Communicating About Single-Use Bottled Water to Mitigate Ecosystem Pollution

Abstract

The use of plastics is ingrained in modern lifestyles, yet plastic waste contaminates ecosystems and harms human health. Single-use bottled water is a substantial contributor to global plastic use despite many sustainable alternatives. The purpose of this study was to examine if knowledge, cognitive beliefs, affective beliefs, attitudes, and perceived access to clean water affected consumers’ intention to purchase single-use bottled water. Data were collected from Florida, Georgia, and Alabama residents using non-probability opt-in sampling. A hypothesized model was analyzed using confirmatory factor analysis and structural modeling in MPlus. Knowledge, cognitive beliefs, and affective beliefs accounted for 34.5% of the variance in attitudes. Additionally, these three constructs, along with the perceived access to clean water and attitudes, accounted for 31.9% of the variance in the intention to purchase single-use bottled water. All direct and indirect paths in the model were significant, indicating that communication targeting knowledge, cognitive beliefs, affective beliefs, attitudes, and the perceived access to clean water may shift the public sentiment. Key findings revealed that increasing consumers’ knowledge and decreasing beliefs about single-use bottled water decreased purchase intentions. The study results highlight the need for communication strategies that challenge emotional attachments and the perceived knowledge of single-use bottled water while emphasizing credible knowledge about the associated environmental and health impacts. Environmental communicators should explore generational differences related to affective beliefs regarding single-use bottled water to target messaging based on aesthetics.

1. Introduction

Plastic waste has been discovered in numerous environments, contaminating ecosystems and harming human health [1,2,3]. Plastic pollution occurs when people directly litter and/or waste management infrastructure is lacking [4], and it becomes difficult to reverse when weathering causes the plastic to fragment into micro- and nanoplastics, making it impossible to remove from the natural environment [4]. These micro- and nanoplastics enter water systems and cause potential drinking water contamination and biological uptake within that ecosystem [5,6,7]. The public can play an immediate role in reducing plastic pollution through proper recycling, decreasing use, and physically cleaning up litter [4,8].

The use of plastics is ingrained in modern lifestyles given numerous societal benefits like comfort, hygiene, and safety [9]. For example, plastics are valuable resources for reducing the risk of disease transmission [9,10]. However, plastic poses serious long-term environmental and human health risks that outweigh its convenience [11,12]. For example, overextraction from aquifers to produce bottled water increases stress on local water supplies, agriculture, and community members [13]. In addition, most plastic produced globally is disposed of after a single use [1]. The integration of plastics into everyday life makes interventions associated with reducing plastic use difficult.

The previous literature has highlighted the need for plastic pollution to be addressed by numerous stakeholders, including governments, industries, and consumers, to effectively curb the excessive use of plastic [14]. However, interventions like improved waste management for recycling plastics and raising consumer awareness are lacking [14,15].

The complexity and widespread impact spanning both social and natural dimensions associated with plastic pollution indicate an urgent need for effective environmental communication to support behavior modification. Effective environmental communication involves strategically translating scientific findings into user-friendly messaging to inform and influence various audiences [16]. Determining the most effective communication strategies, such as which components to include in messaging regarding a specific topic, requires efforts grounded in social science theory. Communication campaigns embedded in social science theory have successfully raised consumer awareness about numerous environmental issues and therefore may be applicable to enhancing stakeholder engagement regarding plastic pollution [15].

1.1. Study Context

Single-use bottled water is a substantial contributor to global plastic use [11]. Single-use water bottles, which are made with polyethylene terephthalate (PET), are a source of plastic pollution that is only expected to worsen over time [17]. The single-use bottled water market has experienced an annual growth rate of 5% per year over the past decade [18], with an expected global market of USD 505.19 billion by 2028 [17]. Yet, the recycling rate for PET bottles and jars in the U.S. was only 29.1% in 2018 [19].

Previous studies have found that a key driver of single-use bottled water consumption is perceived health risks related to tap water, which has been exacerbated by serious but isolated drinking water contamination events in recent years [20,21]. According to Jaffee and Case (2018), “[…] regardless of the reality, perceptions of poor tap water quality have become established in the public consciousness in North America […]” [13], (p. 488). Even so, the U.S. tap water supply is considered one of the safest in the world [20]. Thus, the high use of single-use bottled water may be driven more by perceptions of tap water than perceptions of single-use bottled water, highlighting the importance of water quality in this systematic issue.

This study focused on the residents of three southeastern U.S. states: Florida, Georgia, and Alabama. There is an ongoing dispute over two river basins, the Apalachicola–Chattahoochee–Flint and the Alabama–Coosa–Tallapoosa, shared between the target states. The overarching dispute is due to the allocation of the shared water resources among the states. Georgia is upstream, and water use is continuing to increase as the population in Atlanta skyrockets. Florida and Alabama are downstream and face the potential repercussions of this population growth on their water availability and ecosystem health. The dispute over freshwater availability may cause residents to perceive their access to clean water as limited or uncertain, especially in Florida, which has already faced tangible impacts on oyster fisheries. The regions’ heightened awareness or concern regarding clean water access provides a unique case study to examine single-use bottled water behaviors. Thus, the purpose of this study was to examine consumers’ intention to purchase single-use bottled water to guide environmental communication efforts in geographically similar states.

1.2. Conceptual Framework

The theory of planned behavior (TPB) is frequently used by social scientists to understand volitional environmental intentions, for example, consumers’ intention to use single-use plastics (e.g., [22,23]). The TPB postulates that consumers’ attitudes, subjective norms, and perceived behavioral control affect their behavioral intentions, which are the greatest determinants of behavior [24]. An attitude refers to consumers’ expectations of the outcome of performing a specific behavior. Subjective norms refer to consumers’ perceptions of social pressures to perform a specific behavior. Perceived behavioral control refers to consumers’ perceived ability to carry out a behavior [24].

Studies on single-use plastics often extend TPB models to include additional variables to further explain behavioral intentions. For example, Sun et al. (2017) examined the intention to use plastic bags among consumers in China and found that the TPB explained 67.1% of the variance in intentions, whereas the extended model, which added the consideration of environmental concerns, ethical beliefs, demographics, and convenience, explained 75.4% of the variance. Specifically, attitudes, subjective norms, perceived behavioral control, and convenience all had a positive direct effect on the intention to use plastic bags, whereas environmental concerns and ethical beliefs had a negative direct effect on this intention [23].

Lin and Xu (2021) used the theory of reasoned action, which was an earlier iteration of the TPB, to examine the effect of trust in bottled water advertising, consumer knowledge, consumer beliefs, attitudes, and subjective norms on single-use bottled water purchase intentions. The model explained 53% of the variance in single-use bottled water purchase intentions. Specifically, knowledge had a direct effect on the trust in advertising, cognitive beliefs, and affective beliefs [17].

1.2.1. Consumer Knowledge

Knowledge, whether based in fact or fiction, is often a significant predictor of attitudes. Research had indicated that consumers of single-use bottled water often perceive tap water as unsafe to drink [25]. However, since its widespread adoption in the U.S., single-use bottled water has faced recalls due to potential E. coli contamination [26], the reported levels of bacterial counts have been higher in bottled water than tap water [27], and studies have found PET chemicals leaching into single-use bottled water [25,28]. Previous studies have found that consumers’ knowledge about single-use bottled water has a significant relationship with single-use bottled water purchase intentions [17].

In relation to the TPB, Ateş (2021) examined purchase intentions for eco-labeled foods in Turkey using an extended TPB model that included knowledge, personal norms, self-identity, and the willingness to pay and found that the extended model accounted for 37% of the variance in the purchase intention for eco-labeled foods [29]. Knowledge of eco-labels had a direct effect on attitudes. Yet, empirical evidence regarding consumers’ knowledge about single-use bottled water is mostly absent from the literature [17].

Hypothesis 1.

Knowledge about single-use bottled water will have a direct effect on cognitive beliefs regarding single-use bottled water.

1.2.2. Cognitive and Affective Beliefs

Cognitive and affective beliefs are distinct components of attitudes [30,31]. The exploratory power of the TPB may be increased by incorporating cognitive and affective beliefs into the model [32,33].

Cognitive beliefs refer to consumers’ evaluations, perceptions, and perceived knowledge of anything concrete or abstract [30]. Affective beliefs refer to consumers’ emotions, feelings, or aesthetic preferences towards anything concrete or abstract [30].

Koenig-Lewis et al. (2014) examined the relationship of Norwegian consumers’ intention to purchase products using ecologically responsible packaging with environmental concerns, cognitive beliefs, and affective beliefs. Affective beliefs were examined considering positive and negative emotions and were found to have a direct effect on purchase intentions, with positive emotions having a greater effect than negative emotions. Cognitive beliefs did not have a significant direct effect on purchase intentions but had a significant indirect effect when mediated by positive and negative emotions [32].

Hypothesis 2.

Cognitive beliefs regarding single-use bottled water and affective beliefs regarding single-use bottled water will have a direct effect on attitudes towards using single-use bottled water.

1.2.3. Attitudes

Attitudes are a predictor of behavioral intention in numerous theories, including the TPB. In the context of plastics, attitudes are consistently found to be a significant predictor of intentions in studies about single-use plastic consumption (e.g., [22,23]), the consumption of environmentally friendly products (e.g., [29,34]), and single-use bottled water purchase intentions (e.g., [17]). For example, Raimondo et al. (2022) explored Italian millennials’ intentions and behaviors regarding plastic beverage containers using an extended TPB model and found that attitudes had a direct effect on intentions and an indirect effect on self-reported behavior to reduce the consumption or waste of plastic drinking bottles [22]. Additionally, Wongsaichia et al. (2022) examined Thai consumers’ intention to purchase green food by clustering consumers into two segments, green consumers and non-green consumers. Although there were critical ratio differences for other factors influencing intentions between the two segments, attitudes had a direct effect on intentions for both green and non-green consumers, with no significant difference between the groups [34].

These findings indicate that attitudes are a central construct in shaping behavioral intentions. Therefore, there is a theoretical opportunity to explore attitudes and additional factors that influence attitudes related to single-use bottled water, contributing to the application of constructs in the TPB [35].

Hypothesis 3.

Attitudes towards using single-use bottled water will have a direct effect on the intention to purchase single-use bottled water.

1.2.4. Perceived Access to Clean Water

Previous studies have found that consumers purchase single-use bottled water because they are concerned about the healthiness and safety of their tap water and its organoleptic qualities (e.g., [20,25,36,37,38,39,40,41,42]). For example, Hu et al. (2011) examined the relationship between demographic characteristics, single-use bottled water consumption, and the perceptions of local water quality [38]. The authors found that the perceived quality of groundwater and the perceived safety of drinking water were significant predictors of primary or regular single-use bottled water use [38]. Hu et al. (2011) also explored regional differences in the U.S. and found that respondents in the south (Arkansas, Louisiana, Oklahoma, Texas) and the southeast (Alabama, Florida, Mississippi, Tennessee) were all likely to consume single-use bottled water, regardless of whether they were primary or regular consumers [38]. The study, however, is somewhat dated, and further exploration is warranted in this area as the influence of the perceived access to clean water remains pertinent to this dialogue, especially as climate change impacts tap water availability and quality [43].

Hypothesis 4.

Perceived access to clean water will have a direct effect on the intention to purchase single-use bottled water.

1.2.5. Full Model

Figure 1 showcases a proposed conceptual model for examining additional factors that have the potential to impact attitudes and therefore intentions and behavior modifications in the context of single-use bottled water. While the TPB posits a direct effect of attitudes on behavioral intentions, the previous literature suggests that external variables like the perceived access to clean water can alter this relationship [38]. For instance, Camacho et al. (2025) examined workplace energy-saving intentions using an extension of the TPB. In the model, the organizational culture mediated the effect of attitudes on the intention to save energy in the workplace [44].

In the proposed model, the perceived access to clean water is treated as a mediating variable between attitudes and intentions. Consumers are likely to purchase single-use bottled water regardless of their attitude towards single-use bottled water if they do not feel their tap water is safe (e.g., [38]). As such, the relationship between attitudes and the intention to purchase single-use bottled water depends on the perceived access to clean water. This mediation effect adds to the complexity of the relationship between attitudes and intentions typically found in studies applying a TPB model.

Hypothesis 5.

The perceived access to clean water will mediate the direct effect of attitudes towards using single-use bottled water on the intention to purchase single-use bottled water.

Hypothesis 6.

Knowledge, cognitive beliefs, affective beliefs, and attitudes will have an indirect effect on the intention to purchase single-use bottled water.

2. Materials and Methods

2.1. Participants and Procedure

Data were collected in September 2022 from residents of Florida, Georgia, and Alabama. The selection criteria required a sample of respondents who were aged 18 years or older and representative of the population based on their gender, age, and race/ethnicity. Data were collected using an online questionnaire administered through Qualtrics. Non-probability opt-in sampling methods were used to recruit respondents, who were compensated according to the typical protocol of Qualtrics (Provo, UT, USA). This study was part of a larger panel assessing public perceptions on water-related topics (see [45,46,47]).

The survey instrument was reviewed for content accuracy and face validity by a panel of experts, including faculty members involved in natural resource conservation, survey design, and communication studies. The University of Georgia Institutional Review Board (IRB #00005553) approved the study design. The instrument was pilot-tested to determine content validity with 50 individuals who were representative of the sample. The Cronbach alpha coefficients for all constructs studied were above 0.70, and therefore the scales were deemed reliable and no changes were made following the pilot test [48].

The sample size needed for this study was determined according to the guidance of Cochran (1963) [49].

$$n_0={\displaystyle \frac{Z^{2}pq}{e^2}}={\displaystyle \frac{{1.96}^2\times 0.5(1-0.5)}{{0.04}^2}}=600.25$$

In the equation, $n_0$ is the required sample size based on the specified parameters. The specified parameters are as follows: Z represents the Z-score corresponding to a desired confidence level (1.96 for a 95% confidence level). p is the estimated proportion of the population, and when the true proportion of the population is unknown, it is common to use 0.5. q is 1 − p. e is the margin of error or desired level of precision, which was set as $\pm 4.0\%$ for this sample. Therefore, the minumum sample size needed for these parameters was approxomately 600 respondents.

2.2. Measures

Respondents’ cognitive beliefs regarding single-use bottled water were measured using five Likert-type items adapted from Lin and Xu (2021) on a five-point scale (1 = Strongly Disagree; 2 = Disagree; 3 = Neither Agree nor Disagree; 4 = Agree; 5 = Strongly Agree). Respondents were asked to indicate their level of agreement or disagreement based on their motivations (it tastes good, it is clean, it is from pure and natural water sources, it is safe, and it contains ingredients good for health) for drinking water from single-use plastic water bottles [17]. Respondents had the option to indicate if they never purchased single-use plastic water bottles. An index score for cognitive beliefs regarding single-use bottled water was calculated based on the average of the five items (α = 0.93).

Respondents’ affective beliefs regarding single-use bottled water were measured using three Likert-type items adapted from Lin and Xu (2021) on a five-point scale (1 = Strongly Disagree; 5 = Strongly Agree). Respondents were asked to indicate their level of agreement or disagreement based on if they purchased single-use plastic water bottles for the reasons in each of the following statements: it looks cool to carry store-bought bottled water around, I like the shape/design of the store-bought bottle, and I like the reputation of bottled water brands/companies [17]. Respondents had the option to indicate if they never purchased single-use plastic water bottles. An index score for affective beliefs regarding single-use bottled water was calculated based on the average of the three items (α = 0.87).

Respondents’ perceived access to clean water was measured using four Likert-type items adapted from Grupper et al. (2021) on a five-point scale (1 = Strongly Disagree; 5 = Strongly Agree). Respondents were asked to indicate their level of agreement or disagreement based on how acceptable the tap water in their home was in terms of the following aspects: its taste, smell, appearance, and safety [50]. An index for the perceived access to clean water was calculated based on the average of the four items (α = 0.91).

Respondents’ knowledge about single-use bottled water was established using five true or false questions. Items were adapted from Lin and Xu (2021). Respondents were asked to indicate whether the following statements were true or false: the regulations for producing bottled water are very strict (false), bottled water is originated from spring water in nature (false), most single-use water bottles are recycled into other useful products (false), single-use bottled water is different from tap water (false), and single-use water bottles don’t have much impact on the environment (false) [17]. A knowledge score was calculated based on the number of questions answered correctly and could range from 0 to 5.

Respondents’ attitudes towards using single-use bottled water were measured using seven semantic differential items adapted from Lin and Xu (2021) and Gibson et al. (2021) on a five-point scale. Respondents were asked to indicate how they felt about using single-use plastic water bottles: bad/good, harmful/beneficial, worthless/valuable, unpleasant/pleasant, not acceptable/acceptable, foolish/wise, and not essential/essential [17,51]. An index for attitudes towards using single-use plastic water bottles was calculated based on the average of the seven items (α = 0.96).

Respondents’ self-reported intention to purchase single-use bottled water was measured using 10 Likert-type items adapted from Lin and Xu (2021) on a five-point scale (1 = Strongly Disagree; 5 = Strongly Agree). Respondents were asked to indicate their level of agreement or disagreement based on if they planned to use single-use plastic water bottles in the scenarios in each of the following statements: when I go run errands, when I attend indoor events such as going to a movie or shopping mall, when I exercise in the gym, when I attend outdoor events such as going to a football game or a baseball game, when I exercise outdoors (e.g., jogging or hiking), when I order food at an eatery such as a deli, bakery, coffee shop or cafeteria, when I buy groceries, when I stop at a convenient store, when I am thirsty, and when I go to school/work [17]. An index for the intention to purchase single-use bottled water was calculated based on the average of the 10 items (α = 0.91).

2.3. Demographics

Responses were collected from 907 respondents. Respondents who selected never for the cognitive belief and affective belief sections were removed from the study, resulting in 734 respondents. These respondents were removed because the study sought to determine factors influencing single-use bottled water use. Including respondents who currently or had recently used single-use bottled water allowed us to target the population of interest and test our theory for single-use bottled water users rather than the general population. Demographics were collected based on the U.S. Census. Respondents’ demographics are reported in

Table 1. Demographics of respondents (N = 734).

Baseline Characteristics	F	%
Sex
Male	326	44.4
Female	408	55.6
Age
18–34 years	165	22.6
35–54 years	248	33.8
55+ years	321	43.9
Race *
White	547	74.5
Black	114	15.5
Asian	46	6.3
American Indian or Alaska Native	19	2.6
Other	40	5.4
Ethnicity
Hispanic	125	17.0
Non-Hispanic	609	83.0
Education
Less than 12th Grade	25	3.4
High School Diploma	160	21.8
Some College	165	22.5
2-Year College Degree	93	12.7
4-Year College Degree	186	25.3
Graduate or Professional Degree	105	14.3
Family Income
Less than USD 24,999	158	21.5
USD 25,000–USD 49,999	192	26.2
USD 50,000–USD 74,999	156	21.3
USD 75,000–USD 149,999	175	23.8
USD 150,000–USD 249,999	38	5.2
USD 250,000 or more	15	2.0
State of Residence
Florida	243	33.1
Georgia	256	34.9
Alabama	235	32.0

Note: * Respondents could select more than one race.

. Most respondents were white (74.5%). Similarly, most (77.7%) were older than 34 years of age. Their geographic locations were spread somewhat equally between Florida (33.1%), Georgia (34.9%), and Alabama (32.0%).

2.4. Data Analysis

The psychometrics for the data were assessed using confirmatory factor analysis (CFA) and reliability analysis. All CFA and item loadings were observed to be sufficient, and subsequent analysis was thus undertaken [52,53].

The MPlus (Los Angeles, CA, USA) script can be found in Appendix A. The comparative fit index (CFI = 0.92), Tucker–Lewis index (TLI = 0.91), and the root mean square error of approximation (RMSEA = 0.06; see

Table 2. Model fit statistics.

Model Fit Index	Acceptable Levels [54]	CFA	SEM
CFI	0.90 (though greater than 0.95 is preferred)	0.92	0.92
TLI	0.90 (though greater than 0.95 is preferred)	0.91	0.91
RMSEA	Less than 0.07	0.06	0.06
Chi-Square, χ2	Insignificant p value	χ2(515) = 1909.65, p < 0.000	χ2(516) = 1909.75, p < 0.000

) were used to assess the model fit and deemed adequate given the existing guidelines [54,55]. The proposed path from attitudes to the perceived access to clean water was not statistically significant and was removed from the model.

The reliability was assessed using the convergent validity and discriminant validity. Establishing convergent validity requires an average variance extracted (AVE) above 0.50 or an AVE that is greater than 0.40 with a corresponding composite reliability greater than 0.60 (see

Table 3. Convergent validity of constructs based on their average variance extracted.

Construct	AVE	Composite Reliability
Cognitive beliefs	0.70	0.60
Affective beliefs	0.71	0.59
Attitudes	0.77	0.76
Perceived access to clean water	0.72	0.71
Intention to purchase single-use bottled water	0.43	0.88

[52]). The Fornell and Larcker criterion, which determines if factor correlations are less than the square root of their corresponding AVE, was used to establish discriminant validity (see

Table 4. Discriminant validity of constructs based on the Fornell and Larcker criterion.

Construct	Cognitive Beliefs	Affective Beliefs	Attitudes	Perceived Access to Clean Water	Intention to Purchase Single-Use Bottled Water
Cognitive beliefs	0.84	-	-	-	-
Affective beliefs	0.40	0.84	-	-	-
Attitudes	0.54	0.43	0.88	-	-
Perceived access to clean water	−0.06	0.08	−0.02	0.85	-
Intention to purchase single-use bottled water	0.31	0.23	0.56	−0.11	0.66

Note: Numbers along the diagonal are the square root of their corresponding AVE.

[52]).

A structural model analysis was conducted on the dataset to examine the direct and indirect effects of the variables on the intention to purchase single-use bottled water. The model fit indices indicated an acceptable model fit (χ²(516) = 1909.75, p < 0.000; CFI = 0.92; TLI = 0.91; RMSEA = 0.06; Table 2).

The CFA and structural model were analyzed using MPlus. In the model, having a greater mean for the cognitive belief, affective belief, attitude, and intention constructs indicated a positive evaluation of using single-use bottled water. Having a higher knowledge score indicated higher knowledge about single-use bottled water and having higher perceived access to clean water indicated that respondents had a positive evaluation of their access to clean water.

3. Results

The first hypothesis, that knowledge would have a direct effect on cognitive beliefs, was supported (see

Table 5. Standardized path coefficients for direct and indirect effects on the intention to purchase single-use bottled water.

Effect	M	SD	Direct Effect (95% CI)	Indirect Effect (95% CI)	Hypothesis
Cognitive beliefs	3.57	0.94	-	-	-
Knowledge	2.54	1.45	−0.41 ***	-	H1 (accepted)
Attitudes	3.13	1.29	-	-	-
Cognitive beliefs	3.57	0.94	0.44 ***	-	H2 (accepted)
Affective beliefs	2.46	1.12	0.25 ***	-	H2 (accepted)
Knowledge	2.54	1.45	-	−0.18 ***	-
Intentions	3.11	0.91	-	-	-
Attitudes	3.13	1.29	0.55 ***	-	H3 (accepted)
Perceived access to   clean water	3.77	1.08	−0.10 **	-	H4 (accepted)
Cognitive beliefs	3.57	0.94	-	0.24 ***	H6 (accepted)
Affective beliefs	2.46	1.12	-	0.14 ***	H6 (accepted)
Knowledge	2.54	1.45	-	−0.10 ***	H6 (accepted)

Notes: *** p < 0.001; ** p < 0.01; The indirect effect represents the total indirect effect via all corresponding mediators.

and Figure 2). Knowledge ($\beta$ = −0.39, p < 0.001) had a significant direct effect on cognitive beliefs. Knowledge accounted for 16.7% of the variance in cognitive beliefs (R² = 0.16).

The second hypothesis, that cognitive beliefs and affective beliefs would have a direct effect on attitudes, was supported (see Table 5 and Figure 2). Cognitive beliefs ($\beta$ = 0.44, p < 0.001) and affective beliefs ($\beta$ = 0.25, p < 0.001) had a significant direct effect on attitudes. Knowledge, cognitive beliefs, and affective beliefs accounted for 34.5% of the variance in attitudes (R² = 0.35).

The third hypothesis, that attitudes would have a direct effect on the intention to purchase single-use bottled water, was supported (see Table 5 and Figure 2). Attitudes ($\beta$ = 0.55, p < 0.001) had a significant direct effect on intentions.

The fourth hypothesis, that the perceived access to clean water would have a direct effect on the intention to purchase single-use bottled water, was supported (see Table 5 and Figure 2). The perceived access to clean water ($\beta$ = −0.10, p < 0.01) had a significant direct effect on intentions.

The fifth hypothesis, that the perceived access to clean water would mediate the direct effect of attitudes on the intention to purchase single-use bottled water, was rejected. The perceived access to clean water was removed as a mediator between attitudes and intentions as a result of the CFA. Thus, no mediation analysis was conducted in the structural model.

The sixth hypothesis, that knowledge, cognitive beliefs, affective beliefs, and attitudes would have an indirect effect on the intention to purchase single-use bottled water, was partly supported (see Table 4 and Figure 2). Knowledge ($\beta$ = −0.10, p < 0.001), cognitive beliefs ($\beta$ = 0.24, p < 0.001), and affective beliefs ($\beta$ = 0.14, p < 0.001) had a significant indirect effect on intentions, but the indirect effect of attitudes on intentions was removed as a result of the CFA. Collectively, the predictor variables accounted for 31.9% of the variance in the intention to purchase single-use bottled water (R² = 0.32).

4. Discussion

Effective communication techniques for discussing environmental challenges are needed to encourage consumers to reduce their use of single-use bottled water, considering the associated environmental impacts, like land pollution and the presence of microplastics in freshwater [12,56,57]. Yet, research that builds on social science theory to address single-use bottled water consumption is limited [14,17,41,58] and therefore cannot inform environmental communication practices. This study contributes to the TPB literature by underpinning the role of cognitive and affective beliefs as key antecedents of attitudes and intentions. Overall, knowledge, cognitive beliefs, and affective beliefs accounted for 34.5% of the variance in attitudes. In addition, knowledge, cognitive beliefs, affective beliefs, perceived access to clean water, and attitudes accounted for 31.9% of the variance in the intention to purchase single-use bottled water. The findings indicate that environmental communicators need to address underlying belief systems in their messaging to alter attitudes and intentions. Additionally, the significant indirect effect of knowledge on both attitudes and intentions highlights the importance of adapting theories depending on the topic and audience. The environmental literature often indicates that we need to move beyond knowledge to impact intentions, though this study contradicts that claim by demonstrating that knowledge can still play a critical role in shaping intentions depending on the topic.

4.1. Theoretical and Practical Implications

Knowledge had a negative direct effect on cognitive beliefs, indicating that when consumers were more knowledgeable about single-use bottled water, their positive evaluations or perceptions of single-use bottled water decreased. This finding is consistent with the results of Lin and Xu (2021), who found that knowledge has a direct effect on cognitive beliefs [17]. In addition, knowledge explained 16.7% of the variance in respondents’ cognitive beliefs about single-use bottled water, providing an avenue to decrease positive cognitive beliefs regarding single-use bottled water. Knowledge also had a negative indirect effect on the intention to purchase single-use bottled water, suggesting that knowledge informs single-use bottled water purchasing decisions. This finding implies that knowledge should be examined further with knowledge scales focused on specific dimensions regarding single-use bottled water. For example, one scale may focus on its environmental impacts and another scale may focus on health impacts so that communicators can determine which facet of knowledge to focus their messaging on to decrease single-use bottled water consumption. Moreover, the mean knowledge score of the respondents was low, which may have been related to consumers using organoleptic qualities to evaluate water purity, which leads to a false sense of security in using single-use bottled water [25]. The indirect and direct effects of knowledge in the model imply that communication messages targeted at improving public knowledge about single-use bottled water are needed. Perhaps consumers are not oversaturated with information about single-use bottled water yet, suggesting that environmental communicators have an opportunity to increase consumers’ knowledge levels, but they need to be strategic in their messaging so that concepts are not dismissed.

Consistent with previous studies about the consumption of single-use plastics (e.g., [17,22,23]), attitudes had a positive and significant effect on intentions. In addition, cognitive beliefs and affective beliefs had a direct effect on attitudes, accounting for 34.5% of the variance in attitudes. The finding is consistent with that of Lin and Xu (2021), who found that cognitive and affective beliefs have a direct effect on attitudes, suggesting there are important practical implications regarding the attitude construct in the TPB [17]. In addition to focusing on altering attitudes directly, addressing the antecedents of attitudes (cognitive beliefs, affective beliefs, and knowledge) may be a practical way for environmental communicators to influence intentions related to single-use bottled water. Cognitive and affective beliefs also had an indirect effect on the intention to purchase single-use bottled water, with cognitive beliefs having a greater effect than affective beliefs. These findings differ slightly from those of Lewis et al. (2014), who found that affective beliefs had a direct effect on the intention to purchase products using ecologically responsible packaging and cognitive beliefs did not, but cognitive beliefs had an indirect effect on purchase intentions when mediated by affective beliefs [32]. It is possible that cognitive beliefs play a stronger role in single-use bottled water use due to the perceived health risks associated with tap water or the perceived health benefits associated with single-use bottled water. Environmental communicators should challenge cognitive beliefs by targeting perceived knowledge, considering respondents’ actual knowledge was low. Messaging about the ecosystem impacts of single-use bottled water should provide information about the lifecycle of single-use bottled water, focusing on the limited number of bottles recycled into other useful products (see Appendix B). Messaging about the health impact of single-use bottled water should incorporate the equivalency of tap water and single-use bottled water quality as many respondents believed bottled water originated from spring water in nature and that it was different from tap water.

Though the role of affective beliefs was secondary to that of cognitive beliefs in this study, it should not be overlooked. For instance, the Stanley Cup gained rapid popularity across generations from viral success on social media, including TikTok. The brand has a strong online presence, with both celebrities and influencers working on their marketing. The cups were marketed as a symbol and status item, often purchased as a fashion accessory. It should be noted, however, that the cups were marketed for a variety of beverage types, not exclusively tap water.

The antecedents of affective beliefs were not examined in the present study, but should be considered in future studies due to the predictive nature of affective beliefs regarding attitudes and intentions. Millennials have engaged in a recent trend of accessorizing with reusable water bottles [17], and a social marketing campaign that normalizes this aesthetic may influence the affective beliefs of consumers from different generations. Future studies should examine if different generations are more likely to purchase single-use bottled water so that campaigns can effectively target groups who are frequently consuming single-use bottled water.

Perceived access to clean water had a small negative direct effect on the intention to purchase single-use bottled water. The mean perceived access to clean water indicated that respondents agreed that they had access to clean water, which may explain the small effect. It is possible that separating respondents into groups by their perceived level of access to clean water (e.g., low, medium, high) would change the size of the coefficient and further elucidate the impact of the perceived access on intentions. Another opportunity is to group respondents by their state or geographical location (e.g., coastal or inland) to obtain results with more nuanced practical applications. For example, residents from Atlanta are less likely to view their perceived access to clean water as an issue due to their upstream access to the river. Florida residents, on the other hand, who have experienced reduced flow from the Apalachicola River may have greater awareness of this dispute. Organoleptic qualities may also be playing a key role here as they influence consumers’ perceived access to clean water. Public perceptions of tap water vary based on the geographic location [59]. For instance, coastal tap water may be perceived as lower in quality compared to that in mountainous areas, especially by visitors or residents who did not grow up in that area. A cluster analysis may help address stratification across multiple factors influencing respondents’ perceived access to clean water. We could gain a more nuanced understanding of how respondents’ perceived access impacts behavior modification if the data were stratified or moderated.

The proposed mediation effect of respondents’ perceived access to clean water between attitudes and intentions was removed as a result of the CFA, as the proposed path from attitudes to perceived access to clean water was not statistically significant. It is possible that this occurred because attitudes towards single-use bottled water do not affect perceived access to clean water but rather perceived access to clean water may affect attitudes. Future studies should examine respondents’ perceived access to clean water as a moderating variable because it may change the relationship altogether, rather than adding to the existing relationship.

4.2. Study Limitations

There are several study limitations. First, the study only included residents of Florida, Georgia, and Alabama. Although the findings have important implications for Florida, Georgia, and Alabama, they are not generalizable to the entire U.S. due to regional differences in water access and cultural norms, and this should be considered a limitation of the study.

Next, non-probability opt-in sampling was used for this study. Non-probability opt-in sampling is well accepted in public opinion research, but it poses several limitations. For example, the method limits respondents to individuals with internet access and attracts specific types of people due to the nature of online surveys [60]. Additionally, respondents who opt in to the survey likely do so for a reason, such as having stronger opinions or more access to and comfort using technology. This limits the generalizability of the findings as the sample may not reflect the broader population the study intended to reach.

Another limitation of the study was removing respondents who indicated never for the cognitive belief and affective belief sections regarding single-use bottled water. Specifically, respondents were told the following: If you never drink from a single-use plastic water bottle, please select Not Applicable. Respondents were removed to focus the analysis on single-use bottled water users. However, including respondents who selected never may have provided useful comparative insights into factors that influence non-use, informing targeted communication messages. Additionally, respondents’ perceptions of never using single-use bottled water may not reflect similar behaviors as some non-users are likely strict about their behavior and others may simply be low-frequency users. These nuances were not assessed in the current study.

Using survey data can also result in social desirability bias, where respondents answer questions in a way that they believe will be viewed favorably by others. This may also have influenced respondents who indicated that they never used single-use bottled water. However, the media portrayal of single-use bottled water is generally positive and there may be less of a socially correct answer in the context of single-use bottled water than regarding other topics such as health or water conservation.

5. Conclusions

Public health and ecosystems are threatened by the amount of plastic pollution in our environment and landfills [14]. Environmental communicators are particularly suited to handling complex issues with social and natural dimensions like plastic pollution because of their ability to translate scientific information into digestible consumer content. By exploring the antecedents to regularly studied theoretical variables (like those found in the TPB), the effectiveness of communication messages surrounding important yet unexplored topics, such as single-use bottled water, can be improved, resulting in behavior modifications intended to mitigate environmental impacts [3,17].