.

3.1. Measurement Model Assessment

The measurement model was developed using an iterative process, where indicators were added or removed within each construct, aiming to achieve a strong relationship between the manifest indicators and their corresponding latent construct. Since the measurement model follows a reflective approach, where the latent variable serves as a single predictor of each manifest indicator, multicollinearity is not an issue.

First, in evaluating the reliability of individual items, the factor loadings were found to be relatively high for each construct, surpassing the conventional threshold of 0.70 and indicating a well-fitted reflective model (

Table 2. Individual item reliability: factor loadings.

Latent Variables		Overall	Nature-Oriented Altruistic	More Egoistic-Oriented
Subjective norm	norm_1	0.741	0.714	0.814
	norm_2	0.898	0.885	0.897
	norm_3	0.716	0.744	0.772
Attitude	att_park_1	0.840	0.790	0.834
	att_park_2	0.806	0.800	0.798
	att_park_3	0.799	0.870	0.820
	att_park_4	0.711	0.781	0.752
Perceived Behavioural Control	perc_1	0.740	0.826	0.757
	perc_2	0.619	0.730	0.576
	perc_3	0.896	0.779	0.929
Personal moral Norm	moral_1	0.761	0.834	0.804
	moral_2	0.842	0.835	0.851
	moral_3	0.795	0.860	0.754
Behavioural Intention	int_1	0.886	0.871	0.891
	int_2	0.916	0.908	0.914
	int_3	0.885	0.875	0.861

). This suggests a strong relationship between each manifest indicator and its corresponding latent construct, with more than half of the indicator’s variance being explained by its factor. Although one indicator (perc_2) exhibited a lower factor loading, it still exceeded the threshold of 0.4, suggesting that it may be conceptually somewhat distinct from the underlying latent construct without necessarily casting doubt on the validity of the measurement model.

Second, regarding the internal consistency and reliability of the constructs, the Cronbach’s alpha values were found to fall within the recommended acceptable ranges of 0.70 to 0.90 (

Table 3. Constructs’ internal consistency and reliability: Cronbach’s alpha, composite reliability (CR) and average variance extracted (AVE).

	Overall			Nature-Oriented Altruistic			More Egoistic-Oriented
Latent Variables	Cronbach’s Alpha	CR	AVE	Cronbach’s Alpha	CR	AVE	Cronbach’s Alpha	CR	AVE
SN	0.717	0.831	0.623	0.708	0.818	0.602	0.775	0.868	0.688
PMN	0.720	0.835	0.640	0.797	0.880	0.710	0.711	0.825	0.613
ATT	0.802	0.869	0.624	0.826	0.885	0.658	0.815	0.878	0.643
PBC	0.672	0.792	0.578	0.661	0.776	0.519	0.725	0.820	0.609
BI	0.876	0.924	0.802	0.862	0.916	0.783	0.868	0.919	0.790

). Although the Cronbach’s alpha for the PBC construct slightly fell below the minimum threshold, it aligns with the previous literature. Additionally, the CR index reached the acceptability threshold of 0.80, indicating that the indicators effectively represent the construct.

Third, the AVE values, which serve as an assessment of convergent validity, ranged from 0.578 to 0.802. These values exceeded the suggested benchmark of 0.50, indicating that each latent construct explained more than half of the variability of its indicators. This provided evidence of good convergent validity (Table 3). Fourth, the discriminant validity was verified since the square root of the AVE values in the diagonal cells of the table was always higher than the inter-construct correlations (

Table 4. Discriminant validity.

Latent Variable	Overall					Nature-Oriented Altruistic					More Egoistic-Oriented
	ATT	BI	PBC	PMN	SN	ATT	BI	PBC	PMN	SN	ATT	BI	PBC	PMN	SN
ATT	0.790					0.811					0.802
BI	0.457	0.896				0.426	0.885				0.442	0.889
PBC	0.136	0.510	0.760			0.074	0.366	0.720			0.053	0.440	0.781
PMN	0.326	0.590	0.375	0.800		0.396	0.529	0.223	0.843		0.265	0.678	0.462	0.783
SN	0.377	0.449	0.341	0.484	0.789	0.363	0.382	0.379	0.281	0.776	0.446	0.541	0.347	0.547	0.829

). This implies that each construct captured a unique dimension of the phenomenon that was not represented by any other construct in the model [61].

In summary, the measurement models exhibited strong performance in terms of item and construct reliability, convergent validity, and discriminant validity. All informative criteria met the acceptable thresholds, indicating the satisfactory quality of the measurements.

3.2. Structural Model

The structural model illustrates the relationships among subjective norms, attitudes, perceived behavioural control, personal moral norms, and behavioural intentions towards nature park preservation (Figure 1). Consistent with previous studies [38], which acknowledged the potential inclusion of additional constructs in the TPB approach, we empirically explored personal moral norms as an additional predictor of behavioural intention. Each arrow in the model represents a linear causal relationship between the latent constructs. As a model with mediating effects, the constructs of attitude, perceived behavioural control, and personal moral norms served as mediating variables. Therefore, we considered both direct and indirect effects to obtain a richer picture of the relationships in the structural model.

The informative measures assessing the fit of the structural model to the data were within the acceptable range. The adjusted R² values for the latent constructs, both for the overall model (ranging from 0.77 to 0.89) and the models by subgroup (ranging from 0.63 to 0.93 and 0.71 to 0.89), indicated a strong fit of the structural models. Additionally, the f-square effect size, ranging from 0.16 to 0.21 for the overall model, and from 0.14 to 0.19 and from 0.12 to 0.20 for the separate models, further supported the correct specification of the models.

Table 5. Structural models: direct and indirect effect estimates on behavioural intentions.

Hp	Path	Overall		Nature-Oriented Altruistic		More Egoistic-Oriented
		Path Coefficients	Standard Error	Path Coefficients	Standard Error	Path Coefficients	Standard Error
Direct effects
H1	ATT → BI	0.274 ***	0.070	0.260 ***	0.079	0.226 **	0.109
H2	PBC → BI	0.318 ***	0.072	0.172 **	0.078	0.227 **	0.107
H3a	PMN → BI	0.348 ***	0.091	0.471 ***	0.091	0.357 ***	0.129
H4a	SN → BI	0.069	0.102	0.108	0.133	0.114	0.108
Indirect effects
H4b	SN → PMN → BI	0.169 ***	0.056	0.258 ***	0.066	0.100 *	0.060
H4c	SN → PBC → BI	0.108 ***	0.036	0.060 *	0.032	0.086 *	0.051
H4d	SN → ATT → BI	0.078 **	0.033	0.112 **	0.053	0.062	0.042
H4e	SN → PMN → ATT → BI	0.025	0.018	0.004	0.022	0.020	0.016
H3b	PMN → ATT → BI	0.051	0.035	0.008	0.040	0.072	0.047

* Significant at 10%; ** significant at 5%; *** significant at 1%.

shows the estimates of direct effects (Figure 5) and indirect effects (Figure 6) from the structural models, which were used to test the research hypotheses. The effects of subjective norms, attitudes, perceived behavioural control, and personal moral norms were analysed while controlling for a set of sociodemographic variables, including sex, age, and household economic conditions. These control variables were considered relevant in environmentally-related contexts based on previous research [101,102,103]. However, no significant relationships were found between these control variables and behavioural intentions towards nature park preservation. This finding aligns with a portion of the literature [41,43,93,104] that examined the influence of control variables in the TPB on behavioural intentions towards specific pro-environmental actions.

The results provided support for the first three research hypotheses, which posited direct effects of attitudes (H₁), perceived behavioural control (H₂), and personal moral norms (H_3a) on behavioural intentions. This was evident from the statistically significant direct effect coefficients associated with these hypotheses. In particular, the paths from attitudes (b_H1 = 0.274, p < 0.001), perceived behavioural control (b_H2 = 0.318, p < 0.001), and personal moral norms (b_H3a = 0.348, p < 0.001) to university students’ intentions towards actions aimed at conserving natural park areas were found to be significant and positive.

Regarding H₁, the positive effect of attitudes on behavioural intentions suggested that when university students consider engaging in actions aimed at prevising a natural park area as a good idea, beneficial, pleasant, and wise, they are more inclined to have intentions aligned with nature park conservation. Regarding H₂, the findings indicated that as university students perceive fewer difficulties in the external context, such as resource scarcity, time constraints, and limited opportunities, they are more likely to develop intentions towards nature park conservation. Regarding H₃, the results demonstrated that the stronger the belief among university students that protecting park areas aligns with their personal moral obligation and principles of environmental protection, the more likely they are to form intentions towards nature park preservation.

Conversely, the research hypothesis proposing the direct effect of subjective norms on behavioural intentions (H_4a) did not receive support from the data. This was evident from the non-statistically significant direct effect coefficients associated with this hypothesis (b_H4a = 0.069, p < 0.498), indicating that subjective norms alone do not contribute to the formation of behavioural intentions towards nature park preservation.

Table 5 also shows the indirect effects of subjective norms on behavioural intentions through different mediators, namely personal moral norms (H_4b), perceived behavioural control (H_4c), and attitudes (H_4d) or through personal moral norms and attitudes simultaneously (H_4e). It is worth noting that these indirect effects were found to be statistically significant, unlike the direct effect of subjective norms on behavioural intentions, which was never significant. Specifically, the indirect effects of subjective norms on behavioural intentions through personal moral norms (b_H4b = 0.169, p = 0.0025), perceived behavioural control (b_H4c = 0.108, p = 0.0027), and attitudes (b_H4d = 0.078, p = 0.018) were found to be statistically significant. This means that personal moral norms, perceived behavioural control, and attitudes significantly mediate the effect of subjective norms on behavioural intentions for nature park conservation. However, the indirect effects of subjective norms on behavioural intentions through the simultaneous influence of personal moral norms and attitudes (b_H4e = 0.025, p = 0.165) were not statistically significant. Therefore, the related research Hypothesis (H_4e) was rejected. In the overall model, the proportion of the indirect effects of subjective norms on the total effect was 84.6%. The indirect effects of subjective norms on behavioural intentions can be decomposed into four components, each representing a different mediator: (i) the indirect effect of subjective norms through personal moral norms (44.5%); (ii) the indirect effect of subjective norms through perceived behavioural control (28.4%); (iii) the indirect effect of subjective norms through attitudes (20.5%); (iv) the indirect effect of subjective norms through both personal moral norms and attitudes (6.6%).

Finally, it is worth noting that personal moral norms had a direct influence on university students’ behavioural intentions towards nature park preservation (H_3a). However, the significance of the contribution of personal moral norms to the development of behavioural intentions lost when mediated by attitudes (b_H3b = 0.051, p < 0.145), leading to the rejection of Hypothesis H_3b.

3.3. Multi-Group Comparison

Cluster analysis was performed to explore how behavioural intentions towards nature park preservation, as well as the relationships between their potential determinants (i.e., subjective norms, attitudes, perceived behavioural control, and personal moral norms), may differ among university students with different value orientations. The cluster analysis resulted in the identification of two distinct groups of respondents based on their value orientations. The first group, consisting of 223 students, displayed a strong inclination towards altruistic and biospheric values, with minimal emphasis on egoistic values (referred to as nature-oriented altruistic). The second group, comprising 148 students, exhibited a higher level of egoistic values, alongside consideration for altruistic and biospheric values (referred to as more egoistic-oriented). To further investigate the differences between these subgroups, a multi-group analysis was conducted to compare the structural path parameters. This analysis aimed to determine whether the relationships between the variables differed significantly between the nature-oriented altruistic group and the more egoistic-oriented group of university students.

Before proceeding with testing the differences between the coefficients, it was necessary to verify measurement invariance. Configural invariance was established, as both models (one for each subgroup) were built using the same constructs and indicators. Additionally, compositional invariance was verified, as the factor loadings in the measurement model were invariant across the subgroups. With the presence of measurement invariance, it became possible to compare the structural path parameters between the two subgroups.

As shown in

Table 6. Cross-group differences among structural paths.

Group Comparison	Hypotheses	Path	Estimate	Standard Error
First vs. second	H1	ATT → BI	0.034	0.077
	H2	PBC → BI	−0.055	0.081
	H3a	PMN → BI	0.114 ***	0.039
	H4a	SN → BI	−0.006	0.112
	H4b	SN → PMN → BI	0.158 ***	0.049
	H4c	SN → PBC → BI	−0.026	0.087
	H4d	SN → ATT → BI	0.050 **	0.024
	H4e	SN → PMN → ATT → BI	−0.016	0.014
	H3b	PMN → ATT → BI	−0.064	0.078

** Significant at 5%; *** significant at 1%.

, there were no significant structural differences between the two subgroups of university students concerning the direct effects of attitudes (H₁), perceived behavioural control (H₂), and subjective norms (H_4a) on behavioural intentions. However, some interesting differences emerged. Specifically, the structural paths of personal moral norms on behavioural intentions (H_3a) differed between the two subgroups (b_{H3a(altruistic)} − b_{H3a(egoistic)} = 0.114, p = 0.004), with the nature-oriented altruistic group displaying a stronger impact of personal moral norms on behavioural intentions towards nature park preservation.

Regarding the indirect effects (Table 5), personal moral norms (H_4b), perceived behavioural control (H_4c), and attitudes (H_4d) retained their role as mediators between subjective norms and behavioural intentions for the nature-oriented altruistic group (b_H4b = 0.258, p < 0.001; b_H4c = 0.060, p = 0.061; b_H4d = 0.112, p = 0.035). However, these mediating roles were weaker for the more egoistic-oriented group (b_H4b = 0.100, p = 0.096; b_H4c = 0.086, p = 0.092), and even became insignificant for attitudes (b_H4d = 0.062, p = 0.140). Specifically, there were two significant differences between the nature-oriented altruistic group and the more egoistic-oriented group concerning the mediated paths of subjective norms on behavioural intentions via personal moral norms (b_{H4b(altruistic)} − b_{H4b(egoistic)} = 0.158, p = 0.001) and attitudes (b_{H4d(altruistic)} − b_{H4d(egoistic)} = 0.050, p = 0.037) when considered separately. This suggests that the impact of subjective norms on behavioural intentions via personal moral norms or attitudes is stronger for university students with nature-oriented altruistic values. Lastly, no significant difference was found between the paths of personal moral norms on behavioural intentions via attitudes (b_{H3b(altruistic)} − b_{H3b(egoistic)} = −0.064, p = 0.794).

4. Discussion

The present study provided evidence for the predictive validity of the TPB approach by investigating university students’ intentions to take actions aimed at preserving nature parks. The analysis of reliability and validity of the measurement component, along with the goodness-of-fit measures of the structural component, demonstrated the adequacy of the PLS-SEMs in fitting the data and confirmed their appropriate specification.

The results showed that the TPB constructs, augmented with an additional construct, were good predictors of university students’ behavioural intentions towards nature park conservation. Structural equations analysis proved the direct effect of attitudes, perceived behavioural control, and personal moral norms on behavioural intentions to participate in nature park preservation. However, subjective norms did not exhibit a significant direct influence on the formulation of behavioural intentions for nature park preservation but revealed indirect effects on such intentions.

This study presented evidence supporting research Hypothesis 1 (H₁), which proposed a direct and positive influence of attitudes on university students’ behavioural intentions towards nature park conservation. This finding was in line with previous research that demonstrated a direct effect between young people’s attitudes and intentions to adopt specific pro-environmental behaviours, such as green consumption [105], purchasing of pro-environmental products [13], pollution emissions reduction [39], the choice of transport mode [106], as well as conservation and recycling behaviours [88,107]. Regarding nature parks, our findings were consistent with recent studies [108,109] that highlighted the prevalence of positive attitudes. This can be explained by the fact that while protected areas may entail costs for young people, the numerous benefits they provide are often more widely perceived than the costs incurred. Attitude refers to the individual’s favourable or unfavourable evaluation of the behaviour in question, and a positive evaluation of behaviour and its outcomes increases the likelihood of engaging in that behaviour [110]. In the context of nature parks, we argued that initiatives enabling university students to evaluate the consequences of their actions could effectively enhance their attitudes towards preservation behaviours. University can provide a suitable context for young people to discuss environmental issues and develop awareness, thereby encouraging them to engage in environmental-friendly behaviours. Hence, creating nature-focused citizen science events and sustainability projects can help students understand the importance of nature and biodiversity for their own well-being.

Regarding research Hypothesis 2 (H_s), the findings indicated that perceived behavioural control had a direct and positive impact on university students’ behavioural intentions towards nature park conservation. This finding aligns with previous studies that examined pro-environmental behaviours among students [43], as well as young adults’ intentions to recycle [88] and choose recyclable packaging [90]. Furthermore, it is consistent with research investigating conservation behaviours related to natural resources more broadly [111]. Perceived behavioural control related to individuals’ judgement regarding the feasibility of performing certain behaviours and their ability to exert control over them, taking into account factors such as resources availability, time, and opportunities. Consequently, an individual’s ability to predict outcomes and control the difficulty associated with adopting a specific behaviour influence their final choice. Based on these considerations, a stronger perception among university students of the availability of resources, time, and opportunities facilitates the development of their behavioural intentions towards nature park preservation. It is therefore crucial to create favourable conditions for engaging young people in appropriate actions by removing barriers. The management institutions of protected areas should allocate financial resources to organise events that actively involve young people, raising awareness about the role that nature parks play in biodiversity conservation and protection of endangered wildlife. Additionally, park managers, in collaboration with local institutions and environmental associations, should take the lead in safeguarding open spaces for the collective benefit and actively promoting sustainable development of the regions and their environment. Encouraging increased participation of young people in programs focused on protecting and conserving nature, including activities such as path maintenance, forest improvement, and upkeep of recreational areas, is highly desirable. To ensure youth involvement in these initiatives, accessible information should be provided, especially in educational institutions such as universities.

The additional construct of personal moral norms captured the social altruism and moral awareness among university students regarding the broader societal benefits associated with nature park conservation. In relation to research Hypothesis 3a (H_3a), the study findings revealed a direct and positive relationship between personal moral norms and university students’ behavioural intentions towards nature park conservation. These results align with the findings reported by van Riper and Kyle [112], indicating that individuals who embrace personal moral norms, especially in conjunction with self-interest orientation, are more inclined to participate in nature park conservation behaviours. Furthermore, these findings are supported by more recent research [113] that emphasised the influential role of personal moral norms in fostering a heightened sense of responsibility and encouraging individuals to act in a responsible manner towards nature parks. This implies that university students with higher personal moral norms feel a stronger moral obligation and responsibility to preserve nature parks. Therefore, university students not only consider their attitudes and capabilities for engaging in conservation behaviours but also take into account moral considerations that influence their behavioural intentions. This finding suggests the potential to foster a stronger sense of social responsibility among university students. Moreover, personal moral norms had a lasting impact, as evidenced by previous studies [39,114]. Once an individual develops a moral norm regarding a specific behaviour as an obligation, it tends to persist over time. Therefore, education institutions such as universities could play a crucial role in shaping personal moral norms related to pro-environmental behaviours, including nature park conservation. Similarly, park managers, in collaboration with other public and non-profit organisations involved in environmental management, should provide insights into the value of nature park conservation, thus aiding university students in developing enduring moral norms aligned with nature and the outdoors.

With respect to research Hypothesis 4a (H_4a), the findings of this study rejected the notion that subjective norms have a direct impact on university students’ intentions towards nature park conservation. This finding aligns with Esfandiar et al. [113], who observed a limited influence of subjective norms on individuals’ intention to engage in pro-environmental behaviour in national parks, and Nguyen and Tran [24], who highlighted the role of biodiversity conservation perception. However, it contradicts previous empirical studies that had demonstrated a direct relationship between subjective norms and pro-environmental behavioural intentions [39]. Nevertheless, the results revealed indirect effects of subjective norms on behavioural intentions through personal moral norms, perceived behavioural control, and attitudes, thereby providing support for research Hypotheses 4b (H_4b), 4c (H_4c), and 4d (H_4d), respectively. This suggests that the subjective norms alone are insufficient to directly elicit tangible and satisfactory behavioural intentions towards nature park conservation. This observation could be interpreted within the context of Western people placing less emphasis on collectivism in comparison to their Eastern counterparts [115]. Other studies [19,93] also revealed the indirect influence of subjective norms on people’s intentions to engage in actions aimed at preserving nature park. These studies demonstrated that individuals can gradually internalise subjective norms, perceiving them as ‘personal moral norms’ and developing ‘self-confidence’ in their ability to control and achieve efficacy in their actions.

The present study found that university students’ subjective norms have a direct impact on their attitudes towards nature park conservation. This implies that when university students consider the severe consequences of unfriendly behaviour towards nature parks, subjective norms play a significant role in influencing their behavioural intentions, similar to the mediated effects of personal moral norms and perceived behavioural control.

Drawing on a body of literature that highlights human values as influential guiding principles shaping individuals’ behaviour and interactions with the natural environment [23,51], this study examined the value orientation of university students to gain a deeper understanding of their diverse perspective and approaches to interacting with and managing the natural environment. Individuals vary in the importance they ascribe to each value that guides their behavioural choices. The greater importance individuals place on a specific value, the more decisive it becomes in shaping their evaluations, intentions, and behaviours [116]. Consistent with this notion, a cluster analysis was conducted, leading to the identification of two subgroups among university students: the nature-oriented altruistic subgroup and the more egoistic-oriented subgroup. Each subgroup is characterised by its distinct value heritage. This grouping enabled the testing of research hypotheses that were previously examined with the entire cohort of respondents, now applied to the two subgroups classified based on their value orientations.

By doing so, the study highlighted significant differences concerning the factors that influence behavioural intentions towards nature park conservation. Specifically, the study revealed that personal moral norms have a stronger direct impact on behavioural intentions towards nature park conservation (research Hypothesis 3a, H_3a) among nature-oriented altruistic students. The study also found that both personal moral norms and attitudes play a stronger mediating role between subjective norms and behavioural intentions among nature-oriented altruistic students (research Hypotheses 4b and 4d). This suggests that attitudes and personal moral norms act as a bridge, influencing the relationship between subjective norms and behavioural intentions, particularly for those university students who prioritise nature-oriented altruistic values. These findings support previous research showing the association between human values and behavioural engagement in the context of natural environment conservation [112,117]. Decisions regarding involvement in nature park conservation actions are more extensively processed by nature-oriented altruistic students compared to egoistic-oriented students, drawing upon feelings of moral obligation derived from the awareness of the importance of their actions in mitigating adverse consequences for both non-human species and human beings. The more strongly individuals endorse biospheric and altruistic values, the more likely they are to engage in pro-environmental behaviour [118,119] or exhibit behavioural intentions towards nature park conservation [23,120].

This implies that policies can attempt to steer the university students’ behavioural intentions by reinforcing altruistic and biospheric values and providing accurate information to instil environmental sustainability values, even among egoistic-oriented students. The ultimate goal should be to translate these values into more tangible outcomes rather than leaving them as abstract concepts, recognising that such behaviours may not always bring immediate personal benefits in the short term but undoubtedly benefit others and the environment.

5. Conclusions

This study extended the Theory of Planned Behaviour (TPB) to investigate university students’ intentions to nature park conservation, an area that lacks extensive research compared to other pro-environmental behaviours. The key findings can be summarised as follows.

First, the TPB approach effectively explained university students’ behavioural intentions towards nature park conservation. Attitudes, perceived behavioural control, and personal moral norms were found to be antecedents to these intentions. Subjective norms only had an indirect influence on behavioural intentions via attitudes, perceived behavioural control, and personal moral norms. Second, the study explored university students’ value orientations and identified that those with nature-oriented altruistic values feel a stronger moral obligation due to their awareness of the environmental consequences of their actions. As a result, they exhibit higher behavioural intentions towards conserving nature parks compared to their more egoistically-oriented peers.

The findings have practical implications for park management interventions. Involving university students from local communities in decision-making processes is recommended to preserve nature parks. This aligns with the UN Youth Program for Sustainable Development Goals, which emphasise critical thinking, change-making, innovation, communication, and leadership as key dimensions for effective youth involvement in sustainability decision-making. Young people, as critical thinkers, can analyse existing practices, policies, and power structures, and identify areas for improvement. Young people can act as change-makers by increasing events that actively involve people and fostering cooperation between universities, education departments, and protected area institutions to raise awareness about environmental issues and inspire collective action. As innovators, young people can propose innovative approaches to conservation, sustainability, and visitor engagement within nature parks. They can also serve as communicators, disseminating information about the importance of nature parks and leveraging social media platforms for wider public engagement. Lastly, young people can be effective leaders in park management, bringing positive change to their communities.

Further research should involve a broader sample of young people, extending beyond university students, to enhance the understanding of their propensity towards nature park conservation. Moreover, conducting comparisons across different local contexts and exploring additional constructs within the TPB approach would contribute to a more comprehensive examination of intentional behaviours towards nature park conservation.