This item is only available as the following downloads:
DO CHILDREN WHO KNOW MORE, CARE MORE? ENVIRONMENTAL KNOWLEDGE AND SCOPE OF JUSTICE BY JENNIFER DYER A Thesis Submitted to the Division of Psychology New College of Florida in partial fulfillment of the requirements for the degree Bachelor of Arts Under the sponsorship of Dr. Michelle Barton Sarasota, Florida May, 2009
ii Acknowledgements Thank you to Big Cypress Elementary, the students I intervie wed for this study, and my mom for helping me coordinate the pr oject. I would also like to thank Dr. Barton for her excellent guidance during the thesis process. Thanks to Christina Rogers-Hehr and Kathleen Maxwell for helping check reliability. Lastly, thank you Jasper, for listening to me talk about kids and the environment for hours!
iii Table of Contents ACKNOWLEDGEMENTS ii TABLE OF CONTENTS iii LIST OF TABLES AND FIGURE v ABSTRACT vi INTRODUCTION: Premise of Environmental Education 1 STUDIES COMBINING KNOWLEDGE AND ATTITUDE 3 CHILDRENS ENVIRONMENTAL KNO WLEDGE 6 Perception as a Form of Knowledge 15 CHILDRENS ATTITUDES ABOU T THE ENVIRONMENT 20 Topic-specific Measurements 21 General Measurements 22 SCOPE OF JUSTICE 25 THE CURRENT STUDY 28 METHOD 29 Participants 29 Materials 30 Procedure 33 Reliability 36 RESULTS 36 DISCUSSION 41 REFERENCES 52 APPENDICES Appendix A: Interview Protocol 58
iv Appendix B: Illustrations 67 TABLES 74 FIGURE 86
v List of Tables and Figure TABLE 1: Knowledge Question Topics 74 TABLE 2: Response Categories for What does the word nature mean to you? 75 TABLE 3: Scope of Justice Story Expansion Codes 76 TABLE 4: Knowledge Test Answer Key with Abridged Questions 77 TABLE 5: Examples of Anthropocentris m in Knowledge Test Responses 83 TABLE 6: Frequency of Scope of Justice Expansion Type by Utility Group (and Percentages) 84 TABLE 7: Percent of Participants Per Category of Response 85 FIGURE 1: Per-Participant Knowledge and Scope of Justice Scores 86 with Line of Best Fit
vi DO CHILDREN WHO KNOW MORE, CARE MORE? ENVIRONMENTAL KNOWLEDGE AND SCOPE OF JUSTICE Jennifer Dyer New College of Florida, 2009 ABSTRACT This study explored a potential corre lation between childre ns environmental knowledge and scope of justice (a measure of attitude, defined here as the amount of fairness deserved by the environment). Th irty 5-7 year olds were interviewed. Participants were divided between two groups: positive utility (primed to think of nature as beneficial to humans) and neutral utility (not primed). Participants were told five stories about human-environment conflicts, e ach with a low and hi gh conflict version. In the context of the stories, participants deci ded whether to act in a pro-environmental way, or in a way that benefited humans. Next, a 15-question environmental knowledge interview was completed. The main hypothesis st ated that knowledge scores and scope of justice scores would positively correlate. It was also hypothesized that in high conflict situations participants would ha ve a less inclusive scope of justice, and that participants in the positive utility group would have a more inclusive scope of justice than those in the neutral group. Knowledge and scope of justice scores showed a positive trend. As expected, high conflict situations elicited fewer environmentally inclus ive responses. Contrary to hypotheses, those in the positive utility condition had a less in clusive scope of justice.
vii The results imply a possible relationship be tween environmental knowledge and attitude in children. ______________________________________________ Dr. Michelle Barton Division of Social Sciences
1 Environmental education in creases public awareness and knowledge about environmental issu es or problems. In doing so, it provides the public with the necessary skills to make informed decisions and take responsible action. (U.S. Environmental Protection Agency, 2008) There is a considerable debate on the extent to which the acquisition of knowledge and unde rstanding of environmental issues necessarily leads to mo re positive attitudes towards or behaviour in the natural environment. (Barraza, 2002, p.171) Premise of Environmental Education The above quotes present two opposing ideas about the role of knowledge in forming environmental attitudes. Environmen tal educators, like t hose involved with the U.S. Environmental Protection Agency (2008) often operate under the assumption that increased environmental knowledge can lead to concern for the earth, as well as proenvironmental behaviors. However, some researchers have cautioned that education is not a panacea for inspiring citizens toward either earth-friendly attitudes or behaviors (e.g., Barraza, 2002). Despite this caution, envi ronmental education s eems to be the most popular way to bridge the gap between indiffe rent environmental attitudes and positive ones (Bailey & Watson, 1998; U.S. Environm ental Protection Agency, 2008; Project Learning Tree, 2004). Therefore, at the height of the cu rrent 'green movement' (Orton, 2006) those concerned about the state of the environment are turning to education to help integrate modern living and earth-friendliness. However, to determine if the end result of
2 environmental education is likely to be pro-en vironmental behaviors, it is first necessary to qualify the relationship between knowledge and attitude. This is especially important because these two issues have most often been addressed in separate lines of research. In those projects that have directly compared attitude and knowledge, some research has shown conflicting evidence on the significance of the relationship. For example, in one study Barraza (2002) showed that a higher level of knowledge (through environmental curriculum) may affect children's concern for the environment; in other studies researchers have shown that it does not (B arraza, 1999; Evans, Juen et al., 2007). The knowledge-attitude-behavior association is especially prevalent in literature on childrens education (e.g., Hartsell, 2006; Wals, 1994). Many educators consider young children the ideal audience for pro-enviro nmental training, as their life outlooks are not yet solidified by family beliefs a nd popular media influences (Hess & Torney, 1967, as cited in Pomerantz, 1986; Atkin & Gantz, 1978). Younger children, especially those under the age of 10, are also thought to have less developed understandings of the relationship between humans and the natural world (Palmer & Suggate, 2004). Proenvironmentalists and educators hoping to d ecrease human impact on the earth point to early and middle childhood education as a way to reach this goal. To this effect, Project Learning Tree, a popular environmental education group, aims: To develop students' awareness, appreciation, skills, and commitment to address environmen tal issues [, and to] provide a framework for students to apply scientific processes and higher order thinking skills to resolve environmental problems. (Project Learning Tree, 2004)
3 Although the current study will not address the willingness and ability to resolve environmental problems, it wi ll attempt to address the co re relationship of knowledge and attitude in children. First, these two va riables will be addressed as they have been combined in past research, and then as they have been researched individually. Exploring how these variables have been studied separately will help emphasize measurement issues that may need to be considered for the current study. A differe nt type of attitude construct, scope of justice, will then be discussed and will help structure the current study. Also, as the current rese arch is concerned only with the student-age population, the discussion of research carried out with adults will be limited. Studies Combining Knowledge and Attitudes Arcury (1990) showed that knowledge and attitude directly correlated in an adult sample of Kentucky residents. Three meas ures of knowledge (general environmental knowledge, knowledge about energy, and Kent ucky-specific environmental knowledge), as well as a full version and subscales of the New Environmental Paradigm (Dunlap & Van Liere, 1978) were used in random-dig it dialing phone questionn aires. Participants received one of twelve possi ble combinations of a knowledge scale and an NEP scale. The results of all twelve obs erved combinations showed significant correlations between knowledge and attitudes. This presents clear support for the hypothesis that a relationship does exist, even if just in this spec ific subset of the general population. Similarly, Bradley, Waliczek, and Zaji cek (1999) conducted a study with high schoolers that showed a significant positive correlation between knowledge and attitude. This study incorporated an environmental e ducation program, thereby directly attempting to manipulate the variable of knowledge. This program was a 10-day course that covered a wide variety of environmental science t opics. Knowledge scores were significantly
4 correlated with attitude scores both before the course and after it. Post-course testing showed that knowledge scores had increased by 22%, while attitude scores had increased by only 2%. This shows that the program successfully manipulated knowledge level, though the resulting change in attitudes was not nearly so dramatic. However, the difference in preand post-course attitude scores was significant and the new knowledge and attitude scores also correlated with each other. This suggests that environmental education does have the potential to make attitudes toward nature statistically more positive, though it is unclear from the study whether these are practical differences. Studies between knowledge and attitude have not been as direct with younger children. Some studies have not used genera l knowledge and attitude measurements, or have incorporated comparisons of these variables into larger educational programs. In an example of a topic-specific measurement study, Harvey (1989) investigated whether experiential knowledge of plants increased pro-environmental attitudes. The researcher used a measurement of past experience w ith plants, along with the Pastoralism and Human Dominance attitude scales created by Bunting and Cousins (1985). Experiential knowledge was defined as information gained by direct contact with plants in any context. In Bunting and Cousins (1985) orig inal research, Pastoralism was defined as a positive response to nature, and Human Dominance (originally called Environmental Adaptation by Bunting and Cousins) was defi ned as a lack of concern with humans interference in the natural world. Harvey (1989) conducted this research with 8-11 year olds. The researcher created categories for how plants might be used (for example, as food, as decoration, or as part of a pretend play script ), and asked children if they had ever experienced plants in these ways. Participants then rated how much they had enjoyed those experiences. Then,
5 children completed the Pastoralism and Huma n Dominance sections of the Children's Environmental Research Invent ory (Bunting & Cousins, 1985). Harvey (1989) found that past experiences with vegetation in terms of variety, frequency, and enjoyment were all positively correlated with the measure of Pastoralism, and negatively correlated with Human Dominance. This shows that an experiential understanding of one aspect of nature, vegetation, may be related to more positive attitudes about the environment in general. In a similar study, LindemannMatthies (2005) reviewed Nature on the Way to School a Swiss program for 8-16 year olds intended to increase awareness of the importance of biodiversity. The program's goal was to increase knowle dge of plants and animals in spaces that are not typically c onsidered wild, such as neighborhoods. It was hoped that the end result would be a greater appreciation for these organisms. One of the main activities of this program was to have students place picture frames around favorite plants (or animals, if they would stay still) that were found on the way to school. After framing something, the child then explained why it was interesting to passers-by. Along with the framing activity, in-classroom lessons also taught children about native Swiss plants and animals. Though this was a na tionally implemented curriculum, for the purposes of comparison researchers also obtained control classrooms that did not participate in such activities. Participating students showed a much higher increase in their knowledge of the local environment than student s in the control groups. Al so, preand post-curriculum comparisons of participant appreciation of local plants and animals increased as students ability to identify these organisms increased. Th e increases in apprecia tion were greatest for plants growing in the wild versus la ndscaped plants. These results indicate that
6 increasing knowledge about plan ts and animals increases attitudes towards just those organisms. This is in contrast to Harvey s (1989) study, which showed that increased topic-specific knowledge could positively corr elate with increased general environmental attitude. These studies do provide evidence fo r the knowledge-attitude relationship. However, Arcurys (1990) research was with an adult sample, and will not be directly applicable to the current study. On the other hand, Harvey (1989) and LindemannMatthies (2005) both re searched this phenomenon in children, but concentrated on narrower aspects of knowledge (just plants a nd animals) and attitude (in the case of Lindemann-Matthies). Childrens Environmental Knowledge The definition of knowledge and how it is gauged is very important to understanding the knowledge-attitude relations hip. There are many wa ys to measure the knowledge of a person, but researchers work ing with younger populations can encounter issues with traditional knowledge measuremen ts. For example, it would be practical to administer a written, multiple-choice knowledge test to a group of high school students, or even a group of third graders. The language would have to be adjusted, but the format could remain the same. However, administer ing such a test to students younger than 8 years old presents problems: not only have many younger children not yet mastered the multiple-choice format, some of them may not yet read well. In this overview of how knowledge has been measured, the discussion will start with older students and move subsequently to younger children. Research on each age group provides interesting information about how researchers have tested childrens knowledge, and how children understand their environment and apply that und erstanding to a broad range of situations.
7 After reviewing the conventiona l ways of measuring childr ens knowledge, a different approach to understanding childrens comprehension of the natural world will then be discussed. In an example of research with olde r children, Boyes and St anisstreet (1993) tested 11-16 year olds on their knowledge of a specific topic: greenhouse gases. This topic was chosen for its relevance to curre nt questions of global warming. A multiplechoice test was used, and it wa s not administered in conjunc tion with a special teaching unit. Testing showed that questions concer ning chemicals that worsened the greenhouse effect were clearly difficult for many students. The results also showed that some correct answers might have been products of faulty reasoning. For example, a third of the youngest students identified the effects of ozone correctly. Similarly, younger students seemed to understand that re cycling paper would reduce th e need to cut down trees, which would in turn allow more trees to sequester carbon dioxide. However, the number of students who answered the ozone or carbon sequestering questions correctly decreased in older groups. Researchers suggested that younger students either ar rived at the correct answers by chance or through incorrect logi c. There was no way to test the logic hypothesis with the data that were collected, however. The researchers also found that younge r students thought tang ible sources of waste (such as not recycling paper) would di rectly make the greenhous e effect larger, and that any environmentally-friendly behavior would likely make it smaller. The researchers deemed this problem a "reductionist appr oach" (Boyes & Stanisst reet, 1993, p. 551) in understanding, meaning that students had a di fficult time understanding intangible causes for large-scale issues.
8 This research highlighted some of the frustrations of measuring factual knowledge, not least of which was the inabilit y to have children clarify their answers because of the multiple-choice format. This study also relied on very technical terminology, which was necessary because of the subject matter. There were many poor scores on this test across age groups. Since the only data gathered were multiple-choice answers and not open-ended explanations, it is unclear how these lo w scores should be interpreted: were children merely frustrated by vocabulary or the ta sk, or did they truly not understand cause and effect systems? Ashworth, Boyes, Paton, and Stanisst reet (1995) attempted to break from multiple-choice to measure 13-14 year old's understandings of animal conservation. From pilot data, the researchers de veloped a questionnaire that ga ve children six choices of animals and asked them to rate which animal most deserved to be conserved and why, which deserved to be conserved second most and why, and so on. This way, the measurement had aspects of multiple-choice and free-response. The researchers then coded the answers to th e free-response portion using 20 key reason themes, including danger of extinction, importance to huma ns, and general morality. From these explanations, the researchers concluded that children confuse extinc tion and death of an individual animal, echoing other researchers' findings (i.e., Boyes & Stanisstreet, 1993) that children have a poor unders tanding of global cause and effect chains and tend to think in concrete and small-sc ale ways. Also, the researcher s determined that children often ignore non-mammals in their lists of which animals should be saved. Compared to the previous study, this m easurement seems to allow students more freedom in their answers. It also lets researchers better understand childrens thought processes. This is the most important comp arison between Boyes and Stanisstreet (1993)
9 and this study: the methods represent two different approaches to obtaining knowledge data. It is difficult to compare the success of participants across the two studies because the content was very different. It is possibl e to conclude, though, that Ashworth et al. (1995) were able to gain more insight in to participants logic using an open-ended question component. Other research on topic-specific knowle dge has included a study of rainforests with 11-16 year olds (Greaves, Stanisst reet, Bayes, & Williams, 1993), energy and pollution with 14-18 year olds (Gambro & Switzky, 1999), and the ozone layer with 1112 year olds (Christidou, Koulaidis, & Chri stidis, 1997). Many of the measurements used in these studies have resulted in low mean scores or evidence of faulty logic. This has led researchers to conclude that students have an overall low level of environmental knowledge. Yet, it is difficult to pinpoint why participants had difficulty on these measures. Each research group has had their own interpretation of why participants did not perform well. Ashworth et al. (1995) c oncluded that the age of children studied (1314 years old) did not have the same definition of an animal as scien tists would, and so did not provide examples of animals and reasons for conserving them that fit researchers conservation criteria. Specifically, the pa rticipants thought of only large, usually mammalian animals when asked to provide examples. The researchers suggested that these large and charismatic animals are often the subject of conservationists campaigns, and so exposure effects might have influen ced children to respond in the way they did. Boyes and Stanisstreet (1993) made an ev en bigger generalization about the cause of childrens low performance: they have poor logic and reasoning skills, and cannot conceptually separate global and local problems. However, the researchers did not elaborate whether this was an issue of an inadequate school curriculum, or if the issue
10 was developmental. Some topics, such as the greenhouse effect a ddressed by Boyes and Stanisstreet (1993), have not ye t been addressed in formal e ducational settings, and it is unlikely that children would know about these topics on their own. Or, sporadic informal environmental education programs may have addressed such complex subjects only superficially, which might furthe r confuse the issues in stud ents minds. It is the many possible explanations for poor performance that make creating broadly applicable measurements of environmental knowledge so complicated. The reasons for the low scores coul d be genuinely low subject knowledge, difficult testing procedures, or another unknow n variable. Whatever the reason for low scores, though, the above resear ch highlights a lack of knowle dge in participants. These types of findings do not bode well for researcher s trying to grasp what children actually do understand, versus only what they dont. In order to construct a picture of childrens knowledge from the bottom up, it seems necessary to take a step back to more general measurements, and to expl ore different methodologies. To start with an exam ple of a more general measurement, Leeming, Dwyer, and Bracken (1995) created a scale with the inte ntion of using it to compare environmental knowledge across ages. Though the measure borrowed heavily from a previous environmental knowledge study with adults, age-inappropriate questions from that study were purportedly edited out and new questi ons were created for this Children's Environmental Attitude and K nowledge Scale (CHEAKS). Leeming et al. (1995) incl uded 1st through 7th graders in this research, all of whom took the test twice so that consistency of student response could be measured. The researchers formed two participant groups for the purpose of age analysis: a younger group was made up of 1st through 3rd grader s, and an older group of 4th through 7th
11 graders. Focusing only on the knowledge subscale it is possible to s ee that older children outperformed younger children on both test da tes. Possible scores ranged from 0-180, and older children had a higher mean score on both the first and second administrations ( M = 76.45; M = 81.90) than younger children ( M = 51.34; M = 55.78). Comparing the mean scores of both groups to the potential maximum score of the measurement is discouraging, though. Although th e difference between the two groups was pronounced, neither age group seemed to have a strong grasp on environmental concepts according to their scores. This echoes other researcher 's findings discussed above, that students generally lack knowledge about the environment. It is hard to be satisfied thinking that students have such low levels of environmental understanding. It is possible that while the CHEAKS and other general measures are consistent and convenient meas ures of environmental knowledge, the tests may not give students enough credit for what they do know. Again, this addresses the issue of constructing a picture of childrens knowledge from the bottom up. That is, on measurements such as the CHEA KS children are compelled to choose an answer when they may not understand either the questi on or the possible res ponses. Looking at the CHEAKS questions, some seem written so that a 12-year old could not answer confidently, much less a 6-year old, even t hough both those ages were within the range tested. The questions address sp ecific facts that children might not reasonably learn in the normal course of their schoolwork. For example, a question like "Which of the following groups is most interested in environm ent issues?" with the answer choices the Boy Scouts, the Audubon Society, and the Kiwanis Club immediately put a student at a disadvantage if he or she had never heard of the answer choices. An example of a more factual question was: "Where does most of the garbage go after it is dumped from the
12 garbage trucks?" where landfill, aquifer, the ocean, it is recycled, to make fertilizer were the answer choices. Here, too, the answer choices provide voc abulary that might be too difficult for the lower age range of the sa mple. The researchers did acknowledge the possibility of a floor effect, saying that the difficulty level might have held students back and reflected unfairly on what knowledge they had. Thus, the goal of creating a standardized measurement was appare ntly not met for younger children. Instead, methods for assessing knowledge le vels in that age range might benefit from research done in preschool settings. Th is research offers some alternatives to multiple-choice and other standardized testing procedures. Especially salient in the preschool classroom is the problem of attent ion and engagement, as well as the issue of pre-literacy. For this reason, illustrations, st ory telling and games have become typical tools to use with young children. In an example of using such tools, Palmer (1993; 1994) used pictures and openended questions to let children describe and define environmental concepts in their own words. In this study conducted with 4and 5-year olds, the resear cher focused on childelicited definitions. For example, instead of asking a child what types of things he or she might find in a forest, the researcher showed a picture of a forest and let the child define the place in his or her own words. Then, th e researcher asked what types of things a person might see in a forest, and next what t ypes of things might happen in a forest. With this progressing model, the child was able to build on knowledge that had already been demonstrated. The topics of rainforests, deforestation and enda ngered species, global warming, and waste were addressed in this research, albeit at a very basic level. After the child had described an illu stration, the research er asked about the consequences of human interference in reference to the topic (for example, possible
13 human effects on a forest). From the childs responses, the researcher was then able to construct a concept map to understand how th e child thought about different topics. The dependent measure was a score of elaborati on, defined as the numbe r of steps a child could conceptually move beyond the original topic. For instance, a child who could elaborate on the concept of fo rest to say that trees make up forests, that a forest can provide habitats for other animals, and that if a forest is cut down animals and plants lose their homes, would be given an elaboration score of 3. Elaboration scores in this study ranged from 0-3 in the case of deforestation and endangered species, 0-4 for rainforests, and 0-5 for both global warming and waste. An average of 70% of all students scored a 2 or above across the four topics, and an average of 17% scored in the top two elaboration scores for each question. One topic, waste, had no elaboration scores of 0, and rainforests had only 5% of participants score 0. However, on the topic of deforestation and endangere d species, 47% scored 0, followed by 30% with a score of 1. The scores on these two questions broke from the trend of the other questions, though this was not surprising to th e researcher. Palmer (1994) concluded that the difference in the level of abstraction accounted for some of the difference in scores between the highand low-scoring topics. Fo r example, pieces of waste and rainforests are concrete entities that child ren can create a mental picture of, whereas it is harder to imagine a general picture of an endangered species. Also, waste (in the form of litter) and rainforests are sometimes popular topics in younger classrooms. In contrast, the cause and effect mechanisms of deforestation and e ndangered species might be less likely to be discussed in the average classr oom of 4and 5-year olds. Palmers (1993; 1994) method provides an example of how to create and score a bottom-up knowledge measurement. The method lowered the starting level of knowledge
14 needed by the child to the point at which he or she could be successful. It also let the child build upon existing knowledge easily. If a child could not produce a satisfactory definition of a forest, the child might still be able to give some information about what lives in a forest and why a forest might be important. This methodology might have allowed younger children a chance to show their knowledge in a way that other standardized measures have not made possibl e. The next study that will be reviewed takes a similar approach to testing knowledge by also usi ng illustrations and open-ended questions. Cohen and Horm-Wingerd (1993) were also interested in testing the environmental knowledge of young children in an accessible and fair way. They tested 35 year olds, which is an age range slightly lower than in the research just discussed. However, both projects still aimed to test preliterate children who were unused to formal test settings. Based on writings by Hart and Chawla (1981), Cohen and Horm-Wingerd (1993) proposed that children may not yet have a formal understanding of aspects of the environment and how they relate, because of their lack of both experi ence and traditional schooling. Therefore, the researchers designed three tasks to test the emergence of ecological awareness, versus the explicit environmental knowledge tested for in older students. The tasks used were a picture disc rimination task, a picture arrangement task, and a picture comprehension task, though the picture arrangement task was discarded because of poor reliability. In the discrimination task, children were shown two pictures (for example, a neglected and littered building versus a clean, well-kept one ) and asked, "Which one is nicer?" Responses (gestures to one of the tw o pictures) were scor ed as incorrect or
15 correct. This seemed to be the easiest of the tasks. The comprehension task required participants to explain what was wrong with a picture. An example of a comprehension task picture might be a car with someone throwing litter out of the window. Responses were also scored as correct or incorrect, ac cording to the judgment of the researcher. Though both tasks showed a large range of scores, some patterns emerged from the comprehension task in particular. Thr ee year olds almost never answered the comprehension questions correctly, whereas 4and 5-year olds did often. Though these results were initially thought to support a di fference in comprehension according to age, the researchers cited lower verb al abilities in the 3-year ol ds as a possible alternative explanation for the difference in scores. Furt hermore, 4and 5-year olds answered only 30% of the questions correctly. Cohen and Horm-Wingerds (1993) resear ch did not support a clear timeframe of emergent environmental thought, but the resear chers' methods are st ill worth considering for future studies with young children (potentially age 4 or older). The method made use of pictures to focus participants attention on the subject. It also interchanged questions that required the child to merely indicate a correct response with t hose that required the child to expand on an answer. This is another step toward accessing multiple levels of children's understanding, just as Palmer (1993; 1994) accomplished in her work. These researchers realized that the language used in work with older child ren is inappropriate for pre-literate children or t hose in the beginning stages of literacy, especially when the children are unused to standa rdized testing formats. Perception as a Form of Knowledge. Aside from using non-standard methodologies, there is yet another possible way for researchers to better understand childrens knowledge. As shown previously, pa st research has generalized that students
16 do not know many environmental science facts. Some research has also generalized that young children have low abilities to think in abstract terms and move beyond concrete conceptualizations of the environment. Another branch of research s uggests that the adult perspective of environmental knowledge might also be a concern in improving measurements. That is, some research shows that the perceptual and intuitive knowledge that children hold about their environmen t may be a better indicator of their understanding than factual knowledge. Moreove r, childrens percep tual knowledge of their environment may be very different than adults perceptual knowledge. If it is the case that adults and childr ens environmental perceptions do diverge, then it is possible that te sting childrens environmental knowledge in a way that makes sense to adult educators is not the most valid approach. Of course, it may still be important to measure factual understanding of the environment, as the goal of most environmental science curricula remains to increase this t ype of knowledge. However, if the goal is to understand how children might truly think about their environment, perceptual knowledge may be more useful as it is based on a local awareness of specific natural areas and experiences in them. Research that addresses perceptual knowledge has studied what makes up childrens environmental definitions. To test for a difference between younger and older children in how they defined the environment, Loughland, Reid, and Petocz (2002) conducted a study with 9-17 year old students based on how students interpret the natural world. Participants were asked to complete the sentence "I think the word environment means..." Responses were categorized according to six conceptions: three of these conceptions were object-focused and the ot her three were relati on-focused. Objectfocused conceptions were defined as characteri zations of the environment as a place with
17 things or beings in it. Relation-focused conceptions were defined as characterizations of the environment as doing something for or with other entities. The researchers found that all six conceptualizations were present acr oss age groups, after controlling for language sophistication. However, the majority of younger children ha d an object-focus and not a relation one. For example, younger children do not seem to understand that humans have an impact on the environment, because th ey saw the environment and humans as mutually exclusive entities only occupying the same space. In contrast, older children seem to have developed the idea that hu man actions have implications for anything located in the same space. Again, this proj ect used an open-ended question to measure understanding, but the question was broad e nough that the partic ipants could respond with any number of definitions or affec tive reactions. Therefore, it allowed the researchers a way to compare many possibl e definitions and understandings of the environment, instead of imposing possible answer choices. Looking at environmental definitions fr om a more theoretical standpoint, Heft (1988) argued that children's perceptions are mo re direct than adults because they are not mediated by vocabulary and concepts learned over a longer lifetime. Rather, children perceive their environment directly thr ough the affordances given by their natural surroundings. Heft defined affordances as the inherent uses of objects in the environment based on an individuals per ception. For example, a tree can offer itself as a climb-onable feature to one child, and a swing-on-ab le feature to anothe r. According to the individual nature of affordances, not all children will perceive environmental features identically. Nevertheless, there may be systemat ic differences in how children and adults perceive the environment. Though adults are not unable to recognize the same affordances, according to Hefts model they may be less likely to experience
18 environmental features as children do. For example, a tree may not immediately represent a potential action for an adult, but rather a pa rt of a schema held for a certain landscape. Also, adults more static, schema-based understanding might not be accessible to children, especially when they are young. This is important to keep in mind when asking children about their environmental knowledge, because it shows that their knowledge might be classified as more of an awaren ess of nature, and might not fit the same organizational structure that adults use to think about the environment. Hyun (2005) further studied potential differences between how adults and children perceive the natural world. After collecting individual interviews from 32 parents and children engaged in outdoor activities, Hyun code d interview transcripts for presence or absence of four categories: perception creates thought vs. perception obeys thought, and biophilic reasoning vs. biophobic reasoning The category of perception creates thought is similar to Heft's idea of afford ances, such that children look at the world without many pre-conceptions. The category perception obeys thought is comparable to a schematic approach to pe rceiving an environment. Hyun found that adults and children differed on the language th ey used to describe their surroundings and themselves in the environment. For example, he noted 23 incidences of adults' ideas fitting into the category perception obeys thought whereas there were only 2 incidences of children fitting into this category. On the other hand, 30 in cidences of childrens ideas were observed in the category of perception creates thought and only 9 adult incidences were observed in this category. This research further supports the idea that adults and children do not organize or recall information about the environment in the same ways. Alerby (2000) also considered the possibility that children and adults might not share a common vocabulary about understand ing the environment. Working under this
19 assumption, she eliminated the potential language barrier by incorporat ing drawings into a study with 7-16 year olds. She used the drawings as the dependent measure of children's definitions. She proposed that draw ings were a valid way to understand what children are thinking and feeli ng even if they do not have the technical or emotional vocabulary to talk about those thoughts. Th e prompt for the drawings was a question similar to that used by Loughland et al. (2002): What do you think of when you hear the word environment? After the drawings were completed, Al erby (2000) created themes based on what she found in the children's work. These themes were: the good world the bad world the relationship between the good and bad world and protection and conservation actions The drawings were analyzed for gender a nd age differences on these themes. She found that girls generally showed more emphasis on the good world (i.e., a world dominated by clean, beautiful scenes) than boys. The bad world (a world in which destruction was the norm) was a theme across age groups. Within these categories, younger age groups tended to focus on objects and situations in their drawings that were based on their narrower worldview, whereas ol der students had a broader se nse of global interaction. Age differences found in this study support the presence of a different way of thinking about the environment in younger children. From these drawings, Alerby (2000) concl uded that childrens understandings of the natural world are often intertwined with their affective and personal experiences with that world. Given an open-ended question, child ren do not often give strictly fact-based definitions. This study offers a way to access how experiential, affective, and fact-based types of knowledge are interconnected.
20 In fact, literature on perceptual and expe riential knowledge is often closely related to the attitude component of the knowledge -attitude link. For example, Loughland et al. (2002) described their work as both rela ted to childrens awareness about the environment and also their feelings toward s it. Knowledge and attitude seem to be inextricably linked in other research as well, so that it is sometimes hard to tell if knowledge can be defined without including the attitude aspect. Cohen and HormWingerd (1993), who aimed to understand ch ildrens emerging environmental thought, also provided examples of this type of in ter-relationship. In the picture comprehension task used in their research, it is difficult to discriminate between the role of factual knowledge and the role of attitude in deciding what counts as a correct decision. This is because the task involved asking children what was wrong about someone throwing trash out of a window without really defining the basis for wrong. For example, a correct response could be simply that litter belongs in the trashcan. A correct response may also have been that litter is ugly, or that it w ill hurt the baby animals living in that area (Cohen & Horm-Wingerd, 1993). It is from this gray area between straightforward information and affective reactions that the discussion will move definitively to the next part of the original equation: attitude. Childrens Attitudes ab out the Environment When creating an attitude test, it is impor tant to remember the many factors that can influence the creation of a stable attitude. These can include past experience (Harvey, 1989), ones culture and community (Kahn, 1997), and potentially, knowledge about a particular topic (Kellert, 1985, as cited in Chawla, 1988). Environmental attitude measurements may address only one topic (for example, only animals), or they may be broad and incorporate reactions to many as pects of the environment. For now, the
21 discussion will focus on measurements usi ng specific topics. Then, it will move to examples of general measurements of environmental attitude. Topic-Specific Measurements. Using the 1989 Exxon-Valdez oil spill to frame the discussion, Kahn (1997) examined how childre n think about human intervention in the environment. In particular, the researcher was interested in the diffe rent levels of moral reasoning displayed by children in response to th is disaster. The spill killed marine life, destroyed coasts and affected fishermen and other inhabitants of Prince William Sound. It was unclear from past research whether children would identify all of these affected parties and the extent to which they were harmed. The participants were 2nd, 5th and 8th gr aders in Texas schools. The researcher individually interviewed participants with open-ended questions. The interview included the following types of questions about the event: assessment of knowledge, cause and effect assessment, value judgments, endangered species value judgments, and human versus animal causal judgments. Participants were also asked to explain all of their answers. The interview concluded with a disc ussion on what it means to live in harmony with nature. Results showed that practically all stude nts understood that th e spill was harmful to marine life and the shoreline ecosystem, and that this harm constituted a moral violation. A large proportion of children also understood that the spill was economically and emotionally harmful to the people livi ng along the shoreline, as well as the ExxonValdez oil company. It mattered more to part icipants that the spill hurt marine life (especially endangered species), than that it hurt the shoreline ecosystem. Students also drew a distinction between humans hurting animals and animals hurting other animals, explaining that the former was much less acceptable.
22 When asked to explain why the event had been called a disaster, many students described the damage in anthropocentric term s, as a loss of resources valued by humans. For example, many children noted that an impor tant reason to save animals was so that humans could enjoy observing them in the future. Even older students used many anthropocentric explanations though they also showed a simultaneous increase in biocentric reasoning (that is, they considered nature to have intrinsic value as well). Kahn (1997) concluded that children used an overall high level of moral reasoning when considering the environment. However, childrens reasons behind the high morality may not have been the same acr oss ages. Older children seemed better able to understand the intrinsic value of nature, as well as the va lue of nature to humans. The researcher could support these conclusions based on the detailed explanations elicited during this study. This highlights that it is important for researchers to ask for explanations to responses, or they may never kn ow the variations in l ogic used to come to similar answers. Future studies could addr ess whether the underlying logic difference matters, but for now it may remain an importa nt factor in creating a thorough attitude measurement. General Measurements. Some general measurements of childrens environmental attitudes have been modeled after those used with adults. For example, Manoli, Johnson, and Dunlap (2007) revised the New Envir onmental Paradigm (NEP) (Dunlap & Van Liere, 1978) for use with child ren. The original scale, whic h had been used since the 1970s, consisted of 15 questions that participan ts responded to on a Likert scale. For the children's scale, the researchers simplified th e language of the questions and chose only 10 of the original 15 to use. The measuremen t was administered to 4-6th grade students, and because the age range was so small no comparisons by age were made.
23 The questions asked students to agree or disagree on statements concerning the significance of human impact on the environment. An example statement was: "Nature is strong enough to handle the bad effects of our modern lifestyle." The researchers found that students showed a central te ndency bias for the rating of not sure on four of the statements, and a pro-environmental bias on six of the statements. No question had a strong anti-environmental bias. A total score was based on res ponses to all questions. The scale of possible scores ranged from an an thropocentric worldview to a biocentric worldview. These scores showed that children might have a tendency to relate to and sympathize with the natural world. However, the statements used in this measurement were very broad, and not necessa rily salient to elementary sc hool children. Almost half of the questions showed a central tendency bi as, which could indicat e that students were uncertain what the statements meant, or uncertain about the implications of taking stronger stances. Therefore, the measurem ent might not induce children to think thoroughly about their feelings toward the environmen t and how it is treated. In another example of ge neral attitude measurement, Bunting and Cousins (1985) administered their Childrens Environmental Inventory to 4th-10th graders. The researchers sought to measure attitudes towards all types of environments, not just natural ones. Therefore, three of these outlooks ar e the most relevant to this discussion: Pastoralism, Urbanism, and Environmental Adaptation. Pastoralism was defined as proenvironmental behaviors and nature-oriented interests, Urbanism as a love for things built and structured by humans, and Environmenta l Adaptation as a lack of awareness and concern about the impacts of humans on the natural world. These dispositions were not mutually exclusive, as it was possible to score highly on more than one disposition scale (though the researchers did not correlate scor es on the scales to each other, so
24 relationships between dispositions were not explored). The range of scores for each scale was 20-100, with 60 indicating no strong dis position, and scores above 60 indicating affinity for that disposition. When the researchers administered the I nventory, the disposition with the highest mean score was Pastoralism ( M = 86.06). Urbanism and Environmental Adaptation had the lowest mean scores of all eight di spositions (M = 54.11; M = 50.00, respectively). These scores indicate that th e children had typically pro-envi ronmental attitudes. No age effects were found for the Pastor alism scale, but analysis of gender showed that females generally scored higher than males on the measure. Interestingly, Environmental Adaptation scores decreased significantly w ith age, while younger children showed an optimistic lack of concern about human intervention in the natural environment (Bunting & Cousins, 1985, p. 732). The resear chers hypothesized that Environmental Adaptation scores decreased (as age increased ) because of a simultaneous increase in level of education. Arcury (1990) observed this trend in adul ts: those with higher levels of formal education had significantly highe r attitudes towards the environment. Lastly, Evans, Brauchle et al. (2007) described a way to measure attitudes and self-reported behaviors toward s the environment. The researchers used games derived from the NEP (Dunlap, Van Liere, Mertig, & Jones, 2000) to measure attitudes. Two of the games assessed how participants felt about humans role in the environment. The third game asked participants to state how wo rried they were about environmental issues. This yielded a measure of attitude on an 11 point scale, on which the participants concentrated in the uppe r half of the scale ( M = 7.78). These studies offer many possible me thodological approaches. Kahns (1997) model is an excellent one to use if examining how childre n think of the nature-human
25 relationship in a specific s ituation. Manoli et al. (2007) m easured feelings about this relationship on a more general level, and Evans, Brauchle et al. (2007) converted this measurement into games. Bunting and Cousin s (1985) defined environmental orientations and measured which were most prevalent. While these methods are good ways to generalize the attitudes of indi viduals, there is another way to explore attitudes. This way involves examining the circumstances under whic h an individual will maintain certain attitudes or deviate from them. For example, some items of the NEP (Manoli et al., 2007) suggest that conflict over resources plays a part in forming attitudes. Would participants score the same on the NEP measurement if they were induced to think of themselves in direct conflict with the environment, instead of just humanity in general? On this note, the next attitude measurement to be disc ussed will address potential conditions that matter in attitude formation. Scope of Justice A persons scope of justice is characterized by a willingness to allot resources to and make sacrifices for entities beyond the se lf. It can be summarized as the level of fairness others deserve. It can be applie d to humans (Opotow, republished 2006), as well as animals or ecosystems (Opotow, 1993; Cl ayton & Opotow, 2003). On an individual level, the scope of justice may fluctuate when applied to different enti ties. For example, it is possible to think that animals are deservi ng of resources and fair ness, but that plants are not. The construct of scope of justice can be explained through Opotow's (2006) study on inclusion of non-similar enti ties in considerations of fa irness. The project involved asking city dog-walkers why they did or did not use a dog scoop while walking their dog. Use or non-use constituted a measure of concern for the common environment, because
26 the walkers were using public paths and park s. After the researcher observed whether the walker used the scoop or not, she asked th at person questions about dog walking, the neighborhood, and scooping laws. She found that those walkers who obeyed dog scoop laws felt a closer tie to their community, felt similar to the other people li ving in the community, and thought that dog scooping helped the community by keeping it pr esentable. In essence, these people had a greater attentiveness to th e area than those who did not obey the scooping laws. Those who did not obey the laws did not feel like a part of the community, saw themselves as dissimilar to others liv ing in the neighborhood, and viewed dog scooping as an unnecessary annoyance. In this way, Opotow described some conditions necessary for including another entit y, either human or environmental, in one's scope of justice. These conditions were perceived similarity, utility of the other to oneself, and conflict over resources between oneself and the other. Opotow (1993) created a method that coul d manipulate these three variables for attitudes towards animals. She conducted this research with high schoolers. First, half of the participants received information about a beetle that made it seem very similar to humans in its needs and habits. The other ha lf of the participants received information that suggested beetles and humans were diss imilar. Next, she presented half of each of these groups with different information concerning how useful this beetle could be to humans (either very useful or harmful). Th en, she manipulated the within-participants variable of conflict severity, so that particip ants heard one story de picting a situation in which human need was very high for a space that was also a beetle habitat. Another story implied that humans did not really need the beetle habitat and coul d use another space. After these manipulations, she as ked participants how likely they would be to give up the
27 space in question for the beetle using a 7 point Likert scale, with 1 indicating a narrower scope of justice and 7 indicating a mo re inclusive scope of justice. In her analysis, Opotow (1993) also e xplored responses to manipulation-check questions about how many resources the partic ipant would offer the beetle, how much the participant thought the beetle deserved fair treatment, and how many sacrifices the participant would make for bee tle. She expected that those w ho felt a higher similarity to the beetle and those who felt that the beetle was useful to humans would show a wider scope of justice. Also, she predicted that a higher degree of conflict between the beetle and humans would produce a narr ower scope of justice. Her predictions were correct except for th e condition of similarity, which did not produce any significant differences between groups. Conflict le vel and utility did produce significantly different outcomes Also, manipulating the level of conflict seemed to make participants think th rough decision-making in greater depth: one respondent indicated that when both the beetle and humans needed the same resources, the decision of whether to allocate resources to the beetle becam e more difficult. Ther e were no significant gender effects, once reactions towards insect s were controlled for. The results clearly support that two variable s play into a person's scope of justice: level of conflict and utility. The scope of justice approach may gain a fuller picture of attitude as it unfolds across different situations. This method can be considered along a continuum of attitude measurements: at one end of the continuum are item-agreement measurements like the NEP, which may not require participants to think through their responses in depth. At the other end of the continuum are observational studies of partic ipants acting on attitudes in everyday interactions with the environment. Somewhere between these extremes is the
28 type of method created by Opotow (1993), which manipulates relevant theoretical situations but does not measure those manipulat ions in real life. Even though it is still theoretical, this type of measurement c ould better serve understanding the connection between knowledge and attitude by providing a more thorough measurement of attitude. The Current Study Understanding the pattern of children s environmental attitudes can help educators understand the type of relationship they might have with the environment. Understanding how attitudes are formed, howeve r, is the first step in understanding the total relationship. Level of environmental knowledge is one of the most popular hypotheses about what affects attitudes. As was just discussed, level of conflict and utility between humans and aspects of nature might also affect attitude formation. This study will look at how knowledge and attitude interact. It will also examine whether Opotows (1993) findings about utility and conflict are present in a sample of young children. Looking at these va riables might help build a foundation for studying young childrens everyday relationships with th e environment in future studies. There are three important aspects of th e current study which will address issues observed in past research. Primarily, particip ants in past research have been in upper elementary school grades or beyond, and so here the focus will be on younger elementary children. Looking at the research th at has been done with younger children, knowledge measurements in these studies have often used a format (e.g., multiple choice) or content that may be too difficult for young children. It was hoped that a new measurement, created specifically for this study, would be suitable for young children who do not have extensive reading or testing experience. This measurement was designed to be general enough to gain a broad understanding of envi ronmental knowledge. The questions were
29 influenced by the type of elaboration mode l seen in Palmer's (1994) work, which moves from simple definitions to more abstract concepts. As Palmers work was done with preschoolers, her approach was then app lied to more age-appropriate questions. The third issue involves usi ng an attitude measurement that will give a more detailed picture of children s relationships with the envi ronment. This study used a measurement of scope of justice. This cons truct has been employed previously only with adults and older students (Opotow, 1993). Th e level of conflict between humans and nature was manipulated to examine whether ch ildren would be less pro-environmental if offering help to the environment meant igno ring their own needs and desires. A utility priming statement was used to a similar effect, to examine if thinking that the environment is beautiful and beneficial to hum ans would affect childrens desire to care for it. The bodies of literature lead to explori ng an essential overarching question: Does a childs greater understanding of nature correlate with increased concern for it? This question is important because the logic of many environmen tal education programs rests on the direct relationship between knowledge, attitude, and pro-environmental behaviors. Based on past research with older children a nd adults (e.g., Arcury, 1990; Bradley et al. 1999; Bunting & Cousins, 1985; and LindemannMatties, 2005), it se ems likely that, in younger children, higher levels of environmenta l knowledge will positively correlate with a more inclusive scope of justice towards the environment. Method Participants Thirty 1st grade students participated in this study. All participants were students at a public elementary school in Naples, Flor ida. Ten participants were female and 20
30 participants were male. This distribution was representative of the classroom populations, as the three classrooms sampled from had a combined total of 19 females and 34 males. All students in this study we re 5-7 years old at the time of data collection (Personal communication, L. Dyer, October 20, 2008). Four teen of the particip ants were randomly assigned to be in a positive utility of nature condition, a nd the remaining participants were assigned to a neutral utility of nature condition. Nine hundred students attend this suburb an elementary school. The student population is approximately 50% Caucasian, 36% Hispanic, 5% multi-racial, 4% African American and 3% other ethnicities (District School Board of Collier County, 2008). Three hundred-sixty-four students of this school are classified by the school district as economically needy. Though general ethnicity a nd socio-economic status information is provided here for a better understanding of the school population, this information was not gathered from indi vidual participants. Materials Scope of Justice Stories. Five original short stories we re created for the scope of justice section of this study. The developmen t of these stories was influenced by other scope of justice research (Opotow, 1993). Each story had two versions, and the theme of the story remained constant for both versions. One version presented a high conflict condition, in which the interests of humans and the interests of nature diverged. The other version represented a low conf lict condition, in which the in terests of humans and those of nature were not in direct conflict. Depending on the stor y, nature was represented by animals, plants, or physical spaces. It wa s hoped that by using multiple stories, the possible effects of unfamiliarity with one situation (e.g., campi ng in the first story) could be balanced out by familiarity with other situations.
31 After each story version pa rticipants were asked a yes/no question about how they would resolve the conflict of the given situation. For a complete list of the stories and associated questions, see Appendix A. Each story had two accompanying illustrati ons; one illustration depicted the low conflict version of the story and one depicted the high conf lict version (see Appendix B). The pictures were designed to help child ren focus on the scenario and understand the differences between the stories. The pictures were created using hand drawings, tracings, an Epson scanner and Adobe Photoshop CS3. The pictures were simple colored line drawings. They were kept in clear sheet protectors inside of a binder for durability and ease of use. Though this method was largely infl uenced by Opotows (1993) study on inclusion of animals in the scope of justice, direct comparisons may not be drawn because the studies differ in three key ways. Fi rst, the language of the measurements used by Opotow was meant for use with adults, a nd needed to be modified for young children. Second, a large part of Opotows study was based on validating a method for measuring the scope of justice construct. Therefore, many of the original measurements (such as manipulation checks) were not included in the current study. Third, a measure of how similar a participant felt to nature was excluded as a condition in the current study, because it was found to have no significant impact on how participants widened or narrowed their scope of justice. Lastly, because only one story was used in Opotows research and five are used here, the resulting scope of justice scales have different point ranges. One of Opotows key measurements was a Likert scale that ranged from 1-7 points, while the scale of the current study ranged from 0-20 points.
32 Knowledge Questions. Fifteen questions were used in the knowledge subsection (see Appendix A). These original questions were based in part on a second grade textbook (Second Grade Science: Florida Ed ition, 2007) at the suggestion of the classrooms first grade teachers (Persona l communication, L. Dyer, October 20, 2008). The questions were also influenced by the CHEAKS scale, (Leeming et al., 1995) as well as Palmers (1994) study on young children's emergent environmental understanding. The breadth of topics covered by the questions was influenced by Cherretts (1989) list of the most important concepts in ecology, which was compiled by polling ecologists and environmental scientists of the British Ec ological Society. Table 1 shows the general topic covered by each knowledge question. A second grade textbook was used so th at the level of information would be slightly higher than that al ready presented to the students in their own classrooms. This avoided the potential problem of memorization effects from their own science curriculum affecting the participants' answers. The depth of content, wording, and style used in the questions was largely based on practice questi ons from the first grade big book text, which is used in place of individual texts in younger grade classroom s (Harcourt Science: First Grade, 2007). The information in the questions was also checked for age-appropriateness by one of the cooperating first grade teachers. She compared vocabulary used in the current study with that used in the classroom scie nce curriculum. Some changes to question phrasing and length were suggested by the teach er, and these suggestions were taken into account in the final version of the measuremen t. Specifically, the grammatical structures of some questions were simplified. Other ques tions were divided in to sub-questions so that multi-part answers were not requ ired, though not all questions required
33 segmentation. Specifically, questions that were divided assessed multiple levels of understanding. These questions focused on basic concepts and definitions first, and then addressed larger environmental ideas relate d to those concepts and definitions. The related segments were kept grouped as one overarching question to improve wording and coherency. If a child did not know a vocabulary word tested for in the first segment of a question, but was required to know the definition for the subsequent parts, a standard definition was offered to the student. These standard definitions were adapted from the second grade text (Second Grade Science: Florida Edition, 2007). Some knowledge questions had illustrations to help the participant understand the context (see Appendix B for a sample of thes e). Others did not warrant illustrations because they would have made the conten t of the questions ambiguous or offer the participant an obvious answer. Procedure Each participant returned a signed parent al consent form, and was asked for their signature of assent before starting the interview. Participants were interviewed individually in their classrooms. Interviews were conducted in an area of the classroom removed from other activities, students, and the classroom teacher. To start, each participant was asked, What does the word nature mean to you? This type of question was used in Loughla nd et al. (2002) and Al erby (2000) to better understand how children perceive and define their environment. Seven categories were used in analyzing responses: animals, plan ts, characteristics of place, action, affective value statement, relationship with natu re, or none/inappropriate. Table 2 provides
34 descriptions of these categories. Each participants answer wa s scored for the presence or absence of these categories, and the cat egories were not mutually exclusive. Then, if the participant was in the pos itive utility of nature condition he or she heard a positive statement about the environment: Many people believe that nature is beautiful, and that it can he lp us understand how the world works. Think inside your head... how do you feel about nature?" For the neutral utility condition, the participant was told nothing. These two conditions were mo deled after Opotows (1993) research, in which participants were primed to think of an aspect of the environment as either helpful or harmful to humans. For the first grad e population, the harmful condition was changed to a neutral one. Next, the participant heard both versions of each of the five scope of justice stories. The order of the low conflict and high conflict stories wa s randomized by coin toss, heads meaning the high conflict story woul d be read first with the low conflict story following immediately after. Stor y versions were always read consecutively. Previous to sitting down to the interview, the story illustrations were rearranged so that each participant heard and saw a unique story se quence. After each version of the story, participants were asked one yes/no question concerning how he or she would treat nature in that situation. Except for numb er 4, which was reversed scored, yes answers were coded as 0, and no answers coded as 2. An answer was scored as 1 if the participant mitigated a full yes or no response by finding a way to create an intermediary option (e.g., Yes I would take the shortcut, but I would tip-toe around the flowers as a response to story 5). Uncertain answers (i.e ., I dont know or maybe), were also scored as 1. The range of possible scores was 0-20, and a lower score indicated a narrower scope of justice towards the environment.
35 If the participant did not offer a spontaneous explanation for his or her answer to at least one story, he or she c hose a story to discuss in greate r detail. The qualitative data gained from this discussion were analyzed according to categories created after all students had been tested, because no previous research offered a relevant coding scheme. The categories used to code these explanati ons, called expansions, can be found in Table 3. Following the discussion of at least one story, the participant responded to the knowledge questions. Knowledge questions were scored as correct or incorrect according to an answer key created by the researcher. This answer key was reviewed and approved by one of the cooperating teachers in the study. Each question segment had a maximum of 1 or 2 points, with a maximum of 59 points for the entire measure. Table 4 shows this key and the maximum number of points awarded for each question. Responses to these questions were also analyzed for the presence of an anthropocentric attitude. This attitude was defined as thinking that humans can have more control in the environment than might actually be the case, or that the environment exists for human use (Kahn, 1997). Table 5 gives examples of responses fitting this category. A frequency count of anthropocentric responses was cr eated for each participant. The sequence of scope of justice storie s followed by the knowledge measurement was held constant, following research by Ar mstrong and Impara (1990) on order effects of administering emotion-base d and knowledge scales. In th at study, when children were presented with environmental knowledge measur es first and attitude measures second, they tended to have a more negative attit ude towards the environment. The researchers concluded that this was an effect of frustration; if the child di d not feel he or she had done well on the knowledge test, the child was more likely to channel that frustration into
36 rating the environment negatively. The curren t study addresses this potential problem by always measuring knowledge last. Participants were audio recorded duri ng the whole interview using a handheld recorder. Answers to scope of justice stories and other interview notes were recorded in real time in a notebook. Knowledge measurement answers were scored directly from the audio tapes. After testing was completed, each child was given a sheet of stickers and the chance to pull a small gift out of a grab-bag. Reliability In order to gain a measure of inter-rate r reliability on the kno wledge score, 20% of the tapes were chosen at random and give n to another coder not associated with the study. This coder used the answer key provi ded in Table 4 to score only the knowledge portions of these tapes. Reliability was calculated by dividing the number of point agreements between the researcher and coder by the number of total possible points. This yielded a reliability of 96% agreement. Results Scope of Justice and Knowledge Scores Observed scope of justice scores ranged from 10-20 points, out of a possible 20 points. Participants scores were concen trated in the upper ha lf of the scale ( M = 15.5, SD = 2.7) and were not normally distributed. Ob served knowledge scores ranged from 10-36 points, out of a possible 59 points ( M = 24.17, SD = 6.5). These scores were normally distributed. A one-tailed Spear man correlation coefficient showed that the relationship between these two variables was in the pred icted direction, though the test did not reach traditional levels of significance, r (28) = .29, p = .06 (see Figure 1). Therefore, it seems there may be a tendency for knowledge and scop e of justice scores to vary together.
37 High and Low Conflict Comparisons It was hypothesized that participants would have a lower mean score on high conflict stories. This was based on Opotows (1993) research, in which greater conflict between the environment and humans resulted in a narrower scope of justice towards the environment. A Wilcoxon Z test was used to examine differences on the withinparticipant variable of story conflict level. Participant means were significantly different between high conflict ( M = 7.03, SD = 2.06) and low conflict ( M = 8.27, SD = 1.50), Z = -2.43, p = .02. This difference was in the expected direction. According to these results, when faced with a situation in which a pro-environmental response would mean sacrificing more for the environment, participants were less likely to be proenvironmental. Given that there was a significant differen ce between conflict levels, the scope of justice and knowledge relationship was al so analyzed by this variable. It was hypothesized that high conflict s ituations might require participants to deliberate more when making a decision (Opotow, 1993), and us ing this subscale might better detect a possible relationship between the knowledge and scope of justice scores. High conflict story scores for all participants ( M = 7.13, SD = 2.06) were correlated with knowledge scores using a one-tailed Spearmans correlati on coefficient. Though these results were in the predicted direction, the test did not reach traditional levels of statistical significance, r (28) = .27, p = .08. Utility and Scope of Justice The utility of nature priming statement was used to potentially elicit more generous feelings towards the environment. It was thought th at participants who heard a positive priming statement about the environm ent would be more willing to take the
38 needs of the environment into consideration, ev en if that meant sacr ificing more of their desires (i.e., being pro-environmenta l during high conflict). A Mann-Whitney U test was used to determine if participants in the pos itive utility condition sc ored higher on the high conflict scope of justice questions ( M = 6.56, SD = 2.22) than those in the neutral utility group ( M = 7.79, SD = 1.72). The results of this test were not significant, U = 73.5, p = .10. However, positive utility participants did not score higher on these stories; instead, the neutral utility participan ts did. These results were in the opposite direction than expected. A Mann-Whitney U test for differences in total scope of justice scores showed that those in the positive utility condition ( M = 14.38, SD = 2.83) did not differ significantly from those in th e neutral utility condition ( M = 16.57, SD = 2.59), U = 71.5, p = .09. The means, however, show that partic ipants who heard a positive statement about the environment might have been somewhat less willing to consider the environments needs than those who heard nothing. These re sults were also in the opposite direction than expected. Anthropocentric Responses Anthropocentric responses were defined as answers to knowledge questions that unnecessarily invoked human control over the en vironment. For example, in response to a question about what might help endangered animals repopulate, one child said that the animals could go back to their homes at the zoo so they would not be killed. A post-hoc hypothesis was made that partic ipants who showed a higher fr equency of anthropocentric reasoning might have a lower scope of justi ce score. The frequenc y of anthropocentric responses ( M = 1.70, SD = 1.64) was correlated with scope of justice scores using a two-
39 tailed Spearman correlation coefficient There was no significant relationship between the two variables, r (28) = -.01, p = .98. It was also hypothesized that the two utility groups mi ght differ in frequency of anthropocentric responses. A Mann-Whitney U test was used to examine the potential difference between the positive utility group ( M = .88, SD = .81) and the neutral utility group ( M = .71, SD = .83) participants. This difference was not significant, U = 99.0, p = .56. Expansions Expansions were defined as explanations to scope of justice stories, beyond the required yes or no. For example, one child responde d that no, she would not build a hospital in the pond area because then the animals living there would have no water The last part of her statement was considered an expansion. A post-hoc hypothesis was made that participants in the positive utility c ondition made more expansions. Also, utility groups were compared on the types of expansions given, though there was no specific hypothesis concerning this comp arison. Finally, it was hypothe sized that frequency of expansions would positively correlate with scope of justice scores, as children who made more expansions might be more attuned to the environment. A Mann-Whitney U test was performed for diffe rences in per-participant expansion frequency between utility groups. Th e average number of expansions was not significantly different between positive utility participants ( M = 1.62, SD = 1.31) and neutral utility participants ( M = 1.79, SD = 2.01), U = 102.0, p = .65. There were three mutually exclusive expans ion types, the definitions of which can be found in Table 3. Table 6 shows the fre quencies of the three expansion types per utility group. A 2 (utility) X 3 (expansion type) Fishers Exact test was used to test for
40 differences between the groups. It showed no significant differences in the types of expansions given between utility groups, p = .65. Per-participant expansion frequency, regardless of utility group, was also tested for a potential correlation with scope of ju stice scores. Using a two-tailed Spearman correlation coefficient this re lationship was not significant, r (28) = .089, p = .32. Gender Comparisons It was hypothesized that diffe rences would exist between boys and girls in their responses to the question What does the word nature mean to you? Table 7 shows the observed categorical responses. Percentages ar e used in the presen tation of these data because of the inequality in number of boys versus number of girls. Three gender comparisons warranted further statistical anal ysis: animals, values, and relationships. In each category, girls appeared to mention thes e categories more often than boys. Three 2 (gender) X 2 (presence of cate gory) Fishers Exact tests were used to determine if there were significant differences between ge nders on these categories. There was no significant difference between males and females on the category of animals ( p = .12), or relationships ( p = .09). However, there was a signif icant difference for the category of values ( p = .03). In this category, female s mentioned affective values (e.g., I love it or I think the environment is great ) more often than males. An independent samples t -test was used to determin e a difference in knowledge scores between genders. The results show ed non-significant differences in knowledge scores between females ( M = 25.90, SD = 5.36) and males ( M = 23.30, SD = 6.96), t (28) = 1.04, p = .31. Using a Mann-Whitney U test, females ( M = 14.90, SD = 3.48) and males ( M = 15.80, SD = 2.26) were not statistically different on scope of justice scores either, U = 83.0, p = .50.
41 Discussion Scope of Justice and Knowledge Scores Main Hypothesis. The two variables of environmental knowledge and environmental scope of justice do show a positive trend. That is, the sample neared a correlation coefficient of conventional statisti cal significance in the predicted direction, such that when knowledge increased, scope of justice increased. With such a small sample size, the correlation coefficient obt ained supports further research into the relationship, and suggests a positive relationshi p might exist. Issues pertaining to the two measurements used in this study may provide some insight into the meaning of the observed trend and potential relationship. Scope of Justice. The conditions used in this se ction of the study (positive and neutral utility level; high and low conflic t level) were designed to gain a better understanding of childrens at titudes across different situ ations. Opotow (1993) found significant differences between participant sc ores in these conditions. She also found an interaction between the conditions, such that positive utility did not widen the scope of justice as much in the high conflict condition as it did in the low conflict condition. Data from the current study do not show that utility had the same effect as was observed in past research. In fact, utility had the opposite effect here, so that children who were not primed to think positively about the environmen t had a more inclusive scope of justice. In contrast, the conflict manipulation did have th e intended effect on participants responses: participants were less likely to be inclusiv e toward the environment in high conflict situations than in low conflict ones. It may be the case that the utility mani pulation in the current study was too weak to elicit the intended effect. In Opotows (1993) study, the priming statement used was
42 about the specific aspect of nature addressed in the scope of justice story. In the current study, it was a broad statement about nature in general. In the future, the method might be changed so that there is a relevant priming statement before each story. Using the stories in the current study as examples future research could use a statement about the benefits of protecting gopher tortoises for story one, a statement about the benefits of keeping the pond and habitat thriving in story two, a nd so on. Also, Opotow used two priming statements, a positive one and a negative one, whereas here only a positive one was used. It is possible then that an explicitly negative priming statement is needed to create the contrast in scores between utility gr oups obtained in the original research. Opotows (1993) scope of justice measurem ent used a 7 point Likert-scale, which ranged from 1 (not inclusive of the envi ronment) to 7 (very inclusive of the environment). She performed a 2 (utility) X 2 (conflict) ANOVA and only analyzed the four separate cell means, instead of creating a composite scale.1 Even without a composite scale, the separate cell means did not imply as inclusive a scope of justice as observed in the current study. This is because only one mean in her study (positive utility, low conflict condition) was as high as or hi gher than 3.5 (the midpoint of the scale), whereas all corresponding mean s in the current study reache d or exceeded the midpoint. In fact, even the lowest raw score in this study did not dip below the midpoint. What this comparison of adults and ch ildren cannot answer is whether the high scores found in the current study should be considered typical of a young population. It is possible that the general populati on of 5-7 year olds has a wider range of scores than 1020 points (on a 0-20 point scale), which was the range observed here. Yet, there is 1 These cells corresponded to: high utility, high conflict; high utility, low conflict; low utility, high conflict; low utility, low conflict.
43 evidence from past research to suggest that children truly do have high attitudes towards the environment and great concern for it. Many environmental attitude studies have shown that children often score in the uppe r half of scales an d measurements (e.g., Manoli et al., 2007; Evans, Juen et al., 2007; Bunting & Cousins, 1985), so the scores found here would not be considered atypical. It is also important to consider what it means to be pro-environmental in the context of this measurement. The method used in this study was a simulated type of task: children listened to stories, saw pictures of the situations from the stories, and responded to questions about the stories. Though the situations were m eant to be accessible to the children, they were also like stories that mi ght appear in books. That is, the stories were similar to the childrens lives in many ways, but not necessarily true reflections of how they live on a daily basis. For that reason, the participants might have felt free to respond to the stories as they would to questions about any fictional character or situation. They could respond without worrying th at they would ever have to actually make the sacrifices required in the situations. This might be a reason for the concentration of scores in the upper ranges. Participants knew each stor y was hypothetical, and they may have answered positively regardless of what their decision might be if actually placed in the story conditions. This is always a concern when using imaginary situations to measure attitudes and potential behavior, and it has been addressed in past attitude research (with children, Evans, Brauchle et al., 2007; with adults, Opotow, 1993). Another potential explanation for high scope of justice scores might be the effect of priming and sensitivity to the topic. A ccordingly, it was hypothesi zed that those in the positive utility condition would score higher on high conflict story versions. A distinction was made here between total scope of jus tice scores and the s ubscale of high conflict
44 story scores. All participants heard a high and low version of every st ory, and total scores included responses to all ten versions (5 low conflict stories plus 5 high conflict stories). In contrast, a high conflict score was crea ted using only the res ponses to high conflict story versions. It was thought that increasing the level of conflict between humans and the environment would narrow the scope of justice (Opotow, 1993). However, it was also thought that introducing the variable of positive utility of nature might mediate the effect of conflict, and encourage a par ticipant to be more inclusive than if he or she heard no priming statement. This was based on the idea that if participants were primed to be sensitive to the earth by hearing how useful it was to humans, they would be more likely to put the environment ahead of their ow n desires. As was mentioned previously, statistical testing showed this was most lik ely not the case: participants who heard a positive priming statement actually had lower ov erall scores. Therefore, utility also had no positive mediating effect on conflict level. Knowledge. Though the goal of creating a new knowledge measurement was to allow children to score in the full range of points, the knowledge sc ores still showed a low mean. This same issue of low scores aff ected past studies, such as Leeming et al. (1995). The low scores in this research may ha ve resulted from using material from the second grade text, which was slightly more advanced than what the students had previously encountered. However, a floor effect was not observed in the knowledge scores, which indicates that much of th e content was accessible for the first grade students. Based on past research and the curren t study, one could also conclude that children simply dont have that high a level of environmental knowledge at this age. However, considering the goal of any project measuring knowledge can mitigate that
45 conclusion. Even though it was important to quantify knowledge in the current study, in the end the main hypothesis hinged on whet her a range of knowledge scores was observed in the participants (and there was an observed 25 poi nt range). This range of scores allowed the researcher to assess the differences betwee n participants with comparatively low levels of knowledge and t hose with comparatively high levels of knowledge. In future research, knowledge measurements may have more value if they are used to compare multiple populations or par ticipants across time. These comparisons may have more meaning than evaluating the knowledge of one isolated sample. Anthropocentric Responses and Scope of Justice Scores Anthropocentrism, in the context of th is study, was defined as believing that humans have more control over the environmen t than is the reality. Bunting and Cousins (1985) found anthropocentrism to be more prom inent at this age level than in older children (though in that study, it was referred to as Environmental Adaptation ). This trend has been shown to decrease some what with age, though this study has no longitudinal or age-comparison data that would allow supporting or refuting that finding. Additionally, some knowledge questions in the current study offere d logical opportunities to show anthropocentric reasoning. For exampl e, when asked how plants get food, some participants reasoned (correc tly) that people often feed pl ants to keep them healthy. This is an example of thinking that humans are necessary to the well-being of the earth; as a rule, however, plants do not abso lutely need humans to feed them. Consequently, it was hypothesized th at participants who showed more anthropocentric reasoning would have a narr ower scope of justice. This was because children who had a narrower scope of justi ce towards the environment might be more focused on humans role in the world, and di scount the role and needs of nature. There
46 was no clear evidence that this relationship existed, however Given this non-significant relationship, maybe children who gave these responses simp ly didnt understand what the questions were asking. Also, the overall freque ncy of these responses was low, so there might not be a strong trend for ch ildren to use them in general. Expansions and Scope of Justice Scores Expansions to scope of justice questions were also analyzed, both for content and for frequency. It was hypothesized that positiv e utility participants would have a higher frequency of expansions because they were be tter primed to think about the environment. However, analysis showed that both utility groups offered similar types of expansions. That is, the groups gave an approximately e qual mix of expansions in the categories of not wanting to hurt nature and wanting to benefit humans Larger differences in the types of expansions might be expected if one gr oup gave mostly pro-environmental responses and the other group did not. As discussed pr eviously, though, both gr oups seemed to be inclusive of the environment in their scope of justice. There was also no difference between the two utility groups in how often participants offered these expansions. Any frequency count above one meant that the participant offered spontaneous expansions (only one expansion was explicitly asked for). It was hypothesized that those who had heard the positiv e utility statement would be more attuned to the task a nd therefore more likely to e xplain their reactions to the questions. This was not the case. Also, thos e with a higher frequency of expansions did not have a consistently higher scope of justice score. Maybe those participants who offered a greater number of spontaneous expa nsions were simply more engaged in the task, and the frequency was not dependent on the environmental content of the questions. Also, as mentioned previously, in the current study children were not asked for more than
47 one expansion. It might be possible that when explicitly asked to explain more than one response, there would be a difference in the frequency of expansions given across utility groups, or the frequency of expansions give n between those with high scope of justice scores and low scores. Gender Comparisons Qualitative responses were co llected from the question What does the word nature mean to you? These responses were categori zed for ease of comparison to examine how participants defined nature Since this question was asked before participants heard the positiv e utility statement, it would not make sense to compare responses across utility groups. However, across gender there are some interesting differences in response patterns. Boys and girl s were very similar to each other in their definitions of the environment, except in the categories of animals, value statements, and relationship statements. The percentage of females who mentioned each of these categories was larger than the percenta ge of males who mentioned them. When interpreting these results, though, it is important to note that the researcher did not give any guidelines for the length of response. Therefore, some children may have given a response in one category and stopped, though they might have given more if asked; other children may have talked until told to conclude their response. Lastly, past literature has shown gender di fferences in scienc e-related knowledge, though these studies have not dealt with such a young population. Specifically, boys tend to have higher knowledge scores (Freid & Filho, 1997; Arcury, 1990). Various studies have shown different gender effects on attit ude. Bunting and Cousins (1985) concluded that girls were much more likely to have an attitude of Pastora lism (predisposition to nature) than boys. Other studies have shown no appreciable differences in attitude scores
48 between the genders (Evans, Juen et al. 2007), including scores on a scope of justice measurement (with adults, Opotow, 1993). The participants in the current study did not show any significant differences on either knowledge or scope of justice. It may be that at such a young age, there are no practical differences. Future Research The inferences made from this study ma y be material for future research. Though behaviors were not the focus here, some of the reviewed literature suggests a way to incorporate a behavior measurement without necessitating an observational study. For example, Evans, Brauchle et al.s (2007) study involved games and physical movements to describe the frequency with which children performed certain environmental behaviors. This type of methodology could be important in future research if educators would like to further explore the relationship between all lin ks of the knowledge, attitude, and behavior chain. The possibility that the children in th is study had an overall positive attitude toward the environment could be verified in future studies. It would be interesting to administer other measures of environmental at titude and concern to the same sample and compare the level of pro-environmental at titudes the students displayed on each. However, Opotow (2006) has noted that the scope of justic e construct does not necessarily match up to the results of other t ypes of attitude measurements. This is one of the reasons it is used in this research: it ta kes into account interact ions between variables that can lead to attitude. Conclusion What drives people to incl ude the environment in th eir scope of justice is important to many scientists, policy makers, and educators who feel human impact on the
49 earth has become detrimental. This group of people also feels that humans need to moderate their effects on the environment (Opotow & Cl ayton, 1994). However, because the human population is so large, it would be necessary for a majority of the population to make changes that will have a positive impact on the environment. Those looking for practical ways to inspire this change usually start at a grassroots level, working first to educate citizens in the populat ion instead of expecting govern ment regulation to lead to better environmental practices. Environmental ed ucation becomes the tool for this sort of work (Stoss, 2008). This model of change is the manifestation of the assumption discussed earlier: increase a persons environmental knowledge, and that person will be more likely to consider the needs of the environment. The current study tested the hypothesis that this relationship was supported in a sample of young children. The positive trend seen in the data pa rtially supports a relationship between knowledge and attitude. In fact, the relationship was close to reaching traditional levels of significance in the sample tested. However, it is unclear from th is study whether the relationship strongly exists. It appears that environmental ed ucators would not harm their cause by increasing childrens knowledge. Furthe rmore, even if the relationship between these two variables seems w eak, knowledge still may be the most practical factor to manipulate. In fact, factors implicated by past research as alternative explanations for environmental attitude differences cannot be manipulated at all (i.e., age and gender) (Arcury, 1990). It is also possible that a stronger relationship exists w ithin the general knowledgeattitude one: different elemen ts of environmental knowledge might have different effects on attitude. Though the current study attemp ted to create a general measure of
50 environmental knowledge, topic-based knowledge might be more effective for inspiring pro-environmental attitudes. Such knowledge topics may be participant-specific (e.g., knowledge about the habitat of a favorite animal) versus general (e.g., knowledge about the life processes of a plant). In the case of participant-specificity, salience may also play a part in the effect of knowledge on attitu de, such that information about a favorite animal has a greater effect on attitude th an information a person finds uninteresting. Kellert (1985) proposed a similar domain-s pecific type of rela tionship in his study on knowledge about and attitudes towards anim als. That is, if one knows more about animals, he or she may be kinder to animals; if one knows a lot about the rainforest, he or she may be indignant if others cut it down, ye t may not be as concerned if others destroy different habitats elsewhere in the world. Fu ture research could explore the possibility that specific knowledge is better related to attitude than general environmental knowledge. If that is the case, stronger results might be obtained than in the current study. All things considered, this research does not directly contradict the assumption from which environmental educators work. In terms of immediate application, there are few options available for increasing environm ental scope of justice that are as obvious, popular, and practical as envir onmental education. Future research should not focus on proving this assumption so much as trying to tease it apart and discove r all of its nuances. For example, looking at the types of que stions children answer ed incorrectly in the current study offers some insight into what educators could focus on. For instance, some children had problems with abstract or global concepts, like understanding how an entire species can be endangered, versus just one animal of a species. Also, some children showed low knowledge about topics that might not be often addressed in first grade
51 classrooms, such as environmentally-friendly modes of transportation, or the differences between ecosystems. In any case, it is doubtful that environmen tal education will fall out of favor as a way to widen scope of justice toward the envi ronment. It is important to note that even though scope of justice scores found in this research were quite high, they still showed variation when the condition of conflict was introduced. This suggests that there may still be room to improve how young children feel about the environment: they are not proenvironmental all of the time, ev en in hypothetical situations. The final link in the chain, behavior, is left to explore in future research. The relationship between attitude a nd behavior was not addressed here, and it is still unclear from the conclusions of past studies. As the U.S. Environmental Protection Agency (2008) states, environmental education give s people the awareness and knowledge [, and] necessary skills to make informed d ecisions and take responsible action. Whether or not children will take responsible action ha s yet to be determined, but this study may put future researchers in a good pos ition to answer that question.
52 References Alerby, E. (2000). A way of visualising ch ildren's and young people's thoughts about the environment: A study of drawings. Environmental Education Research, 6 205222. Arcury, T. A. (1990). Environmental attitude and knowledge. Human Organization, 49, 300-304. Armstrong, J. B., & Impara, J. C. (1990). The effects of order of test administration on environmental attitudes. Journal of Environmental Education, 21 (3), 37-39. Ashworth, S., Boyes, E., Paton, R., & Stanisstreet, M. (1995) Conservation of endangered species: What do children think? International Journal of Environmental Education and Information, 14 229-244. Atkin, C.K. & Gantz, W. (1978). Television News and Political Socialization. Public Opinion Quarterly, 42 183-194. Bailey, S., & Watson, R. (1998). Establishi ng basic ecological understanding in younger pupils: A pilot evaluation of a strategy based on drama/role play. International Journal of Science Education, 20 139-152. Barraza, L. (1999). Children's drawings about the environment. Environmental Education Research, 5, 49-66. Barraza, L., & Walford, R. A. (2002). Envir onmental education: A comparison between English and Mexican school children. Environmental Education Research, 8 171186. Boyes, E., & Stanisstreet, M. (1993). The 'greenhouse effect': Children's perceptions of causes, consequences and cures. International Journal of Science Education, 15 531-552.
53 Bradley, J. C., Waliczek, T. M., & Zaji cek, J. M. (1999). Relationship between environmental knowledge and environmen tal attitude of high school students. Journal of Environmental Education, 30 (3), 17-22. Bunting, T. E., & Cousins, L. R. (1985). Environmental dispositions among school-age children: A prelim inary investigation. Environment and Behavior, 17 725-768. Chawla, L. (1988). Childrens concern for the natural environment. Children's Environments Quarterly, 5 (3), 13-20. Cherret, J.M. (ed.) (1989). Ecological c oncepts: The contribution of ecology to an understanding of the natural worl d. Oxford, UK: Blackwell Scientific. Christidou, V., Koulaidis, V ., & Christidis, T. (1997). Children's use of metaphors in relation to their mental models: The case of the oz one layer and its depletion. Research in Science Education, 27 541-552. Clayton, S., & Opotow, S. (2003). Justice and iden tity: Changing perspectives on what is fair. Personality and Social Psychology Review, 7 298-310. Cohen, S., & Horm-Wingerd, D. (1993). Ecol ogical awareness among preschool children. Environment and Behavior, 25 103-120. District School Board of Collier County (2008) Big Cypress Elementary. Retrieved November 11, 2008, from http://www. collier.k12.fl.us/schools/bce.asp Dunlap, R. E., & Van Liere, K. D. (1978). The new environmental paradigm: A proposed measuring instrument and preliminary results. Journal of Environmental Education, 9 10-19. Dunlap, R. K.; Van Liere, K. D.; Mer tig, A., & Jones, R.E. (2000). Measuring endorsement of the new ecological paradigm: a revised NEP scale. Journal of Social Issues, 56 425-442.
54 Evans, G.W., Brauchle, G., Haq, A., Stecker, R., Wong, K., & Shapiro, E. (2007). Young childrens environmental attitudes and behaviors. Environment and Behavior, 39 635-659. Evans, G.W., Juen, B., Corral-Verdugo, V., Corraliza, J. A., & Kaiser, F. G. (2007). Childrens cross-cultural environmental attitudes and self-r eported behaviors. Children, Youth, and Environments, 17 129-143. Freid, B., & Filho, W. L. (1997). Young peopl e's attitudes towards and knowledge about the environment: An analysis based on TIMSS-data. Scientia Paedagogica Experimentalis, 34, 231-244. Gambro, J. S. & Switzky, H.N. (1999). Variab les associated with American high school students' knowledge of environmental issues related to en ergy and pollution. Journal of Environmental Education, 30 (2), 15-22. Greaves, E., Stanisstreet, M., Bayes, E., & Williams, T. (1993). Children's ideas about rainforests. Journal of Biological Education, 27 189-194. Harcourt Science: First Grade (2007). Orlando, FL: Harcourt. Hart, R., & Chawla, L. (1981). The deve lopment of children's concern for the environment. Zeitschrift Fr Umweltforschung, 2 271-294. Hartsell, B. (2006). Teaching toward compa ssion: Environmental values education for secondary students. Journal of Secondary Gifted Education, 17 265-271. Harvey, M. R. (1989). Children s experiences with vegetation. Childrens Environments Quarterly, 6 (1), 36-43. Heft, H. (1988). Affordances of childrens environments: A functional approach to environmental description. Childrens Environmental Quarterly, 5 (3), 29-37.
55 Hess, R.D. & Torney, J.V. (1967). The developm ent of political attitudes in children. Chicago, IL: Aldine. Hyun, E. (2005). How is young childrens inte llectual culture of perceiving nature different from adults? Environmental Education Research, 11 199-214. Kahn, P. H. (1997). Children's moral and eco logical reasoning about the Prince William Sound oil spill. Developmental Psychology, 33 1091-1096. Kellert, S.R. (1985). Attitudes towards an imals: Age-related development among children. Journal of Environmental Education, 16 (3), 29-39. Leeming, C. L., Dwyer, W. O., & Bracken, B. A. (1995). Children's environmental attitude and knowledge scale: Construction and validation. Journal of Environmental Education, 26 (3), 22-31. Lindemann-Matthies, P. (2005). Loveable mammals and lifeless plants: How children's interest in common local organisms can be enhanced through observation of nature. International Journal of Science Education, 27, 655-677. Loughland, T., Reid, A., & Petocz, P. (2002). Young people's conceptions of environment: A phenomenographic analysis. Environmental Education Research, 8 187-197. Manoli, C. C., Johnson, B., & Dunlap, R. E. (2007). Assessing children's environmental worldviews: Modifying and validating th e new ecological paradigm scale for use with children. Journal of Environmental Education, 38 (4), 3-13. Opotow, S., Clayton, S. (1994). Green justice: Conceptions of fairness and the natural world. Journal of Social Issues, 50 (3), 1-11. Opotow, S. V. (2006). Seeking inclusion and pluralism two exploratory studies. Peace and Conflict, 12 349-366.
56 Opotow, S. V. (1993). Animals and the scope of justice. The Journal of Social Issues, 49, 71-85. Orton, D. (2006). The green movement and the deep ecology movement. Retrieved March 15, 2009 from http://home. ca.inter.net/~greenweb/Green_&_DE_ Movements.html Palmer, J. (1993). From Santa Claus to sustainability: Emergent understanding of concepts and issues in environmental science. International Journal of Science Education, 15 487-495. Palmer, J. (1994). Acquisition of environmental subject knowledge in pre-school children: An international study. Childrens Environments, 11 41-53. Palmer, J. & Suggate, J. (1994). The developm ent of children's understanding of distant places and environmental issues: Repo rt of a UK longitudinal study of the development of ideas between the ages of 4 and 10 years. Research Papers in Education, 19 205-237. Pomerantz, G. (1986). Environmental educati on tools for elementary schoolchildren: The use of a popular children's magazine. Journal of Environmental Education, 17 (4), 17-22. Project Learning Tree (2004). Our mission and goals. Retrieved February 10, 2008, from http://www.plt.org/cms/pages/21_19_3.html Second Grade Science: Florida Edition (2007). Orlando, FL: Harcourt. Stoss, F. (2008). If we are so smart, why do we need enviro nmental education? Electronic Green Journal, 26 1-3.
57 U.S. Environmental Protection Agency (2008). Environmental education: Basic information. Retrieved February 10, 2009, from http://www.epa.gov/enviroed/ basic.html Wals, A.E.J. (1994). Nobody planted it, it just grew! Young adolescents perceptions and experiences of nature in the cont ext of urban environmental education. Childrens Environments, 11 (3), 1-2.
58 Appendix A Interview Protocol I. What does the word nature mean to you? II. Utility Group Priming a. Positive: Many people believe that nature is beautiful, and that it can help us understand how the world works. Think in side your head... how do you feel about nature? b. Neutral: Told nothing. III. Scope of Justice Stories Instructions: Now I'm going to tell you a few short stor ies about nature, whic h people also call the environment. I will tell you one version of the story, and then I am going to change the story a little bit and tell y ou another version. Make sure you listen carefully so you can hear the changes, okay? 1. Let's pretend your family is on a camping trip High conflict, presented first: and you need to set up your tent. You look around, and there is only one flat spot without rocks that you could sleep on. You almost put up your tent, and then you see there is a gopher tortoise livi ng on that spot. Gopher tortoise s are like turtles. If you set up your tent right there, the to rtoise nest will be damaged. -Would you put your tent on that spot?
59 High conflict, presented second: and you need to set up your tent. This tim e you look around and there is only one flat spot without rocks that you could sleep on. Exce pt, the gopher tortoise has its nest there. If you set up your tent in that spot, the tortoise nest will be damaged. -What about this time, would you put your tent on that spot? Low conflict, presented first: and you need to set up your tent. There are lo ts of flat spots with no rocks where you could sleep. You see one spot you really wa nt, though, because it is under a big tree. You almost put up your tent, and then you see ther e is a gopher tortoise living there. Gopher tortoises are like turtles. If you set up your tent in that spot, the tortoise nest will be damaged. -Would you put your tent on that spot? Low conflict, presented second: and this time, there are many flat spots without rocks where you could sleep. You see one spot you really want, though, because it is under a big tree. You almost put up your tent, and then you see the gopher to rtoise living there. If you set up your tent in that spot, the tortoise nest will be damaged. -What about this time, would you set up your tent on that spot? 2. A mayor is the leader of a town. Let's imagine the mayor of your town wants to build a new hospital... High conflict, presented first: because the closest one is 100 miles away and it takes a long time for ambulances and other people to drive there. The spot she ha s chosen for the new building is where many animals get their food and water. There is a pond that they come and drink from, and they
60 eat the plants that grow around the pond. If the hospital is built here, then the animals will have to go far away to find a nother place to eat and drink. -Do you think the new hospital should be built on this spot? High conflict, presented second: and this time she wants to build the hospi tal because the closest one is 100 miles away and it takes a long time for ambulances and othe r people to drive there. The spot she has chosen is the same spot I told you about before; it's where many animals get their food and water. If the hospital is built here, th e animals will have to go far away to find another place to eat and drink. -How about now, do you thi nk the new hospital should be built on this spot? Low conflict, presented first: but your town already has a hospital that doesn 't take very long to drive to. The spot she has chosen for the new building is where many animals get their food and water. There is a pond that they come and drink from, and they eat the plants that grow around the pond. If the hospital is built here, the anim als will have to go far away to find another place to eat and drink. -Do you think the new hospital should be built on this spot? Low conflict, presented second: but in this case your town alr eady has a hospital that doesn't take long to drive to. The spot she has chosen is the same one I told you about before; it's where many animals get their food and water. If the hospital is built he re, the animals will have to go far away to find another place to eat and drink. -How about now, do you think the new hospital should be built on this spot?
61 3. Let's pretend that you are a gardener and you grow vegetables... High conflict, presented first: and you use a lot of bug spray, because last year bugs ate all of your vegetables and you didn't grow enough healthy food for your family. But, if you use bug spray, then when it rains the spray will wash off the vegetables into the nearby pond and make the fish that live there sick. -Do you think it is okay to use the bug spra y on your vegetables? High conflict, presented second: and you use a lot of bug spray on your vegetables, this time because last year bugs ate all of your vegetables and you didn't grow enough healthy food for your family. But, if you use bug spray, then when it rains the spray will wash off the vegetables into the nearby pond and make the fish that live there sick. -What about in this case, do you thin k it is okay to use bug spray on your vegetables? Low conflict, presented first: and you use a lot of bug spray on your vegeta bles because you want them to grow really big and win a prize at the fair. But, if you use bug spray, then when it rains the spray will wash off the vegetables into the nearby pond and make the fish that live there sick. -Do you think it is okay to use the bug spray on your vegetables? Low conflict, presented second: and you use a lot of bug spray on your vegetables, this time because you want them to grow really big and win a priz e at the fair. But, if you use bug spray, then when it rains
62 the spray will wash off the vegetables into the nearby pond and make the fish that live there sick. -What about in this case, do you think it is okay to use bug spray? 4. Imagine you are at the beach with your mom or dad... High conflict, presented first: and you see a lot of trash down by the water. There are birds picking through the trash. You go to pick up the trash, but you see that the trashcan is really far away. If you pick it all up and walk to the trashcan, then you won't have any time to play at all. -Would you pick up the trash? High conflict, presented second: and you see a lot of trash down by the water again. There are birds picking through it. You go to pick up the trash, but this time you see that the trashcan is really far away. If you pick up all the trash and walk to the trash can, then you won't have any time to play at all. -Would you pick up the trash in this case? Low conflict, presented first: and you see a lot of trash down by the water. There are birds picking through the trash. The trashcan isn't very far away, so if you picked up the trash you'd still have plenty of time to play. -Would you pick up the trash? Low conflict, presented second: and you see a lot of trash down by the water. There are birds picking through the trash. This time, the trashcan isn't very far away, so if you picked up the trash you'd still have time to play.
63 -Would you pick up the trash in this case? 5. Let's pretend you're on a walk in the woods High conflict, presented first: and you look at your watch. You remember that you need to be at your best friend's birthday party very soon! You see a path that looks like a cool shortcut and you want to take it so you make it to the party on time. But, if you walk on this shortcut, you will step on wildflowers that are growing there and hurt them. -Would you take the shortcut? High conflict, presented second: and you look at your watch. This time, you reme mber that you need to be at your best friend's birthday party very soon! You see a pa th that looks like a c ool shortcut and you want to take it so you make it to the party on time. But, if you walk on this shortcut, you will step on wildflowers that ar e growing there and hurt them. -Would you take the shortcut in this case? Low conflict, presented first: and you look at your watch. You are getting really tired! You see a path that looks like a cool shortcut and you want to take it so you are done walking. But, if you walk on this shortcut, you will step on wildflowers that are growing there and hurt them. -Would you take the shortcut? Low conflict, presented second: and you look at your watch. This time, you are ge tting really tired! You see a path that looks like a cool shortcut and you want to ta ke it so you are done walking. But, if you walk on this shortcut, you will step on wild flowers that are growing there and hurt them. -Would you take the shortcut in this case?
64 IV. Knowledge questions Instructions: Now I'm going to ask you a few questions. Just answer them as best as you can and tell me what you really think. Ready? (Illustration) in front of a question indicates that it had an accompanying illustration. A sample of these illustrations can be found in Appendix B. 1. a. What do you think pollution is? b. What examples of pollution can you think of? c. Does it matter if something is polluted? d. Why? 2. a. What do you think recycling means? b. What can people recycle? c. Why do people recycle things? (Illustration) 3. Suppose you and your mom or dad want to go to the post office close your house. a. Which way of getting there is best for the environment? b. Why? (Illustration) 4. Let's look at the picture of this lake. a. If it's very sunny outside, what do you thi nk will happen to the wa ter in this lake? b. If the water in the lake dries up from the sun, how could more water get back into the lake? (Illustration) 5. This is a picture of a small forest. It is in the middle of a lot of houses. Let's say people want to cut down these trees to build more houses.
65 a. If the small forest was cut down, what do you think will happen to animals and plants that used to live there? b. If the animals' homes are destroyed, do you th ink it will be easy or hard for them to find new homes? c. Why? (Illustration) 6. Here is a picture of a plant that gr ows in a desert. Here is a picture of a plant that grows in a forest, a nd one that grows near the beach. a. Do you think that if we took the forest pl ant to the desert it could easily live there? b. Why? c. Do you think the beach plan t could live in the desert? d. Why? 7. a. Can you think of the names of any plants? b. How about the names of any animals? c. Can you think of any ways that plants and animals are the same? d. Can you think of any ways that pl ants and animals are different? 8. What are some differences you can th ink of between a rock and a plant? (Illustration) 9. Imagine your friend has given you a small plant for your bedroom. Here is a picture of what it looks like. You put the plant on your dresser in a corner of your room and accidentally forget about it. When you look at it the next week, it hasn't grown, its leaves are droopy an d it looks unhealthy. a. What could you do to make it healthier? b. Plants need some of the same things that we do, like food. Where does a plant get its food from? 10. a. What do you think an endangered animal is?
66 b. How do you think an animal could become endangered? c. What could help endangered animals? (Illustration) 11. a. What do you think a habitat is? Here is a picture of a habitat for some plants and animals. b. What kind of place do you think this is? Here are some animals. c. Which animals do you think could live in this desert? 12. Let's say you want to start a garden. But, a ll of the stores are out of seeds. What's another way you could get seeds to start your garden? 13. a. Our cars use gas. Do you think we (t he earth) will ever run out of gas? b. Why? 14. Some plants have thorns on them. Some plants don't taste very good when an animal bites into them. Why would plants have thorns or a bad taste? 15. Let's think of a place where no people live. Now let's think of a place where lots of people live. How do people change nature? Wh at do people do to na ture to make it look different? V. Definitions of words in the knowledge section. These standard definitions are supplied if a child cannot produce a definition within a question. Pollution: Things that people do to make nature/the environment dirty. Habitat: A place where a plant or animal lives. Endangered: There are not very many of that type of animal left in the world. Recycling: To make new things out of old things.
67 Appendix B Illustrations The following pages show monochromatic ve rsions of the scope of justice story illustrations, in order from story 1 to story 5. The low conflict story is placed at the top of each page, and the high conflict story is pla ced at the bottom. Following these is a sample of the illustrations us ed in the knowledge subsection.
74 Table 1 Knowledge Question Topics Question Number General Topic 1. Ecosystem Fragility, Human Effects 2. Conservation of Resources 3. Conservation of Resources 4. Ecological Processes 5. Niche, Human Effects, Competition 6. Niche, Ecological Adaptation 7. Living Organisms 8. Living vs. Non-living 9. Living Organisms 10. Competition, Human Effects 11. Niche, Ecosystem 12. Ecological Processes 13. Conservation of Resources 14. Ecological Adaptation 15. Human Effects
75 Table 2 Response Categories for What does the word nature mean to you? Category Definition Animals Mentions animals in the response Plants Mentions plants in the response Characteristics of place Describes nature, including landscapes, weather, and non-living elements Action Describes activities that one could do in nature, such as play hide and seek Value statement Offers an affective statement about the environment, such as I love it Relationship Describes a spatial or care-giving relationship between humans and the environment, which may or may not be reciprocal: for example, trees are all around us, it is animals and people together on Earth, or you have to take care of nature None/inappropriate Gives a nonsensical response, such as carpet
76 Table 3 Scope of Justice Story Expansion Codes Code Definition 1 Participant does not want to harm the environment 2 Participant wishes to make life ea sier on humans (including themselves) 3 Participant does not know why they ga ve the response, or simply repeats the consequence told in the story
77 Table 4 Knowledge Test Answer Key with Abridged Questions Question Number & Points Possible Acceptable Answers Unacceptable Answers 1. a. (2 pts.) What do you think pollution is? human cause of harm to nature (generic cause) driving, littering, pouring bad things in water (specific cause) factories, cars, trash, (physical object cause) 1. b. (2 pts.) What examples of pollution can you think of? litter bad things in the water smoke/gas/ air pollution 1. c. (1 pt.) Does it matter if something is polluted? yes 1. d. (2 pts.) Why? it harms (any part of) nature it harms humans it causes large-scale problems: The whole planet will be dirty. environmental slogan- We need to keep our Earth clean. 2. a. (1 pt.) What do you think recycling means? lists recyclable objects putting X in recycling bin not sending everything to dump/ trashcan reusing putting X in the trashcan 2. b. (2 pts.) What can people recycle?
78 cans, bottles, plastic, metal, glass, paper if 4 or more objects are listed, award 2 pts. 2. c. (2 pts.) Why do people recycle things? to make new things/ reuse things good for the earth/ environment so recyclables dont go to dump 3. a. (1 pt.) Which way of getting there is best for the environment? bike 3. b. (2 pts.) Why? bike has no pollution/ is cleaner than car bike doesnt use gas/ saves energy so there are fewer cars on the roads helps to keep the earth clean 4. a. (1 pt.) If it's sunny outside, what do you think will happen to the water in this lake? dry up/ go up into the air get hotter/ warmer 4. b. (1 pt.) If the water in the lake dries up from the sun, how could more water get back into the lake? rain humans can refill it 5. a. (2 pts.) If the small forest was cut down, what do you think will happen to animals and plants that used to live there? die need to find new homes wont have resources: water, food, shelter live with people 5. b. (1 pt.) If the animals' homes are destroyed, do you think it will be easy for them to find new homes? hard/ no 5. c. (2 pts.) Why?
79 have to go far to find new homes or resources/ might not find one people are in the habitat/ not enough space for animals might be harmed while looking for new home 6. a. (1 pt.) Do you think that if we took the fo rest plant to the desert it could easily live there? no 6. b. (2 pts.) Why? desert is too hot/ too much sun not enough water different type of dirt plant is not made to live there 6. c. (1 pt.) Do you think the beach plan t could live in the desert? no 6. d. (2 pts.) Why? desert is too hot/ too much sun not enough water plant is not made to live there if responded yes to 6. c., give pt. for saying beach and desert are both hot and/or sunny 7. a. (2 pts.) Can you think of the name of any plants? 1-3 plants, award 1 pt.; 4 or more, award 2 pts. 7. b. (2 pts.) How about the names of any animals? 1-3 animals, 1 pt.; 4 or more, award 2 pts. 7. c. (2 pts.) Can you think of any ways that plants and animals are the same? both are living both need food, water, place to live both grow both are part of nature
80 7. d. (2 pts.) How are plants and animals different? plants stay rooted to one spot animals have bones both eat different things 8. (2 pts.) What are some differences you can think of between a rock and a plant? one is living, one is not/ plants need resources rocks dont have leaves or roots rocks are hard, plants are soft/ plants move in the breeze plants have a scent 9. a. (2 pts.) What could you do to make it healthier? water/ food give it sunlight put it outside/ give it fresh air 9. b. (2 pts.) Where does a plant that lives in a forest get its food from? the plant makes it/ the sun outside from roots/ the ground from people 10. a. (2 pts.) What do you think an endangered animal is? there are not very many of the animal left the animals are harmed by humans the animals negatively affected humans; therefore were killed 10. b. (2 pts.) How do you think an animal could become endangered? people hunt the animal the animal has limited resources: food, water, space they died 10. c. (2 pts.) What could help endangered animals?
81 give animals space to live/ keep them in a habitat ensure access to resources disallow hunting give medical help/ help them survive when hurt help them 11. a. (2 pts.) What do you think a habitat is? a place where animals and plants live where animals and plants have their food and water 11. b. (1 pt.) What kind of place do you think this is? desert 11. c. (1 pt.) Which animals do you think could live in the desert? scorpion, coyote, and snake alligator, polar bear, and whale 12. (1 pt.) What's another way you could ge t seeds to start your garden? take seeds out of other plants/ fruits/ vegetables 13. a. (1 pt.) Our cars use gas. Do you think we will ever run out of gas? yes 13. b. (2 pts.) Why ? we use a lot of it/ there are a lot of cars cant make more of it it goes into the air 14. (2 pts.) Why would plants have thorns or a bad taste?
82 so animals dont eat the plants protection/ thorns will hurt the animals the plants are poisonous 15. (1 pt.) How do people change nature? cut things down build things landscape transportationrelated responses
83 Table 5 Examples of Anthropocentrism in Knowledge Test Responses Question Response If the water in the lake dries up from the sun, how could more water get back into the lake? You could get a hose or bucket and fill it up. If the small forest was cut down, what do you think will happen to animals and plants that used to live there? They could go back to the zoo or They could live with people. Do you think that if we took the forest plant to the desert it could easily live there? Do you think the beach plant could live in the desert? Yes, because someone could water it. or It could live in somebodys house. Plants need some of the same things that we do, like food. Where does a plant get its food from? It gets it from farmers. or People feed plants.
84 Table 6 Frequency of Scope of Justice Expansion Type by Utility Group (and Percentages) Utility Group Expansion Type Positive Utility Neutral Utility 1 5 (31%) 7 (50%) 2 7 (43%) 4 (29%) 3 4 (25%) 3 (21%)
85 Table 7 Percent of Participants Per Category of Response Category Female Male Animals 50% 20% Plants 30% 30% Characteristics of Place 30% 35% Action 30% 30% Value 40% 5% Relationship 30% 5% None/Inappropriate 20% 25%
86 Figure 1 Per-Participant Knowledge and Scope of Ju stice Scores with Line of Best Fit Knowledge and Scope of Justice Scores0 5 10 15 20 25 0510152025303540 Knowledge ScoreScope of Justice Score