5. Follow-up Study on Partial Sample of 12 Schools

The discussion on extrinsic influences on human behavior has never excluded the natural surroundings or other physical forms as factors, however, concrete empirical studies on "who is influenced how by what" are weakly represented.

Moos (1976) describes in one chapter of his book "The Human Context" several physical factors of human behavior - both natural and human made. These factors fall into the categories: weather, architecture, population density and pollution. "Nature" is named as a physical factor but its manifestation is not further defined or addressed. The factors which are explicitly addressed are limited to climate (general), weather (season, air pressure, temperature, wind and humidity) and the phases of the moon. Landscapes, for example, either natural or manipulated through human design and use, similarly, are hardly even mentioned.

Itelson (1974) sums up the relationship between people and landscape as an externalization of culture (the "symbolic" perspective). People and landscape function together, where function is divided into two categories:

1. Functions that satisfy present needs and preferences (the "instrumental" perspective) and
2. Functions that (ought to) satisfy needs and preferences (the "ecological" perspective) of the future.

Both Moos and Itelson maintain that people are influenced by their surroundings to the extent that certain behaviors are even impossible in surroundings that are not conducive to those behaviors or that pose a conflict. Both recognize the school as an important "surrounding" and do not restrict this to mean the school building and classrooms alone. Both are of the opinion that physical manifestations of the architecture and interior design influence perception, behavior and learning significantly. However, neither of the two go into any detail on physical structures and design on the "outside", on the grounds - within the school landscape - as a potential source of conflict for perception, learning and of ultimate interest, here, behavior.

I have found several further gleanings on the influence of environments, in general, on human behavior from the following subjects:

• Spatial density and problem behavior in children (Loo, 1978)
• Emotion-eliciting qualities of environments (Russell, 1978)
• Ambient stressors and human well-being (Campbell, 1983)
• Purposive evaluation of places (Canter, 1983)
• Model of person-environment compatibility (Kaplan, 1983)
• Playground designs and preschool children’s behavior (Brown, 1984)
• Environmental cues and vandalism (Samdahl, 1985)
• School environments and school (disruptive) behaviors (Wiatrowski, 1983)
• Symbolism in building materials (Sadalla, 1993)

Research such as these imply strongly that further research on the effects of the design and use of schools (in the sense of a school "culture") is necessary if we really want to understand more about why and how young people (do not) learn responsible environmental behavior.

Our own research on student environmental motivation has given us insight into the social composition of school student populations and is summarized in the preceding chapter.

Results suggest that not only does the social structure (with regard to environmental motivation) vary from school to school, but that there may well be conditions at schools that effect student environmental motivation in general. In particular, school effects on students tend to be seen most noticeably in their responses to items on coping style, self-competence and social norm.

It is conceivable that if conditions can be enhanced that student preferences and needs in acquiring environmental behaviors may be satisfied better by the schools.

Since all 54 schools in our previous survey report conservation effort s to some degree. This confirms an overall investment in and commitment to conservation. However, our findings have neither detected differences among schools nor been able to explain even in part differences in student environmental motivation. For instance, we have found no evidence to support a popular belief that "more conservation is better" or that a broad variety of activities fosters, for example, student knowledge. It still remains a mystery which possibilities are available to schools that can be implemented for the benefit of students.

In addition to a lack of understanding on the school conditions at play, there appears to be substantial discrepancies between student and administration responses on just what is going on in conservation at their schools. Student responses indicate that they are not aware of many conservation efforts undertaken at their school (especially for energy and water conservation). Conversely, for a number of students at many of the schools there are conservation efforts going on that the administration is not aware of. It appears, then, that schools have greater resources at hand for acquainting students with conservation and practicing conservation skills than they currently employ. But better utilization of school resources - more activity - alone is not expected to bridge the educational gap and promote favorable outcomes in environmental behavior. We presently lack understanding on 1) which school activities are effective in meeting which motivational needs of which students; as well as 2) which perspectives, policies, practices and opportunities in conservation at the school lend authenticity to the cultural landscape.

The evident student differences but shortcomings on information on schools, the above mentioned considerations and a moderate sense of pride have lead to further investigation into the matter. The course I attempt to take here addresses the claim that school environmental conservation is an evident part of environmental education but one which is not reaching the students much of the time. Education for the conservation of nature - or for the protection of the environment - is at least in part a question of culture and cannot be left to a special event, an environmental phase or good intentions to get the job done. Special behaviors require special settings if they are to be performed habitually.

The follow-up study attempts to improve upon the deficits of the study on outcomes of environmental education at school with an extension of the educational model for environmental behavior to include the identification and measure of elements of an "ecological culture" of schools.

The integrated model for environmental action (Rost, 1995) makes up the dependent variable block in both the follow-up study and the survey on environmental education outcomes. See Figures 4-5, 5-1 and 5-2. New items on elements of an ecological school culture have been constructed for the independent variables.

In this follow-up study, as in environmental psychology (Bronfenbrenner, 1979), each culture includes both objects and spaces, on the one hand, and practices or ways of doing things - with objects or with people, on the other. Vygotzki (1987) elaborates upon this by adding a behavioral aspect which defines culture as a continual process of exchange between collective and subjective meanings. Reinterpreted for the "ecological culture" of schools, this would imply that:

Now, it is beyond the scope of this follow-up study to try and capture the process of exchange between states of shared and subjective meaning students go through during the learning process. However, we can try and extract a portion of a collective meaning of conservation at school. This I have represented with the collective learning experiences that are offered to the students as well as by standard practices stated by the administration. This collective meaning, accordingly, cannot be operationalized explicitly as one variable. It ought to be, rather, functionally equivalent with the environmental culture of a school.

The new educational model for environmental behavior here attempts to explain the influence of an ecological culture at school on student motivation to act environmentally in terms of four kinds of independent variables.

• ecological features and environmental modifications: includes the physical and perceptible manifestation and design of the school;
• general environmental practices at school: includes behaviors and use of the school, its resources (energy, water, materials, food/ drink, light and so on as well as persons);
• student participation opportunities: includes educational experiences geared at student participation in conservation activities at school; and
• general standing toward conservation at school: includes items that characterize the school’s role model function in conservation.

It is expected, therefore, that:

1. a school with a marked "ecological culture" has identifiable structural cues; and
2. the "ecological culture" is reflected in its students in terms of knowledge (of conservation at their school), student environmental activity there and, ultimately, environmental motivation such that:

1. environmental practices at school, when a part of the school culture, or habitualized, reduce the aversions and obstacles in performing them;
2. a pronounced ecological culture at school fulfills student needs and preferences thereby promoting the acknowledgment of the necessity of environmental protection, options for solutions, necessary skills, social infrastructure (norms) and, in turn, motivation to act environmentally;
3. participation opportunities in conservation are most likely to be motivational when student participation in the decision and planning process is esteemed and not merely seen as a favor or luxury; and
4. school standing conveys the message that conservation is worthwhile and reinforces practical conservation at the school.

The next sections describe the sample, design and instruments of the follow-up.

5.1. Criteria for the sample

The follow-up study was intended to investigate a way of distinguishing schools which might lead to a plausible explanation for the students differences we observed during the study on outcomes of environmental education at school in 1996.

As it was decided that only twelve of the 54 original schools could be revisited, criteria were needed to choose among them to assure that better information would be collected on their "ecological culture". With the empirical study as my basis, the following criteria were generated for the schools participating in the follow-up study:

• School demography. The central point here is that the kind of location where the school is found can unnecessarily interfere with the sought after influences. In order to compare schools, external "interference" of this kind must be reduced to a minimum. The variation must be held low. Otherwise one runs the risk of measuring a mixed effect. Effects due to unintended and therefore non-describable effects of school location mixed together with the desired effects due to the ecological culture at school. Consequently, the subsample concentrates on twelve schools within cities having a population of at least 75,000 (1) residents in the old federal states (2) only.
• Ecological features and environmental modifications and educational student activity opportunities in conservation at school. Within the framework of the research design, the general condition of environmental education at schools is to be tested for effects on students. Items on this attribute function as independent variables. Within the subsample the variance for this attribute should represent the greatest possible breadth. The only source for this variance we have to go on, albeit inadequate, are previous data on the conservation features and activity opportunities collected in the study on environmental education outcomes.
• School student body composition regarding motivation profiles. The motivation profiles of students and the individual motivation indicators function as dependent variables. Influence due to the ecological culture at school on student motivation competes with other external influences (i.e., the family, media, special events). The relative influence of school in comparison to other sources is expected to be small. Small influences require the measurement of subtle differences in effect and is consequently difficult to detected at all. For this reason, it is important that the breadth of the observed spectrum on student body composition among the schools is upheld. Limited variance for the dependent variables would lead to an underestimation of possible effects.

Table 5-2 compares means scores and range for dependent variables for the follow-up group of twelve schools with the sample of 54.

This table shows mean scores on student motivation of the follow-up group is comparable to those of the sample group. The range is similar for the two extreme motivation profiles but somewhat narrower for the two moderate motivation profiles.

The following section will describe the two instruments (assessment survey and interview) and their development for use in assessing ecological culture at schools. The section that follows will briefly describe the investigative design.

5.2. Investigative instruments for the 12 schools

In addition to measurable health benefits for students and faculty at a school which observes environmentally sustainable practices, I anticipate effects in student learning, motivation and behavior. The results of a study in England on the effects of school grounds (Titman, 1994; Adams, 1990; Department of Education and Science, 1990) suggest evidence significant for student effects on attitudes and behavior. The results of their study indicate that:

• Design and management of school spaces convey messages and meanings to children which influence their attitudes and behaviors;
• School spaces contain "signifiers" for the children which, in turn, reflect the cultural context of their social environment;
• School spaces embody the "hidden" curriculum at schools and has considerable impact on the operation of that school; and
• School spaces and its hidden curriculum is determined by management practices.

"Learning through Landscapes", as Titman’s project is known, suggests that a school and its grounds can be read just as a landscape can. The landscape gives cues on, for example, what behaviors are appropriate and offers students a "training course" of sorts. The students "read" and interpret the school landscape cues and learn.

Whether the school is seen by students or by an observer, the individual elements of the "landscape" result in a unique landscape image for each individual and not in a purely tangible landscape which is identical for all (Hoisl, 1992). Despite this, I require instruments of investigation that can "read" the school landscape in a way that is methodologically sound and practical to use (objective), yet, still relevant to understanding impacts on students. I have found no previously used tool for doing this, so an important stage of this investigation became the development of the necessary assessment instruments.

All data for each school would be collected on site and must be completed within one school day by one person. Students or teachers could not be observed. Two instruments were chosen: an assessment survey and an assessment interview.

The objective behind the assessment survey is to have an instrument which can measure "ecological culture" objectively, quantifiably. It is not the goal of this assessment to be able to say school "X" requires "Y" square meters more green space. Rather this research attempts to identify relevant variables, categorize schools and explain observed differences among students attending these schools. The categories for scoring the assessment survey were chosen to be readily understood, convenient to use and readily transferable for multiple "landscapes". In addition to methodological advantages, the practicality of the assessment is necessary to minimize expenses, effort and time incurred during its use.

But as already mentioned, not only are all cues are not visible cues, they are not necessarily tangible in the sense that they can be seen, felt, heard or smelled (the only means available to the assessment survey). This was the reason for choosing a second instrument, the assessment interview. The interview attempts to get an initial perspective on some non-visible components of the ecological school culture, namely: How are things done there at the school? What are the intended messages that the school administration wants to convey? How do they (intend to) do this? Are there "hidden" messages that may interfere with the intended ones? To ensure that the responses from different schools could be compared, a categorial answer format was provided and strongly encouraged.

Data collected from both new instruments on independent school variables will then be compiled with data collected previously on independent student variables (study on environmental education outcomes) as described in the section on Design.

The next two sections describe the two assessment instruments in more detail.

5.2.1. Assessment survey

If we are allowed to regard the public area beyond the walls of the classroom and within the boundaries of the school property as a defined and designed environment, then this may also be referred to as a man-made habitat or landscape. As a man-made or cultural landscape, the school grounds is the product of part objective part subjective part collective representation. My goal here is to develop an instrument to assess the ecological culture as they are represented at schools.

I have found no documentation on guidelines, previous attempts or instruments for the measurement and evaluation of school buildings and grounds appropriate for this investigation. My concept for the measure and evaluation of school "landscapes" has been developed from work done by Gabriela Leitl et al. (1997) at the "Bureau for Ecological Planning, Environmental Studies and Services" in Bayreuth, Germany.

The interplay and, in specific, the influence of landscapes on people - especially on perception and behavior - is a central theme of the work of city and regional planers. They have the task, among others, to preserve and in some cases restore the uniqueness of landscapes as source for identity and belongingness of the residents and at the same time protect, preserve ("maintain") and develop nature within these landscapes (Abs. 1 Nr. 4 Bundesnaturschutzgesetz, Federal Nature Protection Law).

In 1997 a concept was presented for the measurement and evaluation of landscapes. In general, Leitl’s concept proposes that each landscape has a unique and describable structure. The structure is made up of a set of "carrier" elements more or less bound together in a predictable way characterizing the given landscape. It is these carrier elements that can then be assessed and compiled. On the basis of the assessment, the landscape can then be categorized in terms useful in fulfilling the "main task" or function, in this case, of the city/ regional planner defined by the Federal law described above.

Leitl’s concept is designed to be transferable to various landscapes, but unfortunately not readily so for my purposes. Aside from the fact that the function I am interested in is different, one must bare in mind that the concept from Leitl et al. is applicable to areas generally much larger in scale than school grounds and that consequently, for their purposes, measurement and evaluation takes place at a "lower resolution". A "school" and its property would most likely be recognized and included as an item of measure but not further differentiated. Therefore, an assessment of schools at the level of scale in the follow-up study requires an elaboration on the theoretical structure. (A structure at a level not yet defined in theory and my first challenge.)

In the next few paragraphs I will attempt to make clear how I adapted Leitl’s model for the assessment of schools. The ideas that will be described are:

1. "main task" of a school assessment on ecological culture
2. theoretical structure describing a school "landscape"
3. "carrier" elements and criteria of an assessment survey
4. resulting assessment categories of ecological culture at schools

The first step is to define the "main task" or function of the school "landscape". In search of this definition I found a reference to the "Second Dialog on Environment and Schools in Lower Saxony ("2. Schulische Umweltgespräche in Niedersachsen", Hesse, 1996). Here it was identified that:

This goal, interestingly enough, was previously formulated as the promotion of human and environmental compatibility (Ministry of Education of Lower Saxony, see above). The reformulation records a shift from the goal to preserve environmental and human integrity (details not specified) to the goal to design for a particular setting or habitat. This semantic refinement suggests a definite degree of development of the idea that external physical surroundings are malleable and therefore can be designed intentionally to effect a particular and desirable influence on student behavior.

So, even though one school landscape is much like another in that they both contain similar objects such as sports fields, lawns, paths, vegetation, play equipment or biotic, abiotic, and man-made "micro", "meso" and "macro" structures, they can hardly be considered identical. Because they are not identical, there must obviously exist alternative expressions for these objects resulting in the unique "landscapes" we experience. To "read" how well the school landscape fulfills its task, the individual objects are organized within a structure described in the following section.

5.2.1.2. Theoretical structure of the school landscape

Similar to the idea of "well-being", the purpose of the school grounds is repeatedly described as the fulfillment of student needs or the fulfillment of functions (Brunner, 1997; Moore, 1980, Bastian, 1991). This subject has already been addressed in the Introduction and Historical Background of this report. The school grounds, alternatively, can be viewed as a collection of individual activities which fall into one of several categories. The activity categories that were chosen and serve simultaneously as "carrier" elements of the function of the school landscape are:

• quiet areas for being alone or gathering;
• structured and non-structured activity areas;
• areas for nature and environmental experience; and
• areas for movement between other areas.

Each school is assessed for a structure based on the prevalence and character of these five "carrier" elements.

5.2.1.3. "Carrier" elements and assessment criteria

In contrast to regional planning, it cannot be individual topographical or geographical elements alone that make a (school) landscape what it is. In the case of schools we are primarily interested here in cumulative effect of all elements for the entire student body. The cumulative effect may manifest itself in different ways and result in influencing motivation differently for different kinds of kids.

I have described the school landscape in terms of four "carrier" elements which characterize distinct student activity needs. Table 5-3 typifies the indicators used to identify these types of activities, or "carrier" elements.

It is possible for this assessment survey but hardly helpful to count individual trees or keep track of the square meters of lawn etc. There is certainly, for example, a minimum space requirement for the functionality of areas with respect to number of students or age or... This subject and similar ones go beyond the scope of this research at this time. Instead, the indicators ought to be a measure of the schools ability to meet student needs.

The degree to which the school landscape accommodates each element is a measure of the quality of student experiences in satisfying their activity needs and is ultimately and ideally a measure of the well-being of students there.

"Accommodation" is not just a matter of a checklist indicating that the indicator is there. "I see it" or "I do not". Rather, human perception - of students and assessor alike - includes more than the sense of sight alone and thus the influence of a landscape on people is not merely visual. Not to mention the fact that visual representations of the same element will be perceived differently by different people. Particularly for young people, other sources of stimulation are important. These include the acoustic, olfactory and tactile aspects of experience. For example, it is not likely that the sound of birds has the same effect as traffic sounds or the smell of wild flowers compared with exhaust fumes or the feel of earth compared with concrete. Ideally, the assessment ought to reflect this.

Factors to be considered for the assessment of the four "carrier" elements have been adapted and are depicted in Table 5-4.

In this investigation both individual as well as cumulative, or contextual, effects are regarded as instrumental in creating a landscape and influencing student motivation. This table shows that the elements of school landscapes are assessed for two types of attributes. Included are attributes which assess the element on its own merits or inherent attributes as well as those which assess the element in context or in relationship to things external to the element.

Differences in interpretation are unavoidable, however, they ought to be quantitative and not qualitative in nature and, therefore, to avoid unnecessary differences in interpretation clearly quantifiable categories for assessment are necessary. The assessment criteria are those qualities identified as relevant to the "main task" (see section 5.2.1.1) put in quantifiable terms. Particularly this point has been a special challenge in this investigation.

I am striving at a measure of "things" that for the sake of the kind of statistical analysis I would like to conduct must be recorded in objective and quantifiable terms. Historically, though, these "things" have been recorded - if at all - more often than not subjectively or qualitatively. This data has limited value for empirical statistical analysis - a problem which I am now trying to overcome.

For each attribute there is a defined set of criteria which enables the classification of the indicator into one of three simple categories which characterize its prevalence as "high", "average" or "low". To give an idea of this a very small portion on the list of assessment criteria is shown in Table 5-5.

Each "carrier" element has its own set of criteria for the assessment of the attributes. This table presents only a very small portion of a very extensive list of criteria used to assess the "carrier" elements. During four practice runs it was determined that this strategy was far too intensive, in fact redundant, and hard to manage in relationship to the results it produced. It becomes very complicated, for example, to assess attributes if the elements topographically overlap as is frequent for "non-structured activity" and "nature experience" areas.

To maintain clarity and preserve the informative value of the assessment, it was decided instead to re-focus on six general dimensions of an ecological school landscape. Indeed these dimensions do cover many of the elements and attributes previously described but using dimensions in place of "carrier" elements, I have attempted to reduce the amount of assessment attributes to a minimum and at the same time improve the usefulness of the information. The dimensions of school landscapes which were finally used are:

• Layout
• Design
• Surfaces
• Condition
• Nature experience
• Nature and environmental conservation

I would like to make some overall remarks on the individual on these dimensions or traits as they will later be called.

• Layout: The trait schools have least influence over unless they are in the building stage or are planning large scale renovation. It is unlikely that a school can easily acquire land, remodel or move a building structure.
• Design: This can be a sensitive issue if the school values its image on the outside. Alterations are expensive and time consuming but the results are tangible and noticeable to all. Generally quite a lot of energy and money is put into the maintenance of the grounds. This responsibility lies in virtually every instance not with the school principal or education ministry, rather, with the municipality and is a source of frequent conflict.
• Surfaces: In the late 60’s early 70’s it was commonplace to tarmac the school grounds surface. Generally, only schools with very large grounds survived with more surface types. Only schools who have gotten special additional funding have been able to afford the costly removal of asphalt.
• Condition: It was originally intended that this trait, in part, would reflect a confirmation of student environmental motivation through their caring or uncaring behaviors on the school grounds. However, it was discovered that the large tarmac surfaces and wind prevented accurate measurements of the "indicators" for the various areas of the school grounds. (The mess always ends up trapped in the green stuff - no matter where it was originally littered.) In addition, some schools are frequently used by other individuals and groups after school hours. It was not possible to control for the actions of these people.
• Nature Experience: It was extremely difficult to find satisfactory indicators and categories for this trait. Although I was very conscious about not overvaluing, for example, a group of trees in comparison to, say, one tree, value categories were very difficult to quantify. At some schools it was obvious that kids made the best out of next to miserable natural conditions. It is not possible for an outsider to judge the character or value of student experience accurately. This trait ought to represent experience potential and intention on behalf of the school, not what the kids make out it.
• Nature Conservation: These items reflect typically recommended conservation measures for schools. Although at many schools it is likely to find several of these items, it was continually disturbing to see how little imagination, thought and pedagogical ingenuity has been invested in bringing these events to the minds, hearts and hands of the students
• Environmental Protection: This trait is the most underrepresented of all traits for our sample. It is, fortunately, growing in popularity.
• Other: Some schools have made special efforts regarding their grounds but ones which are not immediately relevant to the ecological culture yet may be worth noting. For example, schools that have made accommodations for student needs they have recognized as important (i.e., a skate boarding arena, graffiti wall, practice course for driver education).

The attributes used to evaluate these dimensions/ traits remain for the most part identical to those worked out previously for the "carrier" elements, only they have been condensed in number and regrouped for the appropriate dimension. A summary of the dimensions and their corresponding attributes is given in Table 5-6. A translation of the original assessment survey is included in the Appendix.

This table depicts the final dimensions and their attributes used in assessing school landscapes. At this stage the assessment is complete. An interpretation of the assessment survey results are discussed in section 5.4.1.

5.2.2. Assessment interview

The assessment survey is concerned with variables on ecological features and environmental modifications of the school grounds, where as the assessment survey is concerned with the remaining variables on:

• general environmental practices at school;
• student participation opportunities; and
• general standing toward conservation at school.

It has been mentioned already mentioned, that in order for it to work, education and learning for responsible environmental behavior ought to be going on everywhere as a normal part of doing things - a standard and not a hobby. But such practices and student involvement take on a variety of forms that are not readily identifiable - not even to the trained eye. Furthermore, the assessment survey is not designed for observing students and faculty directly. Thus, I have not found it possible to "see" general practices or student participation in conservation at school in the sense that I can identify it in a form suitable for the assessment survey (5.2.1). Therefore, a second instrument, the assessment interview, was used in the identification of ecological culture at schools. The interview is intended to collect data on the ecological culture which cannot be observed or might be potentially "overlooked". Examples of questions I cannot answer with the assessment survey are: How are things done there at the school? What are the intended messages that the school administration wants to convey? How do they do this? (through student participation, management and organization, as well as through design) Are there "hidden" messages that may interfere with the intended ones?

In section 5.2.1.1 two conditions were named that would be necessary if student well-being and their taking on responsibilities were to be genuine goals of the school. These goals can only be attained when:

1. the school takes on a positive role and lifestyle which includes the design of the school and person-to-person relationships as well as learning content, methods and technical support; and
2. the school upholds dialogue and cooperation with politicians, education agencies and other institutions on the needs of youth.

School practices, student participation, "lifestyle", student-faculty relationships and dialogue with others, then, formed the core themes of the interview which contained more than 20 in depth items. Questions that summarize the themes addressed by the assessment interview are:

1. Who has initiated conservation efforts concerning ecological features and environmental modifications at the school?
2. Through which means do students experience the meaning of nature and environmental conservation?
3. (How) Are the grounds and the buildings integrated in nature and environmental conservation?
4. What possibilities for conservation are offered to students?
5. What is student behavior in conservation like, in general? Which "effects" of environmental education have been observed in students? Where does work need to be done?
6. How is all this organized (conservation/ student participation) at the school?
7. How is conservation at the school presented to the outside?

A small sample of the assessment interview questions and the answer format is given below. This is an approximate translation. The original questions and answer format are more elaborate. This answer format represents the resulting categories.

The interview was held with the principal of the school or someone appointed by them and was geared to last not longer than one class period (45 minutes). To assure that the responses from different schools could be compared and facilitate completion, the interview was held in a structured format. Questions were pre-formulated, similarly presented and a categorial answer format was offered and strongly encouraged to the interviewees. All responses were recorded on paper immediately and simultaneously tape-recorded. The assessment interview is found in the Appendix.

5.3. Design

Approximately one year after a survey on student outcomes on environmental education at school was conducted, a follow-up investigation took place. The follow-up was not intended to re-test the schools, rather, it was intended to gather more data on the independent variables on traits of an ecological school "culture". Twelve of the original 54 schools were revisited. The schools were chosen from the original sample such that the data from both studies could be compiled together. Significant changes at the school between testing times were accounted for.

Schools were considered for the sample after analysis of three factors 1) demographics, 2) conservation and student activities and 3) motivation subpopulations. A more detailed description is given in section 5.1. I requested the voluntary participation of twelve specifically chosen schools. No school declined cooperation. A day was arranged during early Spring for the assessment survey and interview. The entire visit ran on average five hours.

Students or teachers would not be observed or interviewed in the follow-up study. Instead, four new categories of variables on school traits were under investigation:

• ecological features and environmental modifications of the school grounds: - the "picture". This part attempts to describe a school in its authenticity as a habitat or landscape for conservation and, thus, a landscape for learning. The ecological "culture" of schools is read like a landscape using the "landmarks" (physical objects) and "land use" (behavior) as indicators. Indicators are, for example, natural features or habitats accessible to students in their free time. Data is collected using the assessment survey.
• general environmental practices at school: The data ought to reflect more than the physical indicators and include indicators which characterize the behavioral content of culture, or the "way things are done" there. What type of role model and lifestyle is portrayed at the school? An indicators may be, for example, a system for collecting sorted waste. Data is collected using the assessment interview.
• educational opportunities for student participation in conservation at school - "participation". This variable group tries to characterize to what extent the administration encourages and supports student involvement in conservation at the school through activities for further development of the school grounds or policies or in decision making there. Indicators are, for example, an established contact person or group at the school which handles environmental concerns of students and the school, extra curricular activities, an environmental plan, and other organizational factor. Data is collected using the assessment interview.
• general standing toward conservation at school: includes items that characterize the school’s role model function in conservation as the administration sees it; their opinion on the function of the school grounds in student education of conservation; and the resources they have at hand.

The data on these new variables were collected using two methods (Section 5.2). Data on the ecological manifestations were collected during an assessment "inspection" or survey of the grounds and buildings using an assessment criterion catalogue and during an assessment interview with the school principal.

The criterion catalogue consists of over 40 items (found in the Appendix), is intended to be carried out both indoors and outdoors and includes aspects, such as, layout, design, surfaces, condition, nature and environmental conservation and student access to natural settings.

Data on practices and in particular on student involvement or "participation" as it seems to be referred to more often today, was obtained through more than 20 items during interviews with the school principals or someone appointed by them. The principals were not given the questions prior to the interview. They were, however, given an oral report on our major results from the original study along with a detailed summary of the results of their school.

The design of the assessment is shown below.

In its simplest form, results from individual schools are expected to take the form of one of the quadrants in Table 5-8.

This table depicts a generalized range of results expected from the twelve schools. It is anticipated that schools will tend to score high or low for each of the assessment components (i.e., features or student participation). The resulting profile (one of four) is interpreted as the ecological culture of that school.

The "old" data on the dependent variables on student motivation can then be reinterpreted using, now, the additional, newly collected data on the independent variables on ecological school culture. It is anticipated that the new instruments are more sensitive and the data better such that the ecological school culture can now be reflected in a characteristic make-up of student subpopulations on environmental motivation typical for that ecological culture profile.

For example, it is expected that there are schools that score proportionately high for most or all measures of ecological culture. These "strong" schools, it is postulated, ought to differ from other schools in a number of ways. For instance, proportionately more students at these schools ought to belong to the highest scoring environmental motivation group, have higher than average knowledge of conservation going on at their school and report more experience in environmental activities at (and away from) school than the remaining students.

Or, another category of school is anticipated which scores comparatively low for ecological culture. It is postulated that a significantly larger portion of the student body at these schools will belong more likely than not to either of the lowest scoring motivation groups, score significantly lower for their accuracy on conservation at school and report lower activity at and school than the others.

Similarly there are two potential mixed forms of school ecological cultures that may be encountered. Schools that either score high for features and low for participation. These I have labeled "standard" as I feel this is the constellation characteristic of most schools. Or there may be schools that do not put emphasis on conservation features at all, but nonetheless give the student participation process high priority. These I have named "other" to reflect that these schools perhaps share a similar value for education through participation, but have chosen an alternative "landscape". Instead of "conservation" underpinning participatory activity at school, it could be, for example, "social democracy" or "health".

The next section describes the successes and failures I encountered investigating ecological culture at school and its effect on student environmental motivation using the two assessment instruments and data previously collected.

5.4. Results

Data on schools were collected in 1996 as part of an empirical study on outcomes of environmental education at school and one year later, in 1998, as part of a follow-up study on ecological school culture. The follow-up study comprised twelve of the original 54 German public schools and was intended to produce better data on the independent variables on ecological cultures at schools. Although data on students were not collected during the follow-up survey, the data on and from students collected from the survey in 1996 could be combined with the data collected in 1997. The combined databases include the twelve schools and about 590 9^th graders. The improved database is intended to help explain school effects of ecological culture on students, in particular, on student environmental motivation.

The follow-up investigation was carried out using two instruments: an assessment survey and an assessment interview - both of which were specially developed for this task (section 5.2). The next sections give a general description of the schools in the sample, describe structures found among schools and then propose traits of ecological culture at schools that may explain differences in student motivation.

An outline follows of the results we shall be seeing:

Descriptives:

1. Description of survey variables on features and modifications in conservation and resulting categories; and
2. Description of the interview variables on conservation practices and resulting categories. And on

Structures:

3. Comparison of mean scores for the dependent variables (student activity at school, conservation knowledge and environmental motivation) for two extreme school groups; and
4. Description of four empirically derived ecological cultures and test of the original hypothesis.

A review of the investigative design reminds us that the twelve German public schools in the follow-up study are all found in cities with populations of at least 75,000 in four "old" federal states ("West Germany"). They represent a broad spectrum of conservation practices and student activity opportunities found at public schools here.

School participation was voluntary. No school declined cooperation. A day was arranged at each school for both the assessment survey and interview. The entire visit ran on average five hours. Students or teachers were not observed or interviewed in the follow-up study, rather, data was collected on the ecological culture at schools (independent variables):

Now we will look at the individual variables collected by the assessment instruments.

5.4.1. Assessment survey variables

The assessment survey consists of over 40 items (found in the Appendix), is intended to be carried out both indoors and outdoors and includes items on aspects such as: layout, design, surfaces, condition, nature and environmental conservation and student access to natural settings.

Although data was collected both inside and out (on school grounds and on school buildings), data from the assessment survey inside the schools did not add particularly to my understanding and, as a consequence, have been left out here for the sake of simplicity. Table 5-9 below is an approximate translation of the assignment categories used in evaluating the traits of ecological features at schools.

In this table the three assignment categories should represent a range of prevalence reasonable for each trait. The category level indicates the anticipated value that trait in that form has on ecological culture.

For further evaluation of the criteria on traits, it is important that for each variable all categories are represented in our sample. Traits which have missing categories must be eliminated from the assessment catalogue - at least for this sample. Table 5-10 gives a summary of the number of responses for each response category.

The next step in the evaluation was, for each variable, to assign the schools to groups according to the level at which that variable is present. For traits measured with only one item, the level was predetermined by the categorial answer format. For traits measured with multiple items, scales were built and the upper third was assigned to Level 3, the middle third was assigned to Level 2 and the lower third was assigned to Level 1.

This table shows that each category is represented at least once suggesting that the category levels are reasonable in representing a range of trait prevalence among the twelve schools in the sample.

Due to the small sample size and the effects of aggregation, as has already been mentioned, I cannot expect my data to verify theory. I will not see strong effects from twelve aggregate scores and I cannot rely on significance tests due to the small sample size. I can still, however, compare these schools among themselves and see which traits, if any, seem to be associated with student effects. I can do this by comparing student data (student "effects") for the extreme groups - Levels 1 and 3. The main objective, then, was to build school groups that are distinguishable relative to each other and not to distinguish schools based on a purely quantitative measure or raw score.

A similar procedure was followed for the evaluation of the variables collected with the assessment interview discussed next.

5.4.2. Assessment interview variables

Data on general environmental practices at school and student involvement or "participation" in conservation (as it seems to be referred to more often today), was obtained through more than 20 interview items. The interview was held with the school’s principal or someone appointed by them. The principals were not given the questions prior to the interview. They were, however, given an oral report on our major results from the original study along with a detailed summary of the results of their school.

Almost all principals appreciated this very much and seemed to gain a better understanding for what I mean when I say "conservation", "student participation" and "motivation". I feel this preparation improved our communication for the interview.

The interview lasted between one and two hours depending on how communicative the principal was. Sometimes interruptions caused delays. But for the most part all principals were very cooperative about sharing their experiences and following the prescribed answer format. All twelve interviews were carried through to completion and without hindrances worth mentioning. A very condensed summary of the interview variables has been translated and reconstructed in Table 5-11.

Similar to Table 5-8 on traits of ecological features collected with the assessment survey, this table gives the items and levels for many of the interviewed traits on general practices, student participation opportunities in conservation at school and additional items on the administrations’ standing to school and conservation ("o") in general. This is not the interview and answer format. This is a summary of the resulting categories after responses were evaluated and the three levels had been assigned. The assignment categories, in this case, do represent the range of prevalence for each trait in our sample. The category level indicates the anticipated value that trait in that form has on ecological culture.

The pre-structured answer format was very helpful for the interpretation of the responses as twelve interviewees can often lead to at least thirteen different structures for interpretation. All principals adhered to the given answer format with the exception of one who did not want to be pinned down on anything.

This table depicts the assignment of schools to three levels for traits on general practices (gp), student activity opportunities (sa) and on the general held by the principal on school and conservation (o). Several problems with the data from the interview surfaced. One item was not answered by all school and only two categories were necessary for a second item. Perhaps the most unsatisfactory result is that for many of the items the distribution of the twelve schools among the three categories is poorly balanced. Although these categories may represent school reality, comparing the mean scores of categories whose members are so disproportionate may be unreliable measures of these otherwise reasonable categories.

The next section brings together data from the schools (independent variables) and data from the students (dependent variables).

5.4.3. Comparison of student mean scores for extreme groups

Each school has been assigned to one of three levels for each trait representing ecological culture. From each school there is information from students on 1) their activity preferences at school, 2) their knowledge of conservation at school and 3) their environmental motivation. The percentages for each student activity, motivation and knowledge group per school have been calculated (mean score for group "X") and are compared for the extreme groups "high" and "low". For example,

This table shows that for each trait of the major components (features, practices and student participation and standing) the two extreme groups of environmental culture are compared. The "sample mean" indicates the mean value for the student variable for the entire sample of twelve schools. In general, for each trait (independent variable) if scores for the "high" group are higher than for the "low" group, then the student mean scores improve (dependent variable) with the prevalence of this trait. For example, for "Trait1" the mean score for "low" is 0.3 and increases to 0.5. This result suggests an increase in the number of students who "do a bit of everything" at school as the prevalence of "Trait1" increases. "High" schools have on average 50% more students "doing a bit of everything" than "low" schools for this trait. On the other hand, scores indicate that "Trait2" does not favor students who are generally most "active" in conservation at school.

Table 5-14 shows only a portion of the comparison between the extreme groups of ecological culture traits. The complete table is found in the Appendix.

This table, although enormous, gives us the first impressions of effects of school ecological culture on students. Just a glance reveals that many of the traits are associated with variances in student numbers. For example, both design traits (way up at the top of the table) are associated with an increase in the number of students with knowledge scores in conservation at their school of better than 50% and a decrease in the number of students with scores between 26-50%. The number of students with the poorest knowledge scores remains unaffected. I conclude that this is an indication that the design of school grounds - as an integral part of the ecological culture of a school - may influence student knowledge in conservation at their school positively.

5.4.4. Dimensions of ecological culture and students effects

However, instead of continuing trait for trait for all independent variables, I think it makes more sense to get an overall picture on how student groups (dependent variables) are effected by each of the dimensions of ecological culture.

First, all items for a given independent variable, for example, the two "layout" items for the dimension "features and modifications", were compiled and the deviation from the total mean was calculated for the "high" and the "low" groups each (Table 5-15).

After the deviations for "high" and "low" groups were calculated, the overall effect of the trait could be calculated in terms of the percent change of students belonging to a given group.

This table depicts, for example, that as the design of the school grounds improves the number of students who are least active increases (unfortunately) by 24%. The play surfaces account for a reduction of students primarily involved in clean-up activities of 35%. In general, improved features of the school grounds is associated with a 13% decrease in students doing primarily clean-up activities.

Then this same procedure was repeated for all variables for the remaining dimensions on ecological culture. The four dimensions of ecological culture found at schools and their overall impact on student variables are summarized in Table 5-17.

This table summarizes the general effects on students of the four dimensions used to measure ecological school culture. Ideally, I think one could agree that if ecological culture is a constructive influence on students, then it would be reasonable to expect that the four dimensions of ecological culture ought lead to a decline in the numbers of students in the groups: "little or no", "clean-up", "0-25%", "26-50%", "MOT4" and "MOT3" and an increase for the student groups: "everything", "active", "51-75%", "76-100%", "MOT2" and "MOT1". A description of the actual results follow.

5.4.4.1. Student activity

Effects of the four dimensions on student activity are encouraging but not as straight forward as was hoped. In addition to traits which ought to favor student environmental activity in conservation at school and do, there are other traits with only minimal effects and still others suggest an opposite effect. Expectations were that the four dimensions of ecological culture ought lead to a decline for the groups: "little or no" and "clean-up" activities and an increase for "everything" and "active" student groups. The following table summarizes the observed effects on student activity for each of the four dimensions of ecological culture.

Figure 5-4: Effects of four dimensions of ecological culture on student activity groups

The results can be summarized as follows:

• "Active" student groups, as expected, show an increase due to improved dimensions of ecological culture, except for general activity opportunities at the school. The opportunities reported by the principles appear to lead to a 17% decline in active students. (Why do only these students no longer feel the need to act? There is something about the opportunities that deter only these students; but what?)
• Students who "do a bit of everything" increase in number for all dimensions just as it was expected. This result is quite satisfactory.
• I anticipated that students who find themselves doing more or less only "clean-up" activities would decline with school improvements. This is so for two dimensions: features and modifications and general conservation practices. The remaining two dimensions have next to no effect on the size of this group.
• Finally, it was anticipated that the "least or inactive" students would also decline with school improvements. The two dimensions general activity opportunities and general standing clearly show a suppressing effect. The dimension general conservation practices has a suppressing effect, too, but very weakly so (3%); and features and modifications show a weak but positive effect on this group (5%).

Effects on student knowledge of conservation at their school are evident and as hoped pleasingly straight forward. Expectations were that the four dimensions of ecological culture ought lead to a decline for the groups "0-25%" and "26-50%" (correct responses) and an increase for the student groups "51-75%" and "76-100%". The following table summarizes the observed effects on student knowledge for each of the four dimensions of ecological culture.

The results can be summarized as follows:

The data on effects on student environmental motivation are the most tricky to interpret. It was hoped, as has been suggested many times, that a positive environmental culture would encourage motivation and, therefore, the number of students with the highest motivation scores would increase (MOT1 and MOT2) while the number of students scoring low would decrease (MOT3 and MOT4). Support for this conclusion is not consistent. Rather, the student motivation types responded, in part, unexpectedly to the dimensions of ecological culture as shown in the following figure and described below.

• As anticipated, as the dimensions of environmental culture improve, so does the number of students with the highest motivation scores ("sensitive optimists", MOT1) - except for the dimension general conservation practices at school. This dimension leads to an apparent 9% decline for this motivation group.
• Students of second lowest motivation ("self-doubters", MOT3) are also affected as anticipated - numbers decrease - but this is a weak effect compared to effects on the other three motivation groups. Again the dimension general conservation practices at school produces an effect contrary to expectation. This dimension leads to an 11% increase for this motivation group. This is a significant disappointment when considering the role general conservation practices at school ought to play in "procedural knowledge" and student acquisition of self competence (see Section 3.1).
• The remaining students, those with the lowest motivation ("skeptics", MOT4) and second highest motivation ("go with the flow", MOT2) scores, behave consistently to improvements at school but the resulting groups do not correspond with projections. With increasing ecological culture, the number of students in the second highest motivation group (MOT2) decreases and the number of least motivated students increases ("skeptics", MOT4).

It was originally thought that as the ecological situation at school improves so would motivation, i.e., there would be a "motivational" shift of students to the right - for all groups. Instead of MOT4 shrinking, this is evidently not the case. MOT4 grows significantly, meaning that a number of students from other groups (which?) have been demotivated. This represents a shift in the wrong direction. In favor of the shift is the growth of MOT1. It cannot be proven with this data, but it almost appears as if MOT1 grows at the expense of MOT2, at least in part. Yet the gap made potentially by students moving from MOT2 into MOT1 is not filled by of MOT3 students experiencing a motivational boost. Sure, the number of MOT3 students does decline but not enough. Furthermore, these students most probably did not land in the MOT2 group as it has shrunk so much.

So far I have established that the dimensions chosen to measure ecological school culture do indeed effect students constructively - for the most part. Now would be a good time to ask what this means for schools as a whole. Does an ecological culture exist? Is ecological culture as a whole useful in explaining student differences at schools? Now I will address these questions and conclude the chapter on results.

5.4.5. Four empirically derived ecological cultures and test of the original hypothesis

Results from the previous section do suggest that the measures of ecological culture may indeed influence student environmental activity, knowledge and motivation. Now I would like to return to the major hypothesis proposed by this investigation, namely that:

• design their grounds for ecology, reflect current cultural practices in nature and environmental conservation,
• encourage student participation in related decisions at school and
• have a positive standing toward conservation at their school,

1. greater knowledge about conservation at their school and
2. student activity in conservation activities at school resulting in
3. higher environmental motivation in these students.

For example, it is expected that some schools will score proportionately high for most or all measures of ecological culture. At these "strong" schools, proportionately more students ought to belong to the highest scoring environmental motivation group, have higher than average knowledge of conservation going on at their school and report more experience in environmental activities at school than the remaining students.

At schools whose scores are comparatively low for ecological culture a significantly larger portion of the student body is expected to belong to either of the lowest scoring motivation groups, score significantly lower for their accuracy on conservation at school and report lower activity at school than the others.

Two intermediate ecological cultures remain. But enough on expectations. Let us get on with the actual results. (See the section on Design for a review.)

A cumulative mean score for each of four dimensions on ecological culture was calculated for each school. Each trait is weighted equally. The results of these calculations are recorded in Table 5-18.

The twelve schools were then ranked for each dimension of ecological culture according to the mean score, after which the "better" half was assigned "+" and the "lower" half was assigned "-". Four groups of ecological culture could be identified.

The four types of ecological culture found among the twelve schools in the sample are identified in Table 5-19 below.

• The group labeled "strong" is characterized by having a high ranking for all four dimensions.
• A second group, "eco design" is characterized as having a high score for the dimension on features and modifications on the grounds. This data, and this data alone, was collected with the assessment survey. Data for the remaining three dimensions were collected using the assessment interview.
• Schools belonging to the third group, "eco practices", have at least one high score for a dimension related to "practices" measured with the assessment interview.
• Schools belonging to the final group are labeled "weak" because they do not score high either of the two dimensions.

It certainly would have been possible to group the six schools in the middle differently. This was tried. In addition, the data was cleaned of items that did not conform to desirable effects on student activity and conservation knowledge and environmental motivation. But none of these attempts changed the end results on student mean scores for the four ecological cultures depicted in Table 5-20 dramatically. So, for this table all items for all dimensions were used. A choice for the middle categories was made primarily to look at the difference between schools that score high for the assessment survey and those schools who score low for the assessment survey (but score high for at least one dimension measured with the assessment interview).

Table 5-21 shows these results once again but as deviations from the mean to indicate the overall trend of effect due to the dimensions.

The results indicate that, for example, at schools with a "strong" ecological culture 3% fewer students participated in "little or no" and 5% fewer in "clean-up" activities than the sample means. Conversely, "active" and "a bit of everything" students increased by 5% at these same schools. Although the numbers themselves are not earth shattering, they nevertheless support expectations for schools having a "strong" ecological culture. The following figure depicts the information on all four ecological cultures graphically but for student activity only.

• With respect to the mean for all students in the sample, the eco design group is the only group which has a negative effect on student environmental activity at school. "Little or inactive" students at these schools increase by 8%. Both the strong and the eco practice group lead to an observable decrease. Surprisingly, the weak group did not make much of an impact at all.
• Both the strong and eco design groups lead to an observable decrease of 5% in the numbers of students who feel they only partake in "clean-up" activities. Surprisingly, the eco practice group lead to an observable increase of 8% above the mean. Again the weak group did not make an impact.
• The influence of ecological culture on students who "do a bit of everything" is minimal. Only the strong ecological culture appears to have an effect, and this is positive (4%). No great changes were observed for the remaining three groups
• The most "active" students appear to be encouraged at schools with strong ecological cultures and discouraged by eco design and eco practice cultures. The weak group has no negative influence, data suggests the contrary is perhaps true.

The following figure depicts the information on all four ecological cultures graphically for student knowledge of conservation at their school.

• Students with the poorest conservation knowledge are not affected dramatically by the kind of school they attend, although more of these kids attend schools with a weak ecological culture. There is a minimal tendency for the strong and the eco practice groups to suppress numbers of their students in this category.
• The strong group leads to an observable decrease (10%) in the numbers of students who score between "26-50%" correct. Eco design has no apparent effect. Both eco practice and weak groups show an increase in the numbers of their students in this category (6-7%). This may come as no surprise for the weak group, but it is certainly an undesirable result for schools where practices are emphasized.
• Students scoring between "51-75%" benefit most if they attend a school with a strong ecological culture. Numbers increase by 9%. Eco practice schools experience a slight decline (3%) and weak school show a significant drop (8%) of students in this category.
• Students most aware of conservation at school do so relatively independent of the school they attend. Indications that ecological culture affects these students are minimal. The weak schools show a slight drop (3%) for students in this group.

The following figure depicts the information on all four ecological cultures graphically for student knowledge of conservation at their school.

• The "skeptics" (MOT4) appear to be just this. Their numbers increase at schools with the strongest ecological cultures (by 5%) and decrease at schools where it is weakest (by 7%). An increase of 8% was observed for the eco practice group. The eco design group looses 6% of these students.
• The "self doubters" (MOT3) increase for the eco practice group only (7%). (This is not the desired result.) Their numbers drop but not by much for the three remaining cultures.
• "Go with the flow" (MOT2) students appear not to be favored by ecological culture. Eco design schools produce a minimal effect on these students. The strong and eco practice groups see a decline of these students by 6% each. Only at the weakest schools are these students observed in increasing number.
• "Sensitive optimists" (MOT1) are, for the most part, true to their name, too. Strong and eco design schools are found to have 5-7% more of these students than other schools. Weak schools show a slight decline. However, eco practice schools are found to have 8% fewer highly motivated students.

The following summarizes the anticipated and actual results obtained for student effects for each of the four ecological cultures of the sample.

Table 5-26 summarizes the results on student effects that were observed in the four ecological cultures identified at the schools in the follow-up sample.

This table is a concise summary of the four ecological cultures that were identified in the follow-up study and their distinguishing effects on students. When the actual results found are compared with those that were anticipated, there are a few surprises:

• surprise #1: Student knowledge on conservation at school increases with the prevalence of ecological culture. The "better" the ecological culture is, the better the conservation knowledge scores are. The physical manifestation of ecological cultures on the school grounds appears to have a more significant and positive effect on knowledge scores than do aspects of conservation practices.
• surprise #2: The effect of ecological culture on student activity is more complicated to interpret than for conservation knowledge. (The activity profiles are not to be understood as activity preferences of students.) The design of the school grounds alone does not positively affect students. Students respond better to practices in conservation there or practices in combination with design features.
• surprise #3: Student motivation is definitely more complicated than I had anticipated. A strong ecological culture does not necessarily sustain motivation, in fact it is detrimental for at least some students. There is an indication that something is going on at these schools which has, in addition to a positive influence, a very negative streak for some students. Motivation seems to improve with the prevalence of conservation design, in particular. The most negative effect on motivation has been observed at schools with an emphasis on practices but where design elements are weak.
• surprise #4: Low or in fact no prevalence of an ecological school culture has no observed detrimental influence on either student activity or on student environmental motivation. Students at schools with "weak" ecological cultures show at least comparable, if not better, activity and motivation to act environmentally than students at schools whose ecological culture is somewhat better (eco design and eco practices).

This concludes the section on results. In the next chapter the results will be discussed for their relevance in education and in research.

Home

Contents