International Journal of Environmental Pollution and Remediation (IJEPR)

ISSN: 1929-2732

1. Introduction

In recent decades, there has been an increase in the global consumption of bottled water, particularly in developed countries where tap water is drinkable [1]. The numbers reveal the importance of this market: in the last decade, the annual growth rate was 5% [2]. For instance, in the European Union (EU), where tap water is drinkable in all Member States, presenting high quality values, except for a small minority of countries- Bulgaria, Cyprus, Latvia, Lithuania, and Romania [3], bottled water consumption reached considerably high values, being the EU average of 118 litres per person. In 2019, Italy was the leading market for the consumption of bottled water, with 200 litres of bottled water consumed per person; Germany was the second largest consumer with 168 litres consumed per person, and, in third place, Portugal appears with the consumption of 140 litres of water per person [4]. The high consumption of bottled water, especially in more developed regions that have drinking water from the tap, represents a serious problem for several reasons, namely: i) bottled water requires much more energy in terms of production and distribution; ii) the total energy required for the unit consumption of bottled water ranges from 5.6 to 10.2 megajoules (MJ) per litre, while this number is typically 0.005 MJ per litre for the treatment and distribution of piped water; iii) bottled water has a negative environmental impact, namely plastic waste; iv) plastic bottles are made from petroleum, and most bottles are not recycled, ending up in landfills, forests, lakes and oceans; v) regarding safety and health aspects, bottled water is less regulated than municipal tap water [1], [5]–[8]. Among all these negative impacts associated with bottled water consumption, particularly PET bottles, it is unquestionable that the plastic waste generated is the most harmful environmental consequence. In the coming years, the situation is expected to get even worse. According to Lebreton and Andrady [9], plastic production is expected to double in the next two decades resulting in high amounts of plastic abandoned in the oceans and soils [10]–[12]. In response to this serious environmental problem, a growing number of public and private spaces have been adopting measures to reduce bottled water consumption, namely offering reusable glass, metal, or even reusable plastic bottles; and installing piped water sources or drinking fountains spread throughout the spaces.

HEIs have the potential to shape beliefs and practices and “can play a vital role in developing and implementing an effective intervention to decrease bottled water consumption by increasing tap water consumption” ([13], p.558). Studying the university community can help policymakers to understand perceptions, choices and behaviours towards drinking water and the best initiatives to develop for individuals to reduce bottled water consumption, not only at university but in other contexts, serving as a catalyst potential for promoting practical changes in a broader sphere towards a more sustainable future [1], [14]–[16].

This research study proposes to apply the TPB framework to deepen some of the determinants of individuals` intention and behaviour regarding tap water consumption among the academic community of Coimbra Business School, a Portuguese HEI in the management, accounting and marketing areas. Thus, this study intends to contribute to fill this gap. It is expected to be clearer the influence of some key variables on tap water consumption behaviour, such as attitudes, subjective norms, perceived behavioural control, pro-environmental measures adopted by the HEI on intentions. It is also an objective to understand how individuals` intention to increase tap water consumption influences their behaviour.

The remainder of this study is organised as follows. After an introduction to the research subject, a second section is devoted to literature review. Then, the methodology is presented. In the third section the results are presented and discussed. Finally, the last section is dedicated to the main conclusions, limitations, and future research.

2. Literature Review

2. 1. Bottled and tap water consumption

The literature on water consumption also presents several studies carried out in Higher Education Institutions (HEIs), deepening college students' behaviour regarding bottled and tap water consumption. For example, in a Dutch HEI, van der Linden [17] observed that students bought single use plastic packaging water bottles due to barriers, or perceived barriers, on university campuses. It makes it difficult to use reusable bottles, namely the lack of water fountains on campus facilities. In another study developed at Princeton University, it was observed that providing free reusable water bottles to students significantly reduced the purchase of disposable plastic bottles [18]. At Duke University, after installing several water fountains across the campus, the number of single-use water bottles disposed of in campus trash and recycling streams has reduced considerably [19]. Graydon et al. [13] developed a study to assess community risk perceptions and drinking water choices on the University of South Florida campus. The results revealed that: certain groups (college students and ethnic/racial minorities) drank significantly more bottled water; among political ideologies, liberals drank less bottled water; women reported substantially higher risk perceptions of tap water on campus. Regarding the benefits associated with tap water consumption, students indicated the fact that it is less expensive and better for the environment; and as disadvantages, students reported that tap water has a less pleasant taste than bottled water. At Purdue University, Saylor et al. [20] developed a study to understand students' and staff's beliefs and practices regarding tap water and bottled water. They noted that undergraduate students drank significantly more bottled water than graduate students, faculty and staff, and women drank more bottled water than men. The authors also found the main barriers to tap water consumption: preference for the taste of bottled water, the belief that bottled water is safer and of superior quality, the perception of convenience and ease of use of bottled water. Díez et al. [21] deepen university students' beliefs and environmental attitudes towards bottled and tap water consumption and examined tap and bottled water availability and sales volumes at the University of the Basque Country. The authors observed that students predominately drink tap water and no health or taste issues associated with its consumption were perceived. In this University, most students claimed to use reused plastic water bottles as the most common tap water container. Nevertheless, there was a widespread presence of bottled water sales at university premises, confirming the ubiquity of this commodity. Hence, the authors concluded that a more remarkable shift in sustainable behaviour is needed in the community on and off college campuses.

2. 2. TPB Framework and Hypotheses development

The TPB, first proposed by Ajzen in 1985 [22], is a socio-cognitive model widely applied to explain the individual’s behaviour. A key variable in this model is the individual’s intention to adopt a specific behaviour, which may be defined as a motivational factor to a certain type of behaviour, and according to the TPB, it is influenced by attitude, subjective norms, and perceived behavioural control [22], [23]. Succinctly, attitude may be defined as a positive or negative belief in a particular behaviour; the subjective norm is an individual’s engagement with a specific behaviour due to social pressure; and perceived behavioural control indicates whether an individual’s motivation is influenced by how he or she perceives the level of difficulty or simplicity of a specific behaviour [22]–[24].

Despite its unquestionable contribution to scientific research development, the predictive effectiveness of the TPB has been deeply criticised, mainly due to its reduced number of explanatory variables [25]–[27]. Hence, many researchers developed extended theoretical models, including additional variables to increase the efficiency of its predictive capacity [28], [29].

Following the TPB reasoning, and to deepen the individuals` behaviour regarding tap water consumption, the extended TPB model proposed is presented in Figure 1.

The six research hypotheses proposed are:

• Hypothesis 1 (H1): Individuals` attitude has a positive influence on their intention to consume tap water.
• Hypothesis 2 (H2): Individuals` subjective norms have a positive influence on their intention to consume tap water.
• Hypothesis 3 (H3): Individuals` perceived behavioural control has a positive influence on their intention to consume tap water.
• Hypothesis 4 (H4): Individuals` perceived behavioural control has a positive influence on their tap water consumption behaviour.
• Hypothesis 5 (H5): HEI`s adopted measures to promote tap water consumption have a positive influence on individuals` intention to consume tap water.
• Hypothesis 6 (H6): Individuals` intentions have a positive influence on their tap water consumption behaviour.

3. Materials and Methods

3. 1. Survey Design and Application

From 13 January to 30 March 2022, a representative online survey was conducted among the target population (see [30]) of this study: students and teachers, aged 18 or over, of Coimbra Business School, an HEI which, in the last academic year of 2021/22, had a total of 2857 students and 170 teachers. The final and definitive version of the questionnaire resulted from an extensive literature review on the research subject and from the application of a pilot test on 6 January 2022, using the "think aloud" technique. An initial and provisory version of the questionnaire was distributed to a sample of 10 students and five teachers of the HEI, asking the respondents to think aloud as they completed the questionnaire, verbalising all thoughts that would usually be silent. This procedure took place in the classrooms of the HEI with the presence of one of the researchers in this study. The researcher only watched and took notes, not intervening or influencing the behaviour of the respondents. With this qualitative research technique, it was possible to obtain direct data on the continuous thought process of the interviewees during their activity of answering the questionnaire, enabling them to refine and clarify the questions and, in this way, increase the questionnaire's efficiency [31], [32].

The questionnaire was created and housed online using the Google Forms platform and shared with the school community through HEI's social networks and institutional email. Links to the questionnaire were also posted to a college-wide daily email to all students and college teachers. The same link was posted to multiple student group social media pages, residence hall email lists, and sports teams email lists. Teachers of various courses at the college distributed the link to students in their classes.

The questionnaire consists of two main sections. The first section includes several questions to gather information about the demographic characteristics of the respondents, namely age, gender, academic education, and situation at the HEI. The second section includes different questions on key issues for this research study regarding the respondents' water consumption, namely tap water attitudes, subjective norms, perceived behavioural control, HEI` measures to promote tap water consumption, tap water intention, tap water consumption behaviour.

Using the non-probabilistic convenience sampling method, a sample of 413 valid responses was achieved. After collecting and verifying the obtained database, the data were analysed using the R software, version 4.0.5. To achieve the objectives set out in this study, techniques of descriptive statistics and nonparametric tests were performed. Nonparametric tests were appropriate since these do not require certain assumptions (for example, the normality of the variables) verified in the study. The data were also analysed with the structural equation modelling technique. More precisely, the covariance-based SEM (CB-SEM) approach is considered for the estimation and model validation procedures.

In this research, the following latent variables were used: tap water attitudes (TWA), subjective norms (SN), perceived behavioural control (PBC), tap water intentions (TWI), tap water behaviour (TWB), and measures of the higher education institution (MHEI) [33]–[35]. Each item presented in the questionnaire was measured with a five-point Likert scale, ranging from 1 (strongly disagree) to 5 (strongly agree). Table 1 shows the survey items considered in the questionnaire and the factor loading for each item. The Cronbach alpha values for all measures are also presented (see Table 2) to perceive if multiple-question Likert scale surveys are reliable.

4. Results and Discussion

4. 1. Respondents’ demographic profile

The sample is composed of 413 respondents: 374 students and 39 teachers of the HEI. Around 63% of respondents are female, 63% are aged between 18 and 21, about 73% are undergraduate students, 17% are master's students, and 10% are teachers. Regarding academic education, 62% of respondents have completed secondary education, 29% have a degree, and about 9% hold a master's or doctorate (this is the case of teachers or even students attending a second degree or a second master's degree).

4. 2. Measurement Model Analysis

The measurement model includes six latent factors indicating tap water attitudes (TWA), subjective norms (SN), perceived behavioural control (PBC), tap water intentions (TWI), and tap water behaviour (TWB), and measures of the higher institution (MHEI). To assess the measurement model’s validity, CFA was performed using the covariance-based SEM approach (CB-SEM). The measurement model demonstrated acceptable goodness of fit for the empirical data. The fit indices commonly reported for CFA were calculated and met recommended guidelines [36], [37]. The following values were obtained for the model: CFI = 0.972, TLI = 0.964, both above 0.9 and RMSEA = 0.053 (less than 0.08). Moreover, accordingly to Bentler (2006), the chi-square normalised by degrees of freedom should be less than five, and in this case, it was equal to 1.85.

As described by Gefen et al. [39], the assessment of the constructs involves determining internal consistency reliability, convergent validity, and discriminant validity. To analyse convergent validity, the values of factor loadings and the average variance extracted (AVE) were determined. From Table 1, it can be seen that the factor loadings values of most of the items were above 0.6 (with an exception of two items). Regarding the AVE, it is recommended that it exceeds 0.5 [40]. All the values of AVE verified this condition, as shown in Table 2. These results indicated convergent validity. To assess internal consistency, the composite reliability and Cronbach alpha values were calculated. Table 2 shows that all of the measures were above the recommended value of 0.7, suggesting the high reliability of the scales [37].

According to Hair et al. [37], a rigorous test to assess discriminant validity is to compare the AVE values for any of the constructs with the square of the correlation estimate between these two constructs. The discriminant validity examines the constructs’ variances and uniqueness. This criterion requires that the diagonal value in bold be higher than the values in its row and column, the correlation must be relatively low to exhibit discriminant validity. As depicted in Table 3, this criterion was accomplished for all constructs; therefore, it could be confirmed as the discriminant validity.

From Table 3, it also can be seen that the mean values of tap water attitude, perceived behaviour control and measures adopted by HEI were relatively high (out of 5), and the mean values of the other variables were also relatively favourable.