Development and Validation of the Dust Exposure Reduction Behavior Scale

Abstract

This study was conducted to develop the Dust Exposure Reduction Behavior Scale (DERBS) and to validate it for the general population. The scale was developed based on Williamson’s model of everyday life information seeking. For exploratory factor analysis and confirmative factor analysis, 274 subjects aged 18 or older were recruited in South Korea using convenience sampling between February 1 and March 28, 2021. Content validity, convergence validity, discriminant validity, and internal consistency reliability were examined according to DeVellis’ methodology. The DERBS consists of 12 items with four factors: protection from dust, staying inside, sharing information, and health promotion. Convergent validity with the personal environmental health behaviors scale was supported (r = .63, p < .001), and Cronbach’s alpha was .83. Confirmatory factor analysis showed good model fit (χ² = 106.36, p < .001, Q = 2.22, SRMR = 0.06, RMSEA = 0.06, IFI = 0.94, GFI = 0.94, AGFI = 0.97, NFI = 0.90, CFI = 0.94, PNFI = 0.66, PCFI = 0.69). The DERBS is a valid and reliable scale for measuring dust reduction behaviors under circumstances of indoor and outdoor air pollution. This scale can be utilized in educational interventions aiming to enhance environmental health behavior and promote respiratory health outcomes. This DERBS will be able to objectively measure outcomes in investigations of health behavior under circumstances where air pollution is serious and interventions are needed to target environmental health behaviors. It will also contribute to the promotion of environmental health behaviors related to the reduction of fine dust exposure among community residents.

Plain language summary

This study was conducted to develop the Dust Exposure Reduction Behavior Scale (DERBS) and to validate it for the general population. Currently, as environmental pollution worsens, concerns grow proportionately regarding environmental toxins and their potential detrimental effect on human health. Further, it has been shown that a large proportion of air pollution is inhaled in the form of fine dust, which adversely affects respiratory health. However, there is little research on scales developed for measuring fine dust reduction behaviors, as researchers have only developed tools with a more limited scope appropriate for specific research topics. Therefore, a validated tool is needed to measure changes in health behavior to reduce exposure to air pollution caused by fine dust. Accordingly, we intended to develop a tool to measure fine dust reduction behavior for residents living in communities. This scientific tool will be able to objectively measure outcomes in investigations of health behavior under circumstances where air pollution is serious and interventions are needed to target environmental health behaviors. It will also contribute to the promotion of environmental health behaviors related to the reduction of fine dust exposure among community residents. We believe that our study makes a significant contribution to the literature because the effect was confirmed, so it can be used as a basis for the active application of environmental studies.

Keywords

air pollution dust health behavior brief psychiatric rating scale statistical factor analysis

Introduction

Breathing clean air is a basic need of life. However, it has become clear that global warming, climate change, and air pollution are threatening human health. As environmental pollution becomes more serious, researchers have increasingly focused attention on the harmful effects of environmental toxicants on human health. Environmental pollutants refer to chemicals, heavy metals, micro-dust, electromagnetic waves, and radiation that harm the human body and are present in various components of the environment, such as the water, air, and soil (Manisalidis et al., 2020).

Air pollution can be measured with representative indicators of fine dust, ozone, nitric acid, and sulfuric acid. Fine dust is measured as particulate matter 10 (PM10), which has a diameter of 10 µm or less, and ultrafine dust (PM2.5, with a diameter of 2.5 µm or less (World Health Organization, 2021). In Seoul, the PM10 concentrations in 2011 to 2019 were more than twice that of London, Paris, and Tokyo annually, and PM2.5 levels in 2015 to 2019 were more than 1.6 times higher (Air Korea, 2021). In particular, it has been shown that a large proportion of air pollution is inhaled in the form of fine dust, which adversely affects respiratory health, ischemic heart disease, stroke, and cancer (Bougas et al., 2018; Impinen et al., 2018; World Health Organization, 2021).

Fine dust is mainly measured as PM₁₀. Nitrogen oxides, including nitrogen dioxide, and carbon (black carbon) have been measured as the main pollutants in automobile exhaust gas (Bougas et al., 2018; Clemente et al., 2019; Fioravanti et al., 2018; Madhloum et al., 2019). To measure the concentration of pollutants, satellite air quality information can be used, as well as information on pollutant concentrations in road traffic and the distance of residences from major roads and power plants. A prior study found that the presence of green space within 5 km of a child’s residence was associated with protective health effects, as shown by a decrease in systolic blood pressure by 1.2 mmHg and a decrease in diastolic blood pressure by 1.2 mmHg (Madhloum et al., 2019). The effects of cooking fuel—in particular, the use of firewood or coal in comparison to electric stove use—aggravated health problems in pregnant women and infants (Jiang et al., 2015).

Exposure to air pollutants and fine dust in pregnant women and the use of coal or firewood as fuel during pregnancy were associated with low birth weight. In a study in China, the rate of underweight children was 2.51 times higher among pregnant women who used firewood as fuel than among those who used electric stoves, demonstrating women’s vulnerability to indoor air pollution caused by poverty and housework (Jiang et al., 2015). Nitric acid exposure lowered children’s forced vital capacity, a measure of lung function, and increased their susceptibility to repeated lower respiratory tract infections and allergies. Evidence has also been presented for associations of exposure to air pollutants in children with lower or upper respiratory tract infections and decreased lung function (Bougas et al., 2018). Exposure to nitrate gas in the second trimester of pregnancy increased recurrent lower respiratory tract infections and allergy susceptibility in children aged 7 to 8 years and decreased forced expiratory volume by 25% to 75% (Bougas et al., 2018). Moreover, DNA testing of blood samples from 8-year-old children showed that exposure to nitrate in pregnant women reduced the leukocyte telomere length (LTL), a marker of cellular senescence in white blood cells, by 1.5%. Distance from main roads showed an indirect protective effect on children’s health from environmental pollutants, with the LTL increasing by 1.6% as the distance from the main road doubled (Clemente et al., 2019). The concentration of fluoride compounds in pregnant women was associated with an increase in atopic dermatitis and lung disease in children aged 2 to 10 years, and the relationship of a high concentration of fluoride compounds in cord blood with a high incidence of upper respiratory tract infections and lower respiratory tract infections indicates that exposure to chemicals is associated with respiratory diseases in children. High fluoride concentrations were also found to decrease immunity (Impinen et al., 2018).

The Korean Ministry of Environment has distributed fine dust exposure prevention guidelines to prevent the deterioration of health problems in sensitive groups (Korea Disease Control and Prevention Agency, 2021). People living in areas affected by fine dust can check the daily fine dust level through the national fine dust information service. Williamson’s model of everyday life information seeking (ELIS) is useful for explaining how people seek health information as a health behavior and environmental activities (Savolainen, 2010; Williamson, 1998). Therefore, the ELIS model offers an appropriate framework for developing a scale related to the steps that individuals take to reduce fine dust exposure, especially in an interview for the development of preliminary items (Williamson, 1998).

However, there is little research on scales developed for measuring fine dust reduction behaviors, as researchers have only developed tools with a more limited scope appropriate for specific research topics. Therefore, a validated tool is needed to measure changes in health behavior to reduce exposure to air pollution caused by fine dust. Accordingly, we intended to develop a tool to measure fine dust reduction behavior by applying the tool development methodology of DeVellis (2021) for residents living in communities. This scientific tool will be able to objectively measure outcomes in investigations of health behavior under circumstances where air pollution is serious and interventions are needed to target environmental health behaviors. It will also contribute to the promotion of environmental health behaviors related to the reduction of fine dust exposure among community residents.

Methods

Study Design

This methodological study was conducted to develop a scale for measuring health behaviors to reduce fine dust exposure and to test its validity and reliability.

Subjects

The subjects of this study were 274 people 18 years of age or older living in Korean communities. In the exploratory factor analysis (EFA), the number of subjects was determined to be 150 or more, which satisfied the guidelines for a sample 5 to 10 times the number of preliminary questions (DeVellis, 2021). The minimum sample size for the confirmatory factor analysis (CFA) was also suggested as the number of 150 or more (Holbert & Stephenson, 2002). In total, 277 questionnaires were collected, and 274 copies were finally analyzed, excluding three copies with insufficient responses. The criteria for the selection of subjects were residents who were 18 years of age or older, had Korean literacy, lived in a community, and agreed to participate in the study. The exclusion criteria were those who are currently hospitalized for health problems and those who had difficulty understanding the subject and content of the study.

Measurements

The Personal Environmental Health Behavior (PEHB) scale, which was developed to measure women’s personal environmental behaviors (Kim & Kim, 2021), was used for the convergent validity analysis with permissions from the authors. The PEHB scale has 17 items and consists of four sub-areas: seven items for lifestyle, four items for personal goods, three items for food, and three items for dust. These items are scored using a 5-point Likert scale with 1 point indicating “not at all” and 5 points indicating “strongly agree.” A higher score indicates higher environmental health behaviors. Cronbach’s α of the PEHB was .90 in the original study and .92 in this study.

Research Process

Literature Review

A literature review was conducted from November 11 to 17, 2021, using MeSH keywords and CINAHL headings in CINAHL, PubMed, the Cochrane Library, Embase, DBPIA, and RISS. The keywords used in CINAHL Advanced Search were: “Dust AND Health behavior.” The keywords used in PubMed through the MeSH search were: “((‘Dust’[Mesh]) AND ‘Environmental Exposure’[Mesh]) AND ‘Health Behavior’[Mesh].” The keywords used in the Cochrane Central Register of Controlled Trials were: “dust in Title Abstract Keyword AND health behavior in Title Abstract Keyword—(Word variations have been searched).” The keywords used in Embase search were: “dust:ti,ab,kw AND ‘health beha’:ti,ab,kw.” The Korean literature search used the keywords “fine dust health behavior” and “fine dust reduction behavior” in DBPIA and RISS. As a result of the searches, 17 (CINAHL), 21 (PubMed), 30 (Cochrane Library), six (Embase), 15 (DBPIA), and 29 (RISS) related articles were identified. A Google Scholar search was added with the keywords of “dust reduction health behavior” and “air pollution mitigating behavior,” and nine articles was identified. The authors searched and read the abstracts of all articles, and appropriate articles were selected independently. The numbers of suitable articles were three from CINAHL, four from PubMed, two from the Cochrane Library, two from Embase, seven from DBPIA, and seven from RISS. Five duplicate articles were removed. A hand search was performed in the reference list of searched articles to avoid selection bias, and four articles were added. Finally, 24 articles remained for the development of the preliminary items. Seventy-two items were extracted from the 24 articles by two researchers.

The literature review showed the existed scales for prevention from the dust. The Iranian scale for patients with cardiovascular disease contained cue to action to prevention from exposure dust according to health belief model (Noroozi et al., 2020). The dust exposure protection behavior scale for stone crushing workers in Thailand (Samana & Ketsakorn, 2019) was found. However, they were not for general people but for specific population. The behavioral strategy for pregnant women according to the trans theoretical model (Araban et al., 2017) and personal strategy for minimize the effects of air pollution on respiratory health (Carlsten et al., 2020) were developed to minimize the exposure from air pollution. Korean Ministry of Environment governmental guideline (Korea Disease Control and Prevention Agency, 2021), Government Hong Kong Environmental Protection Department (2022), and WHO global air quality guidelines (World Health Organization, 2021) were useful to prevent exposure from micro-dust and diminish the air pollution. The checklist using items derived from Korean environment governmental guideline was used in the previous researches in Korea (E. Park et al., 2018; M. K. Park & Kim, 2020). However, there were no validated and standardized scales to measure the behaviors regarding reduction from micro-dust for general population.

Dust reduction behaviors correlated with health related concepts. Perceived barrier, enabling factor, perceived susceptibility correlated with protection behavior (Samana & Ketsakorn, 2019). The awareness of dust hazards, sense of responsibility, personal norms, and company behavior were correlated with dust reduction behavior based on the norm activation model (Kaluarachchi et al., 2021). The self-efficacy, decision balance, perceived benefit, and barriers changed with exposure minimization intervention of air pollution (Araban et al., 2017). There were also several studies regarding dust reduction behaviors because of the ambient air pollution in Korea. Health behavior related to particulate matter in older adults correlated with risk perception, attitude, and level of experience of diseases (M. K. Park & Kim, 2020). Respiratory disease patient (Ham et al., 2020) and college students (E. Park et al., 2018) adopted particulate matter related health behaviors with higher knowledge, risk perception, and attitude. Therefore, development of validated scale for general population will contribute to empirically verify the relationship between concepts of the environmental health behaviors in the everyday life in the context of air pollution and climate change.

In-Depth Interview

In-depth interviews for extracting the preliminary items were conducted from December 20 to 30, 2021. The interview participants were selected by researchers in Seoul and Cheonan in Korea using convenience and theoretical sampling. Seven face-to-face and three phone interviews were conducted, and the participants chose the interview modality that they preferred. The interviews ranged from 45 min to 1 hr and 30 min, taking an average of 50 min. The interview locations were the researcher’s office, cafes, and meeting rooms. The interviewees were between 20 and 62 years old, and there were three housewives, two office workers, two teachers, one student, one disability assistant, and one social worker. Three participants were men and seven were women. Semi-structured open-ended questions were constructed as a guide for the research questions through inter-researcher meetings. The theoretical sampling was performed according to relevant variables, including personal characteristics, socioeconomic circumstances, values, lifestyle, and physical environment according to the ELIS model (Williamson, 1998). The research questions were as follows: “Is there anything you do on a daily basis to prevent fine dust exposure?,”“How do you try to protect your health on a fine dust day?,”“Is there anything you do to protect your health from fine dust indoors?,” and “Is there anything you can do to improve your health from fine dust outdoors?” The interviews were recorded after participants signed a consent form and explanation, and the content was analyzed by repeatedly reading the text organized by the transcription program. The researchers independently extracted questions from the interviews, coded them in the Excel program, and extracted a total of 41 preliminary items on fine dust reduction behavior through a researcher meeting. The researchers had more than 15 years of experience in qualitative research, and questions extracted differently among researchers were included in the preliminary items when there was a consensus. It was checked whether the preliminary items were the same as those stated to any of the participants, and no items were dropped. Finally, 41 items were extracted from the interviews by two researchers.

Content Validity

From the 41 items extracted from the interviews and the 72 items extracted from the literature review, 33 duplicate items were excluded, yielding 80 items. In a meeting of the two researchers, 19 inappropriate items were deleted, yielding a final total of 61 items.

The first round of content validation for the 61 extracted items was reviewed by three experts from January 2 to 7, 2022 via e-mail. The experts were two professors of nursing with more than 20 years of experience and one professor of environmental education researching fine dust. The content validity index (CVI) was measured using a 4-point Likert scale, ranging from 1 point (“not relevant”) to 4 points (“highly relevant”) (Lynn, 1986). The average item-level CVI (I-CVI) and the scale-level CVI (average of content validity index for scale, S-CVI/ave) were analyzed. For the I-CVI, the ratio of “quite relevant” and “highly relevant” responses for each item was set to 1.00 or more as appropriate for a CVI professional pool ranging from three to five members, and for the S-CVI/ave, a value of 0.80 or more was set, as has been recommended for a new scale (Polit et al., 2007). The S-CVI/au refers to a scale’s content validity universal agreement and the proportion of items rated as 4 points among all items. An S-CVI/au over 0.80 was considered acceptable (Polit et al., 2007). The I-CVI of items ranged from 0.33 to 1.00, S-CVI/ave was 0.76, and S-CVI/au was 0.75. The 31 items with an I-CVI under 1.00 were deleted from the preliminary items.

The second round of content validation for 30 items was reviewed by the same three experts from January 11 to 15, 2022 via e-mail. Thirty-one items were deleted in the first round of CVI testing. In the second round of content validation, no item was deleted from the 30 items by the same three professionals as in the first round of CVI. The final I-CVI of all items was 1.00, S-CVI/ave was 1.00, and S-CVI/au was 0.92. The Fleiss kappa coefficients were used to measure inter-rater agreement, and 30 items had coefficients of .23 (p = .032). The strength of agreement was evaluated as fair, with Fleiss kappa coefficients ranging from .21 to .40 (Landis & Koch, 1977).

Pilot Test

A pilot study was conducted to test the feasibility from January 22 to 28, 2022. Two participants aged 32 and 60 were asked whether they understood the meaning of the items after reading each of them. The feasibility was evaluated to assess test feasibility by verifying the comprehensibility of the items. The possible scores for comprehensibility ranged from 1 point, indicating “I do not understand at all,” to 5 points, indicating “very well understood,” and the items scored an average of 4.83 points. No item was rated by two participants as being difficult to comprehend, and the average response time was 7 min.

Data Collection

This study was approved by the institutional review board of *** University (KNU-IRB-2021-42). Informed consent was obtained from the participants. The data were collected between February 1 and March 28, 2022. Two researchers collected questionnaires from churches, welfare centers, schools, libraries, companies, and offices across Seoul, Pyeongtaek, Anseong, Cheonan, Asan, Daejeon, and Gongju in South Korea using the convenience sampling method. The subjects of the survey were individuals who lived in Korea in the age range of 18 to 65 years. Due to social distancing measures related to the COVID-19 quarantine, an online survey was administered to 52% of the subjects after an explanation of the study. The other 48% of subjects signed a consent form, subjects completed the questionnaire and returned it face-to-face. Before the start of the survey, the researchers provided a written explanation of the purpose, methods, advantages, and disadvantages of study participation. When interviews were performed, a written explanation was given of the same process to protect participants’ rights, after which the participants signed a consent form. After the interview and survey, the researchers provided gifts to the participants.

Data Analysis

The data were analyzed using IBM SPSS/WIN 26.0 and AMOS/WIN 26.0 (IBM, Corp., Armonk, NY, USA). EFA was undertaken using varimax rotation and principal component analysis. The Kaiser-Mayer-Olkin (KMO) and Bartlett sphericity tests were performed. The factor loading values, scree chart, and variance were comprehensively evaluated. The criteria for item selection were an Eigen value of factor exceeding 1.00, a factor loading value exceeding .40, an item commonality of .30 or more, an item-total correlation (ITC) of .40 or more, and a Pearson correlation between .30 and .80 (Pett et al., 2003). For convergent validity, the Pearson correlation coefficient with the PEHB scale (Kim & Kim, 2021) was analyzed. The fit of the model was identified using CFA, and the fit indices used were the Q-value (chi-square/degree of freedom), standardized residual root mean (SRMR), root mean square error of approximation (RMSEA), incremental fit index (IFI), goodness of fit index (GFI), adjusted goodness of fit index (AGFI), net fit index (NFI), comparative fit index (CFI), parsimonious normed fit index (PNFI), and parsimonious comparative fix index (PCFI). The convergent validity and discriminant validity were analyzed using the average variance extracted (AVE) and composite reliability (CR), which are the average variables of each factor explained by the items corresponding to each factor (Brown, 2015). As a reliability test, Cronbach’s α was calculated to confirm internal consistency, and the criterion was set at .70 or higher (DeVellis, 2021).

Results

General Characteristics of Subjects

EFA and CFA were performed using 274 subjects aged from 18 to 65 years. The average age was 23.05 ± 7.67 years. In addition, 84.7% were women, and 91.6% were unmarried. Most subjects (59.9%) had graduated from high school, and 85.4% were not employed. Most subjects (57.7%) lived in cities. The average distance from a big road with four or more lanes was 412.95 ± 388.83 m (Table 1).

Table 1.

Demographic Characteristics of Subjects (N = 274).

Categories	Subcategories	n (%)	Mean ± SD	Range
Age (year)			23.05 ± 7.67	18–65
Gender	Female	232 (84.7)
Gender	Male	42 (15.3)
Marital status	Married	23 (8.4)
Marital status	Unmarried	251 (91.6)
Education	Elementary school	2 (0.7)
	Middle school	81 (29.6)
	High school	164 (59.9)
	Associate or bachelor’s	27 (9.9)
Employment	Employed	40 (14.6)
Employment	Not employed	234 (85.4)
Residence	City	158 (57.7)
	Country	94 (34.3)
	Urban-rural interface	22 (8.0)
Distance from the nearest large road† (m)			412.95 ± 388.38	3–12,200

Note. ^†A road with four or more lanes.

Exploratory Factor Analysis

EFA was conducted with 274 subjects by applying principal component analysis and varimax rotation to the 30 preliminary items selected based on content validity. The model had a KMO value of .81, which was above the threshold of .60, and a χ² value of 1,203.63 (df = 105, p < .001), which was found to be suitable by the Bartlett sphericity test. Fourteen items with ITC scores of less than .40 were removed. As a result of the factor analysis for the remaining 16 items, one item with a commonality score of less than .30 was removed. The factor analysis of 15 items showed support for five factors. The total cumulative variance explained was 61.87%. However, the correlation coefficients between factors were under .30 and did not show statistical significance. The scree chart marked four factors.

Therefore, the second EFA was conducted by applying principal component analysis and varimax rotation to the 15 items with four fixed factors. The second model had a KMO value of .80 and a χ² value of 1,209.12 (df = 105, p < .001), and the total cumulative variance explained was 64.39%. Two items were deleted because the commonality scores were .25 and .29.

The third EFA was conducted by applying principal component analysis and varimax rotation to the 13 remaining items. The third model had a KMO value of .80, a χ² value of 1,105.02 (df = 78, p < .001), and a total cumulative variance explained of 63.48%. One item was deleted because it was the only item in a factor.

The last EFA was conducted by applying principal component analysis and varimax rotation to the 12 remaining items (Table 2). The final model had a KMO value of .80, a χ² value of 1,065.93 (df = 66, p < .001), and a total cumulative variance explained of 66.84%. The commonality scores ranged from .48 to .68, and the variance explained ranged from 7.71% to 32.49%. The Pearson correlation coefficients between factors ranged from .30 to .76 (p < .001), the correlations of each factor with the total score ranged from .62 to .84 (p < .001) (Table 3), and the ITC scores ranged from .40 to .71 (p < .001).

Table 2.

Exploratory Factor Analysis of the Dust Exposure Reduction Behavior Scale (DERBS) (N = 274).

Factors	Items	Factor loading				Item-total γ	Variance %	ѐ^†
1. Protection from dust	11. I wear a long sleeve clothes to protect myself from dust.	.85	.06	.06	.14	.61	32.49	.73
	10. I wear a hat to protect myself from dust.	.84	.02	.16	.12	.60
	8. I dust off my clothes after going out.	.63	.13	.23	.12	.58
	16. I use less dirty fuel.	.50	.26	−.32	.37	.48
2. Staying inside	3. I avoid exercise on dusty days.	.12	.83	.12	−.07	.45	13.63	.70
	1. I avoid going out on dusty days.	.32	.72	.16	.12	.62
	13. I close the window on dusty days.	−.09	.68	.15	.11	.40
3. Sharing information	29. I check the fine dust guide.	.13	.16	.83	.21	.57	11.01	.86
3. Sharing information	28. I share the fine dust guide with my acquaintances.	.16	.26	.82	.12	.53
4. Health promotion	30. I exercise indoors to promote immunity.	.04	−.04	.11	.84	.62	7.71	.71
	32. I eat healthy food that is good for the respiratory system.	.34	.06	.19	.71	.67
	31. I check health conditions that are associated with vulnerability to dust.	.25	.21	.47	.55	.71
Total							66.84	.83

Note. Eigenvalue = 1.01–4.68, Kaiser-Meyer-Olkin = .80, Bartlett test of sphericity = 1,065.93, Degrees of freedom = 66.

†

Cronbach’s ѐ.

p < .001.

Table 3.

Correlation Between Factors of the Dust Exposure Reduction Behavior Scale (DERBS) (N = 274).

	Factor 1	Factor 2	Factor 3	Factor 4	Total
	r (p)
Factor 1	1
Factor 2	.30 (<.001)	1
Factor 3	.32 (<.001)	.30 (<.001)	1
Factor 4	.47 (<.001)	.30 (<.001)	.76 (<.001)	1
Total	.76 (<.001)	.62 (<.001)	.75 (<.001)	.84 (<.001)	1

The Dust Exposure Reduction Behavior Scale (DERBS) consisted of four factors. Factor 1 was named “protection from dust” with four items, factor 2 was named “staying inside” with three items, factor 3 was named “sharing information” with two items, and factor 4 was named “health promotion” with three items (Table 2).

Convergent Validity

The developed scale showed statistically significant positive correlations with Kim and Kim’s (2021) PEHB scale, proving its convergent validity. The correlations between the DERBS and the total and subscale scores of the PEHB were: total PEHB (r = .63, p < .001), lifestyle (r = .45, p < .001), personal goods (r = .37, p < .001), food (r = .33, p < .001), and dust (r = .63, p < .001). These correlations were all significant and positive, demonstrating convergent validity (Table 4).

Table 4.

Correlations Between the Dust Exposure Reduction Behavior Scale (DERBS) and the Personal Environmental Health Behavior (PEHB) Scale (N = 274).

Category	Subcategories	Dust Exposure Reduction Behavior Scale r (p)
Total Personal Environmental Health Behavior (PEHB)		.63 ( < .001)
	Lifestyle	.45 ( < .001)
	Personal goods	.37 ( < .001)
	Food	.33 ( < .001)
	Dust	.63 ( < .001)

Confirmatory Factor Analysis

CFA was performed using the maximum likelihood method with 274 subjects. All free parameters of the model were statistically significant. The standardized regression coefficients of individual measured variables ranged from .39 to .85, and the correlations between latent variables ranged from .35 to .99. The model fit of the DERBS was shown by the following parameters: χ² = 106.36 (p < .001), Q = 2.22, SRMR = 0.06, RMSEA = 0.06, IFI = 0.94, GFI = 0.94, AGFI = 0.97, NFI = 0.90, CFI = 0.94, PNFI = 0.66, and PCFI = 0.69. The criteria were χ² (p > .05), Q ≤ 3, SRMR ≤ 0.08, RMSEA ≤ 1.00, IFI ≥ 0.90, GFI ≥ 0.90, AGFI ≥ 0.90, NFI ≥ 0.90, CFI ≥ 0.90, PNFI ≥ 0.50, and PCFI ≥ 0.50. All criteria were satisfied except for the significance of χ² (p > .05). Overall, the CFA model fit indices of this tool were good (Brown, 2015) (Table 5).

Table 5.

Confirmatory Factor Analysis of the Dust Exposure Reduction Behavior Scale (DERBS) (N = 274).

Items	Factors	Unstandardized estimates		Standardized estimates	Standard error		Critical ratio	p	AVE		Composite reliability
11	Protection from dust	1.56		0.82	0.05		6.10	<.001
10		1.39		0.81	0.04		7.10	<.001
8		1.15		0.56	0.08		7.11	<.001
16		1.00		0.46	0.09		—	—	0.88		0.96
3	Staying indoors	2.14		0.70	0.09		5.51	<.001
1		2.42		0.85	0.09		5.60	<.001
13		1.00		0.39	0.07		—	—	0.85		0.94
29	Sharing information	1.05		0.83	0.08		10.6	<.001
28	Sharing information	1.00		0.82	0.08		—	—	0.90		0.94
30	Health promotion	0.71		0.52	0.10		9.57	<.001	0.85		0.94
32		0.88		0.69	0.08		7.56	<.001
31		1.00		0.78	0.07		—	—
Fitness index	χ² (p)	Q	SRMR	RMSEA	IFI	GFI	AGFI	NFI	CFI	PNFI	PCFI
	106.36 (<.001)	2.22	0.06	0.06	0.94	0.94	0.97	0.90	0.94	0.66	0.69

Note. AGFI = adjusted goodness of fit index; AVE = average variance extracted; CFI = comparative fit index; GFI = goodness of fit index; IFI = incremental fit index; NFI = normed fit index; PCFI = parsimonious comparative fit index; PNFI = parsimonious normed fit index; Q = chi-square/degree of freedom; RMSEA = root mean square error of approximation; SRMR = standardized root mean residual.

In CFA, the critical ratio values ranged from 5.51 to 9.57, all of which were 1.97 or higher. The AVE values ranged from 0.85 to 0.90. The composite reliability values ranged from 0.94 to 0.96, all of which were over 0.70; thus, the convergent validity was good. The AVE values exceeded the highest squared correlation value (Kline, 2016), which was 0.58 (=0.76²) (Tables 3 and 5).

Internal Consistency Reliability

Cronbach’s α was calculated as a measure of internal consistency reliability for each factor of the developed tool. Cronbach’s α was .83 for the scale overall, .73 for factor 1, .70 for factor 2, .86 for factor 3, and .71 for factor 4; these values were all over .70, indicating acceptable reliability (Table 2).

Discussion

The DERBS developed in this study can be used to measure health behaviors for mitigating dust exposure in the context of air pollution. This scale showed high validity and reliability for assessing people’s environmental behaviors. When people consider how to respond to ambient air pollution, they seek information to reduce exposure to fine dust by acquiring information and changing their everyday life. The ELIS model proposed by Williamson (1998) is appropriate for analyzing how people seek out information and engage in health behaviors to reduce fine dust exposure. Individuals obtain information through purposeful and incidental information seeking. The classic model explains the information network of individuals and the resources of institutions for health behavior in terms of an ecological diagram. The individual is in the center of a concentric circle, which extends outward to intimate personal networks, wider personal networks, mass media, and institutional sources. The peripheral areas that influence an individual’s behavior are the physical environment, individual characteristics, socioeconomic circumstances, values, and individual lifestyle (Case & Given, 2016). The revised ELIS model (Savolainen, 2010) supported that way of life in response to fine dust involved interweaving between actively engaging in fine dust reduction behaviors and passively seeking information from family members, friends, neighbors, social networks, and institutional guidelines. The main type of mastery of life was in accordance with factors of the DERBS in this study. Factor 1 (“protection from dust”) corresponded to “optimistic-cognitive,” factor 2 (“staying inside”) corresponded to “pessimistic-cognitive,” factor 3 (“sharing information”) corresponded to “defensive-affective,” and factor 4 (“health promotion”) corresponded to “pessimistic-affective” (Figure 1). The expanded ELIS model (Savolainen & Thomson, 2021) suggested the information seeking, sharing, and use interact with in the individual life. Optimistic-cognitive mastery of life was defined as reliance on positive health outcome, pessimistic-cognitive domain characterized by a less ambitious way, defensive-affective practice was concerning the solvability of dust exposure, and pessimist-affective behavior adopted the strategy of improvement health (Savolainen, 1995). This reconstructive model also reflected meaning and insight of the DERBS in the context of activities including information seeking, sharing, responding, and health practice (Savolainen & Thomson, 2021).

Figure 1.

Conceptual framework of Dust Exposure Reduction Behavior Scale (DERBS) according to everyday life information seeking (ELIS).

Factor 1 (“protection from dust”) includes items regarding the use of personal protective equipment through wearing a long-sleeve shirt or hat, cleaning dust after going out, and not generating particulate matter from cooking fuel. These items correspond to the behaviors for mitigating dust pollution presented by Kaluarachchi et al. (2021). The Dust Exposure Prevention Questionnaire for Cardiovascular Patients developed by Noroozi et al. (2020) includes an item for wearing a mask in the context of protection from dust. In this study, the mask-wearing item was deleted through EFA due to a low correlation coefficient with the total score of the scale. It should be noted that the environment inside a mask involves high humidity, particles, and a germ-friendly atmosphere, which can be conducive to asthma or skin rashes. N95 masks can protect from dust outdoors, but mask wearing at all times is not recommended (Carlsten et al., 2020). Fuel toxicity depends on the chemical composition of energy sources. Fuels with low gas emissions should be used indoors, because wood, pellets, and charcoal generate large amounts of heavy metals and persistent organic compounds (Sofia et al., 2020).

Factor 2 (“staying inside”) contains items regarding the avoidance of outdoors exercise, not going outside, and closing the window on dusty days. These items are similar to those presented by Noroozi et al. (2020). Outdoor exercise is generally recommended for health, but it increases cardiovascular disease susceptibility under conditions of extreme air pollution. The outdoor activities were risky in the heavy pollution from vehicle, ozone, chemical irritants, and road dust due to lack of information ambient air pollution (Laumbach et al., 2015). Aeration is also recommended by using filters with window-opening and air conditioning if the weather is not extreme under conditions of air pollution. Therefore, dust exposure reduction behaviors should depend on the outside weather information.

Factor 3 (“sharing information”) includes items regarding checking the guidelines related to fine dust and sharing that information with acquaintances. Media warnings are effective for disseminating information to community members widely, easily, and quickly (Noroozi et al., 2020). Governments have proposed national guidelines or notifications through media and social network services (e.g., Government Hong Kong Environmental Protection Department, 2022; Korea Disease Control and Prevention Agency, 2021; United States Environmental Protection Agency, 2014). Sharing information is meaningful from the standpoint of altruistic behavior, unlike the other behaviors in the DERBS. Awareness of air pollution changes can shape human behavior. In the setting of local poor air quality, residents’ environmental affection mediated pro-environmental behaviors (Sheng et al., 2020). Alerts from individual monitors can also promote health behaviors (Carlsten et al., 2020).

Factor 4 (“health promotion”) includes items regarding exercising indoor to promote immunity, eating healthy food that is good for the respiratory system, and checking one’s health conditions that are associated with vulnerability to dust. A healthy diet is important for one’s general health status, and in particular, antioxidants from fruits and vegetables can neutralize free radicals generated from air pollutants (Sofia et al., 2020). Maximizing good health by monitoring and caring for one’s body can help prevent cardiovascular, respiratory, immune, and skin diseases related to air pollution (Carlsten et al., 2020). Alternative health behaviors have been recommended, such as substitution of outdoor with indoor exercise when severe air pollution is forecast (Laumbach et al., 2015). Factor 4 showed that compensatory health behavior (e.g., diet and exercise) was an important way to mitigate stress related to the air pollution. This scale has the advantages of being short and having good CVI, construct validity, convergent validity, and good reliability due to the rigorous scale development. The factors cover the relevance, comprehensiveness, and comprehensibility of dust reduction behaviors in the meaning and conceptual context of health behaviors. Therefore, healthcare professionals can utilize the DERBS to assess health behaviors and can implement health promotion intervention programs among the general population in air-polluted societies.

This study has several limitations. First, data were only collected from inhabitants of seven cities in South Korea, limiting the generalizability of the scale for measuring health behaviors in other populations. In a future study, more diverse participants should be recruited from a broader range of regions, and representative sampling such as quota or systematic sampling should be employed to enhance generalizability. Second, the validity of the developed scale could be enhanced by conducting known-group validation between a group with environmental diseases related to exposure to micro-dust and a healthy group, or by evaluating predictive validity through an analysis of the correlation between the DERBS score and environmental disease occurrence. Lastly, the scale’s reliability could be improved by evaluating test-retest reliability with another sample of the general population after testing the internal consistency reliability.

Conclusion

The purpose of this study was to develop the DERBS for the general population and to test its validity and reliability. The validity and reliability of the measurement tool were confirmed using various metrics, and the DERBS can therefore be used easily and objectively to evaluate health behaviors. The DERBS can be used to evaluate environmental health behaviors among members of the general public. However, a limitation of this study is that it generalized data collected from certain areas in South Korea. The DERBS should be refined in the future by evaluating and tailoring it to various specific regions susceptible to air pollution and diverse patients, especially those with environmental diseases.

Footnotes

Acknowledgements

This work was supported by the National Research Foundation of Korea (NRF) Grant funded by the Korea government (MIST).

Authorship Statement

We acknowledge that all authors meet the authorship criteria according to the latest guidelines of the International Committee of Medical Journal Editors, and that all authors are in agreement with the final version of the manuscript.

Author Contributions

Author responsibility: Study conception and design acquisition—KHK and HSW; Data collection—KHK and HSW; Analysis and interpretation of the data—KHK; Drafting and critical revision of the manuscript—KHK.

Consent to Participate

Informed consent was obtained from the subjects.

Consent for Publication

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This work was supported by the National Research Foundation of Korea (NRF) Grant funded by the Korea government (MIST) (No. RS-2023-00239284).

Ethical Approval

This study was approved by the Institutional Review Board of Kongju National University (KNU-IRB-2021-42).

Data Availability Statement

Please contact the corresponding author for data availability.

ORCID iD

Hyun Kyoung Kim

References

Air Korea. (2021, October 26). Global air pollution. https://www.airkorea.or.kr/web/contents/contentView/?pMENU_NO=127&cntnts_no=4

Araban

Tavafian

S. S.

Zarandi

S. M.

Hidarnia

A. R.

Burri

Montazeri

(2017). A behavioral strategy to minimize air pollution exposure in pregnant women: A randomized controlled trial. Environmental Health and Preventive Medicine, 22(1), 26. https://doi.org/10.1186/s12199-017-0633-8

Bougas

Rancière

Beydon

Viola

Perrot

Gabet

Lezmi

Amat

De Blic

Just

Momas

(2018). Traffic-related air pollution, lung function, and host vulnerability. New insights from the Paris Birth Cohort. Annals of the American Thoracic Society, 15(5), 599–607. https://doi.org/10.1513/AnnalsATS.201711-900OC

Brown

T. A.

(2015). Confirmatory factor analysis for applied research (2nd ed., pp. 31–32). The Guilford Press.

Carlsten

Salvi

Wong

Chung

K. F.

(2020). Personal strategies to minimize effects of air pollution on respiratory health: Advice for providers, patients and the public. The European Respiratory Journal, 55(6), 1902056. https://doi.org/10.1183/13993003.02056-2019

Case

D. O.

Given

L. M.

(2016). Looking for information: A survey of research on information seeking, needs, and behavior. Part 3. Models and theories of information behavior (4th ed., pp. 161–162). Emerald Group.

Clemente

Vrijheid

Martens

D. S.

Bustamante

Chatzi

Danileviciute

de Castro

Grazuleviciene

Gutzkow

K. B.

Lepeule

Maitre

(2019). Prenatal and childhood traffic-related air pollution exposure and telomere length in European children: The HELIX project. Environmental Health Perspectives, 127(8), 87001. https://doi.org/10.1289/EHP4148

DeVellis

R. F.

(2021). Scale development: Theory and applications (3rd ed.). Sage.

Fioravanti

Cesaroni

Badaloni

Michelozzi

Forastiere

Porta

(2018). Traffic-related air pollution and childhood obesity in an Italian birth cohort. Environmental Research, 160, 479–486. https://doi.org/10.1016/j.envres.2017.10.003

10.

Government Hong Kong Environmental Protection Department. (2022). Air quality health index. http://www.gov.hk/en/residents/environment/air/aqhi.htm

11.

Ham

Choi

S. H.

Choi-Kwon

(2020). Factors affecting particulate matter-related health behaviors of patients with pulmonary disease. Korean Journal of Adult Nursing, 32(5), 504–514. https://doi.org/10.7475/kjan.2020.32.5.504

12.

Holbert

R. L.

Stephenson

M. T.

(2002). Structural equation modeling in the communication sciences, 1995–2000. Human Communication Research, 28(4), 531–551.

13.

Impinen

Nygaard

U. C.

Lødrup Carlsen

K. C.

Mowinckel

Carlsen

K. H.

Haug

L. S.

Granum

(2018). Prenatal exposure to perfluoralkyl substances (PFASs) associated with respiratory tract infections but not allergy- and asthma-related health outcomes in childhood, Environmental Research, 160, 518–523. https://doi.org/10.1016/j.envres.2017.10.012

14.

Jiang

Qiu

Zhou

Cui

Lerro

Lin

Zhang

(2015). Exposure to cooking fuels and birth weight in Lanzhou, China: A birth cohort study. BMC Public Health, 15, 712. https://doi.org/10.1186/s12889-015–2038-1

15.

Kaluarachchi

Waidyasekara

Rameezdeen

Chileshe

(2021). Mitigating dust pollution from construction activities: A behavioural control perspective. Sustainability, 13(16), 9005. https://doi.org/10.3390/su13169005

16.

Kim

H. K.

Kim

H. K.

(2021). Development and validation of women’s environmental health scales in Korea: Severity, susceptibility, response efficacy, self-efficacy, benefit, barrier, personal health behavior, and community health behavior scales. Korean Journal of Women Health Nursing, 27(2), 153–165. https://doi.org/10.4069/kjwhn.2021.06.21

17.

Kline

R. B.

(2016). Principles and practice of structural equation modeling. The Guilford Press.

18.

Korea Disease Control and Prevention Agency. (2021). Health guidance to prevent fine dust exposure. Retrieved November 23, 2021, from https://kdca.go.kr/contents.es?mid=a20205070100

19.

Landis

Koch

(1977). The measurement of observer agreement for categorical data. Biometrics, 33(1), 159–174. https://doi.org/10.2307/2529310

20.

Laumbach

Meng

Kipen

(2015). What can individuals do to reduce personal health risks from air pollution? Journal of Thoracic Disease, 7(1), 96–107. https://doi.org/10.3978/j.issn.2072-1439.2014.12.21

21.

Lynn

M. R.

(1986). Determination and quantification of content validity. Nursing Research, 35(6), 382–385. https://doi.org/10.1097/00006199-198611000-00017

22.

Madhloum

Nawrot

T. S.

Gyselaers

Roels

H. A.

Bijnens

Vanpoucke

Lefebvre

Janssen

B. G.

Cox

(2019). Neonatal blood pressure in association with prenatal air pollution exposure, traffic, and land use indicators: An ENVIRONAGE birth cohort study. Environment International, 130, 104853. https://doi.org/10.1016/j.envint.2019.05.047

23.

Manisalidis

Stavropoulou

Stavropoulos

Bezirtzoglou

(2020). Environmental and health impacts of air pollution: A review. Frontiers in Public Health, 8(14), 1–13. https://doi.org/10.3389/fpubh.2020.00014

24.

Noroozi

Eisapareh

Bahadori

Ghahremani

Cousins

Mokarami

(2020). Development and validation of dust exposure prevention questionnaire for cardiovascular patients based on the health belief model. BMC Public Health, 20(1), 1779. https://doi.org/10.1186/s12889-020-09871-3

25.

Park

H.-J.

Kim

S.-H.

Min

(2018). The relationships between particulate matter risk perception, knowledge, and health promoting behaviors among college students. Journal of Korean Biological Nursing Science, 20(1), 20–29. https://doi.org/10.7586/jkbns.2018.20.1.20

26.

Park

M. K.

Kim

G. S.

(2020). Factors influencing health behavior related to particulate matter in older adults. Journal of Korean Academy of Nursing, 50(3), 431–443. https://doi.org/10.4040/jkan.19201

27.

Pett

M. A.

Lackey

N. R.

Sullivan

J. J.

(2003). Making sense of factor analysis: The use of factor analysis for instrument development in health care research (pp. 85–166). Sage.

28.

Polit

D. F.

Beck

C. T.

Owen

S. V.

(2007). Is the CVI an acceptable indicator of content validity? Appraisal and recommendations. Research in Nursing & Health, 30(4), 459–467. https://doi.org/10.1002/nur.20199

29.

Samana

Ketsakorn

(2019). Factors associated with preventive health behaviors regarding dust exposure of workers in stone crushing mills in Saraburi, Thailand. Journal of Public Health, 49(2), 234–248.

30.

Savolainen

(1995). Everyday life information seeking: Approaching information seeking in the context of “way of life”. Library & Information Science Research, 17(3), 259–294.

31.

Savolainen

(2010). Everyday life information seeking encyclopedia of library and information sciences (3rd ed., pp. 1780–1789). Taylor & Francis. https://doi.org/10.1081/E-ELIS3-120043920

32.

Savolainen

Thomson

(2021). Assessing the theoretical potential of an expanded model for everyday information practices. Journal of the Association for Information Science and Technology, 73, 511–527. https://doi.org/10.1002/asi.24589.

33.

Sheng

Dai

Pan

(2020). Influence of air quality on pro-environmental behavior of Chinese residents: From the perspective of spatial distance. Frontiers in Psychology, 11, 566046. https://doi.org/10.3389/fpsyg.2020.566046

34.

Sofia

Gioiella

Lotrecchiano

Giuliano

(2020). Mitigation strategies for reducing air pollution. Environmental Science and Pollution Research International, 27(16), 19226–19235. https://doi.org/10.1007/s11356-020-08647-x

35.

United States Environmental Protection Agency. (2014). Air quality index: A guide to air quality and your health. http://www.epa.gov/airnow/aqi_brochure_02_14.pdf

36.

Williamson

(1998). Discovered by chance: The role of incidental information acquisition in an ecological model of information use. Library and Information Science Research, 20, 23–40.

37.

World Health Organization. (2021). WHO global air quality guidelines: (PM2.5 and PM10), ozone, nitrogen dioxide, sulfur dioxide and carbon monoxide. https://www.who.int/publications/i/item/9789240034228.