3.1. Model Conceptualization and Hypothesis Development
NAM has been widely used by researchers to study individual’s pro-social behavior. It postulates that one’s intention or behavior is influenced by moral norms, which are also referred to as personal norms [
55]. In this theory, norm activation begins with an individual’s awareness of detrimental consequences of the status quo and their ascription of responsibility for not acting pro-socially. This awareness activates a personal norm that determines whether that individual should act in a certain way to prevent or mitigate that harmful consequence [
67]. From a dust pollution perspective, a construction worker’s awareness of its consequence and ascription of responsibility for not resorting to control measures should trigger personal norms that guide the worker’s intention or behavior regarding pollution control. In this model, awareness of consequences deals with whether the worker is aware of the consequences to the fellow workers and the neighboring community when not implementing a specific action to control it [
68]. Ascription of responsibility refers to the feeling of responsibility for not taking action to control pollution [
68]. Personal norms indicate a moral obligation to perform dust pollution control measures during construction operations. In this context, personal norms are referred to as feelings of “moral obligation to perform or refrain from specific actions” [
55], p. 191.
NAM has been applied in different sectors with a view of initiating changes in people’s intention or behavior. One such very popular application is the energy saving behavior of people. Van der Werff and Steg [
26] used NAM to predict factors influencing household energy saving behavior, investigating what motivates people to reduce energy consumption even if that means sacrificing self-comfort. Zhang et al. [
54] used it to understand what motivates an employee’s energy saving behavior in organizations. Unlike in households, there is no financial incentive for employees to save energy in their workplace; they do it purely on altruistic motives, which are heavily underpinned by their personal norms [
47]. A feeling of moral obligation is found to trigger this pro-environmental behavior, which is strongly influenced by the organization’s environmental climate. The findings of Zhang et al. [
54] are very useful for the current study as it deals primarily with worker behavior and also investigates the role of the company in shaping such behavior.
A recent study by Song et al. [
69], which is somewhat related to the present study, investigated the role of China’s notorious haze pollution on people’s energy saving behavior. The undelaying assumption that the purchasing behavior of energy saving appliances could help conserve energy as well as reduce emissions was tested using NAM. The study found strong evidence that personal norms could be used to alter people’s behavior, and the government should be using social networks and social media to propagate the current environmental conditions and the negative consequences of using non-energy saving appliances and its impact on haze pollution. The works of Steg and De Groot [
70] and Shi et al. [
71] are of interest to the current study as their focus was on the willingness of car users to take action to reduce particulate emissions.
Based on the extensive NAM literature, a conceptual model, as shown in
Figure 1, was developed to operationalize this research. It is intended to explain employees’ dust pollution control behavior in construction organizations. According to NAM, an individual’s pro-social behavior is positively influenced by one’s personal norm [
54], hence, we postulate the following hypothesis regarding dust pollution control behavior of construction workers of an organization.
Hypothesis 1. Personal Norm significantly and positively affects the environmental behavior of employees.
In addition to personal norms, there is evidence that an organization’s environmental practices could influence its workers’ environmental behavior [
72]. Whether the construction company has adopted dust control measures in the past and maintains an environmentally conducive work environment will have a bearing on employees’ individual dust control behavior [
3]. Employees of civil engineering projects felt the preparedness and company’s commitment to dust pollution control more than that of building companies, which in turn helps form a positive behavioral intention among managers of civil construction companies towards pollution control compared to those from building companies [
6]. Zhang et al. [
54] found strong evidence to suggest a company can influence its workers’ energy saving behavior in the workplace. It was motivated by a common goal and a sense of moral obligation, which were not based on personal gains [
73]. The construction company’s environmental conduct and credentials were included as a predictor of workers’ behavior in this study. Hence, the second hypothesis of this research is based on the premise that a company’s behavior could influence the workers’ individual behavior as follows.
Hypothesis 2. A company’s behavior significantly and positively affects the environmental behavior of employees.
According to the social exchange theory, when the construction company shows its sense of responsibility to other stakeholders of the environment and takes dust control actions seriously, the employees of that company will foster their awareness of dust hazards and sense of responsibility for dust control because of their identity as organizational members [
69]. Therefore, a construction company can influence not only the behavior but also employees’ personal norms regarding dust pollution control. This aspect is reflected in the third hypothesis developed for this study.
Hypothesis 3. A company’s behavior significantly and positively affects the personal norm of employees.
According to NAM, one’s personal norm is activated by one’s awareness of consequences and ascription of responsibility. Awareness of consequences is defined as awareness of negative consequences to others or to the environment due to an action (or non-action) of a person [
68]. The awareness of health hazards to workers and people living closer to a construction site and other impacts to the neighborhood will trigger personal norms within a worker that dust pollution should be mitigated. Ascription of responsibility is described as a feeling of responsibility when not acting pro-socially to mitigate negative consequences [
68]. This sense of responsibility for one’s action (or non-action) should activate personal norms about the environmental issue in question. The relationship between personal norm and these antecedents is often interpreted in two different ways in the norm activation model.
The popular version out of the two, called the moderator model, defines both awareness of consequence and ascription of responsibility directly impacting personal norms (awareness of consequences and ascription of responsibility > personal norms > behavioral intention/behavior). This version is often used by researchers who investigate pro-environmental behavior of people. The alternative interpretation, which is called the mediator model, uses a sequential relationship in which awareness of consequences affects ascription of responsibility that in turn triggers personal norms [
56,
70]. This version could be portrayed as: awareness of consequences > ascription of responsibility > personal norms > behavioral intention/behavior. The proponents of this version argue that a person tends to be aware of the negative consequences of an action or inaction prior to feeling responsible for it [
74]. However, in the former version, there is no direct link observed between awareness of consequences and ascription of responsibility. Han et al. [
75], using lodging decisions of convention attendees, compared the two versions of NAM and found the sequential model to be more effective in predicting the pro-environmental decision making. This is because only when people become aware of negative consequences do they assign these to themselves and feel responsible for their impacts. As predictors of the personal norm of employees in a construction organization, the study tested three hypotheses as follows.
Hypothesis 4. Awareness of consequences significantly and positively affects the personal norm of employees.
Hypothesis 5. Awareness of consequences significantly and positively affects ascription of responsibility of employees.
Hypothesis 6. Ascription of responsibility significantly and positively affects the personal norm of employees.
3.2. Data Collection and Analysis
A survey method was used to collect data for model testing due to its suitability for obtaining individual beliefs and perceptions. A questionnaire survey was administered among a sample of employees working in construction sites in Colombo, Sri Lanka. A survey is a systematic way of collecting primary data, which can be efficiently utilized to suggest possible reasons for relationships in a model comprised of key variables [
76]. Further, in a quantitative research approach, the problem is best addressed by understanding what factors or variables influence an outcome [
77]. Moreover, a quantitative research approach is suitable for testing objective theories by examining the relationship(s) between different variables [
78]. Therefore, this strategy is well suited for this research, by its very nature to test the relationships postulated in the conceptual model.
The supporting facts of the literature synthesis were used to develop the structured questionnaire survey under two sections. The purpose of
Section 1 is to identify the demographics of the research respondents.
Section 2 of the questionnaire survey was developed to identify the environmental behavior of employees based on the conceptual model. A 5-point Likert scale anchored by “strongly disagree −1” to “strongly agree −5” was given to the respondents to identify the level of agreement with each indicator used to measure the various constructs included as latent variables.
The sample was selected from the SC1 and SC2 graded contracting companies in Sri Lanka. The grading system is administered by the Construction Industry Development Authority (CIDA), which is a government agency entrusted with the regulation of the construction industry. It should be noted that SC1 and SC2 construction firms are the largest in Sri Lanka according to this grading system, which is based on the firms’ financial capacity, human resources, experience, etc. Therefore, the environmental management systems used by those companies are far superior to the lower ranks and their employees are well qualified and capable of answering the questions contained in the study. A combination of purposive and random sampling techniques was used to select the sample from the population of workers based at construction sites of these SC1 and SC2 construction companies. A purposive sample of current projects implemented by those companies in Colombo, the capital city of Sri Lanka, was selected. The questionnaire survey was administered in those sites and the responses were collected using a drop box to make it anonymous. The respondents consisted of project managers, engineers, health safety and environmental managers, site mangers, quantity surveyors, technical officers, tradesmen, plant operators, and unskilled workers, as shown in
Table 2. The reason for selecting these wide-ranging roles is because they have a major influence on the environment as they design and plan project operations, execute and monitor work, and are responsible for project outcomes. Therefore, the sample had both managers and workers, representing both aspects of construction operations. The sample size for this study was based on a rule of thumb generally employed for Partial Least Squares Structural Equation Modeling (PLS-SEM), in which the minimum sample size should be equal to ten times the largest number of formative indicators used to measure a latent construct. By considering these guidelines, 100 respondents were determined as sufficient for the sample.
Data obtained from the survey, which had a 1–5 Likert Scale, were analyzed using descriptive statistics to find behavioral perspectives and passed through the Structural Equation Modeling (SEM) technique for validation of the model. The mean of the descriptive analysis was used with a scale that has been employed in the study of Kazaz and Ulubeyli [
79]. Accordingly, the scale from a difference of 1–5 and intervals with 0.8 was developed to determine the degree of central tendency based on following categorizations:
- -
≤“Strongly disagree” ≤ 1.80;
- -
1.80 < “Disagree” ≤ 2.60;
- -
2.60 < “Neutral” ≤ 3.40;
- -
3.40 < “Agree” ≤ 4.20; and
- -
<“Strongly agree” ≤ 5.00.
Kazaz and Ulubeyli [
78] have also used a similar approach that comprises intervals of the study for investigating the drivers of productivity among construction workers. The rank is defined when there are two or more variables that have the same mean values; the priority is assigned to the variables according to a descending order of standard deviation or coefficient of variation (CV) [
79]. Based on the central tendency, a benchmark mean score of 4.2 was assigned to filter the “strongly agree” factors while 3.4 assigned as a benchmark for filtering “agree” factors.
The questionnaire comprised of a total of 23 items categorized under general information, demographics, variables of environmental behavior (awareness of consequences, ascription of responsibility and personal norms), and environmental behavior of the construction company. To ensure the content validity, the items were developed using literature and definitions of variables were assigned accordingly. Three items were developed to measure the awareness of consequences, one for ascription of responsibility, four for personal norms, three to observe company behavior, and two to represent the employees’ dust pollution behavior. Moreover, Composite Reliability (CR) was used to check the reliability and internal consistency. The survey instrument and model validation tools are elaborated in
Section 4.2.