.‘¢.:‘.‘ ...w.»;:,v' r . , . g... .h‘vglm’g' .1 1 _,.,,..:.~'.:.‘. .'I :1 ""'v|v‘_ ;' .1; v'(: w . ”1"! In. I ‘ 1 115.511- ”fl MIL ._.;1u-4'1\H" “I“ m“? ‘ J-..;.:..';.J:1‘.. .,, .1. , H r ,. 1U; :17! '11 1. ,. .',!,1":,3"‘_‘“ ' . ..,.r .. 2;: u' L, '- "ms?“ .. 1. ,j: I ‘1' I“ 1,. .ann a... :w 1 W." 1;“. 1n ‘ I; ‘ . 1—15.11" ‘ I" . ‘12:; . ""1. a "II 1 ""’ 4 1’; . “'1 . 1 . ' in":;.:r” "" 1 11} 7 4»! - q . 1 1 ’5‘.1‘"‘5-"' I‘M-"'7‘ I 1W A. w» ‘ ~ , 1.1- 1.. 3-1. .1 cm A rm 3.! '3. “,1, 5m. 1 - 1 ,._.1 v- ';..J~ A . .‘ _ .gfi‘L—vhww # A n- fi1'1‘_ 1 81m: -:j' . n -. rq-v-‘WVM *1:- ‘13.”..7 a ”)1 1-" "2:29.11; ." [Lynn rv ‘1'“ ..-: ".3" ”'w'r- I .., "1— J ... ,!.,.T..1.‘-. r r" -* :v ”1.111. v‘ _«..1; '- N; w . .11 «v- W ~— nah—5.1V " 3'4 _,.,,.,. :“i‘vr..--n-z--1;~‘: ._., .1"...,....,:11‘:w1""' r 'm. “.,1.41.1~-ur 1-." -' .~ :31 "Fri €1—12qu ”invn; :'!\‘:" ? . '31 .’.-.1 n1""" 01020 5916 This is to certify that the thesis entitled LANDSCAPE PERCEPTION DIFFERENCES BETWEEN AMERICAN AND TAIWANESE STUDENTS presented by Ying—Hung Li has been accepted towards fulfillment of the requirements for M . ' “fidegree 1n 5C1 ence (:2kagéZ/ép7/r/1géégq Date W X/ /?97/ 0-7639 MS U is an Affirmative Action/Equal Opportunity Institution LIBRARY I Mlchigan State University PLACE IN RETURN BOX to remove this checkout from your record. TO AVOID FINES return on or bdore date due. [MUM DJ L_JL_JC:J VFW L_|l l J l—_|| ____|l I ll iii ll MSU In An Affirmative Action/Equal Opportunity Institution WNW-1 LANDSCAPE PERCEPTION DIFFERENCES BETWEEN AMERICAN AND TAIWANESE STUDENTS BY Ying—thg Li A.THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of MASTER OF SCIENCE Department of Park and Recreation Resources 1994 ABSTRACT LANDSCAPE PERCEPTION DIFFERENCES BETWEEN AMERICAN AND TAIWANESE STUDENTS By Ying-Hung Li The purpose of this exploratory study was to test cultural differences in preferences for and familiarity with selected American and Taiwanese landscapes. In addition, subjects were asked to rate their acceptance of various levels of development in the natural environment . There were 189 American students at Michigan State University and 287 Taiwanese students in Universities in Taiwan who served as subjects. Using color slides as stimuli, significant differences in landscape preferences were observed between these two groups for selected landscape types, mountainscapes and waterscapes. A psychophysical model of landscape quality assessment was employed in this study. American subjects preferred the superior position to View landscapes, while Taiwanese preferred the inferior viewing position to watch scenery. The comparison of acceptable levels of development in natural settings across these two student groups suggested differences as well. Americans tolerated more development in the natural environment than did the Taiwanese . Familiarity scores showed a modest but significant positive relationship with landscape preference ratings. Discrepancy in perceptions were recorded across detailed landscape features such as observer's position, water conditions, and the distance between water and observer. The relationship between landscape preferences and acceptable levels of environmental development was both positive and significant . This study corroborated the results of previous research suggesting that different landscape perceptions are associated with different cultural backgrounds (Zube, 1984, Herbert & Kaplan, 1987) . Moreover, subtle perception differences in landscape perceptions provide useful information for landscape design, planning and development . ACKNOWLEDGEMENTS As a foreign student, it is not easy to complete this thesis in English without help. First of all, I would like to express my sincerest thanks to my major advisor, Professor Joseph D. Fridgen. Without his guidance, encouragement, and patience, I could not have finished this thesis so easily. Second, I wish to thank the other members of my committee. Professor Gaylan Rasmussen provided the basic concepts of landscape design and planning, and Dr. Galen Bodenhausen served admirably as my minor professor in the area of psychology. I also would like to thank Professor Lin and Dr. Hou in Taiwan, and Oumatie, Sherie, and Dr. Ferguson of the Park and Recreation Resources Department, and Dr. Duane Mezga in Department of Landscape Architecture, without access to their classrooms, this survey could not have been completed. In addition, I would like to thank Chai—Jan Liu, Chau-Chang Wang, Kuan—Chou Chen, and Tsao—Fang Yuan for their assistance during my graduate study and oral defense . Finally, I would like to give my deepest appreciation to my parents . Their love and financial support have helped me through the toughest of times. My fiancee, Chiu—Lin Chen, deserves my sincerest thanks. Her true love and understanding inspired and helped me through troubles or problems I faced in my graduate study. iv TABLE OF CONTENTS Page LIST OF TABLES ......................................... Vii LIST OF FIGURES ........................................ ix Chapter I . INTRODUCTION ................................... l The Landscape ................................ 1 Purposes and.Hypotheses ...................... 3 II. LITERATURE REVIEW .............................. 5 Culture and Environment ...................... 5 Perception and Preference .................... 11 Landscape Preference Studies ................. 12 Familiarity" .................................. 17 III . METHODOLOGY .................................... 20 Medels for Assessing Landscape Quality ....... 20 Study Sample ................................. 27 Experimental Stimuli ......................... 27 Procedures ................................... 32 Statistical Analysis ......................... 33 IV. ANALYSIS AND FINDINGS .......................... 35 Sample Profiles .............................. 35 General Findings ............................. 39 Slide Selection .............................. 41 Correlation and Regression ................... 43 Differences.Across Landscape Types ........... 48 Mbuntainscape .............................. 48 waterscape ................................. 49 HDmanscape ................................. 51 V. DISCUSSION AND CONCLUSIONS ..................... 56 Discussion of Results ........................ 56 Conclusions .................................. 61 Application ................................ 61 Future Studies ............................. 63 APPENDICES A" STIMULI ........................................ 66 B. QUESTIONNAIRE .................................. 72 C. FINDINGS FOR EACH SLIDE BY CONDITION ........... 81 D. STATISTICAL RESULTS IN FIGURES ................. 84 LIST OF REFERENCES ..................................... 94 vi Table 3—1 3-2 LIST OF TABLES The evaluation of landscape assessment models ....................................... Comparison of results of regression analysis us1ng informational factors in.preference matrix ........................................ Stimuli types and slides ..................... Profile of each subgroup ..................... Countries visited.by.American and Taiwanese students ..................................... Mean scores of preference, familiarity; and acceptable levels of environmental development Origin of the landscape settings for the landscape conditions ......................... Preference and familiarity across different landscape settings ........................... Correlation.matrix ........................... Regression output table ...................... The mean score of preference for and familiarity with mountainscape ............... 'Fhe'mean.score of preference for and familiarity with waterscape .................. The mean score of preference, familiarity, and acceptable development levels in humanscape ................................... The preference score for each.humanscape dimens1on .................................... Rated acceptable levels of environmental development for each humanscape dimension ... vii 26 30 36 38 40 41 42 44 46 48 50 51 53 55 C—l American subjects ............................ 81 C-2 Taiwanese subjects ........................... 82 viii Figure 2—1 2—2 2-3 D-10 LIST OF FIGURES Page The culture/environment interaction .......... 9 The process of perceiving the environment .... 12 The matrix.of landscape resources evaluation .............................................. 16 The process of environmental perception and resultant preference ......................... 25 The mean score for preference, famdliarityy and acceptable levels of environmental development for each nationality’ ............. 84 The preference score for mountainscape slides .............................................. 85 The preference score for each dimension in the waterscape condition ......................... 86 The mean scores for preference and acceptable levels of development in humanscape slides .............................................. 87 The preference score for the human habitation dimension .................................... 88 The preference score for the road construction dimension .................................... 89 The preference score for levels of development of trail ..................................... 9O Acceptable levels of environmental development for the human habitation dimension ........... 91 .Acceptable levels of environmental development for the road construction dimension .......... 92 Acceptable levels of environmental development for the dimension of trail ................... 93 ix INTRODUCTION Woe Of the physical world, land is a major component on the earth. Land becomes landscape when it is described or seen with its physiographic and environmental characteristics (Laurie, 1975) . Originally, in the Dutch and German languages, the term "landscape" and "Landscahaft" referred to the notions of "territory", and "region" . Later on, this term was used for visible surroundings , and eventually new terms such as "seascape" and "townscape" evolved. German geographers used the term " landscape" not only for representing territorial or visible scenery, but also to represent a broad range of living and non—living phenomena covering the earth' 5 surface . Therefore, landscape includes both natural geographical variation and the human- influenced environment . Zonneveld (1990) gives landscapes the following characteristics: 1 . They each occupy a certain space, a territory: they can be represented on a map. 2. They possess visual forms and patterns: a physiognomy that can be depicted in paintings or photographs . 3. They are " functioning" dynamic systems, consisting of a variety of components and processes that influence each other. 4 . They passed——like all things on earth—~through a sequence of situations, in other words: they have gone 1 2 through an evolution, a history. (p.7) On the basis of this set of characteristics, it appears that " landscapes" include not only natural landscapes but also cultural landscapes. A "natural landscape" indicates no human influences on the natural processes and conditions in the setting. On the other hand, "cultural landscape" refers to human influenced activities. "Cultural landscape" expresses directly the values that derive from human development, modification, change, or replacement of the original scenes of nature at any given time and place. Landscapes reflect different characteristics of particular cultures; for example, landscapes in the East are less systematic, whereas Western landscapes are more geometric. Major systematic analysis of the "landscape" occurred during the decade of the 1960's and in the early 1970's. During this time period, legislation was enacted focusing upon the landscape, resulting in wild and scenic rivers, scenic and recreational trails, and scenic highways. In addition, legislation required environmental impacts for major development projects, including aesthetic impacts . Individuals from many disciplines and professions such as forestry, geography, landscape architecture , psychology, environmental studies, and recreation have concentrated on landscape perception and assessment research. To date, an abundant volume of literature has been published in the area of 3 landscape perception and preference (e.g. , Zube et al. , 1982) . Since culture is one of the major factors that influences landscape formation, cross—cultural studies involving landscape perception and preference have raised the question: to what degree do cultural factors influence landscape preference? Among landscape preference studies, the role of culture as a variable has been examined. One group of researchers has found relatively similar perceptions when people shared similar cultural backgrounds (Zube, 1984 ; Shafer & Tooby, 1973; and Ulrich, 1977) . In contrast, another group provided evidence that culture is a distinct determinant of landscape preference (Sonnenfeld, 1967; Zube & Pitt, 1981; Herbert & Kaplan, 1987; and Yang, 1988) . In the literature over the years, few studies (Nasar, 1984; Berlyne et al., 1974; and Yang, 1988) have compared landscape preferences between Western and non—Western subject groups. In the review of the literature, no study has been found that compares landscape preferences between Taiwanese and American groups . W The main purpose of this study is to compare the visual preferences of Taiwanese and American groups for different landscape types and elements. The results of this study could contribute to a better understanding of cross—cultural differences and similarities in environmental perception. 4 These results could assist landscape designers and managers in project design and planning. Other purposes of this thesis are to compare landscape preferences for environmental development levels across the two cultures; to identify socio-demographic variables that may be related to cultural differences observed; and to assess familiarity and document its relationship to landscape preference. This study will address the following hypotheses: (1) American and.Taiwanese students will report different landscape preferences for and opinions about levels of environmental development. (2) Ratings of familiarity with a landscape are positively and significantly related to landscape preference ratings for each cultural group. LITERATURE REVIEW WW Studies of environmental perception and.preference have indicated that preference for natural settings is remarkably consistent across the general American population (Zube, 1984) . Yet, variations occur between people of different demographic, ethnic and cultural backgrounds, and areas of expertise. From the perspective of different cultures, the studies of preference have suggested that there is a high degree of homogeneity in those of similar cultural backgrounds . For those of different cultural backgrounds , landscape perception differences have been observed (Yang & Terry, 1992). The term "culture" is comprehensively used in social psychology, sociology, and anthropology. A.simple and broad definition of culture was presented by anthropologist Herskovits (1952). He argued that human culture is the human influences portion of the environment. In addition to this simple definition, Altman & Chemers (1984) provide four components of culture: 1. culture refers to beliefs and perceptions, values and norms, customs and behaviors of a group or society; 2. culture is used.to indicate that cognition, feelings, and.behaviors are shared.among a group of people in,a consensual way . 6 3 . culture implies that these shared beliefs, values, and styles of behavior are passed on to others, especially children, and that the socialization and education of new members of the culture help preserve consensus from one generation to the next . 4. a society's values, beliefs, and practices involve more than "mental" and "behavioral" process; culture appears in objects and in the physical environment . (133) Cultural differences appear not only in the organization of society but also in the physical environment such as home design and house form, public buildings, and urban planning. For example, house form is not only the simple result of physical forces or any single casual factor, but also the consequence of a whole range of socio-cultural factors seen in their broadest terms (Rapoport, 1969) . For Chinese culture, influenced by Taoism and Confucianism, "Nature" is termed as "tzu-jan", or 'self—thus' (Liu, 1962) . Chinese people consider themselves as a part of nature, and then imitate nature in their immediate environment. As Liu (1962) stated "Man is advised to submerge his being in the infinite flux of things and to allow his life and death to become part of the eternal cycle of birth, growth, decline, death, and re—birth that goes on in Nature" (p.49) . 7 In western culture, the human is the major object in painting, while natural scenery is in.the background. This emphasizes the anthropocentric concept of Western culture which derived from religion (McAndrew, 1993) . In Western religion, "... humans are not just an ordinary part of nature -- they are made in.God's image and are quite different from the rest of creation" (McAndrew, 1993, p.233) . Those different philosophical notions may be the stimulus for the difference in.people's attitudes toward nature. For example, westerners seem to try to dominate nature, whereas Easterners (especially the Chinese) tend to protect a site rather than endanger it (Rossbach, 1983) . The formal, geometric, and symmetrical patterns of landscape style were developed by Western culture, while the oriental landscape style is based on informal and asymmetriijatterns (Yang, 1988). In.Chinese landscape style, a rock means mountains (yang), which represents the male force; and water represents the sea (yin), which symbolizes the female force (Jellicoe & Jellicoe, 1987). From this point of View, Chinese people conceive of themselves as a part of the whole universe and try to embrace and be near nature in their home courtyards. Physical environment can be divided into two dimensions: the natural environment and the built environment . The natural environment refers to the initial geographical features on earth, whereas the built environment refers to the human influenced environment. Since culture is composed of factors 8 such as the human—influenced part of the environment, culture and.physical environment are closely interrelated. Through a psychological process, culture and physical environment can interact by way of mental and behavioral activities (Altman & Chemers, 1984). "Mental activities" refer to how people see, smell, hear, and interpret the physical environment along with their personal attitudes and beliefs toward the physical environment. "Behavioral activities" refer to what people do overtly in the physical environment. The relationship between culture and environment is termed. "culture—ecology" (vayda, 1969; Berry; 1975). 'Ihe concept of culture—ecology emphasizes the physical environment that plays a determinant role in customs, life-style, social values, and behaviors in different cultures (Altman & Chemers, 1984) . Berry (1975) proposed two types of relationships between environment and culture . The "strong" version suggests the physical environment is the major factor influencing cultural processes. There is a significant one-way causal relationship between.these two variables. In other words, the physical environment plays an important role in influencing cultural formation and development. In contrast, the "weak" ‘version.implies that there are "functional interdependencies" among environment and culture variables; and it is difficult to establish an obvious direction of the relationship or cause and effect (Figure 2—1). (Strong) Culture; Environment: Belief & perception Natural Environment Value & Norm Custom & Behavior Built Environment I? w Psychological Processes: . Mental Activities ‘—‘ Behavioral Activities Eigure_2—_1 The culture / environment interaction (Source: Berry, 1975, Altman & Chemers, 1984) Landscape ecology, another point of view, stresses that the interaction between living and non—living elements already exists (Zonneveld, 1990). There is no specifically strong or weak relationship between culture and environment . Instead, it is combined with these relationships more comprehensively and called "landscape ecology". As Zonneveld (1990) stated: Landscape ecology in its wide sense is involved with cultural aspects as well. Landscape ecology is therefore functioning as a meeting place for many disciplines, a variety of natural sciences as well as humaniora. (p.12) 10 Despite the weak or strong relationship between culture and environment , Zube (1982) argued that there are three components of the human-environment interface . The human component includes past experience, knowledge, expectations, and the social-cultural contexts of individuals and groups. The second component, the environmental component, is dependent upon scale —— ranging from small scale to large elements of the physical environment. The third component is the results of the humanrenvironment interaction. The interaction.results as outcomes in.turn.affect both.the human and environmental components. For example, both the pyramids of Egypt in ancient times and the highly developed metropolitan areas in modern society are interaction results which increase the self confidence of humans in nature and modify the characteristics of the environment. .As a matter of fact, no matter what the relationship is between humans and the environment , the major purpose of cultural and environmental studies is to try to understand what values humans attach to the environment and to document the role humans play'in reflecting the environment..As Laurie (1975) stated: On the one hand, the environment has a definite impact on the individual , and our response may be adapted to the 11 imposed conditions . On the other hand, we are continually manipulating or choosing our physical surroundings in an attempt to make life physically and psychologically more comfortable. (p.153) W Perception and preference are the psychological processes involved when humans judge an object . Perception is viewed as not only dealing with information from environment, but also yielding information about what people are concerned with (Kaplan, 1988) . The process of recording perceived environmental information is done by internal processing of information such as coding, storing, recalling, and decoding the relative location and attributes of objects in the spatial environment (Figure 2-2) . First of all, people obtain information from the real world by way of acquisition. In our mental activities process (internal), we code, store, recall, and decode this information to interpret its meaning to ourselves. As long as getting basic conceptions about the meaning of information, people will try to locate and attribute this information as the functional purpose to their surrounding environment . Through environmental preference , humans not only deal with the outcomes that result from perceiving things and spaces, but also react to them in terms of their potential usefulness and supportiveness (Kaplan, 1988) . It is referred 212 to as positive or negative personal evaluation, feelings, or attitudes toward the environment . The difference between perception and preference is that perception is only the sensing of an object or scene instead of preferring it or not. In other words, we need to perceive the object initially (perception) before we make the judgment as to whether we like it or not (preference) . Internal Processing _ Obtaining Information: Of Information: FunCt1°n3 Acquisition & Sending Coding, Storing Location 5 Attribute Recalling, & Decoding 0f Env1ronment Figure 2-2 The process of perceiving the environment (Adapted from Altman & Chemers, 1984, p.45) W Despite an abundance of literature concerning landscape preferences, only a few studies have dealt with the comparison of landscape preference between Western and non—Western groups (Nasar, 1984; Berlyne et al. , 1974; and Yang & Kaplan, 1990) . The purpose of cross-cultural comparisons is to provide further insights into the nature of environmental preference (Canter and Thorne, 1972; Nasar, 1988) . Two major findings appeared in these landscape preference studies based on inter— cultural backgrounds . 13 The literature on cross-cultural comparison of preference for natural environments shows relatively high agreement when the backgrounds of cultures are relatively similar (Zube, 1984) . For instance, Ulrich (1983) found consistent patterns between Swedes and Americans and argued that studies have shown that these two cultures have more similarities than discrepancies in the perception and evaluation of the environment . Zube and Mill (1976 ) reported high correlation coefficients between two Australian samples (seasonal and year-round residents) and American landscape architecture students in rating Australia landscapes. Zube and Pitt (1981) compared Yugoslavian students, Italian Americans, and various American groups and showed similar high correlation coefficients for scenes of northeastern American landscapes. Shafer and Tooby (1973) also reported consistent preferences in their comparison of Scottish and American samples. The report by Wheeler (1984) presenting a set of preference ratings of scenes by participants in various Western European countries as well as students from the University of Arizona also supports Zube ' s conclusions . Therefore, when the respondents' cultural backgrounds are similar, there is greater agreement in environmental preference . By contrast, among relatively dissimilar cultural groups, the preferences are distinctly less similar (Herbert & Kaplan, 1987) . When culture is diverse (e.g. with respect to life styles, social values, and economic indicators) , and when the l4 landscape styles are strikingly contrasted, the differences in preference are greater. Sonnenfeld (1967) found strong differences in preference among native and non—native.Arctic residents. Zube and Pitt (1981) reported a relatively lower correlation in scenic quality ratings between West Indians and .American. samples 'when. they 'viewed. scenes of the 'Virgin Islands. Yang & Kaplan (1990) found that Western tourists preferred Japanese styled landscape to their own, and Korean people preferred western styled.landscape to Korean.sty1es of landscapes. Zube and his colleague (1985) found very significant differences in perceived urban residential quality when measuring the responses of Saudi.Arabians and Americans. Landscape preference studies in Taiwan are not as well— developed as in the U.S. There is a lack of developed professional and theoretical research methods in this field in Taiwan" Most landscape preference and evaluation studies have been completed by experts whose backgrounds are urban Planning, Horticulture, Forestry, or .Architecture. The assessment methods used were primarily based upon experts' judgment. Since landscape preference studies in Taiwan are just beginning, pervious findings from cross—cultural research in this field are still limited. In a case study, Lin (1979) divided landscapes into two categories: inland.landscapes and.coastal landscapes. For the inland landscapes, three components were used to evaluate their quality; six factors were used to evaluate the coastal 15 landscapes. In addition to these criteria, landscape variety and homogeneity were also major factors used in evaluating the whole landscape. In the case study, he used a grid system to categorize the site into several areas. Using the rating scale which was adopted from Linton' 8 method (1968) , he evaluated the total scores for each grid to decide which area had the highest landscape value (Figure 2-3) . He concluded that this evaluation system could be applicable in recreation planning and landscape evaluation. In addition to the analysis of the physical environment, Lee (1983) added aesthetic attributes to the assessment of landscape preference. She proposed vividness, uniqueness, degree of naturalness, and unity as predictors to evaluate quality of aesthetics. Using a rating scale from 1 (worst) to 5 (best), she asked 224 subjects to rate 12 pictures that were taken in Pin-Lin village. She concluded that the combination of the elements of landscape space and the attributes of landscape aesthetics together made it possible to predict landscape preference by sampling the general population instead of particular groups. Using the SBE (Scenic Beauty Estimation, Daniel & Bolster, 1976) assessment model, Zu (1989) measured landscape preference and paid more attention to climate factors. She proposed the hypothesis that different climate situations would produce different physical visual effects in the same scene. Using 90 color pictures, she asked 102 subjects to rate l6 filililliZfififil as-e 4!¢!#B aaarrnu NILnInuI w i? I liflfi R mu .m x .u 5378‘ IRJKU 53$“ Ilgflli E:IIIG“BZIflI—i:lflllfl-lfifififlfiflllmli B:AOCIIH#ZXI£I'G£IBQ C:Ifll&fl-ING¢HGICI&II 67 The matrix of landscape resources evaluation 1979) (Lin, F. 17 each stimulus from 1 (strongly dislike) to 9 (strongly like) . She concluded that climate condition as a factor has a significant relationship with landscape preference . Furthermore, morning and evening landscapes were the most preferred. Pictures of landscapes on a rainy day were the least preferred. Chu (1986) studied landscape preferences by comparing the difference between color pictures of a site and actual on-site experiences with the Semantic Differential scale. After concluding that there is no difference between these two types of stimuli, he used color pictures to measure the preference of cityscapes in Taipei . He found that the preference scores in commercial and industrial areas were the lowest, whereas governmental and cultural areas had the highest scores . The preference in residential areas varied depending on the degree of development in each sub-residential area. : .1 . . It is reasonable to expect that people prefer the thing they know. They gain familiarity from many circumstances, such as where they live, work, visit, and the culture norms and beliefs they adhere to. Kaplan & Kaplan (1989) stated that familiarity is the major factor accounting for preference differences. Several other cross—cultural studies emphasized the importance of the relationship between preference and familiarity (Hammitt, 1979; Nasar, 1984; Sonnenfeld, 1966; and 18 Yang'&.Kaplan, 1990). Theoretically, individuals frcmlanother country would.be less familiar with scenes taken in the USA. The expectation is that local preferences would.be higher if familiarity had a positive influence, whereas preference would be low if familiarity had a negative influence. In fact, the influence of familiarity on preference has been demonstrated to be quite complex, rather than simply positive or negative (Kaplan & Kaplan, 1982). Past experience and.prior knowledge of specific landscape types often are associated with the highest preference scores. It is also supposed.that excessive familiarity may breed contempt, leading to reduced landscape preferences (Hammitt, 1981). .As mentioned.above, in.environmenta1.preference studies, familiarity has been shown to be an important predictor. Sonnenfeld.(1966) found.that younger persons preferred.exotic natural scenes, while most others preferred more familiar scenes. Hammitt (1979) suggested that the comparison of familiarity ratings with preference ratings for the same landscape showed a positive relationship. Yang & Kaplan (1990) concluded that familiarity served as a positive factor in predicting environmental preferences for both Western visitors and Korean respondents. In contrast, Canter and Thorne (1972) found that residents of Scotland and Australia preferred unfamiliar to familiar urban scenes. Nasar (1984) reported that American and Japanese students prefer foreign scenes (less familiar) to 19 native ones. When rating scenes of the West Australian landscape, Herbert and Kaplan (1987) also reported increased preference for more novel settings and decreased preference for seemingly familiar settings among western.Australian and American students . METHODOLOGY 3 J E . 1i i L J . Contemporary landscape assessment methods have been developed over many years . Each model is designed to deal with specific problems encountered in understanding the values of landscape quality. The purpose of landscape assessment is to provide sensitive and reliable measurements of scenic quality in different landscapes . According to Daniel & Vining' 8 report (1983) , landscape quality models can be divided into five major categories. They are (1) ecological, (2) formal aesthetic, (3) psychophysical, (4) psychological, and (5) phenomenological. Zube and his colleagues (1982) divided these landscape assessments into four paradigms: expert, psychophysical, cognitive, and experimental, where ecological and formal aesthetic categories are the same as the expert paradigm; and psychological categories are analogous to the cognitive paradigm. About 40% of landscape perception articles are in the expert paradigm (ecological and formal aesthetic), and 35% of landscape assessment reports are in the psychophysical paradigm (Zube et al . , 1982) . Three criteria, which are reliability and sensitivity, validity, and utility, were proposed by Daniel and Vining (1983) to evaluate these models. According to these criteria and the valuation of each model, they concluded that psychophysical and psychological models are the most reliable 20 21 and.valid means of evaluating landscape quality (Table 3-1). Table 3-1 The_E3Lal1Latis2n_Qf_Lamlscar_re.ZésseasmenLIYIgrielsa Ecological Formal Psychophysical Psychological Phenomenological Aesthetic Reliability No direct High High reliability High Low reliability, & evidence to re iability and sensitivity re iability & but high Sensitivity proof its between sensitivity in sensitivity reliability experts. Low single and sensitivity due dimension sensitivity to the categorization Validity Difficult or Question: Does High validity High validity More valid in impossible the set of conceptualization to perform formal elements for landscape- capture all of quality the aesthetic assessment than aspects of the in the role of landscape objective landscape features assessment Utility Depends on Widely in use: In the short on how Too sensitive to whether the Forest Service, run, it is not we 1 the be useful development Bureau of Land efficient due to principle of will destroy Management, and considerable psycho ogical critical a number of time and dimensions can wildlife private and expense. In the be tied to environment semipublic long run, it is physical or not. If agencies very efficient features of the yes, it is landscape; and useful. If to related not, the aspects of utility is realistic human still responses to untested the landscape Valuation Natural Can not be Provides precise Provides direct Fails to ecosystem is applied in assessment of measures 0 establish viewed as moderate or public Inman/socral tematic having large-scale perception value and relationship intrinsic studies. Not related to relates between value which easy to relate scenic beauty of landscape psychological separates it these landscape; and quali to responses and from any assessments to can be measured human th landscape reference to economic or as a guide to and features; only other social social values economic cost productivity ints out the values or importance of the lumen human context in welfare which landscapes Note: a: Source: Daniel & Vining, 1983 are encountered 22 In the psychophysical model, a mathematical strategy is used to assess the relationships between geographical characteristics of landscape and the perceptual judgment of Observers. The SBE (Scenic Beauty' Estimation, Daniel & Bolster, 1976) model is representative of this model. This model measures humans' psychological responses specifically from a single dimension of environment such as scenic beauty or landscape preference. The basic assumption in a psychological model is that people's cognitive and affective reactions are evoked by various pieces of landscape information. As Daniel & Vining (1983) stated "Relationships between the perceptual, cognitive, and affective scales and the preference scale are used as a basis for inferences and hypotheses regarding the psychological features of the landscape that determine human landscape pmeferences" (p.66). Kaplan and Kaplan and their associates (1989) have completed a series of studies in landscape assessment based on this model. They provide four constructsz'mystery; coherence, legibilityy and complexity as correlates of landscape preference. In summary, in these landscape assessment models, ecological and formal aesthetic models deal with the interaction between humans and the environment from a physical jperspective. Psychophysical and psychological models consider both physical and psychological attributes together. The phenomenological model concentrates on the detailed 23 psychological process of how people feel in the environment . People's environmental perception and preference can be processed by way of both physical and psychological perspectives. According to Gibson's work (1979) , the physical attributes firstly concern the environmental structures such as the medium, substance, and surface which compose the environment. The medium, such as air and sound, transmits light and vibration from object to object. Substance is a symbol of solid and rigid materials, while surface is the part of substance people can observe. What we can notice is the information available from the interface of substance and medium: the surface . Secondly, the meaning of environmental structure from these physical properties becomes a geographical layout, an event , and the affordance to humans that is derived from the environment . The layout is considered the persisting arrangement of surfaces relative to one another. The event is referred to as any change in an object, place, of surface such as reshaping or repositioning of surfaces; or change in color or texture. Affordances are what the environment can offer to humans and animals in terms of their potential action or use. As Gibson (1979) stated : To perceive is to be aware of the surfaces of the environment and of oneself in it . The interchanging between hidden and unhidden surface is essential to this awareness . These are exiting surfaces; they are specified at some points of observation. Perceiving gets wider and 24 finer and longer and richer and fuller as the observer explores the environment. The full awareness of surfaces includes their layout, their substances, their events, their affordances. Note how this definition includes with perception a part of memory, expectation, knowledge, and meaning -— some part but not all of those mental processes in each case. (p.255) Psychological attributes involve personal traits: norms, values, and beliefs. People obtain information from the external (physical) environment and generate that information by using an internal psychological process (coding, storing, and recalling) to understand the meaning of environment . After perceiving the meaning of environment, people judge this combined information according to their past experience (such as familiarity), and by way of the associated variables that influence preference to decide whether they like it or not (see Figure 3—1). As mentioned before, the main purpose is to compare the landscape preferences of individuals from different cultures. Although the preference predictor could provide certain knowledge by which to understand the cultural differences in landscape preference, not every predictor functions as well as researchers expect. According to the Kaplans (1989) , only the mystery variable was consistency found to be correlated with preference (Table 3-2). Besides, Yang (1988) also failed to 25 test the validity of these predictors in his cross-cultural study of landscape preference. More specific definition of these variables and.better—designed.methods are necessary to evaluate the reliability and validity of these predictors. Physical Attributes Psychological (Spatial Configuration) Attributes . . , . * Demographic Env1ronmental _ Envrronmental Perce1v1ng Features Structure v Meaning Environment ‘ * Beliefs 4 A ' * Values * Norms Medium Layout Surface Event Substance Affordance Past Experience (Familiarity) . V . Environmentall‘ Preference Preference |‘ Predictors: A * Complexity Coherence Mystery Legibility ,Eigure_3;1, The process of environmental perception and resultant preference (Sources: Gibson, 1979; Taylor et al., 1987; Palmer, 1986;.Altman.& Chemers, 1984; and Kaplan & Kaplan, 1989) 26 Table 3—2 gme.,'-oen "3!‘ age. -I- _'10 Study' # Scenes R? F‘ p Complex. Coher. Mys. Legib. (partial correlation coefficient) WOodcock 72 .19 8.28 .001 xx? xx .38 —-° R.Kaplan et al. 59 .19 3.11 .05 ~- -- .31 -— Ellsworth 6O .24 4.35 .005 —— -— .31 -- Herzog, 1987 70 .50 10.64 .001 -- -- .39 xx Gallagher 32 .42 6.75 .001 -- .51 -— xx Harzog et al. 70 .13 3.25 .05 —— .26 34 xx Herzog, 1985 70 .46 9.11 .001 —- .39 .34 xx Harzog, 1985 70 .87 43.70 .001 -- .60 .45 -- Anderson 48 .45 8.92 .001 —— .53 .49 —.43 Herzog, 1984 100 .53 17.42 .001 .33 .21 .33 xx Replication 56 .49 16.76 .001 —.39 .33 .56 xx Nte: a: Kaplan & Kaplan, 1989, p.65 b. Net included in study; c: Net a significant factor Complex: Complexity Coher. Coherent Mys.: Mystery Legib.: Legibility In this research, a psychophysical model is chosen to evaluate landscape preference differences. This method provides precise assessment of public perception related to scenic beauty of landscape that is the major component this study focuses on. Besides, the quantitative technique of this method is looking for the mathematical relationships between the physical characteristics of the landscape and the perceptual judgment of human observers. As Daniel (1990) 27 stated "the psychophysical methods provided a set of tried and tested procedures for translating human responses to environmental stimuli into precise and reliable indexes with known metric properties." (p.634) As mentioned earlier in Table 3-1, this method is highly reliable and valid based on past studies, and this offers another reason to choose this approach . W The populations used in this study were American and Taiwanese students. Subjects in America were chosen at Michigan State University by selecting students in an introductory recreation resources course in the Park & Recreation Resources Department, and students in a Landscape Architecture Department course . Subjects in Taiwan were selected from courses in the Landscape Department in Tunghai University; and in courses in the Horticulture Department in Taiwan University. The survey was conducted both in America and in Taiwan between July and October in 1992. . 1 .1. Using color or black-white photographs to evaluate landscape quality and preference has been shown to have a high degree of reliability and validity in landscape assessment studies (Kaplan & Kaplan, 1989) . In this research, mountainscapes, waterscapes and humanscapes were selected as 28 stimuli to measure subjects' preference, familiarity, and their opinion.about acceptable levels of human.development in natural settings. In many religions and cultures, mountains play a very important role as symbols of the power and dominance within nature. In Chinese, Korean, Iranian,.Asian.Indian, and German cultures, mountains are a central axis of the earth and.serve as a link with heaven (Altman & Chemers, 1984) . Thus, the feeling toward mountains is both negative and positive. In Western cultures, the positive attitudes toward mountains were not distinct until the 19th century, because of the development of transportation (Tuan, 1974). The evolution of transportation vehicles made it more convenient for people to be near or in the mountains to discover the mysterious aspects of mountains. The terms that describe mountains range from "powerful", "respectful", and. "sacred" to "spectacular", "exquisite", and "the essence of natural beauty". waterscapes, such as oceans, lakes, rivers, and waterfalls, are seen. positively' by 'many' cultures. ZMany cultures developed and were shaped by their proximity to water locations. Water provides transportation, food, security, and climate moderation. Many studies have reported that water is one of the major factors influencimg landscape preferences (Palmer, 1986; Zube et al., 1975; and Yang, 1988). Most of these studies were conducted. to evaluate the landscape preference of respondents from Western cultures. While 29 comparing landscape preferences of Westerners and Koreans, Yang (1988) found that water in a landscape scene could strongly enhance landscape preference for both sample groups. Since there is a strong relationship between.environment and culture, human influences on the natural environment could differ, depending upon the characteristics of the culture. The Objective of studying'human.influences in natural environment has been excluded from landscape preference studies (Vining et al., 1984). In addition to understanding preferences across natural landscapes, assessing opinions of acceptable levels of development in a natural area is another major objective of this study; One group of slides presents different degrees of development and is used to address this objective. Thirty four landscape slides were used as stimuli which were taken in America and.in.Taiwani These slides were placed into three types of landscape: mountainscapes, waterscapes, and humanscapes . These categories and categorizations, devised by the experimenter, were reviewed by the experiementer' s thesis committee. In addition to measuring these three types of landscapes, several detailed geographical features were presented.to help in understanding the preference for physical characteristics among different cultural groups (Appendix A). The numbers of slides for each landscape type are shown in Table 3—3. These detailed features are described as followed: 30 Table 3—3 .1. 11.3 landscape Type Slide # Mountainscape (M) Superior (M1) Normal (M2) Inferior (M3) 15 U'IU'IU'I Waterscape (W) Distance dimension (W1) Near category (less than 10 feet, W1) Far category (more than one mile, W2) Combined landscape (W3) Water condition Moving water (W4) Still water (W5) 63 Humanscape (HD) Human habitation (HDl) Road construction (HD2) Trail development level (HD3) Acceptable level of environmental development (HDL) 10b i-PNiih- 10 Total 34 Note: a: The contents of slides in distance dimension were reused in water condition dimensions . b: The same ten slides were reused to assess opinions of acceptable levels of environmental development. * Mountainscape —— The criteria of position of viewer (superior, normal, and inferior) are assessed in this type of landscape. There are five slides to represent each of these three mountainscape dimensions . 1. Superior (M1) 2. Normal (M2) 3. Inferior (M3) 31 * waterscape —— Four dimensions vary in this set of landscape slides. 1. Distance dimension between viewer and.water'varies in the slides from near (W1) to far away (W2): Three slides representing the "near" category have distances between water and viewer of less than ten feet. The other three slides indicating distance between water and viewer of more than one male represent the "far" category. . Combined landscape with mountain & water (W3): Three slides are included in this dimension. . water situation.[moving (W4) & still (W5)]: Four slides are included in moving water sub-dimension, while two slides are represented in the still water sub-dimension. These six slides in this dimension are the same as in.the distance dimension. * Humanscape -- Three dimensions are represented in this type of landscape. 1. HUman habitation -— buildings (low developed to highly developed; HD1): Four slides are in this dimension in which each slide represents one level of human habitation: a single cabin, a single temple, a small village, and an over—developed community. . Road construction (well—paved or rough; HD2): TWo slides are in.this dimension, of which.one is well—paved.and.the other one presents rough road conditions. 32 3. Development level of trail (low to high; HD3) : Four slides are in this dimension. Each slide represents one single level of trail development: a trail along the mountain, a small scale hiking trail, a highly—developed route, and an over-developed route. 4 . Opinions of acceptable level of environmental development (HDL) : In the humanscape condition, subj ects' opinions regarding environmental development issues are compared . All slides in the humanscape dimension will be used in the measurement . In the waterscape condition, six slides used in the distance dimension are re—used in the moving and still water conditions in order to measure another physical attribute. The order of slides is mixed in each landscape type. For example, the 15 slides of mountainscape were randomly mixed together with other slides. Procedures A slide projector was used in a classroom to show the stimuli. Slides were taken with a 35mm standard lens basing normal visual angles. Using a Likert type scale, students were asked to complete a questionnaire asking questions about their preferences for specific slides along with demographic background information. An English questionnaire was used in America, while a translated Chinese questionnaire was used in Taiwan (Appendix B) . 33 Slides were shown to students in a classroom setting. First, the investigator explained the purpose of the study and described the rating scales to be used. Second, all slides were shown briefly to let the subjects have a general idea and understanding of the order of the stimuli to be presented. Third, each slide was presented again for 10 to 20 seconds, the exact display time per slide varied slightly; each exposure was judged by experimenter to give enough time for subjects to evaluate the slide and respond. In the third section, for slides in the humanscape condition, a rating of the level of acceptable development was processed in addition to preference and familiarity ratings. Finally, in the last section of the instrument, students were asked questions concerning their personal background. In this study, the dependent variables are preference (P) , familiarity (F) , and opinions on levels of environmental development (L). The independent variables of interest are mountainscape (M), waterscape (W), humanscape (HD) , and several demographic variables such as nationality and residency. Correlation and regression tests were used to determine which independent variables are associated with the dependent variables. Dummy variables were used in the analysis because some demographic variables were nominal variables. For 34 example, Taiwanese were coded as 0, while Americans were coded as 1 . Using this coding procedure, confusion about the correlation and regression outputs could be avoided. To understand the differences between Taiwanese and American subgroups, t-tests were employed to compare landscape preference, familiarity, and opinions regarding levels of environmental development for the three types of landscapes: mountainscape, waterscape, and humanscape. Comparisons of dimensions within each landscape type were tested with t-tests as well . ANALYSIS AND FINDINGS W There were 189 American subjects and 287 Taiwanese subjects in this study. The total sample size was 476 subjects. Because this survey was conducted in the classroom, and in the presence of the instructor and experimenter, the response rate was very high. Occasionally, some questions were omitted, but no case was found totally invalid in this survey. The frequency distribution by subgroups is shown in Table 4—1. The gender distribution is evenly divided between men and women for both the American and Taiwanese samples. Most of the American subjects lived in suburban areas (62%) whereas half of the Taiwanese subjects lived in urban areas (50%) . This finding could be explained by the fact that Taiwan is a small island where many places have already been developed as urban areas, whereas students in MSU are more likely to come from small towns or rural areas . The age range for each group was primarily from 19 to 22 years old. Many Taiwanese did not have any out-of—country travel experience (71%) , whereas fifty nine percentages of the American subjects had visited one or two countries at least. Most of the Taiwanese subjects who never traveled to other countries were male. This is due to the national military service policy in Taiwan. This policy prohibits males from going out of the country if they have not completed military service. The majority of each group were 35 {KN \nfi‘ . , one»... .0. b n 36 seniors and juniors in college. Table 4—1 W Characteristic American(1) Taiwanese(0) Total # % # % # % Gender Male 97 51 164 57 261 55 Female 92 49 123 43 215 45 Residence Urban 24 13 143 50 167 35 Suburban 117 62 91 34 208 44 Rural 48 25 53 16 101 21 .Academic Major LA.background 77 41 170 59 247 52 an—LA.background 112 59 117 41 229 48 .Academic Status Senior 129 68 95 33 223 47 JUnior 46 24 85 30 131 28 Sophomore 9 5 49 17 58 12 Freshman 5 3 58 20 63 13 Age 19(or below) 14 7 43 15 57 12 20 22 12 59 21 81 17 21 65 34 61 21 126 26 22 45 24 52 18 97 20 23 8 4 24 8 32 7 24 5 3 15 5 20 5 25 4 2 17 6 21 5 26(or above) 22 12 16 6 38 8 Travel Experience 0 23 12 203 71 226 47 1 65 34 29 10 94 20 2 48 25 23 8 71 15 3 13 7 5 2 18 4 4 6 3 7 2 13 3 5 5 3 7 2 12 3 6 5 3 6 2 11 2 7(or more) 24 13 7 3 31 7 37 With regard to travel experience, Americans have a different travel pattern from that of Taiwanese students. In this survey, the American subjects had traveled to 45 countries and the Taiwanese to 30 (Table 4—2) . The countries Americans traveled to most were Canada (148, 89%) and Mexico (61, 38%) . The others are France (30), England (30), Germany (29) , and Australia (24). All of these countries except Australia were in North America or in Europe. Only a few people had been to Asia or the Mid—Eastern countries. Fourteen 'of 15 American students who were landscape architecture students had been to Czechoslovakia. This is because there is a special travel course for students who major in Landscape Architecture; they have the opportunity to visit Czechoslovakia. Of the Taiwanese, only 84 (29%) subjects had travel experiences. Of these, the nationalities and number of trips taken are indicated as follows: Japan (51) , Hong-Kong (25), Singapore (12), U.S.A. (21), Thailand (16), Korea (14), and China (11) . Except for the U.S. , countries visited are in Asia, destinations located near Taiwan and sharing the same cultural background. European countries are the second most popular destination for Taiwanese . France, England, Switzerland, and Holland are the most common European destinations for Taiwanese. Overall, European countries such as France, Germany, England, and Switzerland were the most popular European destinations for both groups, while Japan is the only one in the Asia region preferred by all subjects. 38 Table 4—2 0 ! ‘ V ‘0 0 4.115 a! an. qul‘T‘ 0‘! Country Name American Taiwanese Total Canada 148 1 149 Mexico 61 1 62 ‘U.ShAm 21 21 England (Great Britain) 30 6 36 France 30 9 39 Germany 29 5 34 Australia 24 5 29 Switzerland 18 6 24 Czechoslovakia 15 15 Italy' 12 2 14 Spain 7 7 Ireland 6 6 Scotland 6 6 Greece 6 6 Belgium 6 4 10 Bahamas 6 6 Holland 7 8 15 Denmark 4 1 5 Nbrway 3 1 4 Sweden 3 1 4 Puerto Rico 3 3 Poland 3 3 New Zealand 2 1 3 Fiji 2 2 Israel 1 1 Egypt 1 1 Luxembourg 1 1 Trinidad 1 1 Russia 1 1 Finland 1 1 2 Guatemala 1 1 Barbados 1 1 TUrkey 1 1 Portugal 1 1 Tahiti 1 1 39 Table 4-2 continued Country Name American Taiwanese Total Jamaica 1 1 Venezuela 1 1 Japan 7 51 58 Hong—Kong 2 25 27 Thailand 1 16 17 Singapore 2 12 14 Korea 3 14 17 China 1 11 12 Taiwan 1 1 Philippines 1 6 7 Ioochoo 5 5 Indonesia 4 4 Malaysia 5 5 Nepal 1 1 Cashmere 1 1 India 1 1 Macao 1 1 J E' i . Landscape preferences between American and Taiwanese students differed for mountainscapes and waterscapes (tP1(463)=-2.90, tP2(474)=-5.16, p<.01; Table 4-3) . In the case of humanscapes, the preference scores were the same for both groups . Waterscapes had the highest preference score among the three types of landscape, whereas humanscape had the lowest score for both groups. American subjects preferred mountainscape and waterscape more than did the Taiwanese (p<0.01). 40 Table 4-3 “.2. _ o ‘_ o ' ‘ ‘ ‘0 ‘ an: a 11‘ t ‘9 a! c : ‘,- .cf2Enyircnmental_Deyelonment Stimuli Subjects P1** Fl** P2** F2** P3 F3** HDL** Taiwanese M 3.23 2.89 3.93 2.84 2.52 3.28 2.74 $1 0.56 0.71 0.51 0.68 0.50 0.71 0.46 American M 3.40 2.42 4.19 3.06 2.52 2.50 3.02 551 0.58 0.86 0.59 0.83 0.73 0.81 0.60 mete: PlzPreference for Mountainscape F1:Familiarity with Mountainscape P2:Preference for waterscape F2:Familiarity with waterscape P3:Preference for Humanscape F3:Familiarity with Humanscape HDL: Acceptable Level of Environmental Development Mountainscape:15 slides; Waterscape:9 slides; Humanscape:10 slides; USing t—test, * p<.05, ** p<.01 Famnliarity ratings for all three types of landscapes between these two sample subgroups were significantly different [tm(461)=6.95, te(474)=-3.23, tm(469)=11.08, p<.01]. Taiwanese were more familiar with the mountainscapes and.humanscapes presented than were American students, while Americans reported higher familiarity with waterscapes than the Taiwanese. In addition, the opinions of acceptable level of environmental development between these two cultural groups were significantly" different [tmm(468)=—5.75, p<.01]. Taiwanese agreed less with the levels of human development 41 observed in the natural environment as shown in the landscape slides compared to Americans. The mean scores of preference, familiarity, and opinion of acceptable level of environmental development for the three types of landscape by slide for each nationality are shown in Appendix C. Also, Appendix D illustrates the findings of this study in geographic format. 3] . 3 ES 1 I . The origin of the slides and the distribution of the slides influenced the degree of familiarity in the three types of landscapes. Eighty percent of the slides in mountainscape and 100% of the humanscape slides were taken in Taiwan, while 89% of the waterscape slides were taken in America. Because of this distribution, the degree of familiarity with each landscape type was obviously influenced (Table 4-4) . Table 4—4 0,,91 0, Hr .3110..- ...-r -,‘ .3 -0, h- 3.....- ...-r on. '0 _ Landscape Condition Setting Mountainscape Waterscape Humanscape # % % % Taiwan 12 80 1 11 10 100 American 3 2 0 8 8 9 0 0 42 As seen in Table 4—5, the Taiwanese were more familiar with landscape slides which were taken in Taiwan [t(317.50)= 10.29, p<.01] , whereas Americans were more familiar with landscape slides that were taken.in America [t(321.66)=-2.54, p<.05]. Besides, the.American.subjects liked landscape slides taken in America more than did the Taiwanese [t (365.51) =-4 .32, p<.01]. Table 4—5 ' c ‘,‘a ‘ a0. . i_'ar' .A.res_ I: ,‘_‘n .aIQ“ opc ‘ 'ng- Setting Taiwanese American QM .Sd QM .Sd Landscape in Taiwan? Preference 2.86 0.48 2.97 0.66 Familiarity** 3.08 0.59 2.40 0.75 Landscape in.AmericanP Preference** 3.92 0.49 4.14 0.55 Familiarity* 2.82 0.61 2.99 0.80 Ngte: a: 23 slides were taken in Taiwan b: 11 slides were taken in American Using t—test, * p<.05, ** p<.01 .According to these statistical analyses, the hypotheses of differences between the two cultural groups in landscape preference were supported in mountainscape and waterscape . observed as well in Cross—cultural differences were respondents' opinions of acceptable levels of environmental 43 development in humanscape slides. Based on slide origin, the degree of familiarity with each type of landscape was significantly related to these two cultural groups. W The correlation matrix in Table 4—6 illustrates the relationships between the dependent variables (preference , familiarity, and acceptable levels of environmental development) and the independent variables . The relationship between preference and familiarity was significant and positive (rpl&F1=.22, rP2m=.19, rP3&F3=.25; p<.01) . This indicated that as the degree of familiarity with landscape increased, preferences increased as well . Travel experience had a significant correlation with familiarity in two landscape types (mountainscape and waterscape) (rF1=.11, rF2=.22, p<.01) . Those with more travel experience were more familiar with these landscapes. Residence had a significant and positive relationship with preference scores (rPl=.19, rP2=.22, rp3=.14, p< . 01) . Those who lived in rural areas had higher preference scores for these landscape settings than people who lived in urban areas. With regard to acceptable levels of environmental development (HDL) , nationality (r=.27, p<.01) , residence (r: .14, p< . 01) academic major (major in Landscape Architecture or other, r=.13, p<.01) , and travel experience (r=.11, p<.01) revealed positive correlations with this variable. In addition, acceptable levels of environmental development also 3.on "n 2.8.on "m 5883\er Hmuggyafi mo Hot/3 mHAmumoood o confide "new mommomqgm fies anemone—"g "mm megaphone: cuss >ua8aagm "mm mommomcamufiooz fies 3838mm "E 44 mommomnmcsm mo mosououosm " mm commonsense» mo oodosmmosm " mm mommomsamuufidz mo mocouomonm " E "Mudz do... And. oo. Ave. AmN. AFN. a. ANA. moi Ava. No... AFN. no... AD: Hoi Amv. Avw. AMN. #0. Ana. oo. amoi And. $0.- 50. Amvi do. B And. Amv. Adm. ANN. AmH. AON. ANN. Ho. ho. . And. @Na . Am". mo. Nb 00. Ave. Adm. Amd. oo. ANN. add. No. vo. moi mo. Admi moi flu Avv. Amn. ANN. AoH. ANv. Amm. no. no. u wo. Ava. 00. Ho. 3. mm AmN. ho. Amd. oo. ANv. Anu. 50. mo. . no. I ANN. #0. ANN. mo. Nm AFN. Awn. AON. ANN. Amm. Amw. Umo. vo. I mo. I AmH. mo. Ana. wo. Hm gang MHH. oo. ANN. flea. mo. #0. Moo. no. ANN. i No. Amd. Amm. Audi fig 63. ANH. moo... ao. No. moi ac... v0.1 no. AmN. NHH.- AmN.- AmH. Ava... USA—64 88% mo.: Ana. #0.- #0. $0. moi mo: ANN... AMN. 3N.u Ammf Amn.- add. 038‘ Ava. vo. .. AMH. mo. u Ava. ANN. Amd. No. GHH. u AHN. n mo. AON. vo. gag No.u ho. MNH. mo. co. no. mo. AwH. AmN.» Amm.- mo. mo. no. m2 AhN. Amvi AmH. Adnf .3. ANN. Ana. Awm. AmH. Amni AQN. co. vof E4562 moi Ho. mo. moi NH. mo. mo. Audi ed... 3a. vo. no. vo... Egg gags «Rd. E a: E E E a S E 85:. 62864 owned: gamma 8 852 68 388% m4 magma 45 had a significant and positive relationship with preference scores for humanscape slides (r=.44, p<.01). It is inferred that the more one preferred the humanscape slides, the more agreement one reported for the observed levels of environmental development. Another statistical test, regression, was employed to find. out ‘Which. independent 'variables were significantly related to preference, familiarity, and acceptable levels of environmental development. As can be seen in the Table 4-7, nationality' was found to be a useful variable in the prediction of preference as was familiarity for each landscape type except familiarity' with waterscapes (F2) (Bm;.23, sz=-22: BP3=.20, EFF—.62, BF3=-.96, p<.05) . Residence was another ‘Variable related. to jpreferences and familiarity (except familiarity with mountainscape) (Bp1= .13, BP2=.09, Bm=.10, Bm=.19, Bm=.10, p<.05). Familiarity was confirmed.to be significantly and positively related to preferences in this regression analysis (Bplm=.23, 131225.122 =.11, BP3&F3=.24, p<.05). Only nationality was a significant predictor for acceptable levels of environmental development (HDL) (Ehm=.29, p<.05). Combining the correlation and regression statistical results, it was confirmed that familiarity' had a positive and significant relationship with landscape preference. It was also concluded that nationality and residence not only had significant correlations with preference, but also were good predictors of landscape preference within the regression Table 4—7 Wis Stimuli Factor P1 P2 P3 F1 F2 F3 HDL Multiple R 34 .33 .34 .40 .28 .50 .26 F? .11 .11 .12 .16 .08 .25 .07 Adjust R? .11 .11 .11 .15 .07 .25 .07 F .40 14.34 15.21 43.69 13.08 51.30 34.17 B value Gender -— —- —— -— .14 -- -- Nationality .23 .22 .20 -.62 —— -.96 29 Age -- —- -— —— -— -- -— Residence .13 .09 .10 —— .19 .10 —— Academic -- —- .06 -- -- -— -- Status Academic -- — 12 —- —— —— -- -- Major Travel -- -— —— .10 .09 .09 —— Experience F1 .24 F2 .11 F3 .24 Note: A stepwise regression procedure was used. -- not significant at p<.05 47 analysis. The travel variable contributed to the prediction equation only for familiarity, It indicated that the travel experience would not influence personal landscape preference but would influence the degree of familiarity with the landscape slides. In. evaluating' the lacceptable level of environmental development observed in the slides, only nationality was a significant predictor. Although gender, academic status, and.academ c major had some association with preference and familiarity in selected types of landscape, these results could be attributed to their intercorrelation with personal attributes. Given the purposes in this study, to focus on.cultural differences, only the independent variable, nationality, will be discussed in the detailed analysis to follow. In.summary, the two hypotheses in this study were tested and. confirmed” When. cultural backgrounds were different (Western vs. Eastern), landscape preferences differed in certain types of landscapes. Higher degrees of familiarity with specific landscapes are associated with stronger landscape preferences. In the following section, detailed physical components in each type of landscape slide will be examined and compared across the two groups . 48 DifferencesAcrcsalandscaneJims Momtainscape There were three dimensions under study within the mountainscape slides presented: superior, normal, and inferior positioning of the viewer (Appendix A) . Each dimension was presented in five landscape slides. As seen in Table 4-8, American students preferred slides representing superior and normal Observation positions more than'Taiwanese students did [tm£(467)=-4.06, tmm(470)=-3.09, jp<.01]. There is no difference in landscape preferences for the inferior viewing position. Due to the fact that most of slides in.this type of landscape were taken in Taiwan, Taiwanese students were more familiar with these landscape slides compared to Americans [th(465)=9.55, tm,(470)=4.53, t,,,(465)= 4.89, p<.01]. Table 4-8 Ur U... o c of P ‘_‘ ‘1 ‘ f0 '11.. 0.“ '. ' ' a W Stimuli Subject M1P** M1F** M2P** M2F** M3P M3F** Taiwanese M 3 . 12 3 . 08 3 . 33 2 . 87 3 . 25 2 . 73 Si: 0.69 0.73 0.67 0.67 0.59 0.61 American M 3.40 2.34 3.54 2.54 3.29 2.40 SQ. 0.83 0.92 0.78 0.93 0.75 0.88 Ngte: M1: Superior position; M2: Normal position; M3: Inferior position P: Preference F: Familiarity; Using t—test, * p<.05, ** p<.01 5 slides in each position 49 The normal viewing position had the highest preference score for both groups . The preference sequence from most preferred to least preferred for Taiwanese was normal position, inferior position, and then superior position [tmpmp(282)=-6.37, tmpmp(283)=-4.01, p<.01; and tmpwp(283)= 2.57, P<.05] . For American students, the order of preference was from normal position, superior position, and then inferior position [tM,,W,(183)=-3.11, tmm,(181)=2.88, and tmpwpusz) =5.7, P<.Ol]. W Three geographic criteria were proposed in this type of landscape: distance between viewer and water (near or far away), the landscape which combined mountain and water elements, and water condition (moving or still) (Appendix A) . As seen in Table 4—9, Americans reported higher preference scores in the near distance range, the far away distance range , the combined landscape , and the moving water dimension [tm(474)=-5.44, thp(474)=-3.21, tmp(474)=-3.92, tw,,,(474)= -6.26, p<.01] . Landscape slides with combined mountain and water scenes had the highest scores for both groups. No matter which distance range was varied (W1 or W2) , Americans had higher preference scores than did Taiwanese (p<.01) . In addition, all subjects preferred the near distance range over the far away distance range [tmpmp (286)=6.66, p<.01 for Taiwanese; tmpmp (188)=8.17, p<.01 for Americans). we» 50 lo‘ u-qo - o - 0f Pr‘ - ‘1 c 0 11C 2“. '2 Stimuli Subject W1P** W1P** W2P** W2F W3P** W3F* W4P** W4F WSP W5F** Taiwanese M 3.81 3.21 3.56 2.82 4.42 2.48 3.55 3.11 3.96 2.82 Sd 0.65 0.87 0.71 0.74 0.58 0.92 0.66 0.75 0.72 0.89 American M 4.16 3.59 3.78 2.87 4.63 2.73 3.96 3.08 4.00 3.53 3d. 0.74 0.94 0.76 0.90 0.61 1.17 0.75 0.84 0.72 0.97 mete: W1: Near range of distance (less than 10 feet, 3 slides) W2: Far away ran e of distance (more than 1 mile, 3 slides) W3: combined lan cape (3 slides) W4: Moving water (4 slides) W5: Still water (2 slides) P: Preference F: Familiarity; USing t-test, * p<.05, ** p<.01 In water condition (moving or still), Americans liked moving water conditions better than the Taiwanese (p< . 01) . The Taiwanese had higher preference scores for still conditions than moving conditions [t (286) =—10.57, p<.01] , while there was not much difference in preference scores for the American group. Because eighty percent of landscape slides in waterscape were taken in America, Americans reported higher familiarity scores than did Taiwanese, except for the far away distance range and still water dimension [th(474)=-4-47: tw5F(474)=—8.15, p<.01; th(474)=-2.59, p<.05]. This could be due to the fact that there was one landscape slide taken in Taiwan in both far away distance range and still water dimensions . 51 Himanscape 'Three criteria were compared in.this category: degree of human habitation in the natural environment, condition of road construction, and level of trail development (Appendix A). There were significant differences in. preference scores between the two groups for human habitation and the road construction dimensions [tHDlP(471)=3.08, tHsz(474)=-5.66, p<.01; Table 4-10] . For the human habitation dimension, Americans indicated a stronger dislike for human influenced construction in the natural environment compared to the Taiwanese students (p<.01). Taiwanese disliked landscape slides that showed road construction, while Americans had a higher preference score for this dimension (p<.01) . Table 4-10 Ur U‘go _ o ‘ o ' ‘,‘ ‘0 ‘ on- ' ' or”. : ‘9 o... ‘ WW Stimuli Subject HD1P** HDlF** HDlL HDZP** HD2F** HD2L** HD3P HD3F** HDL3** Taiwanese M 2.59 3.22 2.75 2.70 3.64 2.92 2.36 3.17 2.66 3:1 0.57 0.74 0.53 0.72 0.83 0.83 0.60 0.80 0.61 American M 2.39 2.38 2.82 3.11 3.27 3.46 2.34 2.24 3.01 5d 0.81 0.88 0.72 0.87 1.10 0.74 0.85 0.86 0.85 Ngte: HDl: Human habitation (4 slides); HDZ: Road construction (2 slides); HD3: Development level of trail (4 slides); P: Preference F: Familiarity; L: Acceptable levels of development Using t-test, * p<.05, ** p<.01 52 In regard to familiarity with the humanscape slides, Taiwanese were more familiar with all three dimensions than American were because all landscape slides were taken in Taiwan [th(471)=11.12, tm,(474)=4.17, tm3F(470)=12.04, p<.01] . Regarding acceptable levels of environmental development , significant differences occurred in road construction and the development levels of trail dimensions (tHDZL(474)=-7.31, tm3L(468)=—5.22, p<.01) . Americans accepted these development levels of road construction more than did the Taiwanese. Besides, Americans agreed with these levels of trail development more than did the Taiwanese. In further analysis of the human habitation dimension (Table 4—11, and Appendix A) , based on a single slide for each specific level of landscape character, Taiwanese preferred the landscape with a single cabin in the natural environment more than did the Americans [t (472) =12.07, p<.01] . On the other hand, a large temple in the natural environment was preferred by Americans more than the Taiwanese [t(474)=-4.16, p<.01] . When comparing the preference for a single cabin with a single temple in the natural environment, Taiwanese preferred the single cabin to the single temple [t(284)= 24.33, p<.01), while there was not much difference for Americans. The small village was more preferred for both groups when it was compared with the over—developed community [tmiwanese (286) =6.76, p<.01; tmman (187)=10.64, p<.01). 53 Table 4-11 h- f; ‘ ‘ ‘J. ‘ _ or o o. a !-ll.'....-. v.9: Inc... or Stimuli Subject American Taiwanese Himan.habitation Single cabin (#7)** 2.66 3.83 Single temple (#8)** 2.64 2.26 Small village (#5) 2.43 2.33 Over-developed.community'(#6) 1.84 1.93 Road construction well—paved (#4)** 3.50 3.16 Rough (#9)** 2.71 2.23 Development level of trail .A trail along the mountain (#2)* 2.90 3.09 Small scale hiking trail (#1)** 2.18 2.48 High-developed route (#3) 2.34 2.30 Over—developed route (#10)** 1.95 1.57 mote: Using t—test, * p<.05, ** p<.01 For the road construction dimension, Americans gave higher preference scores to both road construction conditions compared to the Taiwanese (twnremn (474)=-4.01, tnwm(474)=‘ 5.20, p<.01). well—paved road conditions were more preferred for all subjects over the rough condition [tmwanese (286 ) =13 .66; tmncan(188)= 10.64, p<.01]. For the development level of trails, significant differences occurred between the groups for the small scale hiking trail [t(472)=3.34, p<.01]; over-developed route [t(472)=2.09, p<.01]; and a trail along the mountain [t (357.17) =—1.99, p<.05] . When the degree of route development increased (from high—developed to over—developed), the 54 preference scores decreased for both the Taiwanese and Americans [teimse(285)=14.34; tmican (187)=7.22, p<.01)]. Similar patterns were displayed in the reported opinions of acceptable levels of environmental development (Table 4— 12) . In the human habitation dimension, there were significant differences in opinions of acceptable levels in single cabin [t (471)=6.34, p<.01] , single temple [t(474)=-7.18, p<.01] , and small village [t (472)=-4.12, p<.01] slides. A single cabin was more acceptable than a single temple for the Taiwanese group [t (284) =17.76, p<.01] , while there was not much difference for the American group. The small village was more acceptable than the over—developed community regardless of cultural background [tmimse(286)=2.43, p<.05; tame... (186): 7.59, p<.01]. For the road construction dimension, there were significant differences in acceptable levels of development between well-paved and rough surface road for both groups [twen_paved(474)=-4.02, tmugh(474)=—7.47, p<.01] . No matter which road condition, Americans had higher agreement than did Taiwanese. The well-paved condition was more acceptable than the rough road condition for all students [t’I‘aiwanese (286) = 11.02; tmrican(l88)=5.46, p<.01]. For the dimension of trail development, there were significant differences between groups for the small scale hiking trail [t(470)=-2.33, p<.05] , high—developed route [t(473)=-4.95, p<.01] , and the over—developed route 55 Table 4-12 :3 ‘1 : ‘9 a, ‘ r ‘ _ 0, rtvi,01u:o ! mt ‘ amuse .0 a H E' . Stimuli Subject American Taiwanese HUman.habitation Single cabin (#7)** 3.10 3.73 Single temple (#8)** 2.96 2.34 Small village (#5)** 2.93 2.55 Over-developed community (#6) 2.28 2.37 Road construction well-paved (#4)** 3.68 3.32 Rough (#9)** 3.23 2.51 Development level of trail A.trail along the mountain (#2) 3.46 3.46 Small scale hiking trail (#1)* 3.11 2.87 High-developed route (#3)** 3.04 2.61 Over-developed route (#10)** 2.42 1.69 mote; Using t-test, * p<.05, ** p<.01 [t (4'73)=-7.86, p<.01] . When the development level of the trail increased, the opinions of acceptance for both groups decreased. Except the similar responses for both groups in the condition of a trail along the mountain, Americans reported higher development acceptability scores in the other three conditions than Taiwanese did in this dimension. DISCUSS ION AND CONCLUSIONS W In this study, differences in landscape preference and familiarity were tested among American and Taiwanese students, as well as acceptance of various levels of development in natural environments . Significant differences in landscape preferences were observed between Taiwanese and American subjects for selected landscape types: mountainscapes and waterscapes . It is suggested that when cultural backgrounds are truly different (such as language, belief, custom, and behavior), the landscape perceptions and preferences are different . Familiarity scores showed modest, but significant and positive relationships with landscape preference ratings. Although this result supported the hypothesis in this study, slide selection was a major factor influencing the degree of familiarity. This result could provide some indications of why familiarity had either a positive or negative relationship with landscape preference in previously studies . In addition, the comparison of acceptable levels of environmental development across the two student groups suggested differences as well. All subjects, regardless of the cultural background, preferred natural landscapes to the landscapes which had been influenced by humans. It is implied that artificial constructions were not as preferred as natural 56 57 elements in the environment . For mountainscapes, the students preferred the normal position for viewing as seen in the slides. Standing at the top of a hill seemed more attractive than the inferior position for the American group, while Taiwanese had reverse ratings for these two positions. Two explanations are worth considering. One explanation could be the influence of the different geographical features in the students' respective home environments . Mountainscapes are not major geographical features of the Great Lakes region. People who live in these larger flat places do not have much chance to stand in a higher position where the view is wide open. The other reason could be cultural. Since the philosophy of Chinese culture is to prefer nature, respect nature, and imitate nature, standing at the top of a mountain to prove humans' confidence and to overcome nature are not the main themes of Chinese thoughts . From Chinese paintings, we can easily find many examples of this philosophy. Human figures are always painted to be very small in Chinese art. In his book, MW, Lin (1937) described the details of Chinese painting: In a Chinese landscape called 'Looking at a Mountain After Snow,‘ it is very difficult to find the human figure supposed to be looking at the mountain after snow. After a careful search, he will be discovered perching beneath a pine tree -— his squatting body about an inch high in a painting fifteen high, and done in no more than 58 a few rapid strokes. (p. 281) In Chinese art, as illustrated by this example, human subjects are not a main point to be emphasized in natural environments, although humans are major characteristic in the whole story. In contrast, humans are representatives of God in Western religious thought . Examples which emphasis human dominating the nature can be found easily in many religious paintings . In a medieval miniature that depicts Adam naming the beasts, this picture not only indicates Adam (human) has the privilege of naming the beasts but also it puts Adam character in the superior position of the picture (McAndrew, 1993) . With respect to waterscapes, a landscape which combined mountains and water had the highest preference scores . This suggested that a combination of physical elements in a landscape setting was more attractive than a single major geographical feature in a landscape . For the American group, the distance factor seemed more important than the water condition factor. Taiwanese preferred to be near water, and were attracted by the still water condition. It is inferred that people are likely to be near the water to enjoy what experiences water could provide (such as fresh and moist atmosphere, dynamic conditions, and the inspiration of water sounds). Besides, if the water scene was near by, Americans were not concerned about whether the water was moving or 59 still. Although Taiwanese like to be near water, safety and peaceful feelings (still water condition) were most preferred by Taiwanese. In summary, the excitement of being near the water scene was attractive to the Americans, whereas the more peaceful and quiet aspects of the water scenes were more acceptable for Chinese people. In many landscape projects, we see waterfalls or fountains (moving water conditions) as principal objects in Western style gardens, while ponds or small scale lakes (still water conditions) are more prevalent in Chinese gardens to attempt to imitated natural scenery. Humanscape was the least preferred landscape type among the three landscape categories shown to students . Obvious human influences in the natural environment were not preferred by the students . Preference ratings had a positive relationship with reported acceptable levels of environmental development . It was indicated that the higher the preference score for a particular humanscape scale, the more agreement among subjects with respect to the degree of development. Based on slide comparisons, lower levels of development were more preferred than higher levels of development for both Taiwanese and Americans in each dimension of humanscape. Besides , when the degree of human development goes beyond acceptability, Taiwanese reported stronger reactions than did Americans. One possible explanation is that natural resources are very rare and precious in Taiwan. So, Taiwanese pay more attention to or more sensitivity to environmental development 6 0 issues, and are more aggressive regarding these problems . Although most Americans are knowledgeable about environmental development and conservation issues, the American subjects did not report as strong reaction as Taiwanese subjects did to the scale of increased human development observed in the slides . In summary, both American and Taiwanese subjects had common preference patterns in this study such as they all preferred waterscapes among these three types of landscape. When we look into the detailed physical components, it is suggested that the preferences for and opinions about environmental development issues for specific landscape types have been influenced by cultural background variables . Different cultural backgrounds appear to influence people ' s perceptions of landscapes. In the Chinese mind, by association with nature's enormities, a man's heart may truly grow big (Lin, 1937) . The experience of nature for Chinese is not only enjoying the natural scenic beauty but also respecting what nature could give to us. Lin (1937) described what the experience of nature could be: There is a way of looking upon a landscape as a moving picture and being satisfied with nothing less big as a moving picture, a way of looking upon tropic clouds over the horizon as the backup of a stage and being satisfied with nothing less big as a backdrop, a way of looking upon the mountain forests as private garden and being satisfied with nothing less as a private garden. (p.282) 61 1 J . E J' !' There are a number of findings in this study that provided noteworthy insights for the landscape planning and design profession. The finding that cultural differences may influence people's preference in certain types of landscapes should be paid attention to . When designing a landscape project involving cross-cultural issues, such as a resort area for' international tourist, decisions should. .not Ibe subjectively based on one's own cultural background. Viewing perspectives need to be well considered when designing a scenic point for tourists in a recreation area. Providing“ a landscape that combines 'varied. geographical elements such as hills, water, vegetation would be more attractive than a scene containing only one single element . In other words, juxtaposition (a series of arrangements of scenery) is a major consideration in increasing people's attention. The water situation is another main issue in landscape design. In Yang's study (1988) , combining water with vegetation enhanced landscape preference. Since water is a. popular element for people, providing the opportunity to let people to be near water to see, feel, and hear the nature of water would enhance people's landscape preference, and then increase the degree of satisfaction. Natural scenes are more acceptable and preferable than human- influenced landscapes for people. It is suggested that 62 reducing human influences to provide optimum outdoor experiences for people is important. HOwever, some artificial development in a natural environment is unavoidable in our real world. Designing facilities to make them more appropriate in a natural setting through aspects of color, texture, and form is a proper way to provide less interruption to the natural environment , and produce both a high quality of scenic beauty and the satisfaction of human needs in landscape planning'projects. Because slide origin affects degree of familiarity, and because familiarity is associated with landscape preference, the results suggested that selection of materials or symbols of stimuli in a landscape perception study is a very important variable which is related to people's preference. For example, in order to avoid. the influence of slide origin 'upon familiarity, stimuli should be taken from a single country, or choose stimuli which are totally unfamiliar to the subjects. This type of research design would provide results to better understand perceived familiarity and its relationship with landscape preferences. According to a previous study (Vining et al. , 1984) , human influences such as temporary artifacts, feature incongruity, building materials, and residential crowding detracted from.scenic quality in.forest developments. In.this study, several landscapes obviously' influenced. by' human developments (human habitation, road construction, and 63 development level of trail) were presented. They were presented to evaluate preference scores as well as the degree of acceptability of human development in natural environments . It was observed that human intrusion in natural environments was an important issue related to people's landscape perceptions and preferences . Although the criteria we used to measure human intrusion were subjective, these variables could still provide a heuristic "first cut" to further an understanding of human conceptions of the environment (Vining et al. , 1984) . In other words, using subjective criteria to measure human influence in the natural environment still provides useful and meaningful results in landscape perception and preference studies . W Although the proposed hypotheses were confirmed, future studies are still needed to overcome the limitations of this study. Previously, researchers focused upon cross-cultural differences in landscape preference have only paid attention to the differences of physical environment. Discussions of why and how cultural factors influence landscape perception have been limited. Even in this study, theory was limited, with minor references to the strong or weak relationships between culture and environment (Berry, 1975) ; what is needed are more in—depth theoretical analysis of the relationship between culture and landscape perception. 64 Sampling is a very important step in the research process, especially in.studies of cross-cultural differences. In this study, limited by time and budget, American and Taiwanese students were selected, representing only a small portion of American and Taiwanese cultures. FUture studies need to include samples of business people, laborers, the elderly, and children. In Yang's study (1988), he found that age is a major factor associated with preference. Age could not be explored as a variable to estimate preference in this study because the sample was homogenous in terms of age. The landscape types represented in this study included mountainscapes, waterscapes, and humanscapes . These scenes were too general to produce an analysis of detailed differences based.on landscape features such.as land form.and texture. In addition, the aesthetic attributes such as color and degree of brightness, and psychological factors (mystery or complexity) which also play important roles in assessing landscape preference were not measured in this study. For example, a study reported that "The results point to the importance of using different predictor domains, rather than relying exclusively on.any one, since their role in different environmental contexts is likely to vary" (Kaplan et al. , 1989, p.509) . Future studies should focus not only on physical features but also upon systematic differences in aesthetic and psychological attributes in order to understand landscape perception comprehensively . 65 The demographic variables such as residence and travel variables showed significant correlation with landscape preference. In their study, Tips and.Savasdisara (1986) found small differences of opinions related to the residence factor, and travel variables had a minor influence on the visual evaluation of landscape photographs. Since these two variables were not a major consideration in this study, future studies could test the validity of these two variables in detail as predictors of landscape preference. In the world today, viewing scenery is one of the major outdoor recreation activities which is related to landscape perception and preference . Nature provides magnificent scenery and opportunities for people during their leisure time to pursue a recreation experience. The object of landscape preference research is to reveal the charm and interest of nature for people. The role which landscape professionals play is to protect and preserve the integrity of these natural features as well as the cultural and natural visual resources. The research techniques and. theories used, and results presented.in.this study offer useful information to those who ‘manage landscape resources. In.doing so, it provides a chance to better understand visual experiences across cultures. APPENDIX A STIMULI 2%.“. APPENDIX A: STIMULI Mountainscape Slides Superior position (M1) # 4 Americans: 3.13 Americans: 3.65 Americans: 3.37 Taiwanese: 3.13 Taiwanese: 3.09 Taiwanese: 3.07 # 6 # 15 Americans: 3.51 Americans: 3.36 Taiwanese: 3.60 Taiwanese: 2.72.** Normal position (M2) "- ~ . '39.? # 7 # 8 # 9 Americans: 3.67 Americans: 3.63 Americans: 3.44 Taiwanese: 3.32 Taiwanese: 3.80 Taiwanese: 3.14 #11 #13 Americans: 3.16 Americans: 3.81 Taiwanese: 2.97 Taiwanese: 3.47 Inferior position (M3) #2 10 Americans: 4.34 * Americans: 3.41 Americans: 2.99 Taiwanese: 4.27 * Taiwanese: 3 06 Taiwanese: 2.76 '1 #12 #14 Americans: 3.20 Americans: 2.55 ** Taiwanese: 3.31 Taiwanese: 2.88 Note: * the highest preference score ** the lowest preference score m (D waterscape Slides Di stance Dimension (W1 5. W2) Eeancategory (less than 10 feet, W1) '# 5 # 4 # 2 Americans: 4.19 Americans: 3.93 Americans: 4.37 Taiwanese: 3.60 Taiwanese: 3 61 Taiwanese: 4.24 Eat_catcgcr¥ (more than one mile, W2) # 1 # 3 # 8 Americans: 4.23 Americans: 3.64 Americans: 3.48 ** Taiwanese: 4.07 Taiwanese: 3.69 Taiwanese: 2.93 ** # 6 # 7 # 9 Americans: 4.70 * Americans: 4.70 * Americans: 4.51 Taiwanese: 4.57 Taiwanese: 4.58 * Taiwanese: 4.10 ON \0 W (W4 " W5) Milling—Wm (W4) # l 4? 4 # 5 Americans: 4.23 Americans: 3.93 Americans: 4.19 Americans: 3.48 ** Taiwanese: 4.07 Taiwanese: 3.61 Taiwanese: 3.60 Taiwanese: 2.93 ** mum (W5) # 2 # 3 Americans: 4.37 Americans: 4 24 Taiwanese: .64 .69 3 Taiwanese: 3 Note: * the highest preference score ** the lowest preference score 7O Humanscape Slides Human habitation (HDl) # 7 (single cabin) # 8 (single temple) Americans: 2.66 / 3.10 Americans: 2.64 / 2.96 Taiwanese: 3.83* / 3.73* Taiwanese: 2.26 / 2.34 # 5 (small village) # 6 (over-developed community) Americans: 2.43 / 2.93 Americans: 1.84** / 2.28** Taiwanese: 2.33 / 2.55 Taiwanese: 1.93 / 2.37 # 4 (Well—paved) # 9 (Rough) Americans: 3.50* / 3.68* Americans: 2.71 / 3.23 Taiwanese: 3.16 / 3.32 Taiwanese: 2.23 / 2.51 71 level (HD3) # 2 # 1 (a trail along the mountain) (small scale hiking trail) Americans: 2.90 / 3.46 Americans: 2.18 / 3.11 Taiwanese: 3.09 / 3.46 Taiwanese: 2.48 / 2.87 # 3 # 10 (high—developed route) (over-developed route) Americans: 2.34 / 3.04 Americans: 1.95 / 2.42 Taiwanese: 2.30 / 2.61 Taiwanese: 1.57** / 1.69** Note: 1. preference score / opinion of acceptable level 2. * the highest preference score ** the lowest preference score APPENDIX B QUESTIONNAIRE 72 APPENDIX B : QUESTIONNAIRE ENVIRONMEN'I‘AL PERCEPTION AND PREFERENCE -- A CROSS-CULTURAL STUDY -- I am conducting a research on the differences in environmental perceptions and preferences between American and Taiwanese student . I would appreciate your help and participation in this survey. Your answers and opinions will help us understand people's perception of landscapes and evaluate the quality of natural environments in Taiwan and the US . This survey is voluntary, you are under no obligation to participate. But, we hope you will find the survey to be an interesting experience and will consent to participate. Even if you participate, you may choose not to answer certain questions. All results will be treated with the strictest confidence and will remain anonymous in any report or discussion. Ying—Hung Li Graduate Student Department of Park & Recreation Resources Michigan State University East Lansing, MI 48823, USA If you agree to participate in this survey, please fill in these blanks: Date of Survey Place of Survey 73 Instruction In this study, we would like to know your personal preferences and attitudes about the natural environment . There are four parts in this questionnaire. The first two parts ask your'preferences and.familiarity with.two types of landscapes (mountainscape and waterscape) as shown in the slides. The third part asks your personal attitudes toward the level of acceptable development of the natural environment along with your preferences and familiarity for each slide. The final part asks some questions about your personal background and travel experience. The preference scores are given in terms of a liking scale (whether you like it or not). The rating scale is from 1 (do not like it at all) to 5 (like it very'much). The rating of familiarity is measured in terms of the degree of familiarity (are you familiar with this type of landscape) with each slide. The rating scale is from 1 (not familiar) to 5 (very familiar). The attitudes toward the level of acceptable development of the natural environment is measured by way of asking the question "This level of development of the natural environment is acceptable." at each landscape slide. The rating scale is from.1 (strongly disagree) to 5 (strongly agree). Slide will be shown during the first three parts of survey. The investigator will call out slides number to help you correctly match your scores with the slide. 74 For each type of landscape slide, fill in the number that best represent your feeling about the landscape. Part 1 -- In this part, we would like to know your preferences of mountainscape, and the degree of familiarity with each mountainscape slide. Preference : Please rate how much do you like each landscape slide. "I like this landscape . not at all a little somewhat quite a bit very much Familiarity: Please rate how familiar you are with each landscape slide. "I am familiar with this type of landscape." 1 ------------ 2 --------------- 3 --------------- 4 --------------- 5 not very a little somewhat quite a bit very slide # . reference familiari 75 Part 2 -- In this part, we would like to know your preferences of wittceirscape, and the degree of familiarity with each waterscape s l e. Preference : Please rate how much do you like each landscape slide. "I like this landscape n not at all a little somewhat quite a bit very much Familiarity: Please rate how familiar you are with each landscape slide. "I am familiar with this type of landscape." slide # ference familiari l 2 3 4 5 6 7 8 9 76 Part 3 -- In this part, we would like to know your preferences and familiarity of development level, and the level of acceptable of environmental development with each slide . Preference : Please rate how much do you like each landscape slide. "I like this landscape n not at all a little somewhat quite a bit very much Familiarity: Please rate how familiar you are with each landscape slide. "I am familiar with this type of landscape." not very a little somewhat quite a bit very Level of acceptable development: Please rate your personal opinion for each slide concerning this question: "This level of development of the natural environment is acceptable." 1 ------------ 2 ---------------- 3 --------------- 4 ---------------- 5 strongly a little neutral a little strongly disagree disagree agree agree Slide # preference familiarity this level of development is acceptable \OGDQOWU'IIbUJNI-J * H O 77 Part 4 —- Finally, we would like to know your personal background. * 'k * Sex: male female * Nationality Age : Where did you live before the age of 14: urban area suburban area rural area Are you: graduate student senior junior sophomore freshman What is your major: HOW many countries have you visited during your travels? please list those countries you have visited: 78 Z‘Ffiiitfififiiflifl&fifliflfl§§fi% 129.34} ' fifiEEifiEg—IEEBMQTWXitfififlflgi‘fifififififiwig QS‘IEEE‘EFETIE‘IZHFE. ’ aaammammam ’ ffimfifiliéifflfififibfi? fifififiIfi‘lS‘UttfifififififlqTfiE ' D1&i§“$ifi§§tfirfixgflq¥hfi Ekflfifib ’ RUEFa?fi$§fifii§§flfi§ifi§fiiifififi9fi5i§fi ’ €33 ° I‘ififlfi ZKFnfifip‘J’éigfiffiTflfifiA‘i‘iififfiB’fi-fiifllifih} ’ ~$§fififiiflflkgfi (53‘ ° g_~ Eé‘fiififéiflflfiflfifi‘fiflififiBfigfli‘gfgifififififg ° fifififii‘ EEEE 1 ’7} HE’SZ‘EE) EU 5 5% ( 315783232X) ° fifiiflfififfifla 1’7} (% 22:54.55.) Ed 5 ’7} UPSEAE) ° Efiiafii’r} ' mmamamasnasa ffififififi’ifié‘ififfi ’ E816} (seams) £45 ’7} (fcéfifi) ' fiflg‘éfifi} 2%?EEIQBMIAEZI-i’éflfikfimfiifiéfifir° EEEJESEECP ’ fiffififiEE—ififififlifififi’flkfii ’ LXEMIEEEE FfiEfiZ‘EfififlE ° Efflflé‘ififiififiZfifi ' asraaammms ° BM: ff- FJ E1 £111.85: 79 E—E‘M} = fifi'F’i'HE-zfi ’ smasmsaamamaammsaa ° SEE? = ” it aaaamaa ” mamas rages saga an it"s:afix mass = ” if aaaansa ” was mm: .124... .5. as 543% Efififg Efififilfi 2743?? EUEE AVEEE 1 9 10 11 12 13 14 15 $3.8M} = fifiifi‘Ffflfifi ’ 1%$1§¥i7kifiifififl4lEfifigffifiififififiifi ° amaa=”fi smashes” 1. 2. 3. 4. . 5. mamas: Txifi seas as steam asasz”a smashes” 1. 2. 3. 4, 4.3. grams miss: aegis. as. steam. mar: auras may: as): amass mass 1 6 2 7 3 8 4 9 5 8O $381316} = fi‘ifiT‘iUgi—fi ’ Etflffii‘iAfififiWEfii‘Effi ’ nasamas Effififififi’flfifififfi ° afififfi = ” if smashes ” was in sea» 44 as mass = ” 3‘32 Eifififizflafi ” 3E§¥24fi Tigififi aegis. $12.91? fifiaziAE Efiifiififi’flfifififiifi = ” EEEZIXEHT ' fitfifilfifl’flfifififilbfii‘fi 89 ° 1 ° 3 4 5 fiéfiqfilfi Mama ifififi‘fi fififfilfi fifia‘lfilg 271.3% Efifiilfi fii‘fiffi Efifififfi 2313}? Efii‘EE fifiEE 55321313 1 6 2 7 3 8 4 9 5 10 amass: Eff ’ fiflfiTfiE!§ffilAB’J-§z$§fi&fifiififl% ° 1.1$B|J=% it 21%: 3.45%: imsuaum’sem mama : Ffi’i‘fiittE : éfifiTiflqE 5.131.351: = W%$_: 7cm__: 7::—:____: 7::__v x—_____° 631%qu = Lama'masam : —#£§@%§(El2f€= SEWEI8 = APPENDIX C FINDINGS FOR EACH SLIDE BY CONDITION 81 .APPENDIX C: FINDINGS FOR EACH SLIDE BY CONDITION Table C-1 E . Eil' I II ! . 3].: slide #. preference / Sd familiarity / Sd 1 3.13 / 0.99 2.20 / 1.08 2 * 4.34 / 0.78 2.83 / 1.10 3 3.41 / 1.01 2.39 / 1.09 4 3.65 / 1.02 2.44 / 1.09 5 3.37 / 1.10 2.51 / 1.22 6 3.51 / 1.05 2.33 / 1.10 7 3.67 / 0.97 2.53 / 1.20 8 3.63 / 1.05 2.46 / 1.08 9 3.44 / 1.05 2.34 / 1.10 10 2.99 / 1.23 2.32 / 1.18 11 3.16 / 1.13 2.62 / 1.17 12 3.20 / 1.06 2.41 / 1.08 13 3.81 / 0.97 2.76 / 1.13 14 ** 2.55 / 1.10 2.05 / 1.00 15 3.36 / 1.09 2.30 / 1.10 Total (means) 3.40 / 0.71 2.42 / 0.86 Note: * the highest score; ** the lowest score waterscape_Slides slide #. preference / SD familiarity / SD 1 4.23 / 0.85 3.34 / 1.18 2 4.37 / 0.74 3.99 1.12 3 3.64 / 0.93 3.06 / 1.12 4 3.93 / 1.02 3.84 / 1.21 5 4.19 / 0.99 2.93 / 1.14 6 * 4.70 / 0.64 2.76 / 1.28 7 * 4.70 / 0.62 2.73 / 1.28 8 ** 3.48 / 1.24 2.21 / 1.16 9 4.51 / 0.85 2.69 / 1.22 Total (means) 4.19 / 0.59 3.06 / 0.83 Note: * the highest score; ** the lowest score 82 II 3].: slide # reference / SD familiarity / SD this level of development is acceptable / SD 1 2.18 / 1.05 2.05 / 1.04 3.11 / 1.16 2 2.90 / 1.08 2.34 / 1.02 3.46 / 1.04 3 2.34 / 1.05 2.22 / 1.01 3.04 / 0.94 4 * 3.50 / 0.95 3.36 / 1.20 * 3.68 / 0.84 5 2.43 / 1.01 2.40 / 1.04 2.93 / 0.96 6 ** 1.84 / 0.99 2.40 / 1.20 ** 2.28 / 1.12 7 2.66 / 1.22 2.38 / 1.11 3.10 / 1.11 8 2.64 / 1.13 2.33 / 1.01 2.96 / 1.04 9 2.71 / 1.08 3.17 / 1.21 3.23 / 1.02 10 1.95 / 1.01 2.38 / 1.10 2.42 / 1.13 Total (means) 2.52 / 0.73 2.50 / 0.81 3.02 / 0.60 mote: * the highest score; ** the lowest score Table C-2 I . 3 1. I [I ! . 31' 3 slide #. preference / SD familiarity / SD 1 3.13 / 0.95 2.73 / 0.92 2 * 4.27 / 0.81 2.91 / 1.04 3 3.06 / 0.98 2.72 / 1.00 4 3.09 / 0.95 3.44 / 0.99 5 3.07 / 0.97 3.35 / 1.04 6 3.60 / 0.91 3.06 / 1.05 7 3.32 / 0.87 3.26 / 0.98 8 3.80 / 0.88 3.13 / 1.00 9 3.14 / 0.94 2.83 / 0.97 10 2.76 / 1.04 2.98 / 1.07 11 2.97 / 1.16 2.69 / 1.10 12 3.31 / 0.93 2.58 / 0.92 13 3.47 / 1.07 2.47 / 0.97 14 2.88 / 1.10 2.49 / 0.97 15 ** 2.72 / 0.99 2.85 / 0.96 Total (means) 3.23 / 0.56 3.40 / 0.71 bkflze: * the highest score; ** the lowest score 83 W slide #. preference / SD familiarity / SD 1 4.07 / 0.87 3.22 / 1.06 2 4.24 / 0.80 3.03 / 1.08 3 3.69 / 1.02 2.61 / 1.02 4 3.61 / 1.00 3.35 / 1.12 5 3.60 / 0.91 3.24 / 1.09 6 4.57 / 0.65 2.47 / 1.02 7 * 4.58 / 0.77 2.49 / 1.21 8 ** 2.93 / 1.04 2.64 / 1.15 9 4.10 / 0.87 2.48 / 1.01 Total (means) 3.93 / 0.51 2.84 / 0.68 Note: * the highest score; ** the lowest score slide # reference / SD familiarity / SD this level of devel ent is acceptable / SD 1 2.48 / 0.91 2.81 / 1.02 2.87 / 1.04 2 3.09 / 0.92 3.24 / 0.97 3.46 / 0.99 3 2.30 / 0.79 3.24 / 1.01 2.61 / 0.91 4 3.16 / 0.92 3.62 / 0.93 3.32 / 1.03 5 2.33 / 0.91 3.26 / 0.95 2.55 / 0.98 6 1.93 / 0.87 3.72 / 0.97 2.37 / 1.14 7 * 3.83 / 0.89 2.94 / 0.98 * 3.73 / 1.03 8 2.26 / 0.83 2.96 / 0.98 2.34 / 0.84 9 2.23 / 0.92 3.66 / 0.94 2.51 / 1.04 10 ** 1.57 / 0.74 3.41 / 1.04 ** 1.69 / 0.88 Total (means) 2.52 / 0.50 3.28 / 0.71 2.74 / 0.46 Note: * the highest score; ** the lowest score APPENDIX D STATISTICAL RESULTS IN FIGURES Mean Score 84 APPENDIX D: STATISTICAL RESULTS IN FIGURES P1** P2** P3 F1** F2** F3** HDL** American I Taiwanese Eigure_p;l The mean score for preference, familiarity, and acceptable levels of environmental development for each nationality Note: P1: Preference for Mountainscape F1: Familiarity for Mountainscape P2: Preference for Waterscape F2: Familiarity for Waterscape P3: Preference for Humanscape F3: Familiarity for Humanscape HDL: Acceptable Levels of Environmental Development Mountainscape: 15 slides; Waterscape: 9 slides Humanscape: 10 slides; Using t—test, * p<.05, ** p<.01 Preference Score 85 Superior(M1P)** Normal(M2P)** Inferior(M3P) .American I Taiwanese Eigore_p;2 The preference score for mountainscape slides Note: Using t—test, *4 m 1* 0 Single cabin** Small village Single temple** Over—developed community I Taiwanese American Ejgure D—S The preference score for the human habitation Nets: 1- dimension Each level was represented by one slide. The degree of human habitation ranges from one single cabin to an over—developed community. Using t-test, *p<.05, ** p<.01 Preference Score 89 Well—paved** Rough American I Taiwanese Figure D—6 The preference score for the road construction dimension Note: 1. Each scales was represented by one slide. 2. Using t—test, * p<.05, ** p<.01 Preference Score 90 A trail along the mountain* High-developed route Small scale hiking trailM Over—developed route“ American I Taiwanese Figure D—Z The preference score for levels of development of trail Note: 1. Each scale was represented by one slide. The degree of tail development is from a trail along the mountain to an over-developed route. 3. Using t—test, * p<.05, ** p<.01 Opinion of Acceptable Levels 91 2 1 0 Single cabin** Small village** Single temp1e** Over-developed community I Taiwanese American Figure D—8 Acceptable levels of environmental development for the human habitation dimension Note: 1. Each level was represented by one slide. The degree of human habitation ranges from one single cabin to an over—developed community. 3. Using t—test, * p<.05, ** p<.01 Opinion of Acceptable Levels 92 5 4 L 3.68 3 L 2 i l i O .1- ;rruvm “u‘ .a'i: .. Well—paved** Rough** I Taiwanese ' American Figure D—9 Acceptable levels of environmental development for the road construction dimension Note: 1. Each scale was represented by one slide. 2. Using t—test, * p<.05, ** p<.01 Opinion of Acceptable Levels 93 7 346 346 A trail along the mountain High—developed route** Small scale hiking trail** Over—developed route** I Taiwanese American Figure D—IQ Acceptable levels of environmental development for the dimension of trail Note: 1. Each scale was represented by one slide. 2. The degree of trail development is from a trail along the mountain to an over—developed route. 3. Using t-test, * p<.05, ** p<.01 LIST OF REFERENCES 94 LIST OF REFERENCES Altman, Irwin & Chemers, Martin (1984). culture_and enyironment. Monterey, CA: Brooks/ Cole Publishing Company. Berlyne, D. E. Robbins, M. C. & Thompson, R. (1974). cross- -cultural study of exploratory and verbal responses to visual patterns varying in complexity. In D. E. Berlyne (Ed. ), - 0. ‘_ r 0‘, ‘..1.0‘ u. I o. o. . o- _ .0 0J3. 0. v.0 0.. ‘ ‘ .0 0. 0‘5 O 2‘. 0‘ v.99 ‘ o. 0! (pp. 259— 278). New York: Wiley. Berry, J .W. (1975) . An ecology approach to cross-cultural psychology. WM, .10. 51-80. Canter, D. & Thorne, R. (1972). Attitudes to housing: A cross -cultural comparison. WW, .4, 3- 32 Chu, Jung- Shang (1986). WWW ‘__n.o .0 --99 o- 9 -‘ ._- o .."o- Unpublished Masters Thesis, Dan— —Jang University, Taiwan. Daniel T. C. (1990) . Measuring the quality of the natural environment: A psychological approach. American. 25.382211411991511, 45, 633- 637 Daniel, T. C. &Bolster, R. S. (1976). Measuring_landscape -- 9‘ ' Ir- ‘0 014- ‘ no. 09 u. "10.. (Research paper N0. RM—167). Ft. Collins, CO: U. S. Department of Agriculture, Rocky Mountain Forest and Range Experiment Station. Daniel, Terry & Vining, Joanne (1983) . Methodological issues in the assessment of landscape quality. In Altman, I. & Wohlwill, J. (Eds. ), v (pp. 39- 84). New York: Plenum Press. Gibson, James J. (1979). v' perception. Boston: Houghton Mifflin. Hammitt, W. E. (1979) . Measuring familiarity for natural envi ronrments through visual images. In Elsner , G. and Smardon, R. (Eds. ), W J 3 . E . . W helisnaLresources (pp.217— 226) . USDA Forest Service General Technical Report PSW- 35. 95 Hammitt, William E. (1981) . The familiarity— preference component of on— site recreational experience. .Leieure Sciences .4, 177- 193. Herbert, Eugene J. & Kaplan, Rachel (1987). Cultural and sub- cultural comparisons in preference for natural setting. Iandscaneandfirhanllanning .14. 281- 293 Herskovits, M. J. (1952). Manandhiuorks. New York: Konpf. Jellicoe, G & Jellicoe, s. (1987). Thedandscaneioimani 9'... .2 0‘ ‘0 .flll‘! .1“ .0 ‘0 - 0Q 0 0‘ ...‘_‘Q 0.19., New York: Thames and Hudson . Kaplan, Stephen (1988) . Perception and landscape: Conceptions and misconceptions. In Nasar, J. L. (Ed. ), .Enuronmentai ‘ 9‘ ' It 91. ‘ ‘ o .99 -99_ e 99- (pp.45— 55) . Cambridge: Cambridge University Press. Kaplan, S. & Kaplan, R. (1982). ' ' ' w . New York: Praeger . Kaplan. S. & Kaplan. R. (1989) WA psychological_perspectile. New York: Cambridge University Press. Kaplan, S. Kaplan, Rachel, & Brown, Terry (1989). Environmental preference: A comparison of four domains of predictors. W, 2.1, 509— 530. Laurie, Michael (1975) An_'1ntrrr11c_tion_t_o_lands_c_ape architecture. New York: Elsevier. Lee, Su-Sing (1983) . . Unpublished Masters Thesis, Taiwan Univers1ty, Taiwan. Lin. Yan-Jou (1979) Theiandscaneeraluatiortinmreation . Unpublished Masters Thesis, Chung— Shing Univers1ty, Taiwan. Lin, Yutaing(1(9)37.mrtance_o__1ming. New York: The John Day Company. Linton, D. L. (1968) . The assessment of scenery as a natural resource. WM, 84, 218— 238. Liu, James J. Y. (1962 Theammmesemetnc Chicagoz The University (of Chicago Press. 96 McAndreW. Francis T. (1993) . Enyirormentalpsirchology. Pacific Grove, CA: Brook/ Cole Publishing Company. Nasar, J. L. (1984). Visual preferences in urban street scenes: A cross— —cultural comparison between Japan and the United States. W, 15.. 79- 93. Nasar, J. L. (1988). ' W. New York: Cambridge Palmer, James F. (1986). Environment perception. In Smardon, Richard C. et al. (Eds. ), ' analysis (pp. 63— 80) . New York: John Wiley & Sons. Rap0port, Amos (1969). W. Englewood Cliffs, N.J.: Prentice— Hall, Inc. Rossbach Sarah (1983). Eeng_shui_._The_Chines_e_ari—_rrf placement. New York: E. P. Dutton. Shafer, E. L. & Tooby, M. (1973) . Landscape preferences: An international replication. JournalbileisureResearch, 5, 60- 65. Sonnenfeld, J. (1966) . Valuable values in the space and landscape: An inquiry into the nature of enVironmental necessity W, 4, 71- 82 Sonnenfeld, J. (1967). Environmental perce tion and adaptation level in the Arctic. In Iowenthal, D. Ed. ), Enyironmental . Department of Geography, University of Chicago. Taylor, Jonathan G. Zube, Ervin H. & Sell, James L. (1987). Landscape assessment and perception research methods. In Bechtel et al. (Eds. ), MethgrisnrnenirironmentaLand W11 (pp.361—393) . New York: Van Nostrand Reinhold Company . Tips, W. E. J. & Savasdisara, T. (1986 ) . The influence of environmental background of subjects on their landscape preference evaluation. Landscape_Elanning, 13, 125— 133. Tuan, Y. F. (1974). ' ' - ' ' - ' . Englwood Cliffs, New Jersey: Prentice Hall Inc. Ulrich, R. S. (1977) . Visual landscape preference: A model and application Manandjnyironmentjxatem l. 279- 293 97 Ulrich, R. S. (1983). Aesthetic and affective response to natural environment. In Altman, I. and Wohlwill, . (Eds. ), BehayioLandthenaturalenyironment (pp. 85-126).JNew York: Plenum Press. Vayda. A.P. (Ed.) (1969). W New York: Natural History Press. Vining, Joanne, Daniel, Terry C. & Schroeder, Herbert W. (1984) . Predicting scenic values in forested residential landscape. W 15., 124- 135. Wheeler, L. (1984) . The visual impact of 90 scenes. An intercultural comparison. Perceptionpithefinyironment, s, 7- 8. Yang, Byoung- -E (1988).A_cross;c111tnral_comparison_of _. ‘y‘ ‘0 ‘ ‘0 ‘o. .0. “_ ‘ ... A‘» ‘Q. o.._ '9‘ styles. Unpublished Ph. D. Dissertation. University of Michigan. Yang, Byoung- -E & Brown, Terry J. (1992). A cross- -cultural comparison of preferences for landscape styles and landscape elements. W. 24. 47l— 507 Yang, Byoung— —E & Kaplan, Rachel (1990) . The perception of landscape style: A cross- -cultural comparison. , 19, 251— 262. Zonneveld, Jan I. S. (1990) Introduction. In Svobodova, Hana (Ed.) , giltnralasractsoilaridscape (pp.7-12) . Wageningen, Netherlands: Pudoc . Zu, Nien-Tzu (1989). MW preference. Unpublished Masters Thesis, Taiwan University, Taiwan. Zube, E. H. (1984) . Theme in landscape assessment theory. W1, 3, 104-110- Zube, E. H. &Mills, L. V. Jr. , (1976). Cross— cultural explorations in landscape perception. In Zube, E. H. (Ed. ), Amh.e,rst MA: Institute for Man and Environment, University of Massachusetts. Zube, E. H. & Pitt, D. G. (1981) . Cross— cultural perceptions of scenic and heritage landscape. Wing, 8, 69- 87. Zube, E. H. Pitt, D. G. & Anderson, T. W. (1975). Perception and prediction of scenic resource values of the northeast. In Zube et al (Eds. ), Lariscareassessmantilaluesi 98 . (pp.151—167) . Stroudsburg, PA: Dowden, Hutchinson, & Ross, Inc . Zube, E. H. Sell, J. L. &Taylor, G. (1982). Landscape perception: Research, application and theory. Landscape Planning 9.1-33 Zube, E. H. Vining, J., Law, C. &Bechte1, R. (1985). Perceived urban residential quality: A cross- cu—ltural bimodal study. W, 11, 327— 350 I nI CHI an STATE UNI ill HI film;