Effects of Emotions and Ethics on Pro-Environmental Behavior of University Employees: A Model Based on the Theory of Planned Behavior

Abstract

Over the last few years, many activities conducted by university employees have contributed greatly to the global rise of greenhouse gases responsible for the warming of the planet. This phenomenon can be attributed primarily to altering lifestyle, consumption, and portability patterns. Even though there is a significant applicability, understanding the factors determining the pro-environmental behavior (PEB) of university employees is lacking. Thus, this study extended the theory of planned behavior (TPB) to explain PEB among university employees. Ten hypotheses were suggested regarding the relations between the initial TPB constructs, environmental ethics, and emotional intelligence. A self-reported questionnaire was used to collect data from a randomly selected sample of 302 university employees from six Malaysian higher educational institutions listed in the UI Green Matrix 2016. The findings revealed that environmental ethics significantly affected attitudes, perceived behavioral control, and subjective norms. Furthermore, pro-environmental behavioral intentions (PEIs) were directly influenced by attitude and perceived behavioral control, while emotional intelligence mediated the relationship between PEIs and PEB. Moving forward, this study will provide new insights into the literature of university employees’ PEB using an expanded TPB model.

1. Introduction

Universities have over time been the source of social and political change and the institutions of learning. They not only produce most of the world’s authorities and educators, but also expand the limits of knowledge. Apart from that, universities are also notable employers and consumers of products and services, significantly positioning themselves in national and global economic roles. Given the ascribed roles of universities in transforming society and the current environmental and sustainability problems, universities are increasingly being pressured to deal with and adapt to climate change and other issues of sustainable development, as well as the threats and opportunities that come with them. Hence, one of the ways to curtail universities’ negative impacts on the environment is by understanding the pro-environmental behavior (PEB) of their employees.

In essence, PEB signifies the work ethics wilfully showcased by workers at any organizational level and aimed to conserve or better the natural environment and to make the planet a more hospitable environment for future generations. This can be accomplished by drastically reducing resource use, conserving resources, avoiding waste generation, recycling, and conserving water [1]. For the most part, important eco-friendly behaviors usually not mentioned in job descriptions, such as corporate environmental citizenship behavior, frame the environmental sustainability of an organization; however, more often than not, this is not officially in the jurisdiction of the environmental management system [2]. Of late, PEB has garnered attention for its pivotal role in both organizations and the environment [3]. Hence, researchers are studying the efficacy of engaging in environmental volunteering on employees’ PEB [4,5,6].

Despite its importance, PEB is still a new concept for higher education. Thus, understanding factors that influence the university employees’ propensity to engage in PEB is key to the environmental sustainability of universities. Several studies [7] have used the theory of planned behavior (TPB) to provide a framework for understanding employees’ PEBs. According to TPB [7], attitude (ATT), social norms (SNs), and perceived behavior control (PBC) are antecedents of intention-behavior that will predict the actual behavior [8,9]. Meanwhile, several researchers have agreed that the ability of the TPB construct to predict the behavior of a person is more significant when it anticipates the element of nature within the investigation, along with emotional factors and ethical concerns related to the environment [10]. In many studies, emotional intelligence (EI) [11] and environment ethics (EE) [12] have been found to minimize the attitude–behavior gap. Additionally, a review of the environment behavior literature acknowledges that many authors link EI and EE to PEB, suggesting that emotions and ethics may be integral to the PEB of the university employees.

EI is a major component in the formulation of intentions in the sense of environmental behavior, and its role is imperative in environmental decision-making [11,12,13,14]. This is because employees who can regulate their emotions can influence their intention to perform the intended behavior or do the opposite, which is not to perform the behavior. According to Mohiyeddini et al. [15], the mediation role of emotion emphasizes the function of intention as the key predictor of behavior, increasing and enhancing the predictability of the TPB model. Additionally, Arvola et al. [16] stated that one significant criticism of the TPB model has been the lack of inclusion of moral or ethical concerns. Thus, it is imperative to provide university employees with both EE and EI so that they become psychologically empowered, and their attitude and beliefs toward PEB change, ultimately prompting universities toward environmental sustainability. Therefore, this study used an expanded TPB model, which included the constructs of EE and EI, to predict the PEB of university employees. The study’s specific objectives are as follows: (1) to examine the effects of EE on ATT, SN, and PBC; (2) to examine the influence of ATT, SNs, and PBC on pro-environmental behavior intentions (PEBIs) of university employees; and (3) to examine the mediating role of EI on the relationship between PEB and PEBI.

2. Literature Review

2.1. Theory of Planned Behavior

Many who have conducted similar studies on PEB have justified that PEB is mainly steered by self-interest and the rational evaluation of behavioral consequences, using TPB to frame its premise [15,17,18]. At the outset, TPB was initiated to extend the theory of reasoned action (TRA), and it could only account for voluntary and under-control behaviors. In its original form, TRA predicts a person’s behavior where a person’s intention to display a specific behavior is driven by his or her attitude towards the behavior and subjective norm [7,19]. In essence, ATT signifies a person’s positive (favorable) or negative (unfavorable) evaluations of those displaying the behavior, while SNs are the individual’s perception of social pressure when displaying a certain mannerism [1]. Ajzen [7] remarked that a person’s displayed behavior is heavily dependent upon his or her volitional control. It refers to the person’s ability to show how much people are willing to immerse themselves while performing the behavior under the control of power within their beliefs. Hence, TPB is expanded to TRA to address this volitional limitation [19]. TPB argues that an individual’s focus to perform a behavior is curtailed by PBC, ARR, and SNs, which hold a useful conceptual framing to guess and justify human behavior [7].

TPB has been widely adopted in the field of environmental studies to predict human behavior [11,18]. However, several researchers have pointed out its limitations in explaining individuals’ behavioral intentions and actual behaviors [20]. As a result, by broadening the theory, the use of additional variables may improve the predictive power to explain PEBI [21,22]. Many researchers have used PEB and added variables such as “moral norm” [23], “past behavior” [24], “self-identity” [25], “habit” [26], “self-efficacy” [27], “environmental awareness” [28], “sense of community” [29], and “environmental values”[30].

EE and EI are two critical factors to predict, understand, and change PEB [31,32]. Evidence has shown that both EE and EI play a significant role in motivating PEB and a strong relation between EI and ethics in predicting PEB [33]. Therefore, TPB constructs such as ATT, SNs, and PBC associated with EI and EE could explain the integrated and comprehensive voluntary PEBI. To predict PEB, this study included the additional variables of EE and EI.

2.2. Emotional Intelligence (EI)

EI can foresee many pro-social behaviors amongst its staff [13,33,34]; therefore, it might implicate the progress of PEBs. It was initially described as “the ability to track one’s own and others’ emotions and feelings, to distinguish between them, and to use this knowledge to lead one’s thinking and actions” [35] and was eventually summarized to emphasize the perception, usage, understanding, and management of emotions [36].

Empirical studies have shown that EI can positively enhance engagement in PEB [37]. Additionally, other studies have reported direct and indirect effects on PEB. Landa et al. [38] stated that people with EI could easily adapt to their environment, while people who had lower EI levels were not able to adapt easily and their answers might negatively impact their lives. In this regard, we conclude that PEB is influenced not only by cognitive factors, but also by human emotivity.

2.3. Environmental Ethics

EE is efficient in affecting behavior changes [39]. Ethics can be described as the art and discipline of applying concepts and structures to analyze and address complex moral issues [40]. Treviño et al. [41] asserted that ethics is the moral principles that guide what is right or wrong, positive or negative, and suitable or unsuitable in different settings. A person who possesses the correct ethical skills and who pays attention to environmental ethics is considered to be part of the strategy to improve the human standard of living [42]. Furthermore, Kibert et al. [43] stated that EE encompassed the conceptual basis of environmental values as well as more pressing issues framing social perceptions, practices, and policies that protected and sustained biodiversity and ecosystems. Abedi Sarvestani and Shah Vali [44] declared that EE implied that inhabitants of the earth were servers of nature and tasked with the responsibility to protect it.

In previous research, TPB was extended through ethical dimensions [45,46]. Researchers have concluded that mixing the present theories is better and that mixing moral norms and TPB is the best way to explain PEB [40]. Harland et al. [47] discovered that by including ethics, the proportion of explained variance in intention rose.

2.4. Hypotheses Testing

Few studies have investigated the causal relationship between EE and ATT, PBC, and SN in the context of PEB. Yoon [48] supported the roles of moral obligation and justice derived from ethics theories as predictors of the behavioral intentions of individuals to commit digital piracy even when ATT, SNs, and PBC have been taken into account. In the extended TPB model, Conner and Armitage [49] discovered that moral norms explained an additional 4% of the variation in intentions after controlling for TPB constructs. Chang and Chou [50] reported that the ethics perspective and three constructs of TPB explained consumers’ intention to “Bring Your Shopping Bags” when doing grocery shopping in Taiwan. Dowd and Burke [51] demonstrated that having a good moral attitude increased the TPB’s explanatory power dramatically.

Moreover, each of the three TPB constructs is influenced by ethical and moral considerations separately. Ethics was discovered to be linked to ATT constructively [52,53,54,55]. Honkanen et al. [46] showed that ethical values based on environmental and animal protection explained ATT. Sun et al. [56] reported that ethical belief significantly affected consumers’ ATT and intention to use plastic bags in China. On the other hand, Schwartz [57] proposed the norm activation theory, which argued that ethical and moral behavior was the consequence of a personal norm to behave in a certain manner. The moral norm is activated when individuals are aware of the repercussions of their decisions and able to accept responsibility for them. A normative ethical standard may serve as a foundation to shape personal normative beliefs.

In addition to that, Guido et al. [58] illustrated that moral norms or personal beliefs could be considered the main motivator of the purchasing intention of organic food products, and they were, in turn, affected by sns and product personality traits of naturalness and authenticity. Lastly, Park and Ha [59] revealed that personal norms (from the norm activation model) together with PBC influenced the U.S. consumers’ intention to recycle. As a result, the proposed analysis aimed to bridge the gap by combining ethics with all three TPB variables to predict PBI. Based on the findings presented above, this study suggests the following hypotheses:

Hypothesis 1 (H1)

. EE positively affects university employees’ ATT toward voluntary PEB.

Hypothesis 2 (H2).

EE positively affects university employees’ SNs.

Hypothesis 3 (H3).

EE positively affects university employees’ PBC.

Many empirical studies have found links between the three TPB constructs and PEI [60,61,62]. According to Ajzen [63], ATT, SNs, and PBC are predictors of behavioral intention. Lin et al. [64] demonstrated that while SNs and PBC affected citizens’ behavior intentions towards pro-environmental programs in Indonesia, ATT significantly impacted the citizen behavioral intention. Effendi et al. [65] revealed that the PEI of Indonesian students was influenced by SNs, PBC, and ATT. Mufidah et al. [66] also reported that ATT was the strongest predictor and SNs were the weakest predictor that fundamentally affected the consumers’ behavior intention in Taiwan and Indonesia. Based on the previous findings, this study contends that ATT, SNs, and PBC significantly influence the PEI of university employees:

Hypothesis 4 (H4).

ATT positively affects university employees’ PEI.

Hypothesis 5 (H5).

SNs positively affect university employees’ PEI.

Hypothesis 6 (H6).

PBC positively affects university employees’ PEI.

Masud et al. [60] demonstrated the mediating effects of the behavioral intention between ATT, SNs, PBC, and PEB of Malaysian people. Ateş [61] examined the factors influencing the PEB of 340 science teachers in Turkey using TPB and the value identity personal norm model. The results illustrated that ATT and PBC indirectly influenced PEB, while the indirect relationship between SN and PEB was insignificant. In India, Taufique and Vaithianathan [62] also used behavioral intention as a mediator to investigate the determinants of ecologically conscious consumer behavior among young urban consumers. The results suggested that ATT had a significant indirect positive influence on ecologically conscious consumer behavior. Based on previous empirical studies, the following three hypotheses are proposed in this study:

Hypothesis 7 (H7).

University employees’ PEI positively mediates the relationship between ATT and PEB.

Hypothesis 8 (H8).

University employees’ PEI positively mediates the relationship between SN and PEB.

Hypothesis 9 (H9).

University employees’ PEI positively mediates the relationship between PBC and PEB.

EI has been considered a predictor of PEB [67,68,69,70]. Numerous studies introduced EI as a mediator for actual behavior across different contexts and pointed out that EI mediated the relationship between emotional stability and optimism [71,72]. Guil et al. [72] used EI as a mediator of the relationship between self-esteem and anxiety among university students. Furthermore, Di Fabio et al. [71] demonstrated that EI mediated the relationship between personality traits and both hope and optimism. However, empirical studies on how EI impacts university employees’ PEB are lacking. As a result, the proposed framework aims to demonstrate that EI directly affects PEI, and consequently, indirectly affects PEB.

Hypothesis 10 (H10).

EI positively mediates the relationship between PEI and university employees’ PEB.

Figure 1 illustrates a proposed study model focused on the aforementioned hypotheses.

3. Materials and Methods

3.1. Measurements

The self-administered questionnaire was comprised of PEB [30,73], PEBI [74], ATT, PBC, SNs [7,74,75,76], EE [77,78,79], and EI [80], and it was split into five parts. The first section was based on the PEB of university employees, containing seven items, including conserving, avoiding harm, transforming, influencing others through behaviors that motivated green behavior, and taking initiatives. The second section was related to PBEI, with five items that included sample items such as, “I plan to purchase eco-friendly office supplies”. The third section was related to EE, with ten items that included sample items such as “my university has specific environmental policies in place”. The fourth section focused on EI, containing 16 items, including self-emotional appraisal, emotional appraisal, and use of emotions. The fourth section consisted of the TPB constructs, including five items on ATT toward PEB, five items on SNs, and four items on PBC. All items were measured on a 5-point Likert scale ranging from 1 (strongly disagree) to 5 (strongly agree). The last section focused on the background of the employees.

The measurement items were created specifically for this study’s context. Two researchers were asked to review the content validity of these items to determine if they were appropriate to evaluate the voluntary PEI of university staff. Additionally, a pilot study was carried out among Malaysian university employees. A few changes were made to the questionnaire’s terminology and composition as a result of the two procedures. The questionnaire was first created in English and then translated into Malay (Bahasa Melayu) by two bilingual scholars, taking into account the degree of language heterogeneity among respondents. Then, back-translations into English were performed by two separate Malay–English bilingual scholars to ensure the consistency and accuracy of the translations.

3.2. Data Collection

We selected six universities at random based on their success in championing the environment and sustainability on a local and international level, as well as their inclusion in the UI Green Matrix 2016 [81,82]. The six selected universities were Universiti Putra Malaysia, Universiti Utara Malaysia, University Malaya, Universiti Teknologi Malaysia, Universiti Kebangsaan Malaysia, and Universiti Tunku Abdul Rahman. In addition to that, employees who had served at one of the six universities for at least a year made up the target population. Fifty-four items were utilized and all factors (i.e., PEB (7 items), EPEI (5 items), ATT (7 items), SN (5 items), PBC (4 items), EE (10 items), and EI (16 items)) are displayed in

Table 1. Construct reliability and validity.

Constructs	Loadings	α	rho_A	CR	AVE	Constructs	Loadings	α	rho_A	CR	AVE
EE		0.946	0.987	0.953	0.608	1. I intend to reduce the amount of paper I use during my daily work activities.	0.708
1. My university has specific environmental policies in place.	0.807					2. I intend to reduce energy usage in the workplace.	0.856
2. My university has set aside funds for environmental protection.	0.754					3. I intend to avoid using plastic during my daily work activities.	0.755
3. My university integrates environmental programs, strategies, or objectives into its culture.	0.803					4. I intend to purchase office supplies that are not harmful to the environment.	0.856
4. All living things have a moral right to exist.	0.734					EI		0.847	0.856	0.898	0.688
5. The balance of nature is fragile and can easily be destroyed.	0.780					1. Most of the time, I have a good sense of why I have certain feelings.	0.892
6. The earth is like a spaceship with very limited space and resources.	0.797					2. I have a good understanding of my own emotions.	0.871
7. Nature is valuable for its own sake, even if humans acquire no goods or services from it.	0.764					3. I completely understand how I feel.	0.761
8. Breaking the laws of nature will eventually lead to catastrophe.	0.807					4. I am always aware of whether I am happy or not.	0.787
9. Humans can use nature as long as they do not go too far.	0.734					5. I always know my friends’ emotions based on their behaviors.	0.889
10. Human ingenuity ensures that we do not make the earth unlivable.	0.806					6. I am an excellent observer of other people’s emotions.	0.877
AT		0.860	0.878	0.898	0.629	7. I am sensitive to the feelings and emotions of others.	0.801
1. It important to keep the environment clean with green office supplies.	0.719					8. I have a good understanding of the emotions of the people around me.	0.822
2. It is important to prevent pollution in the workplace.	0.836					9. I always set goals for myself and then work hard to accomplish them.	0.827
3. It is important to conserve natural resources in the workplace.	0.815					10. I always tell myself that I am a capable person.	0.885
4. It is important to use environmentally friendly products in the workplace.	0.782					11. I have enough self-motivation to continue.	0.776
7. It is important to reduce harm to the environment in the workplace.	0.838					12. I always tell myself that I should try my best.	0.818
SN		0.867	0.885	0.904	0.654	13. I can manage my anger and deal with problems rationally.	0.712
1. People who are important to me believe that I should reduce my use of paper in the workplace.	0.802					14. I am fully capable of managing my own emotions.	0.870
2. People who are important to me believe that I should conserve energy in the workplace.	0.752					15. When I am angry, I can always cool down quickly.	0.774
3. People who are important to me believe that I should reduce my plastic waste in the workplace.	0.803					16. I have good control of my own emotions.	0.833
4. People who are important to me think I should buy environmentally friendly products in the workplace.	0.836					PEB		0.925	0.932	0.940	0.690
5. People who are important to me think I should conserve resources in the workplace.	0.846					1. Whenever possible, I print in double-sided format at my office.	0.835
PBC		0.881	0.896	0.918	0.737	2. Every time I leave my office, I turn off my computer or notepad.	0.855
1. I have control over performing pro-environment activities in the workplace.	0.820					3. I switch off the lights when no one else is at the office.	0.879
2. I can perform environmental pro-environment activities in the workplace.	0.888					4. At work, I have my own cup that I use for drinking.	0.823
3. I have control over my actions to support the environment in the workplace.	0.849					5. At work, I try to avoid using plastic bags.	0.839
4. I have the ability to carry out pro-environment activities in the workplace.	0.876					6. I am reducing my water consumption at work.	0.832
PEI		0.804	0.834	0.873	0.634	7. In my workplace, I engage in environmentally friendly programs.	0.745

Note. Environmental ethics = EE; attitude = AT; subjective norm = SN; perceived behavioral control = PBC; pro-environmental intention = PEI; emotional intelligence = EI; pro-environmental behavior = PEB.

. The sample size of 378 met the requirements to use structural equation modeling (SEM), as the ratio of sample size to estimated parameters should be at least 7:1 [83]. According to Hair et al. [81], the preferred minimum number of indicators is three. However, in most cases, five to seven indicators are commonly used to represent a construct.

Data collection took place during November and December 2018. A total of 378 questionnaires were distributed using a stratified random sampling technique at these six universities located in Malaysia. Of those, 76 incomplete questionnaires were eliminated, resulting in a final sample of 302 and yielding a valid response rate of 79.89%. Descriptive statistics was used to describe the demographics of the respondents. Partial least squares structural equation modeling (PLS-SEM) was used to approximate the model fit and evaluate the proposed hypotheses and was chosen because Hair et al. [81] advocated for its suitability to test models ranging from simple to complex as well as small to medium sample sizes.

4. Results

4.1. Profile of Respondents

The demographic profile of the selected sample populations included work experience, gender, age, monthly salary, highest educational degree, and ethnicity. Among these respondents, 187 (61.9%) were female and 115 (38.1%) were male. The sample was spread out among all age groups: ages below 18–25 years had one respondent (0.3%), ages below 26–30 years had 24 respondents (7.9%), ages below 31 to 40 years had 121 respondents (40.1%), ages below 41 to 50 years had 124 respondents (41.1%), and 32 respondents (10.6%) were 50 years old and above. Regarding the qualification, the majority of the respondents held a Ph.D. degree (128 respondents; 42.4%), 61 respondents (20.2%) held a master’s degree, 61 respondents (20.2%) had a bachelor’s degree, and 51 respondents (16.9%) had a diploma degree.

Based on the respondents’ experience levels, the majority of them had modest experience. A total of 87 respondents (28.8%) had 1–5 years of experience, followed by 86 respondents (28.5%) who had 6–10 years of experience, 91 respondents (30.1%) having 11–15 years, and 38 respondents (12.6%) having above 15 years of job experience. With regards to race, the respondents were Malay (79.8%, n = 241), Chinese (15.3%, n = 46), and Indian (5%, n = 15).

4.2. Measurement Model

The measurement model was tested using exploratory factor analysis (EFA) and PLS-SEM confirmatory factor analysis (CFA). To obtain the factor structure, all 54 items were subjected to EFA. In the initial stage of EFA, principal components analysis (PCA) with varimax rotation was employed. The Kaiser–Meyer–Olkin Measure of Sampling Adequacy (KMO) value of 0.79 and the results of Bartlett’s test of sphericity (χ2 = 1755.95; p = 0.000) indicated that the data were suitable for EFA. The number of factors was determined using the Kaiser criterion, with eigenvalues larger than one and only items with loadings greater than 0.60 retained. As a result, three items with factor loadings lower than 0.6 (PEI 5, ATT5, and ATT6) were eliminated. The seven factors with the remaining 51 items explained 66.9% of the total variance. The EFA results showed the corresponding eigenvalues for EI (7.78), EE (4.10), ATT (2.37), PEB (1.87), SN (1.61), EPEI (1.35), and PBC (1.05), while the corresponding portions of variance explained were 25.94, 13.66, 7.74, 6.22, 5.33, 4.49, and 3.49, respectively.

PLS-SEM results suggested that the CFA model adequately fit the data (SRMR = 0.067; NFI = 0.908; Hair et al. [81]). Significant correlations were found between the factors when the latent variable correlations were examined (p < 0.01). All factor loadings were above 0.70 and all extracted average variance (AVE) values were greater than 0.5, indicating convergent validity (Table 1). Furthermore, all composite reliability (CR) values ranged between 0.873 and 0.953, indicating internal consistency. Finally, the Dijkstra–Henseler Rho_A values for all constructs were higher than 0.7 [81] (Table 1).

The variance inflation factors (VIF) were used in this analysis to investigate multi-collinearity. A VIF greater than 5 implies multicollinearity. The VIF values in this analysis were lower than the standard parameters, suggesting that there was no multicollinearity issue (see

Table 2. Multi-collinearity test for exogenous latent constructs.

Construct	1	2	3	4	5
1.ATT
2.VPEB					1.947
3.VPEI	1.285	1.947
4.EE	1.000
5.EI			1.000
6.PBC			2.201	1.000
7.SN			1.752	1.000

Note. Environmental ethics = EE; attitude = AT; subjective norm = SN; perceived behavioral control = PBC; pro-environmental intention = EPEI; emotional intelligence = EI; pro-environmental behavior = PEB.

) [84].

Discriminant validity compares the degree of one construct to another using empirical standards. As proposed by researchers to combine several methods, this study applied the Fornell–Larcker criterion with the Heterotrait–Monotrait (HTMT) ratio of correlations [84]. Following the Fornell–Larcker criterion, the results indicated that the discriminant validity was achieved because the square root of the AVE of each construct was higher than the correlation values among any construct pairings. Furthermore, the HTMT values were below the threshold level of 0.85 in all cases, as shown in

Table 3. Measurement model: discriminant validity.

	Fornell–Larcker Criterion							Heterotrait–Monotrait Ratio (HTMT)
Construct	1	2	3	4	5	6	9	1	2	3	4	5	6
1. ATT	0.799
2. PEB	0.380	0.831						0.395
3. PEI	0.454	0.379	0.796					0.539	0.429
4. EE	0.415	0.553	0.265	0.780				0.423	0.595	0.298
5. EI	0.472	0.365	0.602	0.293	0.848			0.539	0.397	0.716	0.318
6. PBC	0.438	0.243	0.480	0.269	0.492	0.859		0.502	0.261	0.565	0.291	0.558
7. SN	0.437	0.242	0.452	0.211	0.502	0.726	0.808	0.507	0.268	0.535	0.231	0.577	0.530

Note. ATT = attitude; EI = emotional Intelligence; VPEB = pro-environmental behavior; EE = environmental ethic; PBC = perceived behavioral control; SN = subjective norm.

. Consequently, these results confirmed the discriminant validity of the study.

4.3. Structural Model

The R², Q², and significance of paths were used to evaluate the structural model.

Table 4. Testing direct relationships.

	Path Coefficient	SD	t Value (Bootstrap)	p Values	BI (2.5%; 97.5%)
H1: EE→ATT	0.415	0.056	7.382	0.000	[0.303;0.528]
H2: EE→SN	0.211	0.066	3.169	0.002	[0.085;0.338]
H3: EE→PBC	0.269	0.063	4.265	0.000	[0.147;0.389]
H4: ATT→PEI	0.280	0.057	4.865	0.000	[0.170;0.397]
H5: SN→PEI	0.148	0.087	1.701	0.089	[−0.014;0.321]
H6: PBC→PEI	0.249	0.083	2.986	0.003	[0.091;0.408]
R2 ATT = 0.172	Q2 ATT = 0.089
R2 SN = 0.144	Q2 SN = 0.026
R2 PBC = 0.172	Q2 PBC = 0.049
R2 PEI = 0.314	Q2 PEI = 0.183
R2 PEB = 0.191	Q2 PEB = 0.117

Note: BI, bias-corrected confidence interval; ATT, attitude; PBC, perceived behavioral control; SN, subjective norm; PEI, pro-environmental intention; PEB, pro-environmental behavior.

displays that all R² values are greater than 0.1. As a result, predictive capacity has been developed. The Q² also confirms the endogenous constructs’ predictive validity. A Q2 value greater than 0 indicates that the model is predictively relevant. The findings demonstrate that there is significance in the prediction of the constructs (see Table 3). In addition to that, the results displayed that the model fit the data, as the standardized root mean square residual value was 0.122, which was significantly higher than the required value of 0.10.

To determine the significance (path coefficient) of relationships between the variables (see Table 4), the bootstrapping technique was utilized [81]. As shown in Table 4, EE significantly impacted ATT (β = 0.415, t = 7.382, p = 0.000). Hence, H₁ was supported. The results revealed that EE significantly impacted SN (β = 0.211, t = 3.169, p < 0.002) and PBC (β = 0.269, t = 4.265, p < 0.001), supporting H₂ and H₃. The findings showed a significant impact of ATT (β = 0.280, t = 4.865, p = 0.000), insignificant impact of SN (β = 0.148, t = 1.701, p = 0.089), and significant impact of PBC (β = 0.249, t = 2.986, p = 0.003) on PEI, supporting H₄ and H₆ and failing to support H₅.

4.4. Mediation Analysis

To determine the mediating roles of PEI and EI, a mediation study was conducted. The results (see

Table 5. Mediation analysis.

	Total Effect		Direct Effect		Indirect Effects
Paths	Coefficient	p Value	Coefficient	p Value		Coefficient	SD	t Value (Bootstrap)	p Values	BI [2.5%;97.5%]
ATT-PEB	0.280	0.003	0.198	0.022	H7: ATT→PEI→PEB	0.106	0.033	3.194	0.001	[0.052;0.178]
SN-PEB	0.148	0.065	0.114	0.087	H8: SN→PEI→PEB	0.056	0.036	1.553	0.121	[−0.006;0.132]
PBC-PEB	0.249	0.006	0.178	0.016	H9: PBC→PEI→PEB	0.094	0.035	2.736	0.006	[0.034;0.165]
PEI-PEB	0.234	0.008	0.106	0.093	H10: PEI→EI→PEB	0.212	0.046	4.578	0.000	[0.122;0.305]

Note: BI, bias-corrected confidence interval; ATT, attitude; PBC, perceived behavioral control; SN, subjective norm; PEI, pro-environmental intention; PEB, pro-environmental behavior.

) showed partially significant mediating role of PEI in the relationship between ATT (H₇: β = 0.106, t = 3.194, p = 0.001) and PBC (H₉: β = 0.094, t = 2.736, p = 0.006), and PEB. Besides that, the results (See Table 5) revealed that PEI did not mediate the relationships between SN and PEB (H₈: β = 0.056, t = 1.553, p = 0.121). EI (H₁₀: β = 0.212, t = 4.578, p = 0.00) fully mediated the relation between PEI and PEB. All results are exhibited in Figure 2.

5. Discussion and Implications

The aim of this study was to add the variables of EI and EE to TPB to examine the factors influencing the PEB of university employees in Malaysia. The study provides a deeper explanation of the relations between EE, ATT, SN, PBC, and EI and university employees’ intentions towards PEB. The results of this study proved that EE significantly influenced ATT toward the PEB, which in turn became the most significant predictor of the PEBI of university employees. The findings are consistent with previous studies that indicate significant relations between value and ethics, ATT, and intention [85,86]. Another significant finding was that EE directly affected respondents’ PBC and SN, which have not been evaluated in the context of PEB in the workplace. As a result, university administrators must be aware of EE, rethink their ethical standards to employees, and assist them in the development of ethical decisions, since EE has been found to significantly impact ATT, PBC, and SNs, affecting university employees’ intention to participate in pro-environmental activities on campus. Besides that, unless universities punish employees for wasting resources, employees might use energy in an inefficient manner because they have no financial obligation to do so, whether or not they ethically believe in sustainability or follow pro-environmental values. Employees can further save energy and participate in any activities relevant to sustainability if their ethical behavior clearly forbids unsustainable behavior, and if their EE promises punishment for unsustainable behavior.

The findings also confirmed that university employees’ ATT directly impacts their intention for PEB. This is in line with previous studies by Bauer et al. [87] and Yusliza et al. [88]. Anthony Swaim et al. [74] reported that employee personal attitude was critical to generate a desirable level of environmental sustainability intention. Hence, any organization needs to ensure that the attitude of their employees is in line with the company’s objective in promoting organizational sustainability. This is because an employee who is in favor of certain behavior or action will put effort into acting. In the context of this study, the employee who is in favor of protecting the environment in their workplace will make an effort to conserve, work sustainably, and avoid harm when preserving the environment in the workplace.

However, some other studies [7,30,74,89] demonstrated that SNs did not significantly and directly affect PEBI among university employees in Malaysia. According to the result of this study, the assumption of social pressure from others for the employee to behave in a certain manner and their motivation to comply with those people’s views to perform a certain behavior or action will not lead to their effort of acting. Mei et al. [90] believe that this situation is due to the declining trend of the Environmental Performance Index Ranking of Malaysia from 51 in the year 2014 to 75 in the year 2018, showing that Malaysia was still lacking a comprehensive method to present the environmental awareness and behavior performance of Malaysians. Since Malaysian environmental awareness and behavior did not positively correlate [90], the universities’ employees who were surrounded by others (superiors, subordinates, and colleagues) who were not acting to protect the environment would influence the former to not to make an effort to conserve, work sustainably, and avoid harm when preserving the environment in the workplace.

Apart from that, PBC directly impacted PEBI. This result is consistent with a previous study [73] that suggests that the increased levels of PBC lead to increased PEBI. Consequently, the employees’ belief in personal control to perform a specific behavior or action will influence their effort to proceed with that action. According to the findings of this study, as employees increase their finances and morale, such as obtaining enough money to purchase renewable goods and being able to minimize paper, energy, and plastic use, they increase their commitment to environmental conservation, increasing their desire to preserve the environment. As a consequence, university administration must plan and implement sustainability initiatives to strengthen employees’ ATT and personal norms, as well as eliminate potential obstacles to ensure pro-environmental lifestyles. Furthermore, the findings of this study also confirmed the mediating role of PEBI, which significantly and positively affected the relationship between ATT and PBC. This result is consistent with other studies [30,40,74]. Therefore, the intention of an employee to perform a specific behavior is stronger when the degree of ATT and perceived ease towards PEB are high. Lastly, this study addressed the numerous constraints and limitations that left room for conceptual and operational improvement in TPB [91]. Existing studies [92,93] revealed that strong intention did not sufficiently lead towards a given behavior, suggesting that adding more constructs would increase the predictability between intention and behavior in TPB constructs [94,95]. Parallel to this, this study emphasized the significant contribution of EI as the mediator between PEBI and PEB. This result is in line with the previous studies [14,96] that suggested that decision-making models such as the TPB should subsume emotions and take greater consideration of the factors that facilitated the intentions.

The employees with elevated levels of EI had more rewarding interpersonal interactions with their coworkers and exhibited more prosocial attitudes and behavior. In other words, the employee can use their intellect and emotional control to assess those environmental issues and decide how to respond to them. As a result, a university employee with a high environmental awareness will use their EI to make environmental protection more desirable, continue with their intention to protect the environment, and contribute to PEB in the workplace. By these particular assumptions, the researcher believes that the anticipated function of EI will contribute to PEB literature by examining other associated variables or antecedents that may reinforce the relationship within the TPB model’s constructs. Thus, the university’s administration would rely on increasing the EI of their employees to meet their environmental sustainability initiatives. This can be accomplished by training, motivational activities, and management–employee engagement. In the future, the management will consider recruiting an employee with a high level of EI so that they can use their ability to regulate their emotions to provide a positive feeling in protecting the environment in their workplace.

6. Conclusions

In conclusion, the study demonstrated that the extended TPB model was effective to study the predictors of the PEB of university employees. According to the findings, EI and EE increase the predictive validity of the TPB model on PEB. The two constructs included in the model also help to clarify their interrelationships, which were critical for advancing the TPB paradigm in the PEI domain. The introduction of the EE and EI constructs allows this to be the first known study that extends the application of TPB for predicting the PEB of university employees in the Malaysian context. Furthermore, the results point to three outcomes. Firstly, the results contributed to the pro-environmental literature by providing additional empirical support for EE in predicting the three constructs of TPB (i.e., ATT, SN, and PBC), along with the significant effects of ATT and PBC in predicting PEI. Secondly, contrary to the results of direct effect, the results display a partially significant mediating role of EI in the relationship between ATT, PBC, and PEB. Thirdly, this study uncovered that EI mediated the effects of PEI on PEB. Overall, the findings offer important insights for policy-makers and university managers in developing countries such as Malaysia, which can assist in better understanding university employees’ PEB and further emphasize the vital role of this context in the TPB.

7. Limitations and Future Research Directions

There were certain limitations to the study, which could be addressed in future research endeavors. Initially, the main variables were assessed using self-report, and part of the covariance between predictor and outcome variables could be attributed to common method variance. As a result, one effective approach is to perform longitudinal research that can demonstrate the causality of the relationships examined in this research. Next, the study did not take into account how much time the university had invested, which was critical in the long-term improvements of employee behavior. As a result, researchers may undertake additional researches to investigate the connection between university sustainability policies and activities and behavioral improvements in university staff toward an equitable and sustainable society.

A further significant limitation is that this research was limited to only one nation, Malaysia. Physical, legislative, and cultural differences can restrict or enhance the organization’s efforts to establish environmental policies and achieve sustainability-related objectives. Furthermore, these cultural contexts can influence how university staff interpret and identify university policies and human resource management (HRM) practices. While this study framed TPB constructs, it also included variables such as ethics to foster PEB in university employees. The authors did not expand the study to determine whether there was an association between particular HRM practices and the degree to which university staff were inclined to participate in PEB. As a result, future studies should determine if there is a connection between individual HRM activities and the degree to which university staff participate in PEB. Future studies should look into this relationship because it will help universities choose where to focus their efforts to maximize their positive impacts on the environment.