• affective factors;
  • engagement;
  • environmental awareness;
  • environmental responsibility;
  • interest in science;
  • science education


  1. Top of page
  2. Abstract
  3. Assessed Factors
  4. Methodology
  5. Results
  6. Discussion
  7. References

This study investigates how affective and self-related factors impact participation in science learning and environmental awareness and responsibility. Using PISA 2006 datasets from Taiwan and Canada having similar level of science competency, the model for this study verifies and expands an earlier model by examining the relationships among science-related interest, enjoyment, self-efficacy, self-concept, leisure time engagement, and future intended interest in science and how these relationships synergistically interact with environmental awareness and responsibility. The most consistent finding revealed that students' science self-concept in both groups was weakly associated with future intended interest and engagement in science learning and with their sense of environmental awareness and responsibility. Reasons for this phenomenon and possible causes underlying why students' science self-concept was weakly connected to their future intended interest in science learning are also presented. Finally, how the results of this study are important to science education instruction and research are forwarded in which students' identity and beliefs about self in science need to part of the next generation of science education reforms. © 2014 Wiley Periodicals, Inc. J Res Sci Teach 51: 1084–1101, 2014

There is wide consensus that one of the major goals of education is to develop responsible citizens who are not only literate in science, mathematics, and reading but are also aware of the natural environment and its associated problems and issues (Birmingham & Barton, 2014; Choi, Lee, Shin, Kim, & Krajcik, 2011; Hungerford, 2009; Hungerford & Volk, 2000; Short, 2009). With this concern in mind, the focus of the 2006 Programme for International Student Assessment (PISA) was on science knowledge and its application in the context of life situations ranging from personal health, natural- and human-induced hazards, and ecosystems to the frontiers of science and technology. PISA, which is administered every three years under the guidance of the Organisation for Economic Co-operation and Development (OECD), is considered the largest international study of comparative performance in the literacy domains of reading, mathematics, and science (Normington, 2002). Starting in 2000, PISA has focused on one of these three literacy domains while also assessing the other two domains (Fleischman, Hopstock, Pelczar, & Shelley, 2010). The goal of PISA is to assess how students 15 years of age use their acquired skills and competencies in reading (emphasized in 2000 and 2009), mathematics (emphasized in 2003 and 2012), and science (emphasized in 2006 and 2015) in real-world scenarios. The results of these assessments produce a ranking of learning among these students within and among the participating countries (Fleischman et al., 2010; Stephens & Coleman, 2007). Thus, the results from PISA 2006, which focused on science literacy, can be used to inform educators and policymakers about the extent to which their students possess the potential to use their knowledge of science in a democratic way toward current science issues (Bybee, 2009).

Published results from various PISA 2006 studies have provided new insights into the interconnections between measures of knowledge, affect, and value as components of interest in science (Ainley & Ainley, 2011); school effectiveness in predicting students' environmental attitudes and awareness toward environmental issues (Coertjens, Pauw, Maeyer, & Petegem, 2010); gender differences in the effects of science interest and environmental responsibility on science aspiration and achievement between genders (Chiu, 2010); and how science interest, self-efficacy, enjoyment, and self-concept are connected to students' future intended interest and engagement in preferred leisure science activities (Lin, Lawrenz, Lin, & Hong, 2013). A common element among all of these studies is how the affective factors of interest and enjoyment are connected to students' science literacy. Results from these large-scale assessments provide educators and policymakers with a window into the broadly defined concept of science literacy, information regarding students' learning engagement, their personal background, and their affective perceptions (Lin, Lawrenz, et al., 2013).

Few studies, however, have investigated how students' future intended interest and engagement in science leisure activities affect their sense of environmental awareness and responsibility within and between two culturally and linguistically different populations. It was for this purpose that we choose to expand and verify the Lin, Lawrenz, et al. (2013) model by adding two additional variables—environmental awareness and environmental responsibility—and then testing its validity within and between Taiwan and Canada—two similarly performing countries but each with different cultural traditions and language backgrounds. In a proposed framework for PISA 2015, Hollweg et al. (2011) stated: “In nations around the world, educational leaders, policy makers, researchers, and educators have expressed the need for data on the status of environmental literacy, particularly as past environmental problems have worsened and new ones have arisen.” (p. 5.24). By expanding and testing the Lin, Lawrenz, et al. (2013) model, the interactions between variables measuring science literacy with those measuring environmental awareness and responsibility could reveal new insights into students' future intended interest in and engagement with science and environmental issues as members of the community willing to engage pressing environmental issues. Thus, the present study seeks to conduct this test on two fronts: verifying and extending the Lin, Lawrenz, et al. (2013) model and then applying the extended model to a dataset from a distinctly different country that demonstrated a similar level of competency.

First, utilizing the PISA 2006 Taiwan dataset, this study seeks to address these challenges by use of the Lin, Lawrenz et al. (2013) model to verify the affective factors of interest, enjoyment, and self-related cognition factors of self-concept and self-efficacy on future intended interest and engagement in preferred leisure science activities of the original model (inside broken-line boundary) and expand the original model to explore if these six factors are related to students' sense of environmental awareness and responsibility (Figure 1). Extending this model will allow investigators to determine if a synergistic relationship exists among these factors and environmental awareness and responsibility where new and stronger relationships are revealed. Researchers have said that few studies investigate the explicit connections between these areas (Littledyke, 2008). Outcomes of this study will provide evidence of the plausibility of this synergistic connection and potential insights for the next generation science education reforms.


Figure 1. The synergistic interactions among affective and self-related cognition factors and environmental awareness and responsibility (Expanded model from Lin, Lawrenz, et al., 2013).

Download figure to PowerPoint

Second, Lin, Lawrenz, et al. (2013) stated: “The data in this study were from one country and may not be representative of people from different countries, therefore, this study should be replicated using data from other countries.” (p. 11). Thus, the present study will duplicate the use of the Lin, Lawrenz, et al. model using the PISA 2006 Taiwanese dataset and applied to test the Canadian dataset, which revealed similar scientific competency. This duplication is considered among measurement specialists as a necessary step in testing the validity of previously discovered inferences originally derived from a model (American Educational Research Association [AERA], 1999). This study is concerned with addressing two research questions:

  1. What are the relationships among affective factors and self-related cognition factors on science engagement, future intended interest, and environmental awareness and responsibility among Taiwanese students?
  2. Do these relationships among the Taiwanese students apply to the Canadian students?

Assessed Factors

  1. Top of page
  2. Abstract
  3. Assessed Factors
  4. Methodology
  5. Results
  6. Discussion
  7. References

Affective Factors

Affective factors are defined as emotional experiences that capture and reflect how a person becomes aware, interprets, and emotionally relates to the environment (Mahn & John-Steiner, 2013). The affective factor of interest has been a focus of many studies within science education (e.g., Dohn 2011, 2013a, 2013b; Dohn, Madsen, & Malte, 2009; Hong & Lin, 2011; Hong, Lin, Chen, Wang, & Lin, 2014; Kang, Scharmann, Kang, & Noh, 2010; Lin, Hong, & Chen, 2013; Lin, Hong, & Huang, 2012; Lin, Lawrenz et al., 2013; Palmer, 2004). Interest has several distinguishing characteristics setting it apart from other factors. Namely, interest is always directed toward an object and plays a significant role in self-motivated behavior, attention, goal setting, and learning strategies in educational settings (Ainley, Hidi, & Berndorff, 2002; Hidi & Renninger, 2006). Students who are interested in science have a greater potential to seek out opportunities that contribute to their public understanding of science (Falk, Storksdieck, & Dierking, 2007). Another important aspect of interest is its two-fold operative nature. Thorndike (1935) stated that interest “acts in a forward direction to dispose the person towards certain behaviors, making him connect situations to responses different from those which would ensure if the interest were lacking.” (p. 58). For example, when students are interested, their motivation to become involved in learning activities is enhanced (Renninger, 2000; Schiefele, Krapp, & Winteler, 1992). Thorndike went on to say that interest “acts in a backward direction to make certain experiences satisfying and so to arouse a confirming reaction which causes the person to continue or repeat the behavior.” (p. 58). As students' interest increases during learning activities, their motivation to become further involved increases (Herwartz-Emden, Schurt, & Waburg, 2007; Schiefele, 2001). Thus, interest could be understood as representing two different states: current interest (i.e., backward direction) or future intended interest (i.e., forward direction). Thorndike stated, “If the behavior is satisfying enough to arouse the confirming reaction, it is continued or repeated.” (p. 58). Therefore, as students' current interest is satisfied, motivation is increased, which propels them forward and provides yet further opportunities for sustainable growth in their learning—an example of synergy.

Enjoyment is viewed as an emotion that is typically short in duration and manifested when a person's perceived skills match the perceived challenges of a particular activity (Fredrickson, 1998; Fredrickson & Branigan, 2005) and reflect a person's emotional beliefs as opposed to how a person thinks (Hartley, 2006). Relevant to this study is how researchers have defined enjoyment in relationship to learning experiences (Goetz, Hall, Frenzel, & Pekrun, 2006; Pekrun, Goetz, Titz, & Perry, 2002). Expressed thoughts such as “I look forward to learning science” reflect the enjoyment one expresses toward learning activities in school; “I enjoy going to botanical gardens” expresses the enjoyment toward preferred or self-selected leisure activities outside of school. Experiencing these emotions toward an area of interest inside or outside of school provides opportunities to develop a generalized sense of enjoyment for that area (Goetz et al., 2006). Enjoyment experienced by students during free-choice, science-related activities (e.g., science museums, astronomy, or robotic clubs) or environment-related activities (e.g., scouting, hiking, or summer camps) also provide opportunities for developing meaningful awareness of science and the environment (Palmberg & Kuru, 2000). These experiences both support and promote interest and engagement in learning—an example of synergy. Thus, the affective factors examined in this study represent students' emotional experiences rather than how they think.

Self-Related Cognition Factors

Self-related cognition represents how a person thinks of himself or herself. One such factor is self-concept, which is generally defined as a collection of beliefs about oneself within academic or social domains that influences different courses of action (Bong & Skaalvik, 2003; Marsh, Trautwein, Ludtke, Koller, & Baumert, 2005). This view of self has been found to be a product of several factors: personal experiences with and interpretations of an environment (Shavelson, Hubner, & Stanton, 1976), reinforcements and evaluations by significant others (Andersen, Glassman, & Gold, 1998), and ascribed expectations of future success and values (Eccles, 1983). As students' positive experiences in science learning feed their interests and provide them with enjoyment while receiving reinforcement from others closest to them, their self-concept about science-related engagement is positively motivated—an example of synergy.

Self-efficacy, another self-related cognition factor, is a measure of a person's self-belief in the ability to be successful in the completion of a task or accomplishment of a goal within a specific domain (Bandura, 1997). There is a general consensus among researchers that measures of science self-efficacy are useful in investigating students' choices of science-related activities, the effort and tenacity they put into the completion of these activities, and the extent to which they are successful (Usher & Pajares, 2008). Self-efficacy, as opposed to self-concept, is future-oriented in representing the extent to which a person is confident in their ability to achieve future success—an example of synergy.

Environmental Awareness and Responsibility Factors

In addition to the assessment of science competencies, PISA 2006 also focused on students' environmental awareness and responsibility. Environmental responsibility is a learned response or action that reflects a sense of ownership or empowerment (Palmberg & Kuru, 2000), and it is associated with the behavioral aspect of a student's sense of environmental literacy (Hsu, 2004). Environmental awareness can involve the affective (e.g., attitudes and beliefs), behavioral (e.g., choices and preferences), and cognitive (e.g., knowledge and responsibility) domains of the student (Balgopal & Wallace, 2009; Littledyke, 2008); it is central to students' self-evaluated sense of environmental literacy. Encouraging this sense of awareness is accomplished through supporting and strengthening positive student engagement in science learning. The Model of Responsible Environmental Behavior (Hines, Hungerford, & Tomera, 1987) suggested that people with a greater sense of personal responsibility, pro-environmental attitudes, or better knowledge of environmental issues are more likely to engage in pro-environmental behaviors. Several other models (Bamberg & Moser, 2007; Blake, 1999; Steg & Vlek, 2009; Tal & Morag, 2013) have explored variables associated with pro-environmental behavior and confirmed the positive relationship between environmental attitude and pro-environmental behavior—an example of synergy. Despite this, there is limited research clarifying how individual and collective affective factors differentially impact environmental awareness and responsibility.

Although environmental issues are included in the science curricula (e.g., food production; Dillon et al., 2005), results suggest that assessing and sustaining the development of students' environmental literacy was more easily accomplished during extracurricular activities than within the science classroom (Dori & Tal, 2000). Such student engagement appears best encouraged through preferred science leisure activities. Assessing students' environmental awareness within natural and informal settings also provides a more meaningful context for evaluating this awareness.

In summary, affective and self-related cognition factors previously mentioned as influential in science literacy also capture and reflect how people become both aware and emotionally and cognitively engaged with their environment. These factors have been used to investigate students' emotions and thinking toward science learning and how they reflect their attitudes toward science literacy. Few studies, however, have investigated how these emotions and self-concepts also relate to students' environmental awareness and sense of responsibility. The focus of this study is to investigate how these affective and cognitive factors toward science learning synergistically interact with students' sense of environmental awareness and responsibility.


  1. Top of page
  2. Abstract
  3. Assessed Factors
  4. Methodology
  5. Results
  6. Discussion
  7. References


Two nationally represented samples, one from Canada (n = 22,646) and one from Taiwan (n = 8,815), were used in this study. Each dataset represented 15-year-old students who participated in PISA 2006; both populations showed nearly identical levels of science competency with scores for Taiwan equal to 532 and Canada equal to 534 (OECD, 2007). In accordance with the national program manager's PISA 2006 manual, each student within the two datasets completed a student questionnaire and a science competencies test.

Factors Measured

A total of eight scales (hereafter factors) were considered in this study. Seven factors represented affective and self-related cognitive components assessed by the student questionnaire: interest in science learning (4 items); enjoyment of science learning (6 items); engagement in science learning (6 items); science self-efficacy (8 items); science self-concept (6 items); environmental awareness (5 items); and environmental responsibility (7 items). The eighth factor was future intended interest derived from attitudinal assessment (18 items) assigned to competency test scenarios within each of the PISA 2006 test booklets, with the number of attitudinal items measured varying from booklet to booklet (Lin, Lawrenz, et al., 2013).

Items used to assess students' affective factors were interest in science (e.g., “How interested are you in learning topics about chemistry?”), enjoyment from science (e.g., “I generally have fun when I am learning topics related to science.”), and engagement in preferred leisure science activities (e.g., “How often do you attend a science club?”). Items to assess students' self-related cognition factors focused on two dimensions: science self-efficacy (e.g., “How easily can you recognize scientific questions related to waste disposal?”) and science self-concept (e.g., “I understand new concepts of natural science easily.”). Items used to assess environment-related factors focused on two dimensions: environmental awareness measured how well students felt they were informed regarding environmental issues (e.g., “How informed are you about the increase of greenhouse gases?”) and environmental responsibility measured how much students agreed with behaviors to help protect the environment (e.g., “How much do you agree with carrying out regular checks on emissions from cars?”). Items used to assess the futuristic factor focused on future intended interest (subsets of items paired with specific content scenarios in the test booklets) measured if the student was interested in further consideration of the specified science ideas contained in the scenarios; for example, the acid rain scenario was followed by 3 items to access the participant's future considerations about (i) knowing which human activities contribute most to acid rain, (ii) learning about techniques that minimize the emission of gases that cause acid rain, and (iii) understanding the methods used to repair buildings damaged by acid rain (Lin, Lawrenz, et al., 2013).

Responses to items for all eight factors were measured using a 4-point Likert-type scale where higher responses (i.e., 4) indicated higher levels of the variable being considered and lower responses (i.e., 1) indicated lower levels. The determination of the psychometrics of the first seven factors used straightforward classical test theory and techniques. However, the design features for the science literacy test booklets and the future intended interest factor required the validity and reliability of these items to be determined differently.


Model building and item construction are not independent, but they must be separately verified using different strategies in order to be considered both valid and reliable. Thus, several different strategies were employed to assess the items in the eight factors of the proposed model.


Strategies utilized were face, construct, and structural validity. Face validity of all attitudinal items was determined by the PISA item panel of experts; translated, back translated, and pilot tested within each of the 21 participating countries; revised and submitted internationally for cross-national validation (Adams, 2009). The PISA 2006 test booklets and student questionnaire were formally administered and pilot tested with a calibration sample of 500 students per country to assess the construct validity and item dimensionality (OECD, 2009b).

Factor analysis was used to explore the structural validity of the measures and confirm the intended design structure. Confirmatory factor analysis using 8,815 Taiwan participants revealed that all items loaded on intended factor and the specific factor loadings for the items related to science interest (0.73–0.81), enjoyment (0.81–0.90), self-efficacy (0.65–0.73), self-concept (0.84–0.89), engagement (0.63–0.80), environmental awareness (0.70–0.83), and environmental responsibility (0.58–0.76). All of these factor loadings were statistically significant (p < 0.01). Similar factor analysis using 22,646 Canadian participants confirmed the item-factor structure with significant (p < 0.01) loadings of all items in their intended factors (0.62–0.75, 0.85–0.92, 0.68–0.72, 0.72–0.82, 0.46–0.81, 0.64–0.76, and 0.66–0.73, respectively).

Collectively, these results support the validity claims about the measure of the seven factors in the questionnaire using reasonably simple approaches. However, the validity of the future intended interest (FII) factor required innovative approaches. Unlike the science interest items on the PISA 2006 student questionnaire, the embedded interest items in the PISA 2006 test booklets explored more specifically students' FII in science and technology within particular real-world scenarios. Fensham (2009) suggested that assessing students' science interest within particular real-world situations explored more specifically students' attitudes toward science and technology—terms that have become vague within many nations and cultures. Adams (2009) reported that the construction and translation of 18 attitudinal items assessing FII embedded with specific science scenarios were rigorous and compliant with the PISA 2006 design standards. These design standards ensured that embedded interest and their support scales (i) are supported by leading interest theorists and current research; (ii) support the cognitive assessment items; (iii) are scalable, reliable, and consistent with the underlying construct of the items; and (iv) make conceptual sense (OECD, 2009a).

The PISA 2006 Technical Report indicated that the embedded interest items had low associations with other measures and were thereby assumed to be unique from the content of the scenarios (OECD, 2009b). Second, Lin, Lawrenz, et al. (2013) did not find any significant pathways between science competency and future interest or engagement in science. Third, a majority of the students felt a strong sense of consciousness toward environmental responsibility; however, this strong sense of responsibility was not associated with their science competency (Paden, 2012). These results support that FII documented a distinct construct of interest independent of science competency.


Reliability estimates of the items in seven factors from the student questionnaire were explored by internal consistencies (Cronbach's α) and by comparing those reported in the PISA 2006 Technical Report (OECD, 2009b) with those determined for the same items in the PISA 2006 datasets for Canada and Taiwan (Table 1). The purpose of this comparison was to confirm that the reliability of the measures for the target factors in this study were reasonable for model building (Martınez, Borko, & Stecher, 2012). Table 1 shows a consistency in the reliability estimates of the affective, self-related, and environment-related factors between the pilot study (n = 500) and those found for the Taiwan and Canada datasets (n = 8,815; n = 22,646, respectively).

Table 1. Collective internal consistency coefficients (Cronbach's alpha) for assessed factors
Assessed FactorsPISA 2006 PilotaPISA 2006 Formal Test
Canada n = 500Taiwan n = 500Canada n = 22,646Taiwan n = 8,815
  1. NR, not reported.

  2. a

    From PISA Technical Report (OECD, 2009b).

Environmental awareness0.770.810.760.81
Environmental responsibility0.820.800.820.79
Future intended interestNRNR0.900.90

Cronbach's α estimates for the seven factors considered were reasonably high and consistent across the pilot and formal studies for both datasets. Collectively, these reliability estimates for the eight factors suggest that they are a reasonably consistent measure of the underlying constructs (Table 1). The reliability estimates for the FII items were estimated for each test booklet and each country. The average internal consistency coefficients for the 13 test booklets for both datasets are shown in Table 1. Adams (2009) stated that the item constructed for PISA 2006 followed rigorous construction and administration procedures that resulted in high-quality data and trustworthy information on which to conduct secondary analyses and build models.

Statistical Analyses

The proposed model (Figure 1) suggests that students' environmental awareness and sense of responsibility toward the environment (dependent variables of pro-environmental attitude and behavior) and future intended interest and engagement in science (dependent variables of science attitude and behavior) are influenced by students' interest in science, enjoyment from science, engagement in learning science, science self-efficacy, and science self-concept (independent variables of science literacy). The theoretical model was tested using structural equation modeling (SEM), more specifically path modeling, with maximum likelihood estimations for path coefficients (Taasoobshirazi & Sinatra, 2011). SEM was chosen because it allows the examination of predicted relationships among a number of independent and dependent variables simultaneously in different constructed scenarios for the purpose of identifying which structure best fits the data being assessed (Schumacker & Lomax, 2010) and produces results that are less biased than linear least squares methods (Nieswandt, 2007).


  1. Top of page
  2. Abstract
  3. Assessed Factors
  4. Methodology
  5. Results
  6. Discussion
  7. References

The results of the data analyses are reported as the descriptive statistics, the model fit conditions, and the models with path strengths.

Descriptive Statistics

The descriptive statistics include the correlation matrix, mean, and standard deviation for the Taiwan dataset (Table 2). Pearson product-moment correlation coefficients (r) are typically used in structural equation modeling (Schumacker & Lomax, 2010) and were used in this study to measure the linear correlation (association) between the ordinal variable pairs in the model. Pearson's r assigns a value between +1 and −1, where 1 is total positive correlation, 0 is no correlation, and −1 is total negative correlation (Streiner, 2005). The intercorrelations for all eight variables in the model were positive and found to be significant for both countries. Using Pearson's r, the dependent variables of environmental awareness and responsibility for the Taiwan dataset are positively correlated with interest (r = 0.35; r = 0.22), enjoyment (r = 0.31; r = 0.20), self-efficacy (r = 0.48; r = 0.23), self-concept (r = 0.28; r = 0.10), engagement (r = 0.28; r = 0.17), and future intended interest (r = 0.25; r = 0.22). These results provide potential evidence on how the addition of the variables of environmental awareness and responsibility extended the predictive pathways of the Lin, Lawrenz, et al. (2013) original model. Furthermore, these results support the previous inferences of Lin, Lawrenz, et al., indicating that the affective and self-related cognition factors effecting students' views of science learning are highly associated with students' future intended interest and engagement in learning science. Clearly, associations among attitudes, self-related and future intended interest, and environmental awareness are higher than the associations with environmental responsibility.

Table 2. Correlation matrix, means, and standard deviations variables for Taiwan
Enjoyment 10.430.570.580.310.200.41
Self-efficacy  10.380.420.480.230.34
Self-concept   10.470.280.100.26
Environmental awareness     10.260.25
Environmental responsibility      10.22
Future intended interest       1
Standard deviation5.233.334.674.063.172.552.9416.63

The same descriptive statistics for the Canada dataset are reported in Table 3. Environmental awareness and responsibility are positively correlated with interest (r = 0.50; r = 0.53), enjoyment (r = 0.49; r = 0.50), self-efficacy (r = 0.64; r = 0.52), self-concept (r = 0.31; r = 0.28), engagement (r = .45; r = .45), and future intended interest (r = 0.20; r = 0.26).

Table 3. Correlation matrix, means, and standard deviations variables for Canada
Enjoyment 10.600.420.610.490.500.36
Self-efficacy  10.360.500.640.520.25
Self-concept   10.310.310.280.17
Engagement    10.450.450.32
Environmental awareness     10.530.20
Environmental responsibility      10.26
Future intended interest       1
Standard deviation5.934.205.957.743.153.614.8616.75

These results are similar to the results shown for Taiwan, which support the original model (Lin, Lawrenz, et al., 2013) and the addition of environmental awareness and responsibility variables to extend their model. However, unlike the Taiwan results (Table 2), the Canada results are mixed. Some associations among attitudes, self-related and future interest, and environment awareness are higher than the associations with environmental responsibility; and some associations among attitudes, self-related and future interest, and environmental responsibility are higher than the associations with environmental awareness (Table 3). Clearly the associations among attitudes, self-related and future interest, and environmental awareness and environmental responsibility for Canada are higher than for Taiwan.

Model Fit Conditions

The theoretical model of science and environmental literacy (Figure 1) was explored by analyzing the paths among the eight factors. A variety of fit statistics were used to measure the fitness of the datasets of both countries to this model. The test results of these model-fit measures included the goodness-of-fit index (GFI = 0.997 Canada, 0.995 Taiwan), the comparative fit index (CFI = 0.997 Canada, 0.993 Taiwan), Tucker-Lewis Index (TLI = 0.976 Canada, 0.932 Taiwan), the standardized root mean square residual (SRMR = 0.009 Canada, 0.017 Taiwan), and the root mean square error of approximation (RMSEA = 0.054 Canada, 0.080 Taiwan). Various sources support the position that these indices demonstrate good model fit (Schumacker & Lomax, 2010).

SEM Results

The SEM coefficients for both Taiwan and Canada are summarized in Table 4. The four independent variables of interest, enjoyment, self-efficacy (SE), and self-concept (SC) are shown (straight arrows) as directly relating to the dependent variables of FII in science and engagement in preferred leisure science activities of the original model and to environmental awareness and environmental responsibility of the expanded model being explored in this study.

Table 4. SEM comparisons between Canada and Taiwan datasets for pathway strengths (β)
Independent variables Dependent variablesCanadaTaiwan
Interest in science learning[RIGHTWARDS ARROW]Environmental responsibility0.230.10
Enjoyment of science learning[RIGHTWARDS ARROW] 0.110.08
Science self-efficacy[RIGHTWARDS ARROW] 0.150.09
Science self-concept[RIGHTWARDS ARROW] 0.01−0.08
Interest in science learning[RIGHTWARDS ARROW]Environmental awareness0.130.12
Enjoyment of science learning[RIGHTWARDS ARROW] 0.010.04
Science self-efficacy[RIGHTWARDS ARROW] 0.500.39
Science self-concept[RIGHTWARDS ARROW] 0.050.04
Interest in science learning[RIGHTWARDS ARROW]Engagement in science learning0.220.15
Enjoyment of science learning[RIGHTWARDS ARROW] 0.310.31
Science self-efficacy[RIGHTWARDS ARROW] 0.160.14
Science self-concept[RIGHTWARDS ARROW] 0.030.14
Interest in science learning[RIGHTWARDS ARROW]Future intended interest in0.300.34
Enjoyment of science learning[RIGHTWARDS ARROW]science learning0.160.21
Science self-efficacy[RIGHTWARDS ARROW] −0.030.16
Science self-concept[RIGHTWARDS ARROW] 0.00−0.06
  Environmental responsibility R20.400.10
  Environmental awareness R20.450.25
  Engagement in science learning R20.430.41
  Future intended interest in science learning R20.180.32

The pathway results for the Taiwan dataset verify generally those suggested by Lin, Lawrenz, et al. (2013), which were confirmed by the pathway results for the Canada dataset with the exception of the negative pathway strength between self-efficacy and FII. Visual models of the SEM results are provided for Canada (Figure 2) and Taiwan (Figure 3). In these two figures, the straight arrows represent paths showing direct effects and curved arrows indicate correlations.


Figure 2. The SEM results for the Canadian dataset.

Download figure to PowerPoint


Figure 3. The SEM results for the Taiwan dataset.

Download figure to PowerPoint

The curved lines and correlation coefficients between pairs of affective and self-related cognitive factors reveal that they are associated and share variance as expected from common features of the constructs and results from the earlier study. The pathways among the expanded model indicate that students' sense of environmental responsibility, environmental awareness, engagement, and future intended interest in science were all affected by their affective and self-related cognitive factors to varying degrees, except the SC-FII pathway of the Canadian dataset. Otherwise, only the pathways between SC and environmental responsibility, enjoyment and environmental awareness, and SC and environmental awareness are small (β ≤ 0.05) and the pathway between SE and FII was negative (β = −0.03) for the Canada dataset. The pathways from the Taiwan dataset were somewhat similarly well behaved and strong, with SC to environmental responsibility being negative (β = −0.08), SC to FII (β = −0.06), and both SC to environmental awareness and enjoyment to environmental responsibility (β = 0.04) being weak.

SEM figures show that four factors had direct effect on Canadian and Taiwan students' environmental responsibility, environmental awareness, engagement in preferred leisure science activities, and future intended interest in science. Among the pathways connected to environmental responsibility, interest in science learning showed the strongest pathway, β = 0.23 (Canada) and β = 0.10 (Taiwan), followed by enjoyment (β = 0.11; β = 0.08), self-efficacy (β = 0.15; β = 0.09), and self-concept (β = 0.01; β = −0.08). This means students who were interested in participating in preferred leisure science activities were more likely to feel a greater sense of environmental responsibility. However, the weakest of these paths suggests that this sense of responsibility is not necessarily a part of the students' personal sense of identity (i.e., self-concept).

The strongest pathway connection was between environmental awareness and science SE with a statistically significant β = 0.50 (Canada) and β = 0.39 (Taiwan), followed by interest in science (β = 0.13; β = 0.12). These results suggest that the confidence students acquired during preferred leisure science activities correlate with self-belief in their ability to be aware of the natural environment. The pathway connection between this belief and SC (β = 0.05; β = 0.05) and enjoyment (β = .01; β = 0.04) are not as strongly connected. These results suggest that students' belief in their personal competency in science and its connection to environmental awareness is not necessarily supported by their personal sense of identity or sense of enjoyment.

Good and relatively stable pathways are seen between engagement and enjoyment (Canada: β = 0.31; Taiwan: β = 0.31), engagement and interest (β = 0.22; β = 0.15), and engagement and SE (β = 0.16; β = 0.14). However, there was a mixed result regarding the pathway strength between engagement and SC (Canada: β = 0.03; Taiwan β = 0.14). This result suggests that engagement in science learning is more connected for Taiwanese students' sense of identity than for Canadian students.

Lastly, the pathway strengths from FII to interest in science learning (Canada: β = 0.30; Taiwan: β = 0.34) and to enjoyment of science learning (β = 0.16; β = 0.21) were statistically significant and showed relatively the same strength for the Canadian and Taiwanese students. However, the pathway strength between FII to SE (β = −0.03; β = 0.16) and to SC (β = 0.00; β = −0.06) revealed mixed and concerning results. The stark difference in pathway strength from FII to SE suggests that Canadian students' SE beliefs derived from their science learning are not necessarily associated with their FII in science. On the other hand, pathway strength from FII in science learning to SE among Taiwanese students indicates a negative connection.

In this study, regression results (R2) indicated the percentage of variance accounted for in the dependent variables (i.e., science future intended interest, engagement, environmental awareness, and environmental responsibility) by the combined independent variables. The R2 results (Table 4 and Figures 1 and 2) represent how the emotions experienced by these students during their preferred leisure science activities can be indicators and predictors of their future engagement in science learning (43% Canada; 41% Taiwan), future intended interest in science (18% Canada; 32% Taiwan), openness toward being informed about environment issues (45% Canada; 25% Taiwan), and behaviors that can protect the environment (40% Canada; 10% Taiwan). Both affective factors and the self-related cognition factors, which effect students' science future interest and engagement, have significant effects on their sense of environmental awareness and responsibility.

The results showing strong pathways between students' current interest in science and future interest in science learning and finding a strong pathway between enjoyment and future engagement in science learning were not surprising. These results from the Taiwan dataset reconfirm the findings by Lin, Lawrenz, et al. (2013) and are cross-validated by results from the Canada dataset. One of the most consistent findings was how students' SC in both groups was weakly associated with the students' FII and engagement in science learning and with their sense of environmental awareness and responsibility. These results suggest that students' beliefs about themselves in science, which was capable of producing positive results, were not associated with their future interest and engagement in science or with their environmental awareness or sense of environmental responsibility. Clearly, self needs to be considered if life-long learners and engaged citizens are goals of science education.


  1. Top of page
  2. Abstract
  3. Assessed Factors
  4. Methodology
  5. Results
  6. Discussion
  7. References

Model formation involves fitting predicted relationships and outcomes to valid and reliable data, which takes on an evolutionary pattern in most cases rather than being a revolutionary change. The present study attempted to replicate this model building–verification–expansion process using the model proposed by Lin, Lawrenz, et al. (2013). First, we verified this 6-factor model among Taiwanese students using the 2006 PISA dataset. Second, we verified the model for Canadian students whose 2006 PISA performance was similar to the Taiwanese students but whose culture and language are different. Our purpose here was to cross-validate the fitness of the six-factor model. The results of these analyses revealed that the original six-factor model was robust for cultures of two different customs and language traditions.

Another purpose was to propose an expanded model that evolved the six-factor model to involve two additional factors of environmental awareness and responsibility. The expanded model was believed to more closely reflect pro-environmental features and to make better use of the PISA datasets. The eight-factor model-fitness results support the proposed relationships among students' environmental awareness and sense of responsibility toward the environment as well as FII and engagement in science and their interest in science, enjoyment from science, engagement in learning science, science self-efficacy, and science self-concept.

Two findings of this study believed to be noteworthy are measures of students' SE and SC, which might differ between Asian and North American cultures. Regarding SE, even though this self-related cognition factor was found to be well connected to environmental awareness (Canada: β = 0.50; Taiwan: β = 0.39), its connections to engagement in science learning (Canada: β = 0.16; Taiwan: β = 0.14), environmental responsibility (Canada: β = 0.15; Taiwan: β = 0.09), and future intended interest in science learning (Canada: β = −0.03; Taiwan: β = 0.16) were disappointing and small. Since SE is future-oriented in representing the extent to which people are confident in their ability to achieve future success and we assumed FII, it would be reasonable to expect to find good connections between SE and FII. SE assessments (e.g., “How easily can you recognize scientific questions related to waste disposal?”) were assessed in the PISA 2006 student questionnaire but were not embedded within a specific situation/context. FII assessments, on the other hand, were subsets of items paired with content scenarios focused on specific, not general, science ideas. Therefore, interest in the specific context/content may be a confounding effect on FII in this SEM analysis of the proposed model. SE experts have warned that if measures of SE were not specifically aligned with criteria tasks being assessed, a “global or generalized” SE measure of attitudes is likely to result (Pajares, 1996). Such a global or generalized measure was viewed by Bandura (1997) as an inappropriate measure of SE.

The second concern is the zero or weak pathway connections from SC to FII, engagement in science, environmental awareness, and environmental responsibility (i.e., Taiwan βs were all small negative or positive values, but Canada βs were small positive values). These results suggest that students do not view science learning and environmental awareness and responsibility as connected to their personal core sense of self. What is critical to students' lifelong learning of science is their science identity and self where science is an object of interest and relevant to their personal and social lives. Such interest transforms students by expanding their emotion-related and value-related awareness of learning in a new, exciting, and meaningful way (Eccles, 2007; Pekrun, 2006; Pekrun et al., 2002). This interaction between affective and self-related cognition factors influencing both attitudes toward science learning and environmental awareness and responsibility is what we define as the synergistic effect of affective factors on student learning outcomes; therefore, we view such student attitudes toward science learning and environmental awareness and responsibility as itself a priority student learning outcome.

No strong consistent cultural differences were apparent for the Canadian and Taiwanese patterns of relationships between self-related factors and FII, engagement, environmental awareness, and environmental responsibility. However, the results of the present study suggest that students' attitudes toward environmental awareness and responsibility may be connected to their positive FII and engagement in learning science and should be connected to their mainstream science literacy involving the recognition and engagement in the public debate about science, technology, society, and environment (STSE) or socioscientific issues (Yore, 2012).

Consistent with the work of previous investigators (Chiu, 2010; Gokmenoglu, Eret, & Kiraz, 2011; Lin & Shi, 2014; Littledyke, 2008), this study emphasized the importance of positioning assessments of environmental awareness and responsibility, which are central to environmental literacy, within the PISA science literacy framework. The synergistic interactions of students' science-related affective and self-related cognition factors with these two aspects of environmental literacy provides educators and policymakers with up-to-date information regarding students' learning engagement and attitudes toward personal responsibility—both of which are important in teaching science and environmental education.

Developing student competencies in science and environmental education while neglecting the leisure science-learning component may prove insufficient for developing the kinds of citizens who are willing to be involved in these areas and necessary public consideration of pressing STSE issues when given free choice. Previous researchers have found competency is not necessarily indicative of students' engagement and understanding of science (Lin, Lawrenz, et al., 2013), nor of a sense of environmental responsibility (Paden, 2012) or awareness. Key to the development of student engagement, understanding, and sense of personal and civic responsibility is the overarching impact that emotions, identity, and self have in connecting science and environmental understandings to awareness and desirable behavior toward the environment. The synergistic interplay among competency, emotion, and responsible engagement appears fundamental to developing and sustaining a deeper understanding and commitment to environmental issues—attributes that are vital to any sustainable effort directed toward mainstream science literacy for citizenship and the democratization of science.

Utilizing the model presented in this study, future educators and investigators will be able to conduct further experimental investigations to define and clarify the specific causal relationships anchoring students' affective and self-related cognition factors with their learning of science and the development of positive environmental attitudes and behaviors. This may require the proposal and verification of multi-tier models and more sophisticated statistical techniques to identify latent and mediating variables with secondary analyses of the 2015 PISA datasets. However, the current results should be used to leverage the emphasis of environmental education in the next generation science education frameworks and standards and to increase the priority and explicit consideration of affective factors, identity, and self in science and environmental education instruction.

The authors would like to thank Marvin G. Connatser and Shari Yore for their incisive editing of this manuscript and the anonymous reviewers for their comments regarding its structure and content. This study was supported by the Taiwan National Science Council, under grant NSC100-2511-S-110-004-MY3.


  1. Top of page
  2. Abstract
  3. Assessed Factors
  4. Methodology
  5. Results
  6. Discussion
  7. References
  • Adams, R. J. (2009). PISA: Frequently answered criticisms. In R. W. Bybee & B. McCrae (Eds.), PISA Science 2006: Implications for science teachers and teaching. (pp. 3945). Arlington, VA: NSTA Press.
  • American Educational Research Association. (1999). Standards for educational and psychological measurement. Washington, DC: American Psychological Association & National Council on Measurement in Education.
  • Ainley, M., & Ainley, J. (2011). A cultural perspective on the structure of student interest in science. International Journal of Science Education, 33(1), 5171.
  • Ainley, M., Hidi, S., & Berndorff, D. (2002). Interest, learning, and the psychological processes that mediate their relationship. Journal of Educational Psychology, 94(3), 545561.
  • Andersen, S. M., Glassman, N. S., & Gold, D. A. (1998). Mental representations of the self, significant others, and nonsignificant others: Structure and processing of private and public aspects. Journal of Personality and Social Psychology, 75(4), 845861.
  • Balgopal, M. M., & Wallace, A. M. (2009). Decisions and dilemmas: Using writing to learn activities to increase ecological literacy. Journal of Environmental Education, 40(3), 1326.
  • Bamberg, S., & Moser, G. (2007). Twenty years after Hines, Hungerford, and Tomera: A new meta-analysis of psycho-social determinants of pro-environmental behaviour. Journal of Environmental Psychology, 27, 1425.
  • Bandura, A. (1997). Self-efficacy: The exercise of control. New York, NY: Worth.
  • Birmingham, D., & Barton, A.C. (2014). Putting on a green carnival: Youth taking educated action on socioscientific issues. Journal of Research in Science Teaching, 51(3), 286314.
  • Blake, J. (1999). Overcoming the ‘value-action gap’ in environmental policy: Tensions between national policy and local experience. Local Environment: International Journal of Justice and Sustainability, 4(3), 257278.
  • Bong, M., & Skaalvik, E. M. (2003). Academic self-concept and self-efficacy: How different are they really? Educational Psychology Review, 15(1), 140.
  • Bybee, R. W. (2009). PISA 2006: An assessment of scientific literacy. Journal of Research in Science Teaching, 46(8), 865883.
  • Chiu, M.-S. (2010). Effects of science interest and environmental responsibility on science aspiration and achievement: Gender differences and cultural supports. Educational Research and Evaluation, 16(4), 345370.
  • Choi, K., Lee, H., Shin, N., Kim, S.-W., & Krajcik, J. (2011). Re-conceptualization of scientific literacy in South Korea for the 21st century. Journal of Research in Science Teaching, 48(6), 670697.
  • Coertjens, L., Pauw, J. B-d., Maeyer, S. d, & Petegem P. V. (2010). Do schools make a difference in their students' environmental attitudes and awareness? Evidence from PISA 2006. International Journal of Science and Mathematics Education, 8, 497522.
  • Dillon, J., Morris, M., Donnell, L. O., Reid, A., Rickinson, M., & Scott, W. (2005). Engaging and learning with the outdoors—The Final Report of the Outdoor Classroom in a Rural Context Action Research Project (pp. 193). Slough, England: NFER.
  • Dohn, N. B. (2011). Situational interest of high school students who visit an aquarium. Science Education, 95, 337357.
  • Dohn, N. B. (2013a). Situational interest in engineering design activities. International Journal of Science Education, 35(12), 20572078.
  • Dohn, N. B. (2013b). Upper secondary students' situational interest: A case study of the role of a zoo visit in a biology class. International Journal of Science Education, 35(16), 27322751.
  • Dohn, N. B., Madsen, P. T., & Malte, H. (2009). The situational interest of undergraduate students in zoophysiology. Advances in Physiology Education, 33, 196201.
  • Dori, Y. J., & Tal, R. T. (2000). Formal and informal collaborative projects: Engaging in industry with environmental awareness. Science Education, 84, 95113.
  • Eccles, J. S. (1983). Expectancies, values, and academic behaviors. In J. T. Spence (Ed.) Achievement and achievement motives. (pp. 75146). San Francisco, CA: W. H. Freeman.
  • Eccles, J. S. (2007). Subjective task value and the Eccles et al. model of achievement-related choices. In A. J. Elliot & C. S. Dweck (Eds.), Handbook of competence and motivation. (pp. 105121). New York, NY: Guilford Press.
  • Falk, J. H., Storksdieck, M., & Dierking, L. D. (2007). Investigating public science interest and understanding: Evidence for the importance of free-choice learning. Public Understanding of Science, 16, 455469.
  • Fensham, P. J. (2009). Real world contexts in PISA science: Implications for context-based science education. Journal of Research in Science Teaching, 46(8), 884896.
  • Fleischman, H. L., Hopstock, P. J., Pelczar, M. P., & Shelley, B. E. (2010). Highlights from PISA 2009: Performance of U.S. 15-year-old students in reading, mathematics, and science literacy in an international context. Washington, DC: U.S. Department of Education, National Center for Education Statistics.
  • Fredrickson, B. L. (1998). What good are positive emotions? Review of General Psychology, 2(3), 300319.
  • Fredrickson, B. L., & Branigan, C. (2005). Positive emotions broaden the scope of attention and thought-action repretoires. Cognition and Emotion, 19(3), 313322.
  • Goetz, T., Hall, N. C., Frenzel, A. C., & Pekrun, R. (2006). A hierarchical conceptualization of enjoyment in students. Learning and Instruction, 16, 323338.
  • Gokmenoglu, T., Eret, E., & Kiraz, E. (2011). The link between environmental awareness and science-related factors: Suggestions for science curriculum, evaluation in education in the Balkan countries. Belgrade, Serbia.
  • Hartley, D. (2006). Excellence and enjoyment: The logic of a ‘contradiction’. British Journal of Educational Studies, 54(1), 314.
  • Herwartz-Emden, L., Schurt, V., & Waburg, W. (2007). Interest, self-efficacy, and achievement evaluations in physics: A study on the effects of single-sex education in German schools. In M. Prenzel (Ed.), Studies on the educational quality of schools: The final report on the DFG priority programme. (pp. 265282). Munster, Germany: Waxmann.
  • Hidi, S., & Renninger, K. A. (2006). The four-phase model of interest development. Educational Psychologist, 41(2), 111127.
  • Hines, J. M., Hungerford, H. R., & Tomera, A. N. (1987). Analysis and synthesis of research on responsible environmental behavior: A meta-analysis. Journal of Environmental Education, 18(2), 18.
  • Hollweg, K. S., Taylor, J. R., Bybee, R. W., Marchinkowski, T. J., McBeth, W. C., & Zoido, P. (2011). Developing a framework for assessing environmental literacy. Washington, DC: North American Association for Environmental Education.
  • Hong, Z.-R., & Lin, H-s. (2011). An investigation of students' personality traits and attitudes toward science. International Journal of Science Education, 33(7), 10011028.
  • Hong, Z.-R., Lin, H-s., Chen, H.-T., Wang, H.-H., & Lin, C.-J. (2014). The effects of aesthetic science activities on improving at-risk families children's anxiety about learning science and positive thinking. International Journal of Science Education, 36(2), 216243.
  • Hsu, S.-J. (2004). The effects of an environmental education program on responsible environmental behavior and associated environmental literacy variables in Taiwanese college students. Journal of Environmental Education, 35(2), 3748.
  • Hungerford, H. R. (2009). Environmental education (EE) for the 21st century: Where have we been? Where are we now? Where are we headed? Journal of Environmental Education, 41(1), 16.
  • Hungerford, H. R., & Volk, T. L. (2000). Changing learner behavior through environmental education. Journal of Environmental Education, 21(3), 821.
  • Kang, H., Scharmann, L. C., Kang, S., & Noh, T. (2010). Cognitive conflict and situational interest as factors influencing conceptual change. International Journal of Environmental & Science Education, 5(4), 383405.
  • Lin, E., & Shi, Q. (2014). Exploring individual and school-related factors and environmental literacy: Comparing U.S. and Canada using PISA 2006. International Journal of Science and Mathematics Education, 12(1), 7397.
  • Lin, H-s., Hong, Z.-R., & Chen, Y.-C. (2013). Exploring the development of college students' situational interest in learning science. International Journal of Science Education, 35(13), 21522173.
  • Lin, H-s., Hong, Z.-R., & Huang, T.-C. (2012). The role of emotional factors in building public scientific literacy and engagement with science. International Journal of Science Education, 34(1), 2542.
  • Lin, H-s., Lawrenz, F., Lin, S.-F., & Hong, Z.-R. (2013). Relationships among affective factors and preferred engagement in science-related activities. Public Understanding of Science, 22(8), 941954.
  • Littledyke, M. (2008). Science education for environmental awareness: Approaches to integrating cognitive and affective domains. Environmental Education Research, 14(1), 117.
  • Mahn, H., & John-Steiner, V. (2013). Vygotsky and sociocultural approaches to teaching and learning. In I. B. Weiner, W. M. Reynolds, & G. E. Miller (Eds.), Handbook of psychology (Vol. 7, 2nd ed., pp. 117146). Hoboken, NJ: John Wiley.
  • Marsh, H. W., Trautwein, U., Ludtke, O., Koller, O., & Baumert, J. (2005). Academic self-concept, interest, grades, and standardized test scores: Reciprocal effects models of causal ordering. Child Development, 76(2), 397416.
  • Martınez, J. F., Borko, H., & Stecher, B. M. (2012). Measuring instructional practice in science using classroom artifacts: Lessons learned from two validation studies. Journal of Research in Science Teaching, 49(1), 3867.
  • Nieswandt, M. (2007). Student affect and conceptual understanding in learning chemistry. Journal of Research in Science Teaching, 44(7), 908937.
  • Normington, D. (2002). Transforming education. Education Journal, (62), 68.
  • Organisation for Economic Co-operation and Development. (2007). PISA 2006: Science competencies for tomorrow's world executive summary. Paris, France: Author.
  • Organisation for Economic Co-operation and Development. (2009a). Assessing scientific, reading, and mathematical literacy: A framework for PISA 2006. Paris, France: Author.
  • Organisation for Economic Co-operation and Development. (2009b). PISA 2006 technical report. Paris, France: Author.
  • Paden, M. (2012). NAAEE releases framework for assessing environmental literacy: Being used in 2015 OECD assessment. Journal of Education for Sustainable Development, 6(1), 1719.
  • Pajares, F. (1996). Self-efficacy beliefs in academic settings. Review of Educational Research, 66(4), 543578.
  • Palmberg, I. E., & Kuru, J. (2000). Outdoor activities as a basis for environmental responsibility. Journal of Environmental Education, 31(4), 3236.
  • Palmer, D. H. (2004). Situational interest and the attitudes towards science of primary teacher education students. International Journal of Science Education, 26(7), 895908.
  • Pekrun, R. (2006). The control-value theory of achievement emotions: Assumptions, corollaries, and implications for educational research and practice. Educational Psychology Review, 18(4), 315341.
  • Pekrun, R., Goetz, T., Titz, W., & Perry, R. P. (2002). Academic emotions in students' self-regulated learning and achievement: A program of qualitative and quantitative research. Educational Psychologist, 37(2), 91105.
  • Renninger, K. A. (2000). Individual interest and its implications for understanding intrinsic motivation. In J. M. Harackiewicz & C. Sansone (Eds.), Intrinsic and extrinsic motivation: The search for optimal motivation and performance. (pp. 373404). San Diego, CA: Academic Press.
  • Schiefele, U. (2001). The role of interest in motivation and learning. In J. M. Collis & S. Messick (Eds.), Intelligence and personality: Bridging the gap in theory and measurement. (pp. 163194). Mahwah, NJ: Lawrence Erlbaum.
  • Schiefele, U., Krapp, A., & Winteler, A. (1992). Interest as a predictor of academic achievement: A meta-analysis of research. In K. A. Renninger, S. Hidi, A. Krapp, & A. Renninger (Eds.), The role of interest in learning and development. (pp. 183212). Hillsdale, NJ: Lawrence Erlbaum.
  • Schumacker, R. E., & Lomax, R. G. (2010). A beginner's guide to structural equation modeling (3rd ed.) New York, NY: Routledge.
  • Shavelson, R. J., Hubner, J. J., & Stanton, G. C. (1976). Self-concept: Validation of construct interpretations. Review of Educational Research, 46(3), 407441.
  • Short, P. C. (2009). Responsible environmental action: Its role and status in environmental education and environmental quality. Journal of Environmental Education, 41(1), 721.
  • Steg, L., & Vlek, C. (2009). Encouraging pro-environmental behaviour: An integrative review and research agenda. Journal of Environmental Psychology, 29, 309317.
  • Stephens, M., & Coleman, M. (2007). Comparing PIRLS and PISA with NAEP in reading, mathematics, and science (Working Paper). US Department of Education. Washington, DC: National Center for Education Statistics.
  • Streiner, D. L. (2005). Finding our way: An introduction to path analysis. Canadian Journal of Psychiatry, 50(2), 115122.
  • Taasoobshirazi, G., & Sinatra, G. M. (2011). A structural equation model of conceptual change in physics. Journal of Research in Science Teaching, 48(8), 901918.
  • Tal, T., & Morag, O. (2013). A longitudinal study of environmental and outdoor education: A cultural change. Journal of Research in Science Teaching, 50(9), 10191046.
  • Thorndike, E. L. (1935). Adult interests. New York, NY: MacMillan.
  • Usher, E. L., & Pajares, F. (2008). Sources of self-efficacy in school: Critical review of the literature and future directions. Review of Educational Research, 78(4), 751796.
  • Yore, L. D. (2012). Science literacy for all—More than a slogan, logo, or rally flag! In K. C. D. Tan, M. Kim, & S. Hwang (Eds.), Issues and challenges in science education research: Moving forward (pp. 523). Dordrecht, The Netherlands: Springer.