1. Introduction
Studies linking awareness of PM2.5 pollution to personal behaviors seeking to avoid such pollution are scant, and there is a lack of focus on this issue in Asia. The East Asia region, due to its economic development, rapid industrialization, and urbanization, has some of the highest levels of air pollution in the world, yet few subnational risk assessment studies on air pollution have been conducted in Asian countries [
1]. Shi et al. [
2] found that few studies looked at behaviors aimed at reducing exposure to PM2.5 in urban among urban residents. In addressing severe haze pollution, it was found that a household’s environmental concern, indirectly and significantly affects occupant’s moral norms and intentions [
2].
This study will attempt to understand the factors influencing anti-PM2.5 behavioral intentions, by adapting the theory of planned behavior framework to investigate the effects of attitudes, subjective norms, and perceived behavioral control on protective behavioral intentions against PM2.5 [
3]. It will explore the effects of attitudes, subjective norms, and perceived behavioral controls on the behavioral intentions of parents with young children. Predicting the behavioral intentions of parents is essential because of the impact of air pollution on individual behavioral intentions and the issue of air pollution is closely related to sustainable development, and to society at large. The application of the Theory of Planned Behavior, in the field of environmental pollution and air pollution, is important [
3,
4,
5]. Air pollution affects human behavior at the micro level, and previous research has investigated the public health risks associated with atmospheric exposure to PM2.5 for different subpopulations, including race and age [
5]. There is a lack of research into variations in parental protection behavior at the family level in Asia [
1]. Through this research, we hope to learn the relationships between behavior and environmental issues, through understanding parents’ prevention behavior and how this plays out through their children who are exposed to air pollution with PM2.5. In doing so, we also seek a greater understanding of the relationship between psychosocial factors and anti-PM2.5 behavioral intentions [
6].
There appears to be general agreement among social psychologists that most human behavior is goal-directed [
7,
8]. According to Ajzen [
9], being neither capricious nor frivolous, human social behavior can best be described as more or less well-formulated plans. Actions are controlled by intentions, but not all intentions are carried out; some are abandoned altogether while others are revised to fit changing circumstances [
10].
There is a lack of research into this topic, and this is made more glaring considering the increasing impetus globally to deal with the problem of air pollution. Of the United Nations Sustainable Development Goals (SDGs) [
11], air pollution is directly or indirectly related to all 17 SDGs to varying degrees [
12] with 169 items within the 17 goals marked as directly related to air pollution and 8 targets explicitly aimed at reducing it. Elder and Zusman [
13] have highlighted air pollution’s fundamental importance to the SDGs and its impact on the diverse issues involved, including affordable clean energy, dignified employment, economic development, industrial innovation and infrastructure, responsible consumption and production, and transport in sustainable cities and communities. Air pollution is also recognized as an important contributing factor to disease globally [
12]. Lode et al. [
14] proclaim the SDGs as an opportunity to fight air pollution globally. Thus, the global promotion of the SDGs and their aimed realization by 2030 are intricately tied to problems resulting from air pollution and are crucial for its reduction [
15].
In Taiwan, air pollution—especially PM2.5 pollution—is a major environmental issue. Although the level of pollution varies greatly in different areas of Taiwan, significant levels do occur in Taipei City and New Taipei City. One reason for such high incidences in these two cities is the high population and population densities there. New Taipei City has the highest number of deaths attributable to PM2.5 in Taiwan at 874 people/year, while Taipei City is relatively lower, but still high at 619 people/year [
16].
According to Lo et al. [
16], the risk assessment of PM 2.5 is important, not only at the global and national levels, but also at the local, county, and city levels, due to its high variability between locations. Understanding the impact of PM2.5 exposure locally can assist central and local governments in the formulation and development of public health and environmental policies. The relationship between local-level analysis, and air pollution prevention and control behavior has been shown to be important [
17], therefore a deeper understanding of local influences and behaviors between Taipei City and New Taipei City is significant, not only for the local and national understanding of the issues and potential solutions available but also for our understanding globally of the factors involved in behaviors and attitudes for the prevention of damage caused by PM2.5 air pollution.
2. Study Area
In this study about PM2.5 air pollution in Taipei City and New Taipei City, living location is a key factor for our investigation. There are differences in the level of urbanization between Taipei City and New Taipei City and this influences the categorization of our study areas. As a result, in accordance with ‘Lo Chi-hon Stratification’ [
18] and Liu et al. [
19], the six districts of Sanchong, Zhonghe, Yonghe, Banqiao, Xindian, and Xinzhuang have been included in our Greater Taipei group rather than the New Taipei City Group-due to the former’s higher urbanization level. In this study, ‘New Taipei City Group’ thus refers to New Taipei City minus the six townships mentioned above, and ‘Greater Taipei Metropolitan Group’ refers to Taipei City and those additional six townships (
Figure 1).
According to the Department of Budget, Accounting and Statistics, Taipei City Government [
20], the population of Taipei City is 2,524,393 with a population density of 9288/km
2 in 2021. According to the Department of Budget, Accounting, and Statistics, New Taipei City Government [
20], the population of New Taipei City is 4,030,954 and the population density is 1964/km
2 in 2021. In this research, the six townships in New Taipei City with the highest population were grouped into the Greater Taipei Metropolitan Group (
Figure 1) due to their population and population density being more similar to that of Taipei City than that of New Taipei City. The six townships and their population and population density are as follows: Banqiao (557,114 and 24,079/km
2), Sanchong (385,328 and 23,615/km
2), Zhonghe (411,214 and 20,414/km
2), Yonghe (219,364 and 38,392/km
2), Xinzhuang (422,653 and 21,413/km
2), and Xindian (303,532 and 2525/km
2). These six townships exhibit an urbanization level similar to Taipei City.
There are numerous definitions of ‘ex-urban’, and the spatial boundaries of such labeled areas are often uncertain [
21]. According to ‘Lo Chi-hon Stratification’ [
18], seven factors are used to identify the spatial boundaries of ex-urban areas mentioned by Ban and Ahlqvist [
21], including population, population density, demographic characteristics, industrial development, public facilities, financial situation, and geographical environment [
18].
When considering the urban geography of Taiwan, ‘Lo Chi-hon Stratification’ [
18] is an often-cited city hierarchy model that is used to determine geographical and socio-economic divisions in Taipei City and New Taipei City. ‘Luo Chi-hon Stratification’ has been widely used in large-scale surveys involving hierarchical sampling in townships and urban areas of Taiwan, such as the “Basic Survey Plan on Social Change in Taiwan”, by the Academia Sinica, and by Hsiao [
22] in conducting a study of “Environmental Awareness and Value in Taiwan”. Since our study focuses on air pollution control behavior, demographic characteristics including population and population density are vitally important, and will help explain how the demographic characteristics of Taipei City and New Taipei City affect the hypothesis that residents in these two regions have different air pollution control behaviors.
For hierarchical regional geographic models of cities, the Central Place Theory by Christaller [
23] can be used to analyze relationships between towns and cities in the same area. However, the Central Place theory has been criticized as too static, as it does not incorporate the element of time into the development of the central place. In addition, whilst the theory works well in the agricultural sector, it does not satisfactorily apply to industrial or post-industrial areas, because of the diversification of their various services and the uneven distribution of natural resources within [
24]. For this reason, this study uses instead the ‘Urban Dense Zone Model’ (Desakota) [
25], which is the opposite of ‘Megalopolis’ as described by Gottman [
26]. ‘Desakota’ is instead described as a ‘City-village Process’, in an Extended Metropolis Region (EMR) and derives from the ‘Theory of Core-periphery’ proposed by the American-regional planning scholar Friedmann [
22], which states that any region has a ‘Core’, together with multiple ‘Peripheral Regions’. In this model, the core area has a decisive influence on the periphery, while there is simultaneously a high degree of economic growth in the overall metropolitan area.
Fann et al. in the United States found that air pollution in different ‘Regional Spaces’ causes different health hazards [
27] and there are large county-to-city differences in PM2.5 in Taiwan. Concentration levels are closely related to the degree of urbanization, industrial development, and of pollutant transmission [
16]. Everard and Longhurst [
15] argue that clean air is a common human asset and that breathing clean air is a universal human right. However, countries with different levels of development have encountered problems in controlling air quality. Underdeveloped countries actually produce relatively low volumes of air pollution, but they necessarily share cross-regional risks and hazards for the air pollution generated by the economic and industrial development in more highly developed countries. Therefore, air pollution is not a problem for any single individual or country, but a globally important issue that affects the common life and health of all human beings, according to the World Health Organization (WHO).
4. Theories and Hypotheses
In the Rational Choice Theory by Ajzen (1991), he utilizes TPB. The TPB is a commonly used research model used to predict BI [
6,
48] and is one of the most commonly used theories in research studies on the prediction of environmental behaviors [
49,
50]. To study behaviors for avoiding air polluted with PM2.5, attitude, perceived behavioral controls, and subjective norms are known to shape such protective behaviors. The TPB model focuses on the controlled aspects of human decision making and information processing of behaviors, which are guided by a conscious self-regulatory process and by goal-oriented behaviors. One’s AT can affect BI indirectly or directly through one’s SN [
51].
Many environmental education studies have found that environmental action cannot only rely on the improvement of knowledge and technology, but also requires the promotion of psychological traits such as affection and attitude [
52]. One example is the recently released “Taiwan Parents’ Air Pollution Awareness and Children’s Impact Survey Report” [
53], which found that 91.4% of parents are aware that air pollution is becoming worse, but that 84.2% of parents have not established the most basic habits for checking air quality. It can therefore be seen that the improvement of awareness of air pollution, and of technology to monitor it, is not enough to promote the occurrence of actions, an individual’s assessment of air pollution risks, and their willingness to take action also influences actions to prevent and control air pollution [
12].
According to the TPB, attitude refers to the degree of a person’s support for a particular behavior [
10]. The definition of subjective norms is that an individual perceives the overall behavioral perception that a group expects to perform [
10], perceived behavioral controls are defined as the difficulty in consciously performing a particular behavior [
10], and behavioral intentions refer to the degree to which people are inclined to engage in a particular behavior [
10]. Personal factor is an individual’s positive or negative evaluation of performing a behavior, this factor is also termed attitude [
10]. The second determinant of intention is a person’s perception of the social pressures put on them to perform, or not perform, a given behavior. Since it deals with perceived prescriptions, this factor is termed subjective norm. Generally speaking, people intend to perform a behavior when they evaluate it positively and when they believe that people whom they deem as important think they should perform it [
10].
Adding to the empirical evidence from previous studies in relation to the study hypotheses, the use of this model may help explain the relationship between social psychology and people’s preventative behaviors related to environmental problems. Despite the many studies assessing the toxicity and impact of air polluted with PM2.5 on humans and the environment, work on the factors that influence people’s behaviors such as AT and SN toward avoiding air polluted with PM2.5 is lacking [
46]. From the perspective of TPB, it has been found that even when accounting for the predictive variables in TPB, understanding past behaviors can help predict future behaviors [
6].
Insight into the factors that affect people’s behaviors, such as AT and SN, for avoiding air pollution is lacking [
54]. In
Figure 2, we tested the following hypotheses.
Hypothesis 1a and 1b (H1a, H1b). The Greater Taipei Metropolitan Group (H1a) and New Taipei City Group (H1b) possess AT that can affect their respective BI [51,55,56,57,58,59].
Hypothesis 2a and 2b (H2a, H2b). The Greater Taipei Metropolitan Group (H2a) and New Taipei City Group (H2b) possess PBC that can affect their respective BI. The PBC of parents of both groups affects their BI directly [56,57,60].
Hypothesis 3a and 3b (H3a, H3b). AT can affect the PBC of the Greater Taipei Metropolitan Group (H3a) and New Taipei City Group (H3b). The results of studies indicate people’s SN and PBC are significantly related [61,62,63].
Hypothesis 4a and 4b (H4a, H4b). The Greater Taipei Metropolitan Group (H4a) and New Taipei City Group (H4b) possess SN that can affect their respective BI. Their SN and PBC affect BI [64,65], we hypothesized that the PBC of parents with young children living in Taipei and New Taipei affect their SN positively. The results of studies have shown that people’s SN and PBC are significantly related [61,62,63].
Hypothesis 5a and 5b (H5a, H5b). The Greater Taipei Metropolitan Group (H5a) and New Taipei City Group (H5b) follow SN that can affect their respective PBC. Their SN affects their PBC directly and affects BI indirectly [60].
5. Methodology
According to Ajzen [
10], numerous studies supporting the theory of rational action have been advanced through various experiments. In order to provide a comprehensive test of the relationship indicated by the theory, it is necessary to present behavioral intention through pilot studies and then use these theories to construct a standard questionnaire. The resultant questionnaire will contain measurements of the following variables:
- (1)
Behavioral assessments, which are assumed to determine attitudes toward behavior;
- (2)
The motivation to comply with subjective norms;
- (3)
Direct measurement of attitudes and subjective norms;
- (4)
The behavioral intention to perform the act [
10].
Our study followed a purposive and stratified methodology [
66] utilizing Liu’s 2018 approach [
3] to develop, deliver, and collect questionnaires. The participating parents were contacted online via the ‘SurveyCake’ platform. The study received 610 questionnaires using an anonymous answering method online. The questionnaire was adapted from previous environmental behavior surveys, especially those involving AT [
67], SN, PBC [
68], and BI [
67]. The questionnaire was reviewed by six experts of environmental education and health promotion.
A five-point Likert scale was used, as well as statistical software (Smart PLS 3.0, SmartPLS GmbH, Oststeinbek, Germany) for statistical and path analyses. Such analyses are critically important for studies such as ours involving small samples of populations, and serious health concerns in vulnerable populations [
69].
For the demographic characteristics of the 610 respondents, there are 180 males and 303 females in the Greater Taipei Metropolitan group, and 56 males and 71 females in the New Taipei City group. The educational attainment of the parents involved is mainly at the university graduate level, with 62.3% and 59.8% in the Greater Taipei Metropolitan Area and New Taipei City, respectively. This is followed by high school graduate level, with 26.3% in the Greater Taipei Metropolitan area, and 33.8% in New Taipei City.
Regarding the entire yearly income of the participating families, 28% and 29.9% of parents earned more than one million Taiwan dollars in the Greater Taipei Metropolitan area and the New Taipei City, respectively; 15.7% of families in the Greater Taipei Metropolitan earned 0.9 to 1 million Taiwan dollars annually; and 18.1% of families in New Taipei City earned 0.7 to 0.8 Taiwan dollars. Regarding the parents’ occupation as a percentage of the total, the service industry was highest (35% and 34.6% in the Greater Taipei Metropolitan area and New Taipei City, respectively), then manufacturing (28.2% and 26.8%), and then housewife (14.5% and 14.2%). The interpretations are based on sound science methodology and the study was data led [
10].
7. Discussions and Implications
According to Gifford and Nilsson [
76], cultural and ethnic differences can affect pro-environmental behaviors. Fang [
48] pointed out that “There are often different environmental concerns between different races and ethnic groups, and cognitive differences can emerge due to cultural differences”. Fang [
48] also suggested that these differences were related to overall thought structures and the logic of different cultures. As with Taiwan, the significance of living location for influencing outlook and norms is important in other Asian cultures [
77]. There is a lack of research on parental protection behavior in Taiwan in relation to different living locations. Therefore, this study is significant and contributes to a theoretical and practical understanding of this issue. Some studies have argued that children, through their learning about the environment from school education, can also transmit environmental knowledge and attitudes to their parents, thus affecting their parents’ environmental behaviors [
78,
79].
Table 9 compares the results of parents living in different locations. In the New Taipei City Group, there is no statistical significance in the SN to BI. Meanwhile, there is statistical significance in the SN to BI of the Greater Taipei Metropolitan Group. The decision-making process of the parents living in Greater Taipei Metropolitan was more dependent on opinions from others and this lead to the statistical significance in the SN to BI, which indicates those parents’ responsiveness to opinions of people close to them [
80]. The higher level of education in Greater Taipei Metropolitan Group might also influence personal attitudes, which could further affect subjective norms [
81].
Both the Greater Taipei Metropolitan and the New Taipei City Group believed they could limit exposure to air pollution with PM2.5, so there is statistical significance in their PBC to BI. The PBC of the Greater Taipei Metropolitan Group was also affected by AT and SN, which suggests that their PBC is influenced by SN and the people around them, but that they can also develop their own AT toward air pollution protection behaviors [
61,
62,
63]. This could be related to higher levels of education and income among the Greater Taipei Metropolitan Group, which has a strong correlation with PBC [
2].
7.1. Influence of Attitude (AT)
Attitude refers to the degree of a person’s support for a particular behavior according to the TPB [
6]. Our study shows that there are differences in the AT of the two groups with respect to protective behaviors related to air pollution with PM2.5. The path model indicates the AT of both groups can affect their BI directly (β = 0.230, t = 4.516 > 3.29 for the Greater Taipei Metropolitan Group and β = 0.295, t = 2.870 > 2.58 for the New Taipei City Group). These results were demonstrated in the four paths of the Greater Taipei Metropolitan Group and three paths of the New Taipei City Group. The study also shows that there is no direct path for AT-BI in the New Taipei City Group.
The attitude of parents living in New Taipei City, which is more rural compared with the Greater Taipei Metropolitan Group, in regard to their anti-PM2.5 behavior intentions aligned with other empirical studies [
2,
60,
82,
83,
84]. In rural societies, positive attitude is the most important factor in influencing behavioral intentions, followed by internality PBC and subjective norms [
81].
7.2. Influence of Subjective Norm (SN)
The definition of Subjective Norms is that an individual perceives the overall behavior that a group expects to perform [
10]. This study looked at the extent to which an individual’s interaction with others, such as family and friends, will affect one’s views and positions on air pollution, as well as in response to the perceived pressure due to public opinion, news media, and private environmental groups.
These results show that the SN of both groups can directly affect their PBC (Greater Taipei Metropolitan Group: β = 0.445, t = 10.155 > 3.29, New Taipei City Group β = 0.632, t = 7.384 > 3.29); however, only the SN of the Greater Taipei Metropolitan Group can affect BI (Greater Taipei Metropolitan Group: β = 0.315, t = 6.201 > 3.29, New Taipei City Group β = 0.192, t = 1.669 < 1.96). This suggests that the BI of both Groups with regard to air pollution is based on the expectations of people around them, especially in the Greater Taipei Metropolitan Group. SN is influenced by the expectations of those perceived as important to the individual [
6].
This is particularly relevant in the context of Asian society, where norms are more greatly valued [
3]. Individuals are usually willing to fulfill the expectations of other individuals or organizations, potentially due to collectivism as a longstanding characteristic of traditional culture. Asian residents are more concerned about the views of other individuals or organizations when deciding whether to take certain actions [
2].
7.3. Influence of Perceived Behavioral Control (PBC)
Ajzen [
85] added the Perceived Behavioral Controls variable (anticipating the likelihood of success or failure of what one wanted to do in the future) to the Theory of Planned Behavior, in order to provide a better interpretation and prediction of behavior, thus forming the framework of planned behavior theory. It is often used to predict behavioral intentions that have positive meaning for individuals or society.
Our research shows that there is statistical significance in the PCB of air pollution prevention behaviors in both groups, greater than SN (Greater Taipei Metropolitan Group: β = 0.334, t = 5.547 > 3.29; New Taipei City Group β = 0.387, t = 2.967 > 2.58). The PBC for the parents of both groups affects their respective BI [
2,
56,
72,
86] associated with air pollution prevention, and that PBC is the main factor influencing their BI.
The PBC of both groups is affected by the same factors in our study. The PBC of both groups was affected by their SN, which has been demonstrated in other research [
46,
47,
48]. The results suggest that the New Taipei City Group possesses a PBC that is affected by SN, rather than AT toward air pollution protection behaviors intention (BI), but that AT can positively affect PBC directly [
62,
86].
7.4. Analysis of Behavioral Intention (BI)
Behavioral Intentions refer to the degree to which people are inclined to engage in a particular behavior [
9]. The behavior is generated by logical thinking of Behavioral Intentions: after the selection, if other decisions contribute to the behavior, then the higher the people’s environmental Behavioral Intentions, and so the higher the actual execution rate of the behavior. In this study, one of the objectives is to find out the degree of willingness to understand the problem of air pollution, cooperate with government policies, and adopt decisive strategies.
The results of the t-test show that the air pollution prevention behavior of the Greater Taipei Metropolitan Group was statistically lower than that of the New Taipei City Group, which means the New Taipei City Group has greater prevention behavior than the Greater Taipei Metropolitan Group toward air polluted with PM2.5.
Only four of the five hypotheses in the New Taipei City Group were supported and formed three paths including AT to BI, AT to PBC to BI, and SN to PBC to BI. The PBC in the New Taipei City Group was based on the SN and PBC around them, which ultimately contributed to the BI of their protective behavior. In the Greater Taipei Metropolitan Group model, the five hypotheses were all supported. Statistically higher scores showed a more diverse and complex model and formed four paths including AT to BI, AT to PBC to BI, SN to BI, and SN to PBC to BI.
In the New Taipei City Group model, four hypotheses were supported. Only the SN to BI path was not supported. The BI generated in the Greater Taipei Metropolitan Group shows a series of more complex and diverse pathways that have significantly lower scores than the New Taipei City Group, which means the decision progress of the Greater Taipei Metropolitan Group is more complex than those of the New Taipei City Group. The Greater Taipei Metropolitan Group can form BI from SN directly. Previous research suggests that higher education level may contribute to the above situation [
26].
The study found a difference between SN and BI. The SN of the New Taipei City Group does not directly affect their BI, but the SN is affected indirectly through the BI [
60] in the Taipei City Group. This means that those parents’ expectations toward the opinions of people whom they care about the most [
6]. The higher level of education in the Greater Taipei Metropolitan Group might also influence personal attitudes, which could further affect subjective norms [
77,
87]. Besides the direct influence of the BI, the New Taipei City Group are influenced by their SN [
64,
65]. This means the BI in both groups is partly derived from meeting the expectations of others around them, not just their own PBC, but via pressure exerted by SN and cultural values.
7.5. t-Test
The result of the independent sample
t-test in
Table 10 found that the AT, SN, PBC, and BI of the New Taipei City Group are statistically higher than the Greater Taipei Metropolitan Group. This means that parents in the New Taipei City Group agree more with the questionnaire statements.
However, SN does not affect BI in the New Taipei City Group and even the scores in the t-test are higher than in the Greater Taipei Metropolitan Group. This may be as a result of both SN and PBC being high in the t-test, thus decreasing the importance of SN in terms of affecting BI.