4.1. Profile of Respondents
Five months of conducting the online survey resulted in 121 total responses (response rate: 91%) (see
Table 1. Due to the challenges the research team encountered in penetrating the management circle and obtaining the relevant respondent contact information, we concentrated on the management level in this study, where the person in charge is accountable for making operational decisions. As a result, the survey takes longer than expected to complete and receive the returned questionnaires. Many follow-ups must be carried out while taking into account their hectic timetable and activity.
Table 1 displays the demographic data of respondents, which includes gender, age, race, degree of education, and monthly income in addition to some generic information.
Male respondents make up more than 96% of the sample, and the majority (33.88%) of them are between the ages of 41 and 50. The respondents are a mix of races, including Malay (34.71%), Chinese (27.27%), Indian (12.40%), and others (25.62%). Another important fact is that 42.98% of respondents, or those who completed the survey, had at least a high school graduation. With 28.10%, SPM graduates are the second-highest group, followed by degree holders (27.27%) and master’s degree holders (1.65%). Significantly, this sample may be able to be generalised into the adaptation of sustainable waste management behaviours within the oil palm industry in the area of Sabah, Malaysia.
4.2. Construct Reliability and Validity
The content validity, convergent validity, and divergent validity of the construct measurements were examined. Before releasing the questionnaire, the Malaysian Palm Oil Board (MPOB)’s customer service division helped us assess the potential list of management levels. These are the individuals who are close to the respondents and have an enormous amount of managerial expertise. The questionnaire was altered multiple times on their expert advice.
The convergent validity test was used to determine whether a latent variable is adequately explained by its observable variables. The standardised loading coefficient, Cronbach’s alpha, composite reliability, and average variance extracted are used to assess construct reliability and convergent validity [
89]. If the standardised loading coefficients of the items on the related constructs are greater than 0.7, measurement scales are considered to have strong convergent validity [
90]. As a result, the observed items with standardised loading coefficients below 0.7 were removed from the estimation model.
Table 2 shows the standardised loading coefficients for each item, demonstrating that all conditions were met. The lowest recorded standardised loading coefficient was 0.882, and the maximum was as high as 0.973.
As shown in
Table 2, all six of the constructions simultaneously produced Cronbach’s alpha values that ranged from 0.967 to 0.981, which is much higher than the required value of 0.7. As stated by [
91], a coefficient greater than 0.8 denotes a high degree of construct dependability and internal consistency. Additionally, all CR values—which range from 0.972 to 0.984—exceed the acceptable cutoff of 0.7. These findings suggest that the constructs have strong convergent validity. Additionally, the AVE indications are higher than the 0.5 criterion, ranging from 0.833 to 0.913. Thus, the measuring scales for the constructs have excellent overall internal consistency, reliability, and convergent validity.
Next, discriminant validity was investigated using the Fornell–Larcker criterion. As claimed by [
92], the square roots of AVE for all latent constructs must be larger than the correlations between any other constructs.
Table 3 reports the square roots of the AVE, shown in the diagonal and in bold, and the off-diagonal elements demonstrate how the constructions are intercorrelated. In every instance, it seems that the on-diagonal AVE square roots are far higher than the inter-construct correlations. This suggests that the model has met the requirements for discriminant validity.
The heterotrait–monotrait ratio (HTMT) of correlations was evaluated to confirm the discriminant validity of the reflective parts of this investigation. This new criterion is considered to be the most difficult evaluation of discriminant validity. The HTMT technique reveals the genuine correlation between two completely reliable latent variables, and a value higher than 0.90 shows a lack of discriminant validity [
83]. However, the confidence intervals of the HTMT must not include the number 1. The HTMT results for this study are presented in
Table 4, which demonstrates that all of the ratios were less than 1, which satisfied the criteria for establishing discriminant validity, and that the measures’ discriminant validity was satisfactory. If the value is less than 1, it is obvious that the two elements are distinct. Therefore, the lower they are, the more divergent they are. All cross-loading correlations are smaller than 1 according to
Table 4. As a result, the discriminant validity is once again strengthened.
As a methodical approach for determining the presence of collinearity and the common method bias, the complete collinearity test developed by [
93] was advised by [
85]. Since all inner VIFs among the latent constructs range from 1.000 to 3.090 and are below the threshold of 3.3, the whole collinearity test produced satisfying VIFs (
Table 5). As a result, the research model is free of usual technique bias.
4.3. Structural Equation Model
The hypothesised structural links between attitude, subjective norms, perceived behavioural control, behavioural intention, and sustainable waste management practises were examined by estimating the structural model after the construct reliability and validity were established (
Figure 3). The model’s predictability was evaluated in terms of the determination coefficient with reference to the R2 value for the dependent latent constructs. Nevertheless, [
80] suggest that R2 should have a minimum value of 10%. According to
Figure 3, the behavioural intention (BI) construct’s R2 values show that up to 91.4% of the construct variability was explained by the attention (ATT), subjective norms (SN), and perceived behavioural control (PBC) constructs. The model’s explanatory factors had explained 89.6% of the variability in sustainable waste management behaviour (SWMB). These parameters define an acceptable level of predictability.
Five of the structural model’s hypotheses were supported (at
p value 0.1) based on the standardised path coefficients,
t-value, and
p-value displayed in
Table 6, however, two other hypotheses were not. Subjective norms, perceived behavioural control, and behavioural intention are all statistically significant positive predictors of sustainable waste management behaviour (H2, H3, and H7), as seen in
Table 6. Concurrently, subjective norms and perceived behavioural control both support the development of behavioural intention (H5 and H6). H1 (=0.034,
t = 0.604,
p = 0.546) and H4 (=−0.025,
t = 0.428,
p = 0.669) are two insignificant hypotheses. The findings show that attitudes have no influence on sustainable waste management behaviours, nor do they affect behavioural intentions.
R2 basically shows how much the variability in the dependent variable can be explained by the predictive variables. The effect size f2 can be quite helpful in determining how much the dependent variable might change if an external variable is removed in this way. The dependent variable in this study is sustainable waste management practises. However, in relation to attitude, subjective norms, and perceived behavioural control, behavioural intention has evolved into an acting dependent variable. In order to determine how much the external variables are affecting the acting dependent variable (behavioural intention), the f2 was also calculated. The F-square is used to calculate effect size (small => 0.02, medium => 0.15, large => 0.35) [
86]. The impact of attitude, subjective norms, and perceived behavioural control on sustainable waste management behaviours has a tiny effect size that is less than 0.02 (f2 = 0.007, f2 = 0.026, and f2 = 0.085, respectively), as indicated in
Table 6 and advised by [
87]. As opposed to this, the influence of attitude, subjective norms, and perceived behavioural control on behavioural intention likewise has a minor effect size (f2 = 0.005, f2 = 0.025, and f2 = 0.104, respectively). It is important to acknowledge that behavioural intention towards sustainable waste management activity has a significant effect size, measuring 4.047 (greater than 0.35).
4.4. Discussion
This study’s findings firstly support the hypothesis that subjective norms (H2) and perceived behavioural control (H3) have a positive effect on behaviour adaption for sustainable waste management in oil palm mill plantations in the Sabah region. The finding that subjective norms had a large impact (H2) is consistent with the study by [
94], who looked at how households manage their garbage. H2 implies that their beliefs are supported by social groups or other companies that work with them [
58]. This demonstrates how external influences including the government, clients, stakeholders, shareholders, as well as the general public and partners, have impacted the palm oil mills to adopt sustainable waste management practises. This result is in line with the practice that has been exercised by estates as reported by [
95], which is that all estates host stakeholder meetings every year, inviting the local population, government agencies, small farmers, and labour unions to discuss their efforts to promote sustainability and to provide a forum for the stakeholders to voice their concerns. In another study that was conducted by [
96] regarding the adoption of sustainable practice adoption amongst smallholders of oil palm farmers in Thailand, the farmers disclosed that the group leader and members influence other farmers to join the group, attend the meetings, and take the trainings necessary to become adopters. This outcome is in line with the findings of [
97], who discovered the significance of individual or group opinions in decision making.
The findings of a significant result on perceived behavioural control towards sustainable waste management behaviour (H3) are in agreement with [
98] study examining the influence of perceived behavioural control and other influencing factors on waste segregation-at-source activity. H3, on the other hand, contrasted with a study by [
99] out on Malaysian paddy farmers, who were examined for their adoption behaviour and intention towards sustainable agriculture. Both results for perceived behavioural control in affecting sustainable waste management behaviour and intention were found to be significant. This study’s findings are not surprising, considering that a significant amount of earlier research has found that an individual’s behaviour when it comes to reducing waste at work is solely determined by their subjective perception of perceived behavioural control [
61]. This demonstrates that top management at the company and the company itself made the decision to implement sustainability at the mill. This is partly in line with the findings of [
97], who discovered that farmers’ intentions to adopt innovations relating to sustainable practice were influenced by their perceptions of having adequate resources which boost their confidence in adoption. A study conducted by [
100], has corroborated this finding. It revealed that smallholder planters who perceive the collection of oil palm residue as easy and possess confidence in their ability to do so are more likely to have the intention to engage in sustainable management practises and, consequently, have a tendency to supply oil palm residue.
This study’s findings support the favourable influence of subjective norms (H5) and perceived behavioural control (H6) as important factors that can affect behavioural intention. The existence of positive responses that the respondents provide has effects on particular persons; hence, good intentions are built up, which explains the positive impact of subjective norms on behavioural intentions. This H5 finding is in agreement with works by [
3,
46,
66,
101], and goes against [
102] study, which represents people as rebellious due to negative feedback to other people’s influences, which causes people to antagonise their influences. In conclusion, it is abundantly obvious that the external factors including the stakeholder for instance are crucial for developing the behaviours of reuse, reduce, and recycling. The important finding of perceived behavioural control (H6) supports the study by [
101], which examined the behavioural intention to use environmentally conscious food packaging. However, it has been found that when it comes to reducing waste at work, an individual’s behaviour intention is solely determined by their perception of behavioural control [
61]. This study demonstrates that for behaviour interventions in waste reduction, an impression of workforce empowerment is essential. When it comes to implementing linked empirical programmes and maintaining environmental sustainability, organisations must also try to provide their staff with a sense of power and flexibility [
20,
62]. Therefore, we believe that a major factor affecting the decision to implement sustainability in oil palm mills is the company and top management. Additionally, attention towards top management and stakeholders should be prioritised. Greater emphasis should be placed on the boards of directors and other key stakeholders in waste management surveys, since they hold significant influence in establishing waste management policies and practises inside organisations. The study on the complex issue of waste management reveals that the actions and perspectives of stakeholders have a substantial influence on the results of waste management [
103]. The boards of directors possess the authority to shape waste management practises by virtue of their supervisory and decision-making responsibilities [
104]. They have the ability to establish waste reduction objectives, assign resources for waste management, and ensure adherence to waste management regulations. Furthermore, the involvement and active participation of stakeholders in waste management governance may improve the efficiency and durability of waste management practises [
105]. It is important to incorporate sustainability into governance and management systems and this demonstrates the significance of board structure, composition, and diversity in fostering sustainability [
106].
Attitudes have led to negligible findings for both behavioural intention (H4) and sustainable waste management behaviour (H1). According to [
3,
36,
37,
107,
108], the attitude towards sustainable waste management behaviour (H1) reported in this study had a negligible impact. One explanation could be that, in the context of evaluating behaviours related to the management of solid waste, attention is typically a key predictor. While oil palm plantations in Sabah appear to be adopting sustainable behaviours due to enforcement and influence from senior management, government, consumers, stakeholders, shareholders, public community, and company’s collaborators, this may be due to the fact that these groups have been enforcing these behaviours. This claim has been reinforced by subjective norms and perceived behavioural control, both of which have had a considerable influence on behaviours and intentions related to sustainable waste management. The fact that PBC is significant means that the top management, including the board of directors and the company itself, which play a significant role in the decision making of sustainable practices in the mills and plantations, has influenced the implementation of sustainable behaviours within palm oil mills in Sabah. As claimed by [
109], the board of directors is made up of a number of individuals with different backgrounds in expertise who work together to coordinate the firm’s operations. Therefore, oil palm plantation mills should concentrate on policies that can improve employee attitudes towards waste reduction and on SWMB implementation tactics. Positive workplace attitudes typically increase the likelihood of successfully implementing SWMB.
According to the findings of this study, the influence of attitude on behavioural intention (H4) was insignificant. The study by [
54] also found that although the relationship between attitude and behavioural intention is significant, it is the weakest of the relationships that were examined. These other relationships included those between attitude, subjective norms, and perceived behavioural control towards sustainable waste management behaviour. The relationship between behavioural intention and perceived behavioural control has the greatest strength, followed by that between behavioural intention and subjective norm. However, this discovery almost implies the outcome indicated by H4. The results from [
110], where attitude exerted the second-lowest relationship strength towards behavioural intention, are almost in agreement with this. This supports the earlier finding that although people have positive opinions, they do not practice waste separation, while perceived behavioural control was the one that showed the highest relationship in that study. As explained by [
111], who have a different perspective on attitude, it is frequently challenging to determine whether these factors have any influence on waste management because people’s attitudes and behaviours are highly subjective and unpredictable. This is why there are conflicting results regarding attitude.
Additionally, H7 revealed that behavioural intention significantly affected behaviours related to sustainable waste management. As demonstrated by [
112], positive attitudes and positive normative issues result in higher intentions to implement sustainable activities. The stated result (H7) validates these findings. This showed that human beings are expected to act on their intentions when opportunities are available. Negative attitudes and a lack of concern for normative issues both seem to have a role in limiting the willingness of individuals to adapt. Another study that strengthens the result of H7 was conducted by [
113] who found that household behavioural intention has a large impact on how much people recycle, reuse, and reduce in their homes. The partial least squares approach was used to analyse the data, which were gathered from 670 Malaysian households. H7 was further supported by [
114], who discovered that behavioural intention significantly affects how fruit and vegetable waste (FVW) is managed. The results of further research addressed the influence of behavioural intentions. As stated by [
66], the theory also takes into account the behaviour-changing effects of purpose. Meanwhile, [
115], indicate evidence that behavioural intentions affect waste reduction behaviour by assessing the attitudes and behaviours of 491 workers at a construction site in Shenzhen, China. [
116] studies citizens’ recycling behaviour using the TPB and finds that citizens’ behavioural intentions affect their recycling behaviour using a sample of 2004 respondents in South African metropolitan areas. [
19] indicate that the proof that behavioural intentions have a positive effect on reducing waste and obtain insight into consumers’ behaviour towards food waste. Additionally, the importance of this behavioural intention finding for sustainable waste management behaviours has demonstrated that company intentions might affect how sustainability practices are implemented in the mills.