Get a 10 % discount on an order above $ 100!
Critiquing Sampling Strategy and Sample Size in a Research Article
Critiquing Sampling Strategy and Sample Size in a Research Article
- Critique the sampling strategy used. What are the strengths and limitations of the study due to the sampling strategy used?
ORDER ORIGINAL, PLAGIARISM-FREE ESSAY PAPERS HERE
- Critique the sample size used.
- Did the author(s) justify the sample size? If not, what is the problem with that?
- a 2- to 3-paragraph critique of the sampling strategy and sample size used in the article. Include an analysis of how sampling strategy can strengthen or weaken a quantitative research study.
lable at ScienceDirect
Journal of Environmental Psychology 31 (2011) 314e322
Contents lists avai
Journal of Environmental Psychology
journal homepage: www.elsevier.com/locate/jep
Living in grey areas: Industrial activity and psychological health
Sibila Marques*, Maria Luísa Lima Centro de Investigação e Intervenção Social (CIS/ISCTE-IUL), Ed. ISCTE, Av. das Forças Armadas, 1649-026 Lisboa, Portugal
a r t i c l e i n f o
Article history: Available online 7 January 2011
Keywords: Industrial activities Physical contexts Place perception Psychological health
* Corresponding author. Tel.: þ351 217903079; fax E-mail address: firstname.lastname@example.org (S. Marqu
0272-4944/$ e see front matter � 2011 Elsevier Ltd. doi:10.1016/j.jenvp.2010.12.002
a b s t r a c t
The main goal of this paper was to explore the relationship between living in industrial areas and individual’s level of psychological health. Using a quasi-experimental design main findings suggest that, regardless of the type of industry that is operating, there was a significant association between living in industrialized areas and decreased levels of well being, optimism and use of active coping strategies. However, results on anxiety and depression were especially high in areas associated with air pollution. Moreover, there was also a significant association between more subjective meanings of place and psychological health. According to a reality-orientation criterion, evidences showed that when individ- uals live in industrial areas perceptions of their places as industrial are associated with lower depression, anxiety and psychiatric symptoms. Critiquing Sampling Strategy and Sample Size in a Research Article
� 2011 Elsevier Ltd. All rights reserved.
We live in an age of heavy industrialization. The rise of the industrial revolution brought many important achievements, leading to the overall development of countries’ economy and wealth. However, the massive creation of factories and new cities also had important social and environmental impacts turning many traditional green landscapes into what are perceived as ugly grey portraits. This dichotomy between a pure and beautiful countryside versus the dirty and black city is not new and probably no one has explored this idea better than the famous English writer Charles Dickens. For instance, in the novel Great Expectations (1860/1861), the main character Pip is consistently clear in his idyllic descrip- tions of the country as a calm and beautiful place, whereas the city was described, in his own words, as an “ugly, crooked, narrow, and dirty” place (p. 153).
The idea that living in different physical settings may have differential effects on individuals has been explored in a more scientific background. Until now, most of these studies have been particularly interested in testing the health impacts of living in urban vs. rural places. Some evidences suggest that higher urban- ization rates are related with environment-related morbidity both in low income (von Shirnding, 2002) and advanced countries (Sclar, DArch, & Carolini, 2005). For instance, in support of this prediction, Haynes and Gale (1999) showed clear differences in mortality and deprivation in health among rural and urban residents in England.
: þ351 21790396. es).
All rights reserved.
The results of this study showed that rural wards had mean values of mortality and morbidity lower than national average values, while those in Inner London and other metropolitan cities were less healthy.
More recently, some studies have also presented compelling evidences that the level of industrialization (and not merely urbanization) is also related to poorer health (Downey & Van Willigen, 2005; Evans & Kantrowitz, 2002). A good example of research showing the effect of industrial contexts on health is the large-scale study conducted by Boardman and colleagues (Boardman et al., 2008). In this study, and in agreement with expectations, results showed a positive correlation between living close to industrial activities and stress levels, even after controlling for the effect of several demographic variables such as gender and level of income.
Studies that explore the relationship between industrial activi- ties and health are especially important because there seems to be an unequal distribution of physical sites according with several demographic variables. In this sense, some studies suggest that poorer people, from underprivileged minorities, are the ones who end up living in the most industrialized and polluted places (Adeola, 1994; Brulle & Pellow, 2006; Lima, 2008). For instance, in one recent study conducted in England, Walker, Mitchell, Fairburn, and Smith (2005) showed an unequal distribution of industrial sites in England, with sites disproportionately located in deprived areas and near deprived population. In a similar vein, other evidence showed that industrial and hazardous areas are particularly occupied by Blacks and Hispanics (Szasz & Meuser, 1997). This kind of “social injustice” has been particularly explored in the US, covering several issues such as ethnicity, class, income, age and population density
S. Marques, M.L. Lima / Journal of Environmental Psychology 31 (2011) 314e322 315
(Bryant, 2003; Davidson, 2003). Evidences showing that living in more industrial sites may have a significant and direct effect on one’s psychological health clearly emphasize the type of social injustice that some individuals in our societies are exposed to.
Given the importance of this topic, we believe the effect that neighborhood physical contexts – more rural or more industrial – have on individual’s health is still an issue under exploration. In fact, studies addressing this topic still follow a narrow perspective, focusing mostly exclusively on a biological model of humane environment interaction (Evans,1982). First of all, it seems that this research is often particularly concerned with the effects that living in these contexts may have on humans’ physical health (e.g., Dunn & Kingham, 1996; Elliot et al., 2001; Pless-Mulloli et al., 1998; Walker et al., 2005) with moreor less disregard to other type of impacts such as psychological consequences of living in these type of places. In fact, as far as we know, apart from the Boardman et al. (2008) study just cited, there are only few studies exploring the consequences of this type of exposure to psychological mental health and most of the times these studies are limited in the number of psychological impacts they address (Arnetz, 1998; Downey & Van Willigen, 2005; Weiss, 1998). Second, they also tend to adopt a clearly more “objective” perspective of impact assessment, focusing mainly on the physical characteristics of neighborhoods and neglecting the possible mediating role that more “subjective” variables (e.g., place perception) might play in the determination of health impacts. Given the evidences suggesting that the way people perceive their environment affects in a significant manner their overall level of well being and health (e.g., Cavalini, Koeter-Kemmerling, & Pulles, 1991; Chattopadhyay & Mukhopadhyay, 1995; Staples, 1996; Steinheider & Winneke, 1993), this perspective offers a reduc- tionist view on this issue. In this sense, we believe that a psycho- logical perspective on environment and health may broaden our knowledge regarding this relationship. In fact, according with psychological models of stress (Evans, 1982; Lazarus & Folkman, 1984) we should expect effects of exposure to environmental physical contexts on health outcomes that go beyond the typical mortality and morbidity rates considered in traditional epidemio- logical studies. In fact, mental health symptoms such as irritability, depression and anxiety should also be affected by environmental quality and should be given an appropriate status (Evans, 1982). Moreover, effects of on health should not only be determined by the objective exposure to environment stimuli; subjective appraisals of environment conditions should also assume a main role in the prediction of health outcomes.
There are good reasons to expect differences in overall levels of psychological health among individuals living in industrial and non-industrial neighborhoods. Industrial contexts should be asso- ciated with higher exposure to noise and air pollution, which are often associated with poor mental health. These environmental stimuli may affect individuals in a direct manner (e.g., Babisch, Ising, Galacher, Sweetnam, & Elwood, 1998; Brunekreef & Holgate, 2002); however, most often, health is influenced by the contexts of exposure and by individual’s perceptions (for reviews on this matter please see Bronzaft, 2002; Evans & Jacobs, 1982; Passhier- Vermeer & Passchier, 2000). Typically, perceiving a certain place as more industrial should be related with poor psychological health. Often perception of industrial activities take the form of annoyance which is generally defined as “a feeling of dissatisfaction associated with any agent or condition that is believed to affect individuals in an adverse way” (Steinheider & Winneke, 1993, p. 353) and that has been associated with the increase of stress (e.g., Evans & Jacobs, 1982). Specifically, annoyance regarding noise has usually been identified as a source of low psychological and phys- ical well-being in more general terms (e.g., Ouis, 2001; Staples, 1996) and annoyance regarding air quality, although less studied,
has also been associated with harmful effects to psychological health (e.g., Cavalini et al., 1991; Chattopadhyay & Mukhopadhyay, 1995), especially when it is associated with a sense of environ- mental threat (Lima, 2004; Lima & Marques, 2005).
However, there may be certain situations where perceptions of place as industrial in general may be related with better mental health. This may happen, for instance, when people perceive regional socio-economic benefits linked to the increase in economic activity and employment (see Downey & Van Willigen, 2005 for a discussion on this issue). In this case, people’s percep- tions are not limited to annoyance, but also to the perception of positive benefits. In support of this prediction, Boardman et al. (2008) showed better mental health among men and women who lived near industrial activities, who did not have children, and that worked in this type of activity.
On the other hand, we believe that there are other factors that may also link perceptions of a place as industrial and better psychological health. In fact, for someone living in an objectively industrial area, it may be more realistic to consider their place as industrial than as non-industrial. In turn, this realistic vision seems to be an important determinant of adequate levels of mental health.
The link between realistic perceptions and mental health has been widely explored within the domain of health psychology, and several authors have discussed the amount of realism that is actually good for one’s health (Colvin & Block, 1994; Taylor & Brown, 1988, 1994). In an influential paper, Taylor and Brown (1988) made an important claim in favor of the use of illusion strategies. According with these authors, people tend to use in a pervasive, enduring and systematic manner certain positive illu- sions (e.g., exaggerated perceptions of control and mastery; unre- alistic positive self-evaluations) that serve an important role that help bring about and maintain psychological well-being. However, one should be cautious to interpret the meaning of “positive illu- sions”. According to a later paper by Taylor and Brown (1994) it is not true that the undifferentiated use of positive illusions is related with adjustment in mental health. For instance, and in accordance with the authors, “if a small group of individuals persist in believing that they can cure themselves of indisputably advancing, chronic, or life-threatening diseases, we might find that these individuals are maladjusted, as is sometimes the case” (Taylor & Brown,1994, p. 23). After extensive elaboration regarding this topic, Taylor and Brown (1994) end up clearly defending the idea that at extreme levels, the use of illusions may indeed be directly linked with poor psychological health. This moderate perspective on the use of positive illusions is more in line with traditional perspectives on mental health, that place a great emphasis on “reality orientation” as a criterion of adjustment and well-being (Colvin & Block, 1994; Jahoda, 1958). In this sense, for someone that lives in an indus- trial area, it may actually be more realistic and adapted to perceive his or her place as it is: an industrial place. Critiquing Sampling Strategy and Sample Size in a Research Article
1.1. Overview of the present study
In the present paper we are especially interested in exploring the consequences for psychological health of living in industrial versus non-industrial areas. Using a vast array of health measures, we hope to show that the neighborhood’s level of industrialization is associated with a much broader array of psychological impacts than has been traditionally assumed. To measure psychological health we included several indexes widely used as measures of psychological health in the psychological literature: psychological well-being, dispositional optimism, anxiety and depression, psychiatric comorbidity and coping strategies. Our perspective allowed us to tap a more conventional symptomatic view of mental health (i.e., depression, anxiety and psychiatric comorbidity), as
S. Marques, M.L. Lima / Journal of Environmental Psychology 31 (2011) 314e322316
well as measures of positive functioning (i.e., psychological well being and dispositional optimism) and also indication of some underlying psychological processes, in particular coping strategies. Measuring coping strategies allow us to evaluate the type of usual cognitive and behavioral responses that people use to manage distress and address the problems of daily life. This is an important measure because the use of coping strategies has been often associated with overall levels of psychological health and well being (Folkman & Moskowitz, 2004). The most influential authors in the field have argued that coping processes are not inherently good or bad (Folkman & Moskowitz, 2004; Lazarus & Folkman, 1984) and that they need to be evaluated in the specific context in which they occur. For this reason, different coping strategies were considered.
In this study we have two main goals. The first is to test the association between living in industrial neighbourhoods and psychological health. Our hypothesis is that, in agreement with previous studies (Boardman et al., 2008) living in industrial areas will be associated with lower psychological well-being, lower dispositional optimism, higher anxiety and depression rates, higher susceptibility to psychiatric problems and lower use of adaptive coping strategies than living in non-industrial areas.
Second, in this paper, we are also interested in exploring the relationship between “perception of industrial activity” and levels of psychological health. We assume that people living in areas classified as industrial would perceive their areas as more “indus- trial” than people living in areas classified as “non-industrial”. Moreover, and based on the idea that realism is an important determinant of overall adaptation and mental health, we hypoth- esize that, for people already living in the industrial areas, perceiving their place as “industrial” should actually be associated with better psychological health.
In order to test our hypothesis we used a quasi-experimental design where we compared psychological health of individuals living in four different areas. Three of these areas are objectively classified as industrial, whereas one is classified as a non-industrial neighborhood. The choice of the areas included in the study was done based on the classification made by the Directorate-General for Spatial Planning and Urban Development, part of the Ministry of Equipment, Planning and Territorial Administration (DGOTDU, 2000). The three industrial areas vary in the type of industrial activity: one area is occupied by a mixture of several type of industries and is especially affected by air quality issues (odor) (Ind 1); the second area is characterized by the activity of chemical industry and is affected mostly by air pollution (smoke and parti- cles) (Ind 2); and the last areas is occupied by textile industry and is affected mostly by water quality issues (Ind 3). The less industri- alized sample (Non-ind) is mostly a residential neighborhood.
Table 1 Socio-economic characteristics of participants in the four samples.
Area N Age Man (%) Married (%) Level of educ
Ind 1 100 46.8a (18.9) 52.0 64.0 57.0 Ind 2 111 44.3a (16.9) 49.5 55.0 53.2 Ind 3 116 44.7a (15.4) 50.0 75.0 56.0 Non-Ind 75 49.7a (16.7) 49.3 82.7 50.7
Note: means in a column that do not share the same subscript are significantly differen *p < .05.
Based on data obtained from DGOTDU (2000) we were able to choose the four areas of interest. It is important to refer that, in order to control for possible differences due to country region, the four areas were all located in the Northern part of Portugal and shared similar demographic characteristics with one important difference: percentage of people employed in the industrial sector in the Non- Ind area was lower (34%) than in the remaining areas (Ind 1: 49.08%; Ind 2: percentages ranged from 43.32% to 57.55% according to specific locations; Ind 3: percentages ranged from 45.29% to 63.08%). Moreover, we did not find significant differences among the four areas regarding age, gender and level of education (INE, 2001).
The interviews were conducted face to face by trained inter- viewers. They took place at respondents’ house and took approxi- mately 30 min. This technique of data collection was used because it maximizes the response rate on controversial issues and allows peoplewith lowlevelsofeducation to answer thequestionnaire. Two criteria were considered to define the characteristics of the sample: the parishes included in the target area and the educational level of the resident population. In each location, the houses to sample were randomly chosen and in each house the interviewee was also randomlychosene thelastadulttohavehisorherbirthday,provided that he or she consented to be part of the study. First contact was always done in the presence of the interviewee but, in some cases, some telephone calls were needed to ensure the interviews.
The interviewers received specific training concerning the inter- views’procedures andthestructureof theinterviewprotocol. Inorder to perform a quality control of the interviews by a member of the team,anIDcodenumberwasgiventoeachoftheparticipants,andthe name, address, and phone number of the interviewed was collected. Critiquing Sampling Strategy and Sample Size in a Research Article
402 participants took part in the survey distributed across the four areas (Table 1). We did not find significant differences between the four samples regarding gender, age, marital status and level of education. However, we found some significant differences between the samples regarding mean years of residence, F (1, 398) ¼ 883.80, p < .00, h2p ¼ .69, percentage of active population, c2 (3, N ¼ 402) ¼ 8.58, p < .05, and level of income, c2 (3, N ¼ 402) ¼ 73.24, p < .001. Analysis of the overall pattern of results showed that individuals in the Non-Ind sample have been living in the area for fewer years than those in the three industrial areas and that they have higher level of income. To control for possible confounding effects of these socio-economic factors on psychological health, we entered these variables as covariates in posterior analysis.
2.4.1. Psychological well-being To measure well-being we used the scale proposed by Ryff (Ryff,
1989; Ryff & Keyes, 1995). Developed in the context of
ation (%) Years of residence Active (%) Level of income (%)
>4 years <1250 V >1250 V
43.0 31.1a (20.6) 4.0* 96.0* 4.0* 46.8 33.0b (19.9) 6.3 91.9* 8.1* 44.0 29.9b (20.7) 6.0 79.3 20.7 49.3 24.4c (16.1) 6.7 49.3* 50.7*
t according with the Sheffé test.
S. Marques, M.L. Lima / Journal of Environmental Psychology 31 (2011) 314e322 317
a developmental perspective of self throughout the life course, this scale aims to evaluate multiple dimensions of positive psycholog- ical functioning including positive relations with others, autonomy, environmental mastery, purpose in life, personal growth and self- acceptance. This scale has been used in several international studies particularly to evaluate effects on well-being due to important life changes (Heidrich & Ryff, 1993). In the present study, we used the 18 items version of this scale adapted to Portuguese by Novo, Duarte Silva, and Peralta (1997). Participants were presented with the 18 items and asked to indicate the degree in which they agreed with each one of them (1 ¼ Strongly disagree until 6 ¼ Strongly agree). Results were summed up in a final score to calculate the overall level of psychological well-being such that higher scores indicate higher psychological well-being. In agree- ment with previous findings (Novo et al., 1997), the scale showed good psychometric qualities (Cronbach alpha ¼ 0.78). Critiquing Sampling Strategy and Sample Size in a Research Article
2.4.2. Optimism Dispositional optimism refers to individual’s typical tendency to
anticipate favorable events and it was measured using the Life Orientation Test (LOT), a 12 items scale developed by Scheier and Carver (1985) and translated by us based on the opinion of two independent specialists. LOT is composed by 8 items that aim to measure dispositional optimism (4 negative and 4 positive items). Scores are obtained by summing up the answers to the 8 items, after inverting negative items. Previous studies showed good internal consistency (Cronbach alpha ¼ 0.76) and test-rest accuracy (0.79 for a four week interval and 0.72 for a period over 13 weeks). In the present study, internal consistency of the scale was 0.70.
2.4.3. Anxiety and depression This was assessed using a short version of hospital anxiety and
depression scale (HADS), a 14 item scale developed by Zigmond and Snaith (1983). HADS was first developed to assess the psychological state of clinical samples in a hospital setting, but is considered as quite appropriate for community surveys in which there is no intention of producing a clinical individual diagnosis (Loewenthal, 1996). The scale is divided in two subscales, one to measure anxiety and the other to measure depression, each composed by 7 items. Results in each subscale are obtained by summing up the items after recoding of negative items and may vary between 0 and 21. The cut-off point to consider psychological symptoms is 12 but Moorey et al. (1991) refer that after 8 there may be light signs of psychological disturbance. In the present study, we found good psychometric results for the two subscales (Cronbach alpha anxiety ¼ 0.83; Cronbach alpha depression ¼ 0.77).
2.4.4. Psychiatric comorbidity Eventual presence of psychiatric comorbidity was assessed
using the short version of the General Health Questionnaire (Goldberg, 1992). The twelve-item scale was developed to detect in the general population the presence of non-psychotic psychiatric illnesses. Scores may vary between 0 and 12 and 8 is considered the cut-off to detect psychiatric illnesses. GHQ has been validated in several countries and it generally an easy to use scale, revealing overall good psychometric qualities. In the original sample, its ability to detect psychiatric cases showed an overall sensitivity of 93.5% and a specificity of 78.5%. We conducted a preliminary study using the GHQ, revealing good psychometric qualities (Cronbach alpha ¼0.87). In the present study internal consistency of the scale was 0.84. Critiquing Sampling Strategy and Sample Size in a Research Article
2.4.5. Coping strategies To evaluate the use of dispositional coping strategies (i.e., typical
tendency to stressful events) we used the short version (16 items)
of COPE, a scale developed by Carver, Scheier, and Weintraub (1989) and translated to Portuguese by Marques-Pinto (2000). The short version of COPE is composed by 4 of the initial 11 subscales: 2 measuring positive coping strategies (active coping and planning) and 2 measuring negative coping strategies (denial and behavioral disinvestment). These categories were identified in the study con- ducted by Marques-Pinto (2000). In COPE, participants are asked to answer in a 4-point scale (1 ¼ I don’t usually do that until 4 ¼ I usually do that) the degree in which they feel their reactions pre- sented in the items match their usual behavior in stressful situa- tions. The scores of each subscale are given by summing up the respective items and they may vary between 4 (null use of that coping strategy) and 16 (intensive use of that coping strategy). Results in each subscale indicate the degree in which each type of coping strategy is used. Previous studies regarding the psycho- metric qualities of COPE showed good results, with values of internal consistency higher than 0.60 and test-retest reliability between 0.42 and 0.89. According to Weinman, Wright, and Johnston (1995) these values point towards a “reasonably stability” (p. 10). The study conducted in Portugal also revealed good psychometric qualities with the values of internal consistency for the four subscales varying between 0.61 and 0.78. In the present study, Cronbach alpha for the four subscales were as following: active coping: 0.75; planning: 0.72; denial: 0.56; behavioral disin- vestment: 0.58.
2.4.6. General perception of the neighbourhood We evaluated the level of annoyance regarding neighbourhood
noise during the night (2 items e.g.,“To what level do you feel annoyed by noises during the night?”, Cronbach alpha ¼ 0.78) and during the day (2 items e.g.,“To what level do you feel annoyed by noises during the day?”, Cronbach alpha ¼ 0.71), air quality (2 items e.g., “To what level do you feel annoyed by dust or smells?”, Cron- bach alpha ¼ 0.90) and overall perceived environmental quality (9 items e.g., “The landscape in this area is very beautiful”, Cronbach alpha ¼ 0.70).
2.4.7. Perception of place as industrial Participants answered two questions aimed at characterizing
their perception of their neighborhood environment (“Are there any factories or industrial units in your area of residence?”; “Can you see any close chimney, factory or industrial unit from your window?”). We considered that participants perceived their place as industrial if they indicated “yes” to these two questions. We considered that participants perceived their place as non-industrial if they answered “no” to these two questions. We did not consid- ered in the analysis participants that only answered “yes” to one of these questions.
2.4.8. Personal identification Participants were also asked to fill some personal data regarding
gender, age, educational level, marital status, occupational status (actively working or unemployed), level of income and number of years living in the area.
3.1. General perception of the neighborhood
The analysis of results showed that participants in the Non-Ind area revealed lower environmental annoyance than individuals in the three industrial areas related with daytime noise, F (3, 398) ¼ 27.48, p < .001, h2p ¼ .17 night time noise, F (1, 398) ¼ 37.40, p < .001, h2p ¼ .22, and air quality, F (1, 398) ¼ 57.33, p < .001, h2p ¼ .30. Finally, they also rated overall perceived environmental. Critiquing Sampling Strategy and Sample Size in a Research Article
Table 3 Adjusted mean, Standard deviations and Analysis of Covariance (ANCOVA) results for Psychological Health Indexes.
Variables Ind 1 (n ¼ 100)
Ind 2 (n ¼ 111)
Ind 3 (n ¼ 116)
Non Ind (n ¼ 75)
F (3, 395)
Well-being M 67.67a 68.63a 75.23b 84.39c 53.99*** 0.29 (SD) (9.00) (8.96) (8.83) (9.44)
Optimism M 39.81a 39.16a 41.23b 46.48c 28.71*** 0.18 (SD) (5.20) (5.16) (5.06) (5.46)
Anxiety M 7.45a 7.39a 4.46b 4.54b 15.11*** 0.10 (SD) (4.20) (4.00) (4.09) (4.42)
Depression M 6.71a 7.38a 4.41b 4.67b 14.01*** 0.09 (SD) (3.90) (3.90) (3.77) (4.07)
GHQ M 2.00a 2.49a,b 1.13a,c 1.74a 5.46*** 0.04 (SD) 2.6 2.63 2.58 2.68
Active coping M 8.63a 10.49b 11.85c 13.05d 56.14*** 0.30 (SD) (2.30) (2.32) (2.26) (2.42)
M 8.78a 10.26b 11.54c 11.83c 30.19*** 0.19 (SD) (2.40) (2.42) (2.37) (2.51)
Coping Beh Desinv
M 6.72a 9.92b 8.74c 8.56c 18.20*** 0.12 (SD) (3.20) (3.16) (3.12) (3.38)
Coping Denial M 8.23b 10.22 ab 9.82 ab 10.70 ab 19.61*** 0.13 (SD) (2.30) (2.32) (2.26) (2.42)
Note: means in a column that do not share the same subscript are significantly different according with the Sidak test. ***p < .001.
S. Marques, M.L. Lima / Journal of Environmental Psychology 31 (2011) 314e322318
quality of the area better than individuals in the three industrial areas, F (1, 398) ¼ 70.42, p < .001, h2p ¼ .35 (Table 2).
3.2. Perception of place as industrial
In agreement with our expectations, results revealed that indi- viduals in the Non-Ind area referred significantly less the presence of factories nearby their residence, c2 (3, N ¼ 402) ¼ 100.28, p < .001, and indicated less often that they could see chimneys from their windows, c2 (3, N ¼ 402) ¼ 81.13, p < .001 (Table 2).
3.3. Area and psychological health
To evaluate differences between the four areas regarding psychological health first we performed a MANCOVA with Area as a between-subjects factor and the psychological health measures as dependent variables. Area refers to the four neighbourhoods included in this study (Ind 1, Ind 2, Ind 3 and Non-ind). To control for possible confounding effects of demographic variables, partici- pant’s occupational status, years of residence in the neighbourhood and level of income were entered as covariates in the analysis. Using Pillai’s trace, there was a significant effect of Area on overall psychological health measures, V ¼ 0.53, F (24, 1170) ¼ 10.41, p < .001, h2p ¼ .18. Univariate ANCOVAS for each dependent vari- able were conducted as follow-up tests to the MANCOVA. Using the Bonferroni method for controlling Type I error rates for multiple comparisons, each ANCOVA was tested at the 0.006 level (Table 3). First of all, the overall analysis of results revealed better psycho- logical health for individuals living in the Non Ind neighborhood. In fact, participants in the Non Ind area showed significantly higher Psychological Well-being, Optimism and use of Active Coping Strategies than individuals in the other three industrial areas. Moreover, residents in the Non Ind neighbourhoods scored higher than Ind 1 and Ind 2 in Planning coping (it is interesting to see that results regarding the use of active coping strategies and planning are the ones where effect sizes are higher). Hence, these results seem to show that living in industrial places, regardless of the specific type of industry that is operating (at least for Well-being, Optimism and Active Coping Strategies), may have a hindering effect on positive psychological health. However, the pattern of results regarding more negative aspects of psychological health is not so clear-cut. In fact, in these indexes, although generally we found worst results in Ind 1 and Ind 2 areas when compared with the Non Ind area, we also found some similarities between psychological health of individuals living in the Non Ind area and individuals living in Ind 3. For instance, results show that depres- sion and anxiety levels are worst in individuals from Ind 1 and Ind 2 than in individuals in Ind 3 and Non Ind areas. Moreover, results regarding the GHQ measure show that individuals living in Ind 2 have significantly higher probability of having psychiatric symp- toms than individuals in Ind 3. These results suggest that specifi- cally negative impacts of living in industrial areas vary as a function of the specific type of industry that is operating. Hence, they seem to indicate that industrial areas with high impacts on air quality are