Cross-section analyses of attitudes towards science and nature from the International Social Survey Programme 1993,2000,and 2010 surveys

Abstract

This paper explores public attitudes towards science and nature in twelve countries using data from the International Social Survey Programme environment modules of 1993, 2000, and 2010. Analysis of attitude items indicates technocentric and pessimistic dimensions broadly related to the Dominant Social Paradigm and New Environmental Paradigm. A bi-axial dimension scale is utilized to classify respondents among four environmental knowledge orientations. Discernible and significant patterns are found among countries and their populations. Relationships with other substantial variables in the surveys are discussed and findings show that the majority of industrialized countries are clustered in the rational ecologist categorization with respondents possessing stronger ecological consciousness and optimism towards the role of modern institutions, science, and technology in solving environmental problems.

Keywords

dominant social paradigm environmental attitudes knowledge orientations new environmental paradigm

1. Introduction

The attitudes of people towards modern and scientific underpinnings of environmental approaches are examined through the analysis of environmental attitudes utilizing International Social Survey Programme (ISSP) survey data from twelve countries within the last two decades. The paper begins with the objectives of the study, followed by a discussion of the contextual concepts drawn from previous literature on environmental attitudes. It then presents results of factor analysis on public attitudes pertaining to science and nature and its applicability to represent two dimensions. The 2-factor structure is used to support a fourfold typology of knowledge orientations which are then examined through a series of correlations with other substantive survey variables and ANOVA comparisons. The last section discusses the main findings and conclusions of the study.

2. Objectives of the study

This study explores public attitudes towards science and nature. At the outset, given the limitation of only eight environmental attitude variables and twelve countries that are consistently available for three ISSP surveys, it aims to uncover recurring environmental attitude dimensions. Subsequently, an attempt is made to apply a typology of environmental knowledge orientations in the categorization of respondents and countries, to present a more adequate conceptualization of the Dominant Social Paradigm and New Environmental Paradigm in contemporary society. Finally, a comparative understanding of environmental attitudes is hoped to be derived between countries, in changes across time, and in the differences based upon knowledge orientations and their implications.

3. Conceptual context

Environmental attitudes and paradigms

Adopting Milbrath’s (1984) framework of having two main value groups in modern society, Aoyagi-Usui (1999: 125) in her research on environmental values classified corresponding ISSP questions on “attitudes towards science and nature” as expressing either Dominant Social Paradigm (DSP) or New Environmental Paradigm (NEP) ideas. Among the fourteen questions utilized in her research, eight were available in the ISSP 1993, 2000, and 2010 modules and are primarily utilized in the cross-section analyses of this paper to represent public environmental attitudes.

Categorized as DSP items:

People worry too much about human progress harming the environment. (WorryTooMuchProgHarmsEnv)

We worry too much about the future of the environment and not enough about prices and jobs today. (WorryTooMuchEnvOverPriceJobs)

Modern science will solve our environmental problems with little change to our way of life. (ModernScienceSolveEnvProb)

In order to protect the environment (respondent’s country) needs economic growth. (EnvNeedsEconGrowth)

Categorized as NEP items:

Economic growth always harms the environment. (EconGrowthAlwaysHarmsEnv)

Almost everything we do in modern life harms the environment. (ModernLifeHarmsEnv)

Overall, modern science does more harm than good. (ModernScienceMoreHarm)

We believe too often in science, and not enough in feelings and faith. (ScienceVsFeelingsFaith)

The DSP and NEP worldviews

The concept of the “dominant social paradigm” was first developed by Pirages and Ehrlich (1974), who described it as “the prominent world view, model, or frame of reference through which individuals or collectively, a society, interpreted the meaning of the external world” and “a mental image of social reality that guided expectations in a society” (p. 43). Milbrath (1984) elaborated further on the DSP as characteristically representing the traditional or conventional values that subscribed to beliefs in endless economic growth, technological progress, abundance of nature, and human dominion over the environment.

With the growing concerns over environmental problems and the attitudes that contributed to environmental degradation in the 1970s, Dunlap and Van Liere (1978) developed the “new environmental paradigm”, focusing on themes concerning the existence of ecological limits to growth, maintaining the delicate balance of nature, and the rejection of anthropocentric views towards nature’s being primarily for human use. The NEP highlighted the ability of humanity to disrupt ecosystems and of modern industrial societies to exceed the capacity of limited natural resources.

Worldviews as dichotomous or bi-polar

Pirages and Ehrlich (1974) had observed that “countless people were questioning many basic beliefs inherent in the DSP, as these beliefs were no longer useful in successfully interpreting their social reality”. They proposed the idea of having an environmental paradigm to replace the prevalent ruling social worldview, claiming “[t]he persistence of any society is threatened when its DSP no longer offers valid guidance for survival” and warning of the dangers of the “uncritical acceptance of an environmentally unsustainable and outmoded DSP” (pp. 44–47).

The DSP with its emphasis on progress, technology, materialism, development and separation of humans from nature, was in contrast to that of Dunlap and Van Liere’s (1978) NEP which recognized humans as part of nature, and the existence of limits to physical growth and development. Moreover, the “NEP scale took account of the wider view of the relationship between modern societies and the environment, and acted as a measure of the proposed shift in people’s worldviews” (Van Petegem and Blieck, 2006: 625). Herein, “Technocentrism is considered as an extension of the DSP, and Ecocentrism as part of the NEP as the reaction to it” (Barrett et al., 2002: 238–239). Early environmental discourse and interpretations of socio-environmental system interactions had been influenced by this essential dichotomy.

Worldviews as bi-axial dimensions

Dunlap (2008: 9) recounts that even back in the late 1980s, there was an increasing number of people “unwilling to accept the opposing DSP and NEP as a realistic depiction of the choices facing them, due to the enormous appeal of sustainable development – offering a synthesis of the traditional growth-oriented and new environment-oriented paradigms”.

The development of the “New Ecological Paradigm” by Dunlap et al. (2000) dealt with the weaknesses of the original NEP scale by adding new facets to the ecological worldview and accommodating discussions on the multidimensionality of the scale. Although Dunlap (2008: 11–12) asserts the NEP scales are coming to “replace the older mechanistic, reductionistic, and technologically optimistic worldview that guided science a few decades ago” (p. 15), he still noted less consistency in their application for less developed countries such as Eastern European and Latin American nations. Van Petegem and Blieck (2006) have observed in their research on children, that Zimbabwean respondents “believed more strongly that nature can withstand modern industry, suggesting a DSP worldview”, in contrast to Belgian children. Interestingly, however, “Zimbabwean children also appear to maintain a NEP worldview as their responses support the idea that humans must obey the laws of nature instead of mistreating the environment”. Their research indicated that these children “both possessed an ecological and a utilitarian view of the environment” (Van Petegem and Blieck, 2006: 631–632).

Similar instances of this dualism were also observed in other studies conducted in Brazilian and Mexican communities (Bechtel et al., 1999, cited in Van Petegem and Blieck, 2006; Corral-Verdugo and Armendáriz, 2000). It was suggested that “acceptance of the NEP implied a clear rejection of the anthropocentric views of the DSP in industrialized societies”, whereas the “distinction between the two worldviews was not as clear-cut in less industrialized societies” (Van Petegem and Blieck, 2006: 633). These then implied a “holistic view of the human–environment relationship” for the developing countries, and “elaborates on the belief of the necessity of balance between environmental protection and satisfying human needs, that fits well with many definitions of sustainable development” (Goodland, 1995, cited in Van Petegem and Blieck, 2006; Corral-Verdugo and Armendáriz, 2000).

Literature addressed these aspects through the development of alternative non-dichotomous and integrative orientations such as the “New Human Interdependence Paradigm” proposed by Gärling, Biel, and Gustafsson (2002), that was subsequently operationalized by Corral-Verdugo et al. (2008) in surveys conducted in France, Italy, Mexico, and India. Furthermore, Bogner and Wiseman (2002) through factor analyses of questionnaire batteries determined a 2-factor structure of environmental attitudes encompassing “preservation” and “utilization”. They then proposed the two-dimensional “Model of Ecological Values” wherein ecological values are determined by the position of a person on two orthogonal dimensions: “A biocentric dimension that reflects conservation and protection of the environment (Preservation); and an anthropocentric dimension that reflects the utilization of natural resources (Utilization)” (Wiseman and Bogner, 2003: 787; see also Milfont and Duckitt, 2004).

Given instances when certain groups of people exhibit contrasting combinations of anthropocentric, technocentric, and ecocentric worldviews; myriad studies that have uncovered and utilized bi-axial structures for understanding attitudes; that certain environmental approaches such as sustainable development are drawing on both DSP and NEP facets; and the consideration of the availability of ISSP survey data that focus on “public attitudes on science and nature”, then, rather than treating DSP and NEP as bi-polar or dichotomous, it may be worth considering the representation of these worldviews as bi-axial dimensions in order to allow for more versatility in understanding people’s environmental attitudes.

Environmental scientific knowledge and environmental knowledge orientations

Cohen (2000) proposed that two axial dimensions condition individual propensities in interpreting ecological information and could be utilized to evaluate environmental knowledge orientations. When assimilating environmental data, people are guided by values, emotions, and philosophies of the good life rather than by a dispassionate manner. The first orienting dimension – ecological consciousness, is described as “varying from weak to strong, dependent on the person’s willingness to make trade-offs in order to protect the environment, their intellectual emphasis, loyalties, affections and outcomes” (Cohen, 2000: 81). Individuals with strong ecological consciousness are “very keen to preserve the integrity of ecosystems and give precedence to lifestyles consistent with responsible environmental stewardship” (Barrett et al., 2002: 240). As such, this paper treats this dimension as “value-based aspects of environmental commitment” (Cohen, 2000: 83), a prioritization between either ecocentric or technocentric values, which is more analogous to the traditional representation for environmentalism and its depiction of the DSP.

Arguing that “people do not simply premise their relationship to the environment on a moral code, but simultaneously condition it by a preference for a particular epistemology”, Cohen offered a second orienting dimension – epistemological commitment. This is described as the particular organizational schema used by people, recognizing that “the unavoidable ambiguity of ecological information and the ways in which people absorb it is a function of the specific interpretive filter they use to differentiate the consequential from the inconsequential” (Cohen, 2000: 82–83). Acknowledging the existence of a virtually limitless number of knowledge systems, Cohen classifies the various frameworks for interpreting environmental data into two broad categories: on one hand, rational-scientific epistemologies “influenced by the intellectual revolution of the seventeenth century and [that rely] on the accumulation of purportedly objective facts”; on the other hand, numinous and aesthetic epistemologies “composed of disparate forms of knowledge with origins ranging across extensive geographic and historical distances, and [that include] folk and indigenous wisdom, local and tacit knowledge, religious knowledge traditions and knowledge movements in art, music, literature, and philosophy” (2000: 84).

In this paper, in addition to the two epistemological categories that represent the long-standing dichotomy in knowledge systems, epistemological commitment is also treated in relation to the NEP via the two faces of contemporary environmentalism, by considering the preference of people to adopt scientific and modern approaches to environmental issues. People then have either an optimistic view of the modern, scientific, technological, mainstream, and institutional methodologies, or a pessimistic perception, opposed to the aforementioned tenets, carrying “romantic ruminations and critique of industrial society” (Cohen, 2000: 77), with such people being more amenable to alternative, traditional, or local knowledge systems.

Cohen’s typology of environmental knowledge orientations

Contrasting orthogonally the two axial dimensions of ecological consciousness and epistemological commitment, Cohen developed a fourfold typology. A corresponding “archetypal environmental knowledge orientation that provides a distinct organizational schema for interpreting ecological information” is represented by each quadrant (2000: 85–86). Figure 1 depicts this typology along with the corresponding relationship with the DSP and NEP bi-axial dimensions, similarly adopting that of Aoyagi-Usui’s (1999) results from the factor analysis of her survey questions.

Figure 1.

Typology of environmental knowledge orientations (Cohen, 2000) combined with Aoyagi-Usui’s concepts (1999) and application by Barrett et al. (2002).

Rational Ecologism combines a high resolve for both science and the environment, attempting to harmonize these two commitments. Through this interpretative filter, ecological information is tempered by an implicit understanding that rational knowledge will provide the intellectual resources from which it will prove possible to address ecological concerns. It is described as optimistic, technologically focused, environmentally aware, and views progress in salutary terms (Cohen, 2000: 86). Prometheanism is described as having a strong disposition toward science and a weak ecological consciousness. It is favored by fervent technological enthusiasts as a perspective used to discount anxiety about environmental issues. Although acknowledging minor, localized problems from time to time, it contends that our greatest resource is our capacity as human beings to apply our ingenuity to the management of the planet. Arcadianism integrates numinous-aesthetic knowledge with a strong ecological consciousness. This interpretive filter in its various guises was used by social movement organizations to shape public environmental sensibilities in the 1960s and 1970s. This orientation generally expresses strong skepticism about science and technology and harkens back with nostalgic longing for more traditional lifestyles. The last perspective, Ecocidal Mysticism, combines numinous-aestheticism with weak ecological consciousness. It is associated with certain “Millenarianist or Chiliastic religious belief systems”, and “disregard[s] the need for measured behavior with respect to nature” (Cohen, 2000: 86). As such, ecocidal mysticists ascribe to dominion over nature and possess low prioritization for ecological concerns.

Though the four quadrants proposed by Cohen described varieties of views found in the developed world, Barrett et al. (2002: 241) still caution against uncritical acceptance of any typology, noting the inherent problems pertaining to where boundaries are drawn and explanations on how categories blend together. It is also acknowledged that individual typologies are often an eclectic mix of different ideas.

4. Data

Analyses are based on data from the International Social Survey Programme (ISSP). The ISSP is a continuing annual program of cross-national collaboration on surveys of currently 48 member countries, covering topics important for social science research and developing annual surveys focused on a single topic (see www.issp.org). The ISSP environment modules measure attitudes to environmental protection and preferred governmental approaches. The ISSP employs a questionnaire method on highly representative national random samples. The standard sampling procedure is a stratified, multistage random sample considering region, household and persons within the household. The target population is adults, with differing age limits in each country. Sample size is about 1000 in most countries with the number of respondents in the twelve countries of the three surveys used in this study ranging from 958 to 2560 (Table 1). A 12-country subsample is used as only these countries consistently had the eight “attitudes towards science and nature” variables. The countries covered by this study span three continents, with four post-communist, two Asian, and the rest considered as western developed countries. ISSP datasets are provided by the Zentralarchiv für Empirische Sozialforschung, University of Cologne and can be found at the GESIS Data Archive (http://zacat.gesis.org/).

Table 1.

Survey characteristics.

	ISSP 1993				ISSP 2000				ISSP 2010
	N	% male	% female	Year of fieldwork	N	% male	% female	Year of fieldwork	N	% male	% female	Year of fieldwork
Great Britain (GB)	1261	42.3	57.7	1993	972	41.8	58.2	2000	928	44.4	55.6	2010
United States (USA)	1557	42.6	57.4	1993	1276	43.9	56.1	2000	1430	42.4	57.6	2010
Norway (N)	1414	49.7	50.3	1993	1452	51.7	48.3	2000	1382	49.6	50.4	2011
Czech Republic (CZ)	1005	47.7	52.3	1993	1244	40.4	59.6	2000	1428	47.9	52.1	2010
Slovenia (SLO)	1032	46.2	53.8	1993	1077	44.4	55.6	2000	1082	45.5	54.5	2011
Bulgaria (BG)	1183	47.8	52.2	1994	1013	46.7	53.3	2000	1003	42.1	57.9	2011
Russia (RUS)	1931	41.9	58.1	1993	1705	45.6	54.4	2000	1619	34.5	65.5	2010
New Zealand (NZ)	1271	44.3	55.7	1993	1112	48.5	51.5	2000	1172	46.5	53.5	2010
Canada (CDN)	1467	46.3	53.7	1993	1115	47.8	52.2	2001	985	48.8	51.2	2011
Philippines (RP)	1200	50	50	1993	1200	50	50	2000	1200	50	50	2009/2010
Japan (J)	1305	45.5	54.5	1992	1180	44.2	55.8	2000	1307	47.9	52.1	2010
Spain (E)	1208	48.1	51.9	1994	958	34.8	65.2	2000	2560	49.8	50.2	2010
Pooled total	15834				14304				16096

All surveys were national in scope, with respondents aged 15 and over.

Consistent topics covered in all three ISSP environment surveys were: 1) “Left-right Dimension”, 2) “Attitudes towards Environment, Science and Nature”, 3) “Respondent’s Behavior and Environmental Protection”, 4) “Environmental Efficacy and Skepticism”, 5) “Environmental Policy”, and 6) Demographics. The question on postmaterialism was modified. As some questions were removed for “Dangers of Specific Environmental Problems”, while “Scientific and Environmental Knowledge” also became optional in the 2010 survey, variables within the above-mentioned topics that are present in all three surveys are primarily utilized for this paper.

5. Investigation results and analysis

Attitude systems in the survey countries

As in previous research done on environmental attitudes, values, concerns and behavior (Aoyagi-Usui, 1999; Aoyagi-Usui et al., 2003; Franzen and Meyer, 2010; Kelly et al., 2003; Motherway et al., 2003; Van Petegem and Blieck, 2006), explorative factor analysis via the principal component method was conducted on three ISSP datasets using the pooled samples from twelve countries focusing on eight ISSP variables of “attitudes towards science and nature”.

Questionnaire items were rated by respondents on a 5-point Likert-type scale ranging from 1 (strongly agree) to 5 (strongly disagree); these were recoded as to represent the scores in an increasing manner, 1 being the lowest and 5 being the highest. Direct Oblimin rotation was utilized to arrive at a non-orthogonal (oblique) solution, allowing the factors to be correlated. Factor loadings derived from the pattern matrices contain the coefficients representing unique contributions to variance explained for each factor.

The three datasets each had two factors with eigenvalues greater than 1. Together, these two factors account for over 45% of the variance observed in the data. For the pooled samples, this study follows the criteria of most studies and statistical software packages that use exploratory factor analysis in determining the number of principal components to retain based on the “Kaiser criterion” or “K1 rule” of eigenvalues greater than one and the “Scree test” where eigenvalues are plotted against the number of components and as the plot moves towards later components there is a sharp decrease in eigenvalues which then levels off (Tabachnick and Fidell, 2006: 607–675).

In the 1993 pooled samples, variation is explained by two factors: the first factor is “Environmental Pessimism” while the second factor had loadings with negative relationships to “Technological/Economic Centered” ideas, and is thus interpreted as prioritizing Ecocentrism. In both the 2000 and 2010 pooled samples, the first factor is identified as “Technological/Economic Centrism”, and the second factor as “Environmental Pessimism”. Referring to Aoyagi-Usui’s (1999: 126) interpretation, an “Environmental Pessimist considers modern human activities, including science, as inevitably harmful to the environment”, with high positive loadings on the four variables: EconGrowthAlwaysHarmsEnv, ModernLifeHarmsEnv, ModernScienceMoreHarm, and ScienceVsFeelingsFaith. Those who possess an Ecocentric view do not believe that science and economic growth are means of solving environmental problems (high negative loadings on the four statements WorryTooMuchProgHarmsEnv, WorryTooMuchEnvOverPriceJobs, ModernScienceSolveEnvProb, and EnvNeedsEconGrowth). On the other hand, those with a “Technological/Economic Centered” view prioritize science and economic growth, and subscribe to the belief that they can address environmental issues (high positive loadings on the four statements WorryTooMuchProgHarmsEnv, WorryTooMuchEnvOverPriceJobs, ModernScienceSolveEnvProb, and EnvNeedsEconGrowth). Overall, variables grouped together in the three pooled samples of the datasets and were consistent with the DSP and NEP categorization used by Aoyagi-Usui in 1999. Table 2 shows the factor loadings and their consistency, considering adjustments for variances explained and eigenvalues.

Table 2.

Factor loadings of pooled samples for 1993, 2000, 2010.

	1993 loadings		2000 loadings		2010 loadings
	Ecocentrism	Pessimism	Technocentrism	Pessimism	Technocentrism	Pessimism
	Factor 2	Factor 1	Factor 1	Factor 2	Factor 1	Factor 2
WorryTooMuchProgHarmsEnv	−.692	.162	.702	.106	.680	.151
WorryTooMuchEnvOverPriceJobs	−.664	.127	.690	.158	.690	.149
ModernScienceSolveEnvProb	−.653	−.075	.627	−.092	.586	−.100
EnvNeedsEconGrowth	−.634	−.102	.615	−.077	.609	−.119
EconGrowthAlwaysHarmsEnv	.026	.710	−.042	.709	−.112	.731
ModernLifeHarmsEnv	.148	.740	−.145	.709	−.157	.679
ModernScienceMoreHarm	−.124	.673	.117	.673	.188	.682
ScienceVsFeelingsFaith	−.074	.480	.084	.482	.298	.493
Eigenvalue	1.393	2.219	2.109	1.438	2.246	1.390
Total variance explained cumulative	45.157%		44.338%		45.451%

Extraction method: Principal Component Analysis. Rotation method: Oblimin with Kaiser normalization. Additional screening for factor retention done via Monte Carlo Test of Parallel Analysis.

The groupings of the technocentric DSP variables together can also be seen in a study made by Franzen and Meyer (2010) wherein these correspond to 4 of the 5 items of their second factor. They considered this grouping as encompassing “rational consideration of the likely influence of science, economy and the individual on environmental quality” (Franzen and Meyer, 2010: 223–225). These technocentric–ecocentric and optimistic–pessimistic dimensions also bear slight similarity with Bogner and Wiseman’s (2002) 2-factor structure of environmental attitudes encompassing “utilization” and “preservation”.

Average factor scores of each country from the 1993, 2000, and 2010 datasets are compared in Figure 2, which depicts the variation between countries and shows the changes over time. Respondents from Norway, Canada, New Zealand, and the United States have consistently remained environmental optimists and ecocentric. Whereas, factor score averages of Bulgaria and the Philippines depict environmentally pessimistic and technocentric attitudes in all three ISSP surveys.

Figure 2.

Attitudes towards science and nature: average factor scores in each country.

To apply Cohen’s typology (2000) in the analysis of these observations, the respondents’ technocentric and pessimism factor scores are utilized to classify individuals into four environmental knowledge orientations. Ecocidal mysticists are people with technocentric prioritization and pessimistic attitudes towards science, Prometheans possess technocentric and optimistic attitudes, Arcadians are ecocentric and pessimistic, while rational ecologists have ecocentric priorities and optimistic attitudes. To confirm the validity of these groupings, canonical covariance analysis (CCA) using discriminant functions (Tabachnick and Fidell, 2006: 384–405) on the eight variables on environmental attitudes from the combined datasets was conducted. The data points were significantly and correctly classified within the four environmental knowledge classifications derived from the three principal component analyses; CCA on the pooled samples of the three combined datasets resulted in the first discriminant function having an eigenvalue of 2.120 explaining 57.7% of the variance and a canonical correlation of .824. The second discriminant function had an eigenvalue of 1.551 explaining 42.2% of the variance and a canonical correlation of .780. All groupings were found to be significantly different from each other (Wilks’ Lambda = .125, P < 0.001).

Among the twelve countries in all three datasets, within the rational ecologist categorization, Norwegians always demonstrated the highest optimism towards modern institutions and science, while Canadians exhibited the highest ecocentric perspective. Conversely, in the ecocidal mysticist classification, Filipinos had increasing pessimism – with the highest levels for 2000 and 2010, along with slight increases in technocentrism. Bulgarians had noticeably decreased pessimism and technocentrism from 1993 to 2000, though maintained almost the same levels from 2000 to 2010. People from the Czech Republic had a Promethean knowledge orientation for 1993 and 2000, though increased pessimism situated it slightly within an ecocidal mysticist position by 2010. Slovenians had occupied an ecocidal mysticist position in 1993 and 2000, gradually decreased their technocentrism and eventually exhibited a slightly more Arcadian attitude by 2010. People in Great Britain possess a slight predisposition towards rational ecologism, though becoming Arcadian in 2000. Japan consistently remained in the Arcadian quadrant from 1993 to 2010, and noticeably decreased in pessimism in 2010.

By considering Cohen’s knowledge orientations along with the framework of Barrett et al. (2002), it may be possible to further infer, for instance, that people in four western developed countries (Canada, New Zealand, United Kingdom, and the United States) had attitudes favoring a rational ecologist knowledge orientation more consistently. These countries share relatively close levels of environmental optimism and ecocentricism among their peoples, and this alignment of attitudes towards science and nature may be considered as being derived from sets of common institutions, historical influences, and religious and cultural beliefs among these countries.

Though the results for each country may seem to reinforce the intuitive categorizations made by Cohen in terms of national character, the various origins of the surveys still had to be considered. Hence the four items of each of the factors extracted were assessed for their internal consistency by means of Cronbach’s alpha (Field, 2009). Results show relatively higher alphas for respondents from western industrialized countries, with the exception of Spain and Japan, as compared to post-communist and developing countries in the three surveys (Table 3). These may be partially explained in the observations of previous studies arguing that scales are “products of a certain organized space and time with respect to their conceptual content and formulation of issues in selected statements” (Erdogan, 2009: 1029). The use of unmodified scales in different cultures can be problematic in terms of reliability and validity, with “modern research on environmental values [being] predominantly western in origin” (Erdogan, 2009: 1029) utilizing theoretical concepts such as the DSP and NEP formulated and developed in the United States (Gooch, 1995). Varying results on the reliability and dimensionality of these constructs may also be attributed “to the contextual (cultural and historical) character of the study population” (Erdogan, 2009: 1029).

Table 3.

Cronbach’s alpha reliability statistics of factors in each country and pooled samples.

	ISSP 1993		ISSP 2000		ISSP 2010
	Four ecocentric items	Four pessimistic items	Four technocentric items	Four pessimistic items	Four technocentric items	Four pessimistic items
Great Britain (GB)	0.65	0.59	0.69	0.61	0.62	0.53
United States (USA)	0.63	0.55	0.63	0.49	0.57	0.55
Norway (N)	0.67	0.61	0.65	0.59	0.63	0.56
Czech Republic (CZ)	0.49	0.55	0.49	0.55	0.50	0.63
Slovenia (SLO)	0.58	0.62	0.58	0.60	0.54	0.65
Bulgaria (BG)	0.41	0.50	0.45	0.59	0.19	0.51
Russia (RUS)	0.25	0.48	0.32	0.49	0.38	0.51
New Zealand (NZ)	0.64	0.63	0.64	0.64	0.62	0.56
Canada (CDN)	0.61	0.61	0.64	0.52	0.63	0.53
Philippines (RP)	0.46	0.57	0.48	0.55	0.51	0.56
Japan (J)	0.51	0.50	0.53	0.44	0.56	0.54
Spain (E)	0.47	0.38	0.53	0.40	0.49	0.47
Pooled	0.59	0.56	0.58	0.55	0.57	0.58

Significant correlations with other ISSP variables

In ISSP 1993, higher DSP factor scores indicate ecocentrism, while in ISSP 2000 and ISSP 2010 increasing factor scores indicate technocentrism. Hence, as ecocentrism is the reverse of technocentrism, the positive correlation coefficients for DSP 1993 factor scores are treated as negative, and vice versa for the purpose of analysis. Nonparametric bivariate correlations (Spearman’s Rho) of the derived factor scores corresponding to both the DSP (technocentrism) and the NEP (pessimism) had consistently positive correlations with each other (.188, .158, .158) and on fatalistic environmental efficacy (EfficFatalism). Additionally, a derived environmental knowledge scale (EnvKnowledge) along with certain environmental behaviors such as signing petitions and giving money to environmental organizations, and educational degree obtained (educational level) were found to have negative correlations. Overall, the DSP factor scores had higher correlation coefficients with other ISSP variables as compared to the NEP factor scores. The following sections further discuss results for the above-mentioned variables. ANOVA with Tukey’s Post Hoc test was conducted to detect the significant differences (Tabachnick and Fidell, 2006: 331) between the four environmental knowledge classifications.

Scientific and environmental knowledge

Six variables defined by ISSP as “scientific and environmental knowledge” are utilized from the 1993 and 2000 datasets. Respondents chose from a scale ranging from “definitely true” to “definitely not true” on the following statements:

Antibiotics can kill bacteria but not viruses. (correct)

Human beings developed from earlier species of animals. (correct)

All man-made chemicals can cause cancer if you eat enough of them. (incorrect)

If someone is exposed to any amount of radioactivity, they are certain to die as a result. (incorrect)

The greenhouse effect is caused by a hole in the earth’s atmosphere. (incorrect)

Every time we use coal or oil or gas, we contribute to the greenhouse effect. (correct)

As with the treatment of these variables in previous research by Franzen and Meyer (2010) as well as that of Hadler and Haller (2011), each response was recoded with a low value indicating an “incorrect” answer and a high value a “correct” answer to obtain a composite scale. It must, however, be noted that “correctness” in such cases does not necessarily indicate only intellectual aptitude for scientific knowledge, but corresponds more with the agreement or affinity of a respondent with the prevailing views in science – the orthodoxy of people conforming to generally accepted or authorized theories, doctrines, and practices. The EnvKnowledge scale represents a respondent’s mean value across all items and ranges from 1 to 4 with a higher value indicating higher agreement with prevalent scientific and environmental knowledge. The answer “Can’t choose/don’t know” was coded as “1” (incorrect). To prevent the loss of cases due to missing values for some of the questions, for those with at least one response, the sum of the valid items is divided by six.

Nonparametric correlations of EnvKnowledge for ISSP 1993 and ISSP 2000 were found with variables on educational level (.319, .309), while there were negative coefficients for environmental efficacy variable EfficFatalism (−.280, −.284), the DSP factor scores (−.292, −.239), and the NEP factor scores (−.279, −.239). These are followed by the coefficient of the environmental behavior of signing a “petition” (.234, .221), and can be increased in correlation when using a scale, PublBehavScale (.254, .251) measuring all public environmental behavior responses.

On the basis of the average scores for environmental and scientific knowledge (EnvKnowledge) in each country for ISSP 1993 and ISSP 2000, respondents classified as rational ecologists had the overall highest average scores (2.765 in 1993 and 2.739 in 2000) for environmental and scientific knowledge; they also had the highest scores in every country for both datasets with the exception of Russia and the Philippines in 2000 where Promethean and Arcadian respondents in these countries respectively had a higher average score. Respondents classified as ecocidal mysticists seemingly had the lowest average scores among countries, with the exception of Bulgaria in 1993, Russia in 2000, the Philippines in 1993 and 2000. Overall, a very small percentage decrease (-0.22%) in environmental and scientific knowledge is seen from 1993 to 2000.

A parallel may be drawn to the findings of Allum et al. (2008) in their meta-analytic review of the relationship between public attitudes and public knowledge of science and technology. Their research also observed a “small positive correlation between general attitudes towards science and general knowledge of scientific facts” (p. 35), and furthermore also found “little cross-cultural variation in the correlation between knowledge and attitudes, with the variation explainable with a single country-level indicator of the proportion of the population going on to higher education” (p. 51).

Comparing environmental knowledge orientations, rational ecologists had the highest means for EnvKnowledge, while ecocidal mysticists had the lowest means. ANOVA with Tukey’s Post Hoc test shows significant differences between rational ecologist and other groups, although Prometheans and Arcadians did not have significant differences from each other for EnvKnowledge in 1993 and 2000.

Environmental efficacy and environmental behavior

Previous research on environmental attitudes found personal and societal efficacy to be relevant to environmental attitudes and behavior across cultures and nations (Aoyagi-Usui, 1999; Ester et al., 2004). To explore this aspect, a statement on environmental efficacy is utilized from ISSP 1993, 2000, and 2010 where respondents choose whether they “Strongly agree”, “Agree”, “Neither agree nor disagree”, “Disagree”, or “Strongly disagree”. The variable is recoded from 1 to 5 in order to indicate increasing agreement with the statement “It is just too difficult for someone like me to do much for the environment”.

For this statement reflecting fatalistic sentiment towards environmental efficacy, nonparametric correlation indicates positive correlations to DSP technocentrism (.384, .319, .324) and NEP pessimism scores (.222, .191, .227), along with negative coefficients with educational attainment (−.256, −.221, −.178), and environmental knowledge (−.280, −.284).

On the basis of average scores of fatalistic perceptions (EfficFatalism) in each country by environmental knowledge orientation for ISSP 1993, 2000, and 2010, respondents classified as ecocidal mysticists, with technocentric and environmental pessimistic attitudes, seemingly had the highest average scores among all twelve countries for all three surveys with the exception of Bulgaria in 2000; they also had the overall highest average scores (3.32 in 1993, 3.26 in 2000, and 3.27 in 2010), while rational ecologists had the lowest average scores (2.25 in 1993, 2.34 in 2000, and 2.29 in 2010). Generally among countries, rational ecologists and Arcadians had lower scores than Prometheans and ecocidal mysticists, but had increased in fatalism among groups at 1.79% and 5.08% respectively from 1993 to 2010. Comparing orientations, ecocidal mysticists had the highest scores for EfficFatalism, followed by Prometheans, Arcadians, and rational ecologists. ANOVA with Tukey’s Post Hoc test shows significant differences between groups.

Public environmental behavior such as “signing a petition about an environmental issue in the last five years” and “giving money to an environmental group in the last five years”, has the strongest inverse correlations with DSP and EfficFatalism for 1993, 2000, and 2010; this is followed by environmental knowledge having positive correlation for 1993 and 2000. These findings seem to agree with Hadler and Haller’s (2011: 324–325, 328) observation of scientific/environmental knowledge being substantial for becoming active, but with “knowledge only affecting public behavior – probably due to knowledge scales including general problems, but little about concrete problems in the immediate life context of a respondent”.

The percentage of respondents in a particular environmental knowledge orientation who engaged in at least one form of public environmental behavior (e.g. membership of environmental groups, signing petitions, giving money, or joining protests) yielded the following results: For industrialized countries classified as rational ecologists, the Arcadian and rational ecologist segments have the highest participation percentages. New Zealand has the highest percentage of respondents engaging in public behavior for all three surveys with Arcadians leading among the orientations followed by rational ecologists. Overall, countries classified as ecocidal mysticists have the lowest percentages of participation from all of their environmental knowledge orientation segments. These findings also lend support to Aoyagi-Usui’s assertion that environmental pessimists in Japan are least likely to engage in public environmental action when compared to environmental pessimists in other rational ecologist industrialized countries (1999: 128, 133). Comparing orientations, ANOVA with Tukey’s Post Hoc test shows ecocidal mysticists and Prometheans did not have significant differences from one another, but had significant differences (P < 0.001) from Arcadians and rational ecologists for 1993, 2000, and 2010.

Educational level and environmental concern

The level of educational attainment of respondents was recoded to range from “pre-secondary”, “secondary”, and “post-secondary” for the pooled samples of the twelve countries in 1993, 2000, and 2010. Negative correlations were found for both environmental attitude factor scores of DSP technocentrism (−.210, −.189, −.159) and NEP (−.209, −.190, −.184) environmental pessimism. Consistent positive correlations are found with environmental and scientific knowledge (.319, .303).

To explore education as a country-level variable, this paper also utilized the United Nations Development Programme’s “Education index”, an international human development indicator based on the mean years of schooling of adults and expected years of schooling of children (UNDP, 2011). A simple correlation of each country’s education index scores in 2000 and 2010 against its average DSP factor score for technocentrism and NEP factor score for pessimism, results in strong negative correlations with r values greater than −.7 and significant at the 0.01 level, thus indicating that countries with higher education index scores have more ecocentric and optimistic environmental attitudes.

Average scores for respondents’ willingness to make trade-offs or sacrifices for the environment are represented by a scale composed of responses to statements of “paying much higher prices”, “much higher taxes”, and “accepting cuts in their standard of living” in order to protect the environment. The scale is the average of each respondent’s answers, ranging from 1 to 5, that corresponds to their increasing willingness to make personal sacrifices for the environment. For 1993 and 2000, all industrialized countries categorized as rational ecologists had Arcadians as having the willingness to make trade-offs. Unfortunately, from 1993 to 2010 there was an overall decrease of −13.53%, with Arcadians having the greatest decrease in willingness among the four environmental knowledge orientations at −17.89% and rational ecologists having the least decrease at −8.50%.

It is observed that ecocidal mysticists predominantly compose the respondents at pre-secondary level, for eleven countries in 1993, 2000, and 2010, with the exception of Norway. For the post-secondary level, Japan had Arcadians as the majority, while all other industrialized countries classified as rational ecologists have the post-secondary level predominantly composed of rational ecologists.

Other observations

In the cross-national investigation of ISSP 1993 and 2000 by Franzen and Meyer (2010: 227–229), they found that between-country differences (15 per cent of the total variance) were best explained by the countries’ wealth as measured by gross domestic product (GDP) per capita based on purchasing power parity (PPP). Wherein, on average “populations in richer nations have higher levels of environmental concern than inhabitants of poorer countries” (Franzen and Meyer, 2010: 229). Their findings supported the affluence or prosperity hypothesis (Diekmann and Franzen, 1999) that claims that “as a population becomes wealthier, the demand for higher environmental quality should rise, which should result, in the aggregate, in a positive correlation between a nation’s wealth and its level of environmental concern” (Franzen and Meyer, 2010: 221). Similar findings were also found in multilevel analysis of ISSP 2010 data (Franzen and Vogl, 2013).

As such, this paper also explores “wealth” as a country-level variable defined as the country’s GDP per capita based on PPP (World Bank, 2008), of the same year that the ISSP environment survey was conducted. Similar to the results of Franzen and Meyer (2010: 226), results of this study show that wealth is positively correlated with DSP factor scores corresponding to attitudes towards ecocentrism. The highest levels of ecocentric attitudes are observed in countries with high levels of GDP per capita based on PPP such as Canada, Norway, New Zealand, USA, and Japan. Low levels can be observed in countries like Bulgaria, Russia, and the Philippines. In 1993, a simple correlation of a country’s GDP per capita based on PPP and its average DSP factor scores for ecocentrism results in a strong Pearson correlation of r = .86, P < 0.0001 and follows the observations of previous research in the subject matter (cf. Diekmann and Franzen, 1999). Interestingly, for both ISSP 2000 and ISSP 2010 where DSP factor scores pertain to technocentric attitudes, there are strong negative correlations (r = −.81, P = 0.0013 in 2000; r = −.82, P = 0.0012 in 2010) correspondingly indicating that the above-mentioned countries with high levels of GDP per capita based on PPP have the lowest levels of technocentrism among the twelve countries.

Wealth is found to seemingly have negative correlations with NEP factor scores that correspond to environmentally pessimistic attitudes for ISSP 1993, 2000, and 2010. The lowest levels of pessimistic attitudes are observed in countries with high levels of GDP per capita based on PPP such as Canada, Norway, and USA. High levels of pessimism can be observed in countries like Bulgaria and the Philippines. Although a simple correlation of a country’s GDP per capita based on PPP and its average NEP factor scores for pessimism resulted in Pearson correlations with mixed strengths of r = −.57 for 1993, r = −.55 for 2000, and r = −.91 for 2010, it was only in 2010 that correlation is significant at the 0.01 level.

6. Discussion and conclusions

Across twelve countries, the distributions of respondents among the orientations indicate that even among the wealthiest and most technologically advanced nations, a substantial part of their populations are still considerably ecocidal mysticist or Arcadian in their disposition towards science and nature. Hence, this implies that attitudes of people can adhere to a traditional perspective like the DSP but at the same time adopt a less anthropocentric worldview such as those espoused by the NEP.

Prometheanism and Arcadianism, as Cohen stated, serve as the “warp and weft of modernity”, that is “public faith in science, technology, and continual progress had regularly been tempered by the critical scrutiny of more circumspect traditions” (2000: 98–99). Those classified as Arcadians indeed exhibit environmentally conscious scrutiny as the group with the highest percentage to engage in at least one form of public environmental behavior. The results also show Prometheans and Arcadians as having significant differences in altruistic perspectives of environmental efficacy. As such, backlashes resulting from policy incongruence with personal environmental attitudes are most likely to be expressed by Arcadians. Notwithstanding, results of the surveys suggest that those who possess faith in modernity and science such as Prometheans and rational ecologists, possess higher environmental and scientific knowledge, along with a less fatalistic disposition towards environmental efficacy as compared to the ecocidal mysticists and Arcadians. Overall, however, the most stark contrasts in terms of knowledge, willingness to make trade-offs for the environment, fatalism, and engagement in public behavior are between those classified as ecocidal mysticists and those as rational ecologists, highlighting the importance of cultivating ecocentric and optimistic environmental attitudes.

The observed trends in environmental discourses in recent years which heavily favor scientific approaches and modern institutions, such as ecological modernization and green governmentalism (cf. Bäckstrand and Lövbrand, 2007), relate well with the ISSP datasets that show increased ecocentrism and optimism in science and modern life in the most developed countries, particularly from 2000 and 2010. Furthermore, among the four orientations, rational ecologists tend to exhibit among the samples the least decreases in willingness to make sacrifices for the environment and in altruism, and have comparable percentages to Arcadians in their participation in public environmental behaviors such as donating money to environmental groups and signing petitions for environmental causes.

It may be argued that current environmental discourses that appeal to people and countries with environmentally conscious and optimistic attitudes are in an advantageous position. For instance, the arguments for reconciling technological progress and environmental protection through “ecological modernization” or “sustainable development” endeavors become tenable for those adopting these attitudes. While rational ecologists have faith in science, modern institutions, and technology, at the same time they are willing to make personal trade-offs, and more importantly continue to show support and provide funding for the cause of environmental protection. Specifically for environmental issues difficult to perceive such as climate change, this bodes well as the correlations seem to suggest they would have increased awareness of the dangers of rising world temperatures, along with greater scientific knowledge of the atmosphere. Moreover, the willingness of people possessing this orientation to support initiatives despite their having to make sacrifices such as paying much higher taxes and prices, and to change their standard of living, will most likely be a major contribution to environmental programs that require significantly higher initial outlays and public support in their formative stages.

In the aggregate, as the richest countries in terms of GDP per capita based on PPP are consistently classified as rational ecologists, this may also translate into something that Cohen foresaw as “continued momentum to build internationally around an eco-modernist agenda, at least among the world’s most advanced countries” (2000: 100). That the qualities, or in this case, attitudes that are embodied by most of the contemporary environmentalists in the developed world, preeminently describe the “strong resolve for rationalism (in terms of optimism) and moderation (in terms of ecocentrism)” (Cohen, 2000: 100). Moreover, barring unforeseeable or exogenic factors within their national contexts, these predominant environmental attitudes of people enhance their own country’s capability to successfully chart a path towards a future roughly consistent with the principles of ecological modernization and eventually together with other like-minded nations towards multinational concerted efforts of green governmentalism.

Using environmental knowledge orientations may also allow an understanding of the obstacles in addressing global issues that are challenged by scientific uncertainty and the North–South divide. Among rational ecologists and Arcadians, as individuals or even collectively as countries, doubt or uncertainty as to the scientific evidence of anthropogenic causes and contribution to global warming may affect even their actual knowledge of environmental issues. As for Prometheans and ecocidal mysticists, outright skepticism or indifference towards environmental degradation may be a bigger reason for lower knowledge scores as they prioritize societal progress and economic development rather than intangible environmental concerns. Thus, a further challenge is to find innovative ways to engage individuals and countries in relation to their ecocentric orientations, educational levels, and beliefs in fatalistic environmental efficacy, in order for them to learn and accept scientific and environmental knowledge.

It is also believed by the author that a glimpse of the clearer picture becomes possible when considering environmental politics drivers such as political opportunity structures, environmental non-governmental organizations, social movements, environmental conditions and dramatic events, along with cultural aspects within the countries to provide contextual explanations (Brulle, 2010). This paper recognizes that more rigorous analysis is required before any firm confidence can be expressed in the validity of this analytical approach and caution should be exercised when applying it to different cultures. Aside from the low internal consistency of most countries for the two factors, this study was only able to utilize exploratory factor analysis as attempts at confirmatory factor analysis proved difficult for the datasets with their limited availability of attitudinal variables, yielding root mean squared error of approximation of over .05 and comparative fit indices scoring poorly with less than .95 (Gable and Wolf, 1993) for the majority of the country subsamples.

As such, future studies may apply multilevel and country-specific investigations that require additional data sources containing other variables and which are customized to a particular cultural context. Doing such may further explain overarching relationships within the results of these three ISSP datasets and open possibilities for predictive analysis and empirical establishment of causality. This particularly becomes salient when determining the reasons why certain countries would fall in seemingly counter-intuitive categorizations or explaining why people may have differences or changes in the dominant factors in their attitudes over time. This also becomes of relevance to global issues such as climate change when governments of countries such as Canada, New Zealand, and the United States – having comparably more ecocentric and optimistic populations along with higher environmental knowledge – start to pursue discordant and non-committal approaches to environmental governance that favor anthropocentric objectives, seemingly incongruous to the attitudes even of their own peoples.

Overall, this paper demonstrated the application of bi-axial dimensions and environmental knowledge orientations using environmental attitudes with survey data empirically showing that these classifications have significantly different patterns among populations of various countries in the last two decades. Thus, this may provide future research with an additional tool to understand environmental attitudes in terms of prioritization between the environment or economic progress, and also in terms of people’s disposition towards modern society, scientific knowledge, and environmental issues.

Footnotes

Funding

This research received no specific grant from any funding agency in the public, commercial, or not-for-profit sectors.

Author biography

Joseph Anthony L. Reyes is a PhD student at the Graduate School of Environmental Studies at Tohoku University and is a recipient of a Japanese Ministry of Education, Culture, Sports, Science, and Technology (MEXT) scholarship.

References

Allum

Sturgis

Tabourazi

Brunton-Smith

(2008) Science knowledge and attitudes across cultures: A meta-analysis. Public Understanding of Science 17(1): 35–54.

Aoyagi-Usui

(1999) A comparative analysis of citizens’ values and pro-environmental behavior. Asian Geographer 18(1–2): 123–134.

Aoyagi-Usui

Vinken

Kuribayashi

(2003) Pro-environmental attitudes and behaviors: An international comparison. Human Ecology Review 10(1): 23–31.

Bäckstrand

Lövbrand

(2007) Climate governance beyond 2012: Competing discourses of green governmentality, ecological modernization and civic environmentalism. In: Pettenger

(ed.) The Social Construction of Climate Change. Aldershot: Ashgate, pp. 123–147.

Barrett

BFD

Kuroda

Miyamoto

(2002) Ecological modernisation, environmental knowledge and societal change: Attitudes and behaviour of young people in Japan. International Research in Geographical and Environmental Education 11(3): 237–261.

Bechtel

Verdugo

De Queiroz Pinheiro

(1999) Environmental belief systems: United States, Brazil, and Mexico. Journal of Cross-cultural Psychology 30(1): 122–128.

Bogner

Wiseman

(2002) Environmental perception: Factor profiles of extreme groups. European Psychologist 7(3): 225–237.

Brulle

(2010) Politics and the environment. In: Leicht

Jenkins

(eds) The Handbook of Politics: State and Civil Society in Global Perspective. New York: Springer, pp. 385–406.

Cohen

(2000) Ecological modernisation, environmental knowledge and national character: A preliminary analysis of the Netherlands. Environmental Politics 9(1): 77–106.

10.

Corral-Verdugo

Armendáriz

(2000) The “New Environmental Paradigm” in a Mexican community. Journal of Environmental Education 31(3): 25–31.

11.

Corral-Verdugo

Carrus

Bonnes

Moser

Sinha

(2008) Environmental beliefs and endorsement of sustainable development principles in water conservation: Toward a New Human Interdependence Paradigm Scale. Environment and Behavior 40(5): 703–725.

12.

Diekmann

Franzen

(1999) The wealth of nations and environmental concern. Environment and Behavior 31(4): 540–549.

13.

Dunlap

(2008) The New Environmental Paradigm Scale: From marginality to worldwide use. Journal of the British Sociological Association 40(1): 3–19.

14.

Dunlap

Van Liere

(1978) The “New Environmental Paradigm”: A proposed measuring instrument and preliminary results. Journal of Environmental Education 9(4): 10–19.

15.

Dunlap

Van Liere

Mertig

Jones

(2000) Measuring endorsement of the New Ecological Paradigm: A revised NEP scale. Journal of Social Issues 56(3): 425–442.

16.

Erdogan

(2009) Testing the new ecological paradigm scale: Turkish case. African Journal of Agricultural Research 4(10): 1023–1031.

17.

Ester

Simões

Vinken

(2004) Cultural change and environmentalism: A cross-cultural national approach of mass publics and decision makers. Ambiente & Sociedade 7(2): 45–66.

18.

Field

(2009) Discovering Statistics Using SPSS. London: SAGE.

19.

Franzen

Meyer

(2010) Environmental attitudes in cross-national perspective: A multilevel analysis of the ISSP 1993 and 2000. European Sociological Review 26(2): 219–234.

20.

Franzen

Vogl

(2013) Two decades of measuring environmental attitudes: A comparative analysis of 33 countries. Global Environmental Change, https://dx-doi-org.web.bisu.edu.cn/10.1016/j.gloenvcha.2013.03.009

21.

Gable

Wolf

(1993) Instrument Development in the Affective Domain: Measuring Attitudes and Values in Corporate and School Settings. Evaluation in Education and Human Services series. Boston: Kluwer Academic.

22.

Gärling

Biel

Gustafsson

(2002) The new environmental psychology: The Human Interdependence Paradigm. In: Bechtel

Churchman

(eds) Handbook of Environmental Psychology. New York: Wiley, pp. 85–94.

23.

Gooch

(1995) Environmental beliefs and attitudes in Sweden and the Baltic States. Environment and Behavior 27(4): 513–539.

24.

Goodland

(1995) The concept of environmental sustainability. Annual Review of Ecology and Systematics 26(1): 1–24.

25.

Hadler

Haller

(2011) Global activism and nationally driven recycling: The influence of world society and national contexts on public and private environmental behavior. International Sociology 26(3): 315–345.

26.

Kelly

Kennedy

Faughnan

Tovey

(2003) Cultural Sources of Support on which Environmental Attitudes and Behaviours Draw. September. Dublin: Environmental Protection Agency.

27.

Milbrath

(1984) Environmentalists: Vanguard for a New Society. Albany: State University of New York Press.

28.

Milfont

Duckitt

(2004) The structure of environmental attitudes: A first- and second-order confirmatory factor analysis. Journal of Environmental Psychology 24(3): 289–303.

29.

Motherway

Kelly

Faughnan

Tovey

(2003) Trends in Irish Environmental Attitudes between 1993 and 2002. First report from the Research Programme on Environmental Attitudes, Values and Behaviour in Ireland. Department of Sociology, University College Dublin.

30.

Pirages

Ehrlich

(1974) Ark II: Social Response to Environmental Imperatives. San Francisco: W.H. Freeman.

31.

Tabachnick

Fidell

(2006) Using Multivariate Statistics, 5th edition. Boston: Allyn and Bacon.

32.

United Nations Development Programme (2011) Education index (expected and mean years of schooling). International Human Development Indicators. Human Development Report Office. Available at: http://hdrstats.undp.org/en/indicators/103706.html

33.

Van Petegem

Blieck

(2006) The environmental worldview of children: A cross-cultural perspective. Environmental Education Research 12(5): 625–635.

34.

Wiseman

Bogner

(2003) A higher-order model of ecological values and its relationship to personality. Personality and Individual Differences 34(5): 783–794.

35.

World Bank (2008) International Comparison Program database. World Development Indicators. Available at: http://data.worldbank.org/indicator/NY.GDP.PCAP.PP.KD/countries?display=default