Documente Academic
Documente Profesional
Documente Cultură
James F. Palmer
James E Palmer is Associate Profes- Abstract: Reliability is measured by whether an investigation will obtain similar results
sor and Undergraduate Curriculum when it is repeated by another party. It is argued that reliability is important at the level of
Director of the Landscape Architec- individuals. While all landscape assessments are based on individual judgments, they are fre-
ture Program at the State University quently aggregated to form co~npositejudgments. The use of inter-group, intra-gwup and inter-
of New York’s College of Environmen- rater measures of reliability in the landscape perception literature is reviewed. This paper inves-
tal Science and Forestry in Syracuse, tigates the reliability of assessing various visible landscape qualities using data primarily from
New York. He holds an M.L.A. degree previously published studies. The results indicate that there is reason for concern about the reli-
and Ph.D. degree from the University ability of rating scales used in thisJ~eld, and suggest actions for both research and practice.
of Massachusetts in Amherst. His Landscape studies have joined the shifting sands of scholarly inquiry. The empirical pos-
research focuses on perceptions of itivist tradition continues to hold that there is an objective landscape that can be studied
visible landscape qualities. He is par- through the careful application of scientific principles. Post-positivists, among whom I include
ticularly interested in developing GIS myself, stand back from the stranglehold that science recently held on legitimate knowledge.
models to predict landscape percep- They acknowledge other ways of knowing, but favor empirical methods of discovery. In response
tions and the interaction between to positivism, post-modern landscape scholars with diverse viewpoints are investigating new
landscape change and changes in per- ways to explore the personal meaning that landscapes hold for us. For instance, Norberg-Schulz
ceptions. He also hosts LArch-L, the (1979) has contributed to the rising importance of genius loci, Potteiger and Purinton (1998)
electronic discussion group for land- are exploring landscape narratives, and Brook (1998) is adapting Goethe’s approach to scien-
scape architecture. He may be tific inquiry through direct experience. In this flurry of activity some foundational attributes of
reached by email at applied everyday experience are lost or ignored. One such attribute is reliability, which is the
zooey@mailbox.syr.edu. subject of this paper.
Reliability and the Human Condition The poor boy hopes to recover This bud of love, by summer’s
from his blunder: ripening breath,
Palmer 167
residents, suggesting that they "may but not the average reliability for Schroeder (1984) reports the
in fact have different perceptions of individuals (Brown et al. 1988; Brown inter-rater reliability of attractive-
scenic quality in the rural land- and Daniel 1987; Daniel et al. 1989; ness or scenic beauty from ten stud-
scape." The need for further study is Hetherington et al. 1993; Rudis et al. ies, as well as one study of visual air
indicated. 1988). Typically these reliabilities are quality, two studies of enjoyableness,
Buhyoff and his colleagues above .90 for scenic beauty. and another of safety for urban parks.
(1979) compare the perception of Reports of the intra-group relia- The attractiveness studies have a
insect damage to forest scenes by bility of ratings other than scenic mean reliability of .530. Patsfall and
foresters, environmentalists, and the beauty or preference are less com- his colleagues (1984) report single-
public. There was an overall similar- mon. Herzog (1987) used ratings of rater reliabilities of.23 for SBE rat-
ity among the groups, but ratings identifiability, coherence, spacious- ings of vistas along the Blue Ridge
were affected by awareness of the dis- ness, complexity, mystery, texture, Parkway. Anderson (1976) extended
ease. Groat (1982) compared percep- and preference to characterize sev- Zube’s (1974) study and found an
tions of modern and post-modern enty natural mountainous, canyon, inter-rater reliability for individuals of
architecture by accountants and and desert scenes. Using groups of .69 for scenic value judgments of 212
architects. She found that the from thirteen to twenty-six students, scenes. He also used the Spearman-
accountants were not sensitive to he obtained intra-group reliabilities Brown prophecy formula to deter-
post-modern principles. for mean scores that ranged from .69 mine that a group of four raters
Some of these and other studies for coherence to .97 for preference. would achieve a composite reliability
find evidence that there is a strong Intra-group reliabilities are for scenic value of approximately .90.
similarity among ratings by the pub- generally high, and the more raters Craik’s research group used
lic, students, and professionals in there are in a group, the greater will four scenes to evaluate inter-rater
many instances. In other studies, be the group-mean reliability. The reliabilities for fifteen attribute rat-
expert knowledge appears to change reliability of group-means is impor- ings, primarily those used in the
the ratings by professionals and stu- tant if they are going to be used as Bureau of Land Management’s visual
dents, sometimes quite significantly. variables in predictive research mod- resource assessment procedure
However, none of these group com- els or for making environmental deci- (Feimer et al. 1979). Even after
parisons sheds light on the reliability sions. However, they do not reflect extending the analysis to 19 scenes,
of ratings by individuals, whether the reliability of individual raters. they found that "the reliability of rat-
they are environmental professionals Inter-rater reliability. In practice, ings tends to be low for single
or members of the public. landscape assessments conducted in observers and hence it is advisable to
Intra-group reliability. Most the field rarely involve judgments by use composite judgments of panels of
reports of rater reliability are for more than one or perhaps two trained independent observers" (Feimer et al.
group-mean ratings rather than for professionals. Assessing slides or 1981, p. 16). In a later report on this
the average reliability for an individ- other representations makes it more same work, Craik (1983, p. 72) rec-
Ual rater within a group. For instance, convenient to use a panel of evalua- ommends that "given present rating
Gobster and Chenoweth (1989) iden- tors. However, it is still unusual for systems, panels of at least five mem-
tified thirty-four of the most com- more than a few professionals to bers, rendering independent judg-
monly used landscape descriptors apply any of the agency evaluations ments, are required to achieve ade-
representing physical, psychological systems to a series of slides (Smardon quate levels of composite reliability."
and artistic attributes. In two experi- et al. 1988; USDA 1995; USDI 1980). The basis of this recommendation
ments, using groups of thirty and The only common occurrence of a comes from the application of the
twenty-two raters, they report relia- large group evaluating landscape Spearman-Brown prophecy formula
bilities of .64 to .99 for group-mean scenes is to create mean ratings for a to determine the composite reliability
ratings. single attribute (e.g., scenic beauty or for a group of any size based on the
Extensive research on forest preference), normally for research mean intra-group reliability. Kopka
landscape aesthetics has developed purposes. and Ross (1984) also evaluated the
during the past thirty years. Much of inter-rater reliability of BLM’s "level
this research uses the scenic beauty of influence" procedure applied to
estimation (SBE) method (Daniel existing scenes. The findings from
and Boster 1976), a statistical tech- these two independent studies con-
nique that standardizes each rater’s cerning the assessment reliability of
scores in order to control the varia- formalistic design qualities in the
tion in how a rater "anchors" the rat- landscape are summarized in Table 1.
ing scale. The reliability of group- They fall substantially short of
mean ratings is frequently reported,
Juneau, AK B/W offset press ’86 16 Rating-9 Residents Random ’86 406 Palmer & Smardon 1989
Juneau, AK B/W offset press ’86 t6 Rating-9 Public meeting Random ’86 41 ibid.
Dennis, MA Color photos ’76 56 Q-sort-7 Registered voter Random ’76 68 Palmer 1983
Dennis, MA B/W offset press ’76 56 Q-sort-7 Town list Random ’87 34 Palmer 1997
Dennis, MA Color slides ’76 56 Rating-10 Resident Available ’96 31 unpublished
Dennis, MA B/W offset press ’76 56 Q-sort-7 Env. Prof. Selected ’83 5t Palmer t985
Dennis, MA B/W offset press ’76 56 Q-sort-7 Env. Prof. Selected ’85 67 Palmer 1985
White Mtn, NH Color web press ’92 64 Rating-10 Residents Random ’95 77 Palmer 1998
White Mtn, NH Color web press ’92 64 Rating-10 USFS prof. Random ’95 205 ibid.
White Mtn, Nit Color web press ’92 64 Rating-10 Opinion leaders Census ’94/5 97 ibid.
US impact pmrs Color web press n/a 32 Compare-100 Austrian students Available ’90 59 Palmer et al. t990
US ~mpact pmrs Color web press n/a 32 Compare-100 French students Available ’87 99 ibid.
US ampact pmrs Color web press n/a 32 Compare-100 German students Available ’88/9 47 ibid.
US tmpact pmrs Color web press n/a 32 Compare-100 Hong Kong stds. Available ’90 53 ibid.
US ~mpact pmrs Color web press n/a 32 Compare-100 Italian students Available ’87 26 ibid.
US ~mpact pmrs Color web press n/a 32 Compare-100 Japanese students Available ’87 52 ibid.
US ampact pmrs Color web press n/a 32 Compare-100 Korean students Available ’87 128 ibid.
US ampact pmrs Color web press n/a 32 Compare-100 Puerto Rican stds. Available ’87 14 ibid.
US ~mpact pmrs Color webpress n/a 32 Compare-100 Spanish students Available ’87 100 ibid.
US ampact pmrs Color web press n/a 32 Compare-100 Utah students Available ’88 40 ibid.
US ~mpact pmrs Color webpress n/a 32 Compare-100 Yugoslav students Available ’87 47 ibi&
US ~mpact pmrs Color web press n/a 32 Compare-100 Central NY stds. Available ’87 59 ibid.
Palmer 169
Table 3. Inter-group reliability of scenic ratings for Dennis, Massachusetts.
Citizen ’76 Citizen ’87 Citizen ’96 Env. Prof. ’83 Env. Prof. ’85
a random sample of residents and forests, and environmental profes- In the White Mountain
attendees at a public workshop to sionals stationed in National Forests clearcutting study, the inter-group
study the human-use values of wet- in the northeastern quarter of the correlation of citizens with opinion
lands in Juneau, Alaska. The survey United States. The final study leaders is .980, and with Forest Ser-
included sixteen photos representing involves twelve groups of college stu- vice professionals it is .978. The cor-
the range of local wetland types and dents from around the world (Palmer relation between opinion leaders and
conditions. The second study began et al. 1990). They evaluated sixteen Forest Service employees is .974. The
as part of a community effort to matched simulations from the north- correlations among the four groups of
develop a comprehensive plan for eastern and southwestern United opinion leaders are shown in Table 4,
Dennis, Massachusetts. In 1976, a States portraying pre- and post- with an average correlation of .894.
random sample of registered voters impact conditions. Citations for these Table 5 shows the correlations among
evaluated fifty-six photographs repre- data-sets and the general characteris- the seven groups of Forest Service
senting the town (Palmer 1983). Resi- tics of the respondents and simulation environmental professionals. Their
dents evaluated the same scenes in media are summarized in Table 2. average correlation is .971.
1987 and 1996 (Palmer 1997). These Results. The correlation between In Table 6 are the correlations
ratings are compared to those from the mean scenic ratings of Juneau between twelve student groups from
employees of the U.S. Army Corps of residents and workshop attendees is around the world. Even with such
Engineers which were gathered in .971 for scenic ratings of sixteen diverse respondent groups, the aver-
preparation for a training course in diverSe wetlands. Table 3 shows the age inter-group correlation is .804.
landscape aesthetics (Palmer 1985). correlations among groups evaluating The lowest correlation is .496
The third study evaluated simula- the Dennis scenes. The average cor- between students from Japan and
tions of different harvesting intensi- relation among the five groups is Germany, while the highest is .969
ties, patterns, and patch sizes of .937. The highest correlation is .992 between the Austrian and German
clearcuts in the White Mountain between the two groups of environ- students. The inter-group correla-
National Forest (Palmer 1998). mental professionals, and the lowest tions from these four studies indicate
Respondents included a random sam- is .895 between 1996 citizens and why the quality of landscape assess-
ple of regional.residents, opinion lead- 1985 professionals. ments enjoys such a high reputa-
ers in the management of the area’s tion-most measures of reliability
meet the highest standards, and all
but a very few meet standards of
acceptability.
Table 4. Inter-group reliability among opinion leaders’ scenic ratings for clearcutting alternatives in the
White Mountains, New Hampshire.
Appalachian Forest Resources North Country Roundtable on
Trail Council Steering Committee Council Forest Law
Appalachian Trail Council -- -0.934 0.890 0.881
Forest Resources Steering Com. 0.934 -- -0.902 0.890
North Country Council 0.890 0.902 -- -0.868
Roundtable on Forest Law 0.881 0.890 0.868 --
However, the inter-rater relia- If landscape assessments are and scenic quality. His landscape
bilities in Table 7 are much less made primarily by individuals and not dimensions were all measured from
encouraging. The average inter-rater by large panels of evaluators, then USGS 1:24,000 topography or Massa-
correlation is .307 for Juneau resi- these results indicate that the relia- chusetts MapDown land use maps.
dents, and .355 for workshop atten- bility of scenic assessments is unlikely Palmer (1996) used a similar
dees. That is approximately one-third to be reaching acceptable levels. The approach to validate a GIS model of
the inter-rater correlation between next sections will consider the relia- spaciousness. A regression analysis
the two groups. The average of the bility of other visual qualities. found landscape dimensions
intra-group correlations for the five explained approximately half of the
Dennis study groups is .608. Again, Inter-group and Inter-rater Reliability of variation in Zube’s perceived scenic
this is a substantial drop in reliability Landscape Dimensions value and Palmer’s perceived spa-
from the average inter-group correla- Zube (1974) defines landscape ciousness.
tion of .937. The average inter-rater dimensions "as physical characteris- Shafer (1969) employed a dif-
correlation is .554 among the three tics or attributes of the landscape ferent approach to measure land-
major groups in the White Mountain which can be measured using either scape dimensions. He divided an eye-
study. This is down from an average normal ratio scales or psychometric level photograph into foreground,
inter-group correlation of .977. The scaling." Examples of such dimen- middle ground and background. Then
average of the twelve inter-rater cor- sions include: percent tree or water the area and perimeter of content
relations from the international study cover, length or area of the view, rela- areas were measured in each zone.
is .427, down from an average inter- tive elevation change, and various Examples of content include water,
group correlation of .804. edge and contrast indices. These trees, buildings, ground cover, and
dimensions bear a remarkable resem- pavement. This approach to measur-
blance to those employed by quantita- ing landscape dimensions also
tive landscape ecologists today accounts for approximately half the
(Tnrner and Gardner 1991). Zube variation in visual preference.
investigated the relationship between
twenty-three landscape dimensions
Palmer 171
Table 7. Inter-Rater Reliability of Scenic Ratings from Four Studies.
Location Respondents n Mean 95%-ile Median 5%-ile
The approaches developed by Methods. Respondents are thirty Naturalism refers to aspects of the
Zube and Shafer use physical tools to advanced landscape architecture or landscape that could exist without
measure the landscape’s dimensions. environmental science students at human care. Nature is an expres-
Human judgment can also be used to State University of New York’s Col- sion of how much vegetation is in a
estimate these measurements, for lege of Environmental Science and view, how organic are its elements
instance the relative area of a view and patterns, and how uncontrolled
Forestry in 1997 and 1998. They eval- are the natural processes.
covered by forest. However, when uated offset printed photographs of Development refers to aspects of
people are used as the measuring Dennis, Massachusetts taken in 1976. the landscape that are human cre-
device, more complicated constructs They were instructed to identify the ations. Development is an expres-
can also be measured, such as natu- highest, lowest, and intermediate sion of human control over natural
ralism or spaciousness. It is the relia- quality scenes and describe the crite- processes or patterns, and the dom-
bility of using people to measure ria for their decisions. Using these inance of structures, such as build-
landscape dimensions that is tested in scenes as anchor points on a seven- ings, roads, or dams.
this section. point scale, they sorted the remaining
scenes among the seven rating levels.
Each quality was evaluated on a dif-
ferent day. The four landscape dimen-
sions were described as follows:
Palmer 173
Mystery in the landscape is the (USDI 1980; Smardon et al. 1988). Line is a path, real or imagined,
result of incomplete perception. It More recent manuals developed in that the eye follows when perceiv-
describes the extent to which fur- Great Britain build on the work of ing abrupt differences in color or
ther information is promised to the Dame Sylvia Crowe and demonstrate texture, or when objects are
obsetwer if she were to walk deeper aligned in a one-ditnensional
how a more complete palette of aes-
into the scene. This is not a prom- sequence, described by its boldness,
thetic factors can be used to describe
ise of surprise, but of information complexity, and orientation. It is
that has continuity with what is and evaluate landscapes (Lucas 1990; usually evident as the edge of forms
already available. Bell 1993). in the landscape. Bold vertical lines
Methods. Respondents were which interrupt the skyline tend to
These definitions are based on twenty-five professionals in a two-day dominate weak horizontal lines.
descriptions given by the Kaplans and visual assessment continuing educa- Texture is small forms or color
their graduate students (Kaplan and tion course taught in early-December mixtures aggregated into a contin-
Kaplan 1989; Herzog 1989; Herzog, 1997 in Albany, New York. The indi- uous surface pattern. The aggrega-
Kaplan and Kaplan 1982). vidual contrast ratings are the maxi- tion is sufficient that the parts do
Results. Table 9 lists the reliabil- mum contrast from the land/water, not appear as discrete objects in
ities obtained for the informational the composition of the scene. Tex-
vegetation, or structures components
attributes. These reliabilities are tures are described by their grain,
of the landscape. The ratings involve density, regularity, and internal
highest for the exploratory variables two visual principles, contrast and contrast. Coarse and high-contrast
complexity (.315) and mystery (.262). dominance. Visual elements are the textures tend to dominate fine-
The three-dimensional variables, source of visual contrast in the land- grained textures of low internal
legibility (.214) and ~nystery (.26), scape, creating the patterns that we contrast.
are higher than the two-dimensional see. An object may differ from its set- Scale is the relative size of an
variables. The lowest reliability was ting or other objects in one or more object in relation to its surrounding
found for the two-dimensional under- element. When there is significant landscape. The scale may be in
standing variable, coherence(. 186). contrast in one or more of the ele- relation to the landscape setting as
However, all of these intra-group reli- a whole, the proportion of field-of-
ments, one object may dominate
abilities are unacceptably low. view, or other distinct objects.
other parts of the landscape. The con- Large, heavy, massive objects
trast or dominance of the following within a confined space dominate
Inter-Group and Inter-Rater Reliability of six visual elements are evaluated: small, light, delicate objects in
Compositional Elements more expansive settings.
A common approach to assess- Color is the major visual property
of surfaces attributed to reflected Space is the three-dimensional
ing visual impacts involves evaluating arrangement of objects and voids.
light of a particular intensity and
the amount of change in the scene’s wavelength. Described by its hue Compositions are described as
visual composition. Litton (1968; (tint or wavelength), value (light or panoramic, enclosed, feature, focal,
USDA 1973) initiated what became or canopied. Position of objects or
the common use of form, line, color, darkness), and chroma (saturation view in the landscape is relative to
and texture as the attributes most or brilliance). Lighter, warmer, topography. Backdrop is the sky,
brighter colors tend to "advance", water, or land background against
commonly used to describe landscape
while darker, cooler, duller colors which objects are seen. Objects
character and change. In particular; which occupy vulnerable positions
procedures have been developed that tend to "retreat" in a scene. Dark
next to light tends to attract the within spatial compositions, which
evaluate changes in contrast associ- are high in the landscape, and/or
eye and this contrast becomes a
ated with forIn, line, color, texture, as visual focal point. which are seen against the sky dom-
well as scale and spatial dominance Form is the mass or shape of an inate in the scene. The sum of the
object or objects which appear uni- contrast and dominance ratings is
fied. Forms are described by their used to create an index of visual
geometry, complexity, and orienta- impact severity.
tion in the landscape. Forms that
are bold, regular, solid, or vertical
tend to be dominant in the land-
scape.
(.503), and form (.563) are minimal. utes. Landscape preference and the is not so great as to nullify the useful-
Only texture contrast (.620) has a compositional elements form, line, ness of evaluations by random sam-
minimally acceptable reliability. colo~; and texture may be in a gray ples of the public, nor can it be used
These results are comparable to area between connotative and deno- to justify evaluations by only one or
those from the studies summarized in tative meaning. The reliability of two professionals.
Table 1. The reliability of scale domi- coherence, complexity, mystery, and There are several possible rea-
nance (.561) and spatial dominance legibility is sufficiently low to suggest sons that occur to the author for the
(.376) is also low. that they have highly connotative generally poor results reported here.
The sum of these five contrast meaning. One possibility is that photographs
ratings forms a contrast index that is The relatively poor reliability of
more reliable (.630) than any of its the information variables may con-
individual components. The reliabil- tribute to the discussion of nature
versus nurture, whether there is a
Palmer 175
scape assessments are most com-
Table 11. Inter-rater reliability of scenic ratings by professionals and the public.
monly made by a single professional.
Dennis, MA White Mtns. Published evaluations of the reliabil-
ity of a single evaluator (i.e., intra-
Professionals .668 .619 group reliability) for various land-
Public .568 .512 scape qualities do not ever meet pro-
fessional standards (i.e., greater than
provide insufficient information or For some time, the practice of .9) and normally fall below minimally
that different people fill in missing landscape assessment has been domi- acceptable levels (i.e., .7). Table 9
information differently. This is a nated by methods that use rating shows the number of evaluators
question of the validity of photo- scales or checklists. The mixed find- needed for each of the landscape
graphic representations. Many stud- ings reported here suggest that it qualities considered in this paper.
ies have reported that photographs may be appropriate to investigate the The size of these evaluation panels is
appear to be valid representations, reliability of other landscape assess- determined by applying the Spear-
but recent work by Hoffman (1997; ment methods. Those who used rat- man-Brown Prophecy Formula (Nun-
Hoffman and Palmer 1994) suggests ing scales did so because they sought nally 1978; Feimer et al. 1979; Ander-
that there are serious validity con- reliable results; they never made son et al. 1976).
cerns under some circumstances. claims to uncover deep meaning. In Several recommendations seem
Another possible explanation is that contrast, other researchers chose to appropriate given these findings.
the instructions describing the land- develop qualitative methods to search Research involving landscape assess-
scape attributes to be evaluated were for deeper meaning and to gain a ments by the public or professionals
must include: (1) field validations of
Table 12. The number of assessors needed to obtain minimally and professionally reliable ratings of landscape attributes.
Number of assessors to reach:
Attribute Tested reliability .9 reliability .7 reliability
Scenic value .582 7 2
Naturalism .796 3 l
Development .762 3 1
Spaciousness .715 4 1
Complexity .315 20 6
Mystery .262 25 7
Legibility .214 33 9
Coherence .186 40 11
Color Contrast .503 9 3
Form Contrast .563 7 2
Line Contrast .423 13 4
Scale Contrast .280 24 6
Texture Contrast .620 6 2
Scale Dominance .561 7 2
Spatial Dominance .376 15 4
Palmer 177
Norberg-Schulz, C. 1979. Genius Loci: Towards Patsfall, M. R., N. R. Feimer, G.J. Buhyoff, and Smardon, R. C., S. R.J. Sheppard and S. New-
a Phenomenology of Architecture. New York: J. D. Wellamn. 1984. "The prediction of man. 1981. Prototype Visual hnpact Assess-
Rizzoli International Publications. scenic beauty from landscape content ment Manual. Syracuse: SUNY College of
Nunnally, J. C. 1978. Psychometric Theory. New and composition."J0urna/ofEnvironmen- Environmental Science and Forestry.
York: McGraw-Hill. talPsychology 4(1): 7-26. --,J. F. Palmer, A. Knopf, K. Grinde,J. E.
Palmer,J. F. 1983. "Assessment of coastaI wet- Potteiger, M. andJ. Purinton. 1998. Landscape Henderson and L. D. Peyman-Dove.
lands in Dennis, Massachusetts." In The Narratives. New York:John Wiley & 1988. Visual Resources Assessment Procedure
Future of Wetlands: Assessing Visual-Cul- Sons. for US Army Corps of Engineers. Instruction
tural Values of Wetlands. Montclair, New Robinson, W. S. 1950. "Ecological correlations Report EL-88-1. Vicksburg, Mississippi:
Jersey: Allanheld, Osmun Co. and the behavior of individuals." Ameri- U.S. Army Engineer Waterways Experi-
¯ 1985. "The perception of landscape can Sociological Review 15(3): 351-357. ment Station.
visual quality by environmental profes- Rudis, V. A.,J. H. Gramann, E.J. Ruddell, and Turner, M. G., and R. H. Gardner (eds.). 1991.
sionals and local citizens." Syracuse: J. M. Westphal. 1988. "Forest inventory Quantitative methods in landscape ecology: the
Faculty of Landscape Architecture, and management-based visual prefer- analysis and interpretation of landscape het-
SUNY College of Environmental Sci- ence models of southern pine stands." erogeneity. New York: Springer-Verlag.
ence and Forestry. Forest Science 34(4): 846-863. U.S. Department of Agriculture, Forest Ser-
--. 1996. Modeling Spaciousness in the Dutch Schroeder, H.W. 1984. "Environmental percep- vice. 1973. National forest management. Vol.
Landscape. Report 119. Wageningen, The tion rating scales: a case for simple 1. Agriculture Handbook No. 434.
Netherlands: Agricultural Research methods of analysis." Environment and Washington, D. C.: U.S. Government
Department, Winand Staring Centre. Behavior 16(5): 573-598. Printing Office.
--. 1997. "Stability of landscape percep- Shafer, E. L.,J. E. Hamilton and E. A. Schmidt. --. 1995. Landscape Aesthetics: A Handbook for
tions in the face of landscape change." 1969. "Natural landscape preferences: a Scenery Management. Agriculture Hand-
Landscape and Urban Planning 37(1/2): predictive lnodel.’Journal of Leisure book No. 701. Washington, D. C.: USDA
109-113. Research 1(1): 1-19. Forest Service.
--. 1998. Cleareutting in the White Mountains: Shakespeare, W. 1992. Romeo and Juliet. In Great U.S. Department of Interior, Bureau of Land
Perceptions of Citizens, Opinion Leaders and Literature: Personal Library Series. 1992. Management. 1980. VisualResource
U.S. Forest Service Employees. [NYCFRD (CD-ROM) Parsippany, NJ: Bureau Management Program. Washington, D. C.:
98-0i] Syracuse, NY: New York Center Development, Inc. U.S. Government Printing Office.
for Forestry Research and Develop- Shrout, P. E. andJ. L. Fleiss. 1979. "Intraclass Zube, E. H., D. G. Pitt and T. W. Anderson.
ment, SUNY College of Environmental correlations: uses in assessing rater reli- 1974. Perception and Measurement of Scenic
Science and Forestry. ability." Psychological Bulletin 86 (2): Resources in the Southern Connecticut River
--., S. Alonso, K. Dong-hee,J. Gury, Y. 420-428. Valley. Pub. No. R-74-1. Amherst: Insti-
Hernandez, R. Ohno, G. Oneto, A. tute for Man and His Environment,
Pogacnik, and R. Smardon. 1990. "A University of Massachusetts.
multi-national study assessing perceived ,J. L. Sell andJ. G. Taylor. 1982. "Land-
visual i~npacts." Impact Assessment Bulletin scape perception, research, application
8(4): 31-48. and theory." Landscape Planning 9(1):
--, and R.C. Smardon. 1989. "Measuring 1-35.
human values associated with wet-
lands." In Intractable ConJlicts and their
Transformation. Syracuse, NY: Syracuse
University Press.