Perceived haze,stress,and negative emotions: An ecological momentary assessment study of the affective responses to haze

Abstract

The aim of this study was to examine the mediating role of stress in the association between people’s perceived haze and negative emotions in daily life. Using ecological momentary assessment, 95 college students reported their perceived haze, stress, and negative emotions twice a day over the course of 2 weeks. The results showed a positive relationship between perceived haze and negative emotions. More importantly, this association was significantly mediated by levels of stress. Findings suggested that people who perceived more severe haze may report higher stress levels, which in turn may lead to increases in negative emotions.

Keywords

mediating effect negative emotions perceived haze stress

Introduction

As awareness of haze resulted from PM 2.5 (the most dangerous of the criteria air pollutants currently regulated by the Environmental Protection Agency) has risen these days, researchers focused on identifying the actual and potential severity of haze. In China, for instance, studies have demonstrated a drastic rise in the severity of air pollution since 2013. According to the records of a haze warning system, Beijing, the capital of China, experienced 18 consecutive haze days in November 2015, which has been the highest number recorded since 1951 (http://news.cntv.cn/2016/01/12/ARTImzbv3jOsZcCKugUzIDqb160112.shtml). Nevertheless, concerns over air pollution and its harmful consequences have not been limited to industrialized nations or zones. Even in rural Alaska, 328 air quality surveys conducted within seven communities over a period of 2 years have identified a rising potential for exposure to air pollutants such as airborne particulate matter (PM; Ware et al., 2013). In addition, efforts have been devoted to the investigation of the cause and effects of haze. However, there were few studies on affective responses to haze. And how possible mediating variables influence the relationship between perceived haze and mental health has not been studied carefully. We discussed in detail affective responses to haze and its relationship with stress.

Effects of haze on humans

Exposure to air pollution, such as a haze, can lead to a multitude of undesired outcomes. Physical health problems such as a variety of respiratory and cardiovascular health outcomes are normally the result of excessive exposure to haze (Shiga, 2004). In addition, a number of epidemiological studies have provided neurobehavioral evidence on the association between exposure to ambient air pollutants and reduced cognitive performance in people of various age groups (Calderón-Garcidueñas et al., 2011, 2012; Chen and Schwartz, 2008).

Affective responses to haze

Haze may also elicit negative affective responses in those who are affected.

People’s affective responses, or feelings and emotions toward haze, include both attitudes and emotional experiences. Findings from conventional survey-based studies have shown that people acknowledged their concern about air pollution and the concern increased over time (Johnson, 2011; Smith et al., 1964). Prasetyo et al. (2013) found through an analysis of more than 26 million tweets during the 2013 Southeast Asian haze crisis that Singaporeans expressed primarily shock, anger, and other negative emotions on social media during haze days. On the day where haze was the most severe, the majority of tweets were characterized by words reflecting negative emotions (e.g. anxiety and anger), which included swear words. There was also a general increase in reports of emotional problems among people who were experiencing haze (Hsien-Jin and Ling, 2008; Song et al., 2015). These emotional problems included feeling depressed, lacking in energy and general motivation, feeling anxious and afraid, sleeping difficulties, and feelings of agitation and irritability (Hsien-Jin and Ling, 2008). Dan and Juan (2016) speculated that the positive relationship between haze and emotional problems could be explained by the reduction in sunlight during the occurrence of haze, which in turn disrupts the body’s regulation of melatonin levels—a hormone that has been commonly associated with mood changes (Lewy et al., 2006).

Perceived haze and affective responses

Although the level of chemical air pollution is, of course, an important factor whether annoyance and health symptoms will be elicited or not, air pollution was not found to directly influence annoyance and symptoms. The impact of air pollution on these outcomes was mediated by perceived pollution (Claeson et al., 2013). While odorants perceived unpleasant typically evoke annoyance, worry, anxiety, nervousness, aggression, depression, and disgust, odors that are perceived pleasant have the opposite effect (Kirk-Smith et al., 1983; Knasko, 1992, 1995; Ludvigson and Rottman, 1989; Schiffman and Williams, 2005; Sucker et al., 2008). People’s anxiety and fear increased significantly when visibility declined and people experienced a strong smell of the haze (Kim et al., 2013). To conclude, perceived haze plays an important role in predicting environmentally induced emotions.

Haze and stress

Besides negative emotions, the experience of air pollution, in general, has also been related to increases in psychological stress (Krupat, 1985; Lazarus and Cohen, 1977), especially since exposure to air pollutants is considered a stressor, which threatens one’s well-being and health. Furthermore, Ho et al. (2014) found evidence suggesting a positive association between a haze crisis and reports of mild psychological stress. Considering the associations between stress and haze exposures is critical in accurately understanding their interactions. Perceived severity of pollution also plays an important role since the levels of stress-induced physiological activity in the autonomic nervous system and brain are conditional on people’s perception of air quality (Alaoui-Ismaïli et al., 1997a, 1997b; Ehrlichman et al., 1997; Lorig and Schwartz, 1988; Miltner et al., 1994b). Haze relates to stress and, in turn, psychological stress can influence susceptibility and may potentiate the effects of air pollution in mental health (Clougherty and Kubzansky, 2009).

Stress and affective responses

Our proposition of stress as a mediator of the relationship between perceived haze pollution severity and affective functioning in daily life was grounded by Clougherty and Kubzansky’s framework (Clougherty and Kubzansky, 2009). This framework posits that the stress process can be deconstructed into following three phases: (a) the presence or identification of the stressor (i.e. any event, condition, or external stimuli posing a physical or psychological challenge), (b) the appraisal of the stressor (i.e. how one experiences, perceives, or interprets the event), and (c) the response to the perceived stress (e.g. psychological and physiological consequences). Accordingly, haze would be a stressor, and the perceived severity of the haze would directly influence the appraisal of this stressor. The experience of haze may be especially stress-inducing given that most people view it as something that is beyond their control (Rankin, 1969). The appraised stress, in turn, shapes the stress response, which also includes affective outcomes. In line with our proposition, chronic stress has long been associated with negative emotional states (e.g. depression, anxiety, anger), with associations so consistent that some researchers have proposed depression as a form of stress response (Kiecolt-Glaser et al., 2002; Sternberg et al., 1992). Moreover, Fiedler et al. (2005) found that stress predicted anxiety symptoms and cortisol, outweighing any potential interaction with pollution.

This study

In this study, we directed our attention from the relatively well-documented effects of haze on physical health (e.g. Baxter et al., 2007) to its effects on psychological well-being, namely, affective functioning and stress. Using an ecological momentary assessment (EMA) design, we examined the relationships between perceived haze, and the experiences of stress and negative emotions in daily life. Whereas previous research has relied mostly on aggregate self-report assessments to examine associations between air pollution and mental health effects (Ware et al., 2013), EMA assesses daily psychosocial experiences through the immediate reporting of experiences in the natural environment rather than reconstruction of information from memory (Bolger et al., 2003; Stone et al., 1996). This minimizes the likelihood of inaccurate descriptions of behaviors, affect, or cognitions, which cannot be controlled for in designs that utilize global, retrospective reports (Bradburn et al., 1987; Stone et al., 1998; Thompson et al., 1996).

More importantly, we tested the role of stress as a mediator of the association between perceived haze and negative emotions. We focused exclusively on perceived haze because it was found to be the mediator of air pollution and annoyance and symptoms. The level of air pollution was not directly related to the symptoms (Claeson et al., 2013).

Based on prior research and theory, we hypothesized that higher perceived severity of haze would be associated with more negative emotions (e.g. depression, anxiety) in daily life. We further predicted that this association would be driven by stress levels, in that higher stress levels would predict increased negative emotional responses.

Materials and methods

Participants

With the exclusion of one student whose participation was withdrawn by choice, our final sample comprised 95 students from a Chinese university with an average age of 21.40 years (standard deviation (SD) = 2.92 years). Fifty-eight of them (61.1%) were females. The average age of them was 21.4 years (SD = 2.92 years). As for education, 49 of them (51.6%) were graduate students, and 46 of them (48.4%) earned a master’s degree. None of the participants reported any experience of long-term meditation or any history of mental disorders.

Procedure

This study was approved by the Ethics Committee of School of Psychology, Nanjing Normal University. Upon agreement to participate in the study, participants first attended a laboratory session where they were given a detailed explanation of the study. Our survey was programmed according to a stratified interval scheme: Participants were required to complete a short questionnaire, which contained measures of their perceived severity of haze, stress, and subjective emotions, twice a day (at 9:00 a.m. and 5:00 p.m.) for 14 consecutive days. This questionnaire was accessed through a web link, which the researchers sent at 9:00 a.m. and 5:00 p.m. each day through the WeChat application (a popular Chinese social media and messaging platform). To ensure that participants do not miss the text message containing the link, all participants were required to allow notifications from the WeChat application and to complete the questionnaires within 1 hour of receiving the link. Upon the completion of each questionnaire, participants were offered the option of entering the “lottery” (by clicking the “lottery” button) where they received a chance of winning 1–2 RMB.

Measures

Perceived haze was assessed by a single item: “In your opinion, what is the level of haze at this moment in your opinion?” Items were rated on a 101-point Likert-type scale with anchors 0 = No haze at all and 100 = Extremely serious.

Stress was assessed with three questions: “At the moment, I feel that I cannot handle the most important things in my life,” “At the moment, I feel nervous and stressed,” and “I know something must be finished now.” Items were rated on a 101-point Likert-type scale with anchors 0 = Not at all and 100 = Very much so. Cronbach’s alpha of this questionnaire was 0.85, indicating that those items had good internal consistency.

Negative emotions were assessed with five items assessing anxiety, depression, anger, and discomfort, respectively (e.g. “Do you feel depressed at the moment?”). Items were rated on a 101-point Likert-type scale from 0 = Not at all to 100 = Very much so. The Cronbach’s alpha of this questionnaire was 0.83.

Analysis using hierarchical linear model

Since we proposed the relationship among perceived haze, stress, and negative emotions, an analytic strategy was chosen to address the effect. Given that a traditional hierarchical linear model (HLM)-based approach to testing multilevel mediation is prone to confound within-group mediation effects with between-group mediation effects (Zhang et al., 2009), we followed Zhang et al.’s (2009) multilevel mediation approach of “centered within context with reintroduction of the subtracted means,” also termed CWC(M), to test the mediation hypotheses. Specifically, we analyzed the relationship of perceived haze (independent variable), stress (mediator variable), and negative emotions (dependent variable) by testing 1-1-1 multilevel mediation model because all the variables in the study were measured at Level 1 (within-person). Based on Zhang et al.’s (2009) 1-1-1 CWC(M) approach, hypothesis-testing was completed in three steps.

We first tested the relationship between the perceived haze and negative emotions (e.g. depression score) by group-mean centering the independent variable (centered at each person’s mean to ensure that it featured within-person fluctuations instead of between-person differences) and including the group mean of the independent variable as a control variable in the Level 2 equation.

Model 1

Level 1

N E_{- 2 t i} = π_{0 i} + π_{1 i} * (H A Z E_{t i}) + e_{t i}

(1)

Level 2

\begin{array}{l} π_{0 i} = b_{00} + b_{01} * (H A Z E ._{i}) + u_{0 i} \\ π_{1 i} = b_{10} \end{array}

(2)

The prediction of perceived haze for negative emotions was divided into two levels: Level 1 was the prediction of HAZE_ti for NE_−2ti. To obtain greater insight into mediational mechanisms and investigate the predictions dynamically, NE_−2ti was created on the basis of NE_ti denoting the next-time (2t) momentary negative emotions of person i, which is predicted by coefficient b₁₀ of HAZE_ti representing the this-time (t) perceived haze of person i. Level 2 was the prediction of HAZE_i (the average value of each individual’s perceived haze) for NE, the predicting coefficient b₀₁ demonstrating how levels of individual’s overall perceived haze predict negative emotions. The error terms e_ti and u_0i capture the residual variance at Level 1 and Level 2, respectively.

In the second step, we examined the relationship between the perceived haze (independent variable) and a stress (mediator variable) by group-mean centering the independent variable and including the group mean of the independent variable in the equation at Level 2.

Model 2

Level 1

S T R E S S_{t i} = π_{0 i} + π_{l i} * (H A Z E_{t i}) + e_{t i}

(3)

Level 2

\begin{array}{l} π_{0 i} = b_{00} + b_{01} * (H A Z E ._{i}) + u_{0 i} \\ π_{1 i} = b_{10} \end{array}

(4)

Similar to the equations of Model 1, the predictions of perceived haze for stress were also divided into two parts: the predicting this-time (t) perceived haze for the next-time (2t) stress and the predicting coefficient b₀₁ representing the prediction of individuals’ overall perceived haze for stress.

Step 3 was to include both the independent variable (perception of haze) and the mediator variable (stress) to demonstrate the dependent variable by group-mean centering the independent variable and the mediator variable and adding the group means of the independent variable and the mediator variable to the equation at Level 2.

Model 3

Level 1

\begin{array}{l} N E_{2 t i} = π_{0 i} + π_{1 i} * (H A Z E_{ti}) \\ + π_{2 i} * (S T R E S S_{ti}) + e_{t i} \end{array}

(5)

Level 2

\begin{array}{l} π_{0 i} = b_{00} + b_{01} * (S T R E S S ._{i}) \\ + b_{02} * (H A Z E_{- i}) + u_{0 i} \\ π_{1 i} = b_{10} \\ π_{2 i} = b_{20} \end{array}

(6)

While the predicting coefficient b₀₁ represents the prediction of stress for negative emotions and coefficient b₀₂ represents the prediction of haze under the control of stress for negative emotions, the coefficients b₁₀ and b₂₀ which exclude the individual differences mean the prediction of this-time (t) perceived haze and stress for next-time (2t) negative emotions, investigating the dynamic relationships between perceived haze, stress, and negative emotions. The results from our analyses (Models 1 and 3) are reported in Table 1.

Table 1.

CWC(M) analyses for each emotion, predicted by perceived haze (Model 1) and perceived haze and stress (Model 3) at between-person level (Level 1) and within-person level (Level 2).

Emotion	Level	Model 1 (perceived haze)			Model 3 (perceived haze and stress)
		Perceived haze			Perceived haze			Stress
		B	SE	p	B	SE	p	B	SE	p
Depression	Level 1	0.091	0.016	0.000	0.36	0.058	0.000	0.104	0.018	0.000
Depression	Level 2	0.718	0.085	0.000	2.923	0.537	0.000	0.106	0.012	0.000
Anxiety	Level 1	0.067	0.015	0.000	0.36	0.01	0.023	0.103	0.013	0.000
Anxiety	Level 2	0.871	0.154	0.000	0.238	0.099	0.018	0.218	0.018	0.000
Anger	Level 1	0.014	0.008	0.085	−0.002	0.009	0.818	0.044	0.012	0.000
Anger	Level 2	0.347	0.062	0.000	0.139	0.066	0.036	0.073	0.011	0.000
Discomfort	Level 1	0.112	0.018	0.000	0.082	0.013	0.000	0.084	0.019	0.000
Discomfort	Level 2	0.165	0.026	0.000	0.417	0.107	0.000	0.131	0.028	0.000

CWC(M): centered within context with reintroduction of the subtracted means; SE: standard error.

Results

In Model 1, we found that people’s perceived severity of haze pollution at the current time point predicted increased negative emotions, such as depression, anxiety, and discomfort at the next time point at Level 1. At Level 2, people’s general level of haze perceived was also a significant positive predictor of negative emotions; individuals who generally perceived haze pollution as higher in severity tended to feel more negative emotions in daily life.

Similarly, the results of the second step (Model 2) revealed a significant positive relationship between perceived haze severity at the current time point and stress at the next time point at Level 1(B = 0.36, standard error (SE) = 0.058, p < 0.001). At Level 2, individuals who generally perceived more severe haze were also more likely to report higher stress levels in daily life (B = 2.923, SE = 0.537, p < 0.001).

Finally, the findings of the third step (Model 3) demonstrated a significant mediation of stress on the relationship between perceived haze pollution severity and negative emotions on both levels. At Level 1, each time people perceived higher level of haze, they expressed more stress, and next time they experienced more negative emotions. In other words, the relationship between perceived haze and negative emotions was mediated by stress dynamically. At Level 2, the results revealed that the individuals’ stress significantly mediated the relation between overall perceived haze and negative emotions.

Discussion

The objective of this study was to investigate the associations between perceived haze, stress, and negative emotions in the daily life of university students in China. Consistent with previous studies, our findings showed that perceived severity of haze positively predicted negative emotions (Briere et al., 1983; Rotton, 1983) as well as stress (Ho et al., 2014). The more severe one perceived the haze to be (at one time point), the more likely one was to experience negative emotions, such as depression, anxiety, anger and discomfort, and stress (at the next time point). Moreover, we found that people’s momentary feelings of stress positively predicted their experiences of negative emotions. More crucially, we demonstrated that stress significantly mediated the relationship between perceived haze and negative emotions. That is, the higher the severity of haze perceived, the higher the degree of stress experienced, which in turn predicted higher degree of negative emotions experienced. We proposed that this pattern of finding may reflect an adoption of an emotion-focused strategy to cope with the stress (Lazarus and Folkman, 1984) that arose from perceived haze pollution.

From a theoretical perspective, these findings supported the use of top–down processing model for understanding the impact of haze on emotions. This mechanism involves higher order neutral processing in the brain that affects emotions, physiology, behavior, and health (Claeson et al., 2013). Heavy top–down processing was a characteristic feature of olfaction (Smeets and Dalton, 2005; Stevenson and Boakes, 2003), in this context expressed as perceived haze that influenced affective symptoms’ development. The individual’s perceptions of the odorous source were utilized for an adequate response to the chemical exposure. Thus, if the individual perceived more serious haze, this would result in negative emotions to guide the individual to avoid the exposure.

The findings had significant implications for theories that attempted to explain the affective responses to haze. First, this study was one of a few that has used EMA to investigate the day-to-day psychological responses to perceived haze. As literature on the negative effects of air pollution has been generally atheoretical and lacking in sophisticated methodology (Evans and Jacobs, 1981), the current findings, which were derived from more naturalistic and repeated measurements, provided complementary evidence to the existing research. Second, this study was also the first, to the best of our knowledge, to demonstrate the role of momentary stress as a mediator of the relationship between haze and affective responses. The current findings can also be effectively applied into mental health awareness programs in both clinical and organizational contexts, particularly in areas with severe air pollution. For instance, people could receive trainings to reduce or manage stress levels as a strategy to minimize the negative emotional consequences of air pollution.

Despite the strengths and implications of this study, there were some limitations that should be considered. First, care should be taken in the generalization of our findings since our sample comprised only university students in China. While this study has only examined the role of stress as one potential mediator of the association between perceived haze and affective responses, the mechanisms underlying this association were complicated. Future investigations should therefore consider a number of other potential mediators of the above-mentioned association.

In conclusion, this work demonstrated that one’s level of stress was one driver behind the relationship between perceived haze and experience of negative emotions in daily life. These results offered a more nuanced understanding on the psychological consequences of air pollution.

Footnotes

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 research was supported by the Ministry of Education (MOE), in China, Liberal Arts and Social Sciences Foundation (project no. 14YJC190001); the 13th Five-Year Plan of Educational Science in Jiangsu Province (project no. C-c/2016/01/13); and China Postdoctoral Science Foundation (2017M610333).

References

Alaoui-Ismaïli

Robin

Rada

, et al. (1997a) Basic emotions evoked by odorants: Comparison between autonomic responses and self-evaluation. Physiology & Behavior 62(4): 713–720.

Alaoui-Ismaïli

Vernet-Maury

Dittmar

, et al. (1997b) Odor hedonics: Connection with emotional response estimated by autonomic parameters. Chemical Senses 22(3): 237–248.

Baxter

Clougherty

Paciorek

, et al. (2007) Predicting residential indoor concentrations of nitrogen dioxide, fine particulate matter, and elemental carbon using questionnaire and geographic information system based data. Atmospheric Environment 41(31): 6561–6571.

Bolger

Davis

Rafaeli

(2003) Diary methods: Capturing life as it is lived. Annual Review of Psychology 54(1): 579–616.

Bradburn

Rips

Shevell

(1987) Answering autobiographical questions: The impact of memory and inference on surveys. Science 236(4798): 157–161.

Briere

Downes

Spensley

(1983) Summer in the city: Urban weather conditions and psychiatric emergency-room visits. Journal of Abnormal Psychology 92(1): 77–80.

Calderón-Garcidueñas

Engle

Mora-Tiscareño

, et al. (2011) Exposure to severe urban air pollution influences cognitive outcomes, brain volume and systemic inflammation in clinically healthy children. Brain and Cognition 77(3): 345–355.

Calderón-Garcidueñas

Mora-Tiscareño

Styner

, et al. (2012) White matter hyperintensities, systemic inflammation, brain growth, and cognitive functions in children exposed to air pollution. Journal of Alzheimer’s Disease 31(1): 183–191.

Chen

Schwartz

(2008) Metabolic syndrome and inflammatory responses to long-term particulate air pollutants. Environmental Health Perspectives 116(5): 612–617.

10.

Claeson

Lidén

Nordin

, et al. (2013) The role of perceived pollution and health risk perception in annoyance and health symptoms: A population-based study of odorous air pollution. International Archives of Occupational and Environmental Health 86(3): 367–374.

11.

Clougherty

Kubzansky

(2009) A framework for examining social stress and susceptibility to air pollution in respiratory health. Environmental Health Perspectives 117(9): 1351–1358.

12.

Dan

Juan

(2016) Depression resulted from haze. Journal of Clinical Psychiatry 2: 111.

13.

Ehrlichman

Brown Kuhl

Zhu

, et al. (1997) Startle reflex modulation by pleasant and unpleasant odors in a between-subjects design. Psychophysiology 34(6): 726–729.

14.

Evans

Jacobs

(1981) Air pollution and human behavior. Journal of Social Issues 37(1): 95–125.

15.

Fiedler

Laumbach

Kelly-McNeil

, et al. (2005) Health effects of a mixture of indoor air volatile organics, their ozone oxidation products, and stress. Environmental Health Perspectives 113: 1542–1548.

16.

Zhang

, et al. (2014) Impact of 2013 south Asian haze crisis: Study of physical and psychological symptoms and perceived dangerousness of pollution level. BMC Psychiatry 14(1): 81.

17.

Hsien-Jin

Ling

(2008) Psychological and physical impact of the haze amongst a Malaysian community. Jurnal Sains Kesihatan Malaysia 6(2): 23–34.

18.

Johnson

(2011) Acculturation, ethnicity, and air pollution perceptions. Risk Analysis 31(6): 984–999.

19.

Kiecolt-Glaser

McGuire

Robles

, et al. (2002) Emotions, morbidity, and mortality: New perspectives from psychoneuroimmunology. Annual Review of Psychology 53(1): 83–107.

20.

Kim

Jeong

, et al. (2013) Autophagy deficiency leads to protection from obesity and insulin resistance by inducing Fgf21 as a mitokine. Nature Medicine 19(1): 83–92.

21.

Kirk-Smith

Van Toller

Dodd

(1983) Unconscious odour conditioning in human subjects. Biological Psychology 17(2): 221–231.

22.

Knasko

(1992) Ambient odor’s effect on creativity, mood, and perceived health. Chemical Senses 17(1): 27–35.

23.

Knasko

(1995) Pleasant odors and congruency: Effects on approach behavior. Chemical Senses 20(5): 479–487.

24.

Krupat

(1985) People in Cities: The Urban Environment and Its Effects, vol. 6. Cambridge: Cambridge University Press.

25.

Lazarus

Cohen

(1977) Environmental stress. In: Altman

Wholwill

(eds) Human Behavior and Environment. New York: Springer, pp. 89–127.

26.

Lazarus

Folkman

(1984) Stress, Appraisal and Coping. New York: Springer.

27.

Lewy

Lefler

Emens

, et al. (2006) The circadian basis of winter depression. Proceedings of the National Academy of Sciences of the United States of America 103(19): 7414–7419.

28.

Lorig

Schwartz

(1988) Brain and odor: I. Alteration of human EEG by odor administration. Psychobiology 16(3): 281–284.

29.

Ludvigson

Rottman

(1989) Effects of ambient odors of lavender and cloves on cognition, memory, affect and mood. Chemical Senses 14(4): 525–536.

30.

Miltner

Matjak

Braun

, et al. (1994) Emotional qualities of odors and their influence on the startle reflex in humans. Psychophysiology 31(1): 107–110.

31.

Prasetyo

Gao

Lim

, et al. (2013) Social sensing for urban crisis management: The case of Singapore haze. In: Jatowt

Lim

E-P

Ding

, et al. (eds) Social Informatics. New York: Springer, pp. 478–491.

32.

Rankin

(1969) Air pollution control and public apathy. Journal of the Air Pollution Control Association 19(8): 565–569.

33.

Rotton

(1983) Affective and cognitive consequences of malodorous pollution. Basic and Applied Social Psychology 4(2): 171–191.

34.

Schiffman

Williams

(2005) Science of odor as a potential health issue. Journal of Environmental Quality 34(1): 129–138.

35.

Shiga

(2004) Smog clogs arteries: Pollution does lasting harm to blood vessels. Science News 166(24): 372–373.

36.

Smeets

Dalton

(2005) Evaluating the human response to chemicals: Odor, irritation and non-sensory factors. Environmental Toxicology and Pharmacology 19(3): 581–588.

37.

Smith

Schueneman

Zeidberg

(1964) Public reaction to air pollution in Nashville, Tennessee. Journal of the Air Pollution Control Association 14(10): 418–423.

38.

Sternberg

Chrousos

Wilder

, et al. (1992) The stress response and the regulation of inflammatory disease. Annals of Internal Medicine 117(10): 854–866.

39.

Stevenson

Boakes

(2003) A mnemonic theory of odor perception. Psychological Review 110(2): 340–364.

40.

Stone

Schwartz

Neale

, et al. (1998) A comparison of coping assessed by ecological momentary assessment and retrospective recall. Journal of Personality and Social Psychology 74(6): 1670–1680.

41.

Stone

Turkkan

Bachrach

, et al. (1996) The Science of Self-Report: Implications for Research and Practice. Bethesda, MD: National Institutes of Health.

42.

Sucker

Both

Bischoff

, et al. (2008) Odor frequency and odor annoyance. Part II: Dose–response associations and their modification by hedonic tone. International Archives of Occupational and Environmental Health 81(6): 683–694.

43.

Thompson

Skowronski

Larsen

, et al. (1996) Autobiographical Memory. Mahwah, NJ: Lawrence Erlbaum.

44.

Ware

Lewis

Hopkins

, et al. (2013) Sources and perceptions of indoor and ambient air pollution in rural Alaska. Journal of Community Health 38(4): 773–780.

45.

Hengyan

Xitong

, et al. (2015) The effect of haze on the pregnant women’s psychological status. China Journal of Health Psychology 23(7): 1085–1088.

46.

Zhang

Zyphur

Preacher

(2009) Testing multilevel mediation using hierarchical linear models: Problems and solutions. Organizational Research Methods 12(4): 695–671.