Gregory J. Buhyoff

Psychophysical Models: An Example with Scenic Beauty Perceptions of Roadside Pine Forests

Relationships among environmental characteristics and scenic beauty perceptions/judgments can be assessed through application of psychophysical modeling methods. Interpreting psychophysical relationships helps establish model validity, identifies testable hypotheses, and generates information useful to future model development, theory building, and design efforts. The paper discusses the psychophysical modeling approach in detail, and presents scenic beauty models that were developed for roadside views of pine forests in the southeastern United States. The results indicate that great potential for scenic beauty management exists in this region. There is an optimal stand density (number and size of trees) for scenic beauty which is dependent upon tree age. As one would expect, older stands are generally more scenic and can “support” more stand density without harming scenic quality perceptions. Interactions among tree age, number of trees per acre, tree diameter, and variability in tree diameter were also found.

The prediction of scenic beauty from landscape content and composition

Abstract Two studies were conducted to examine distance classes of vegetation (foreground, middleground and background) and scene composition (presence of vegetation in left, center or right section of the image) as predictors of perceived scenic beauty. In study one, 41 students rated 63 landscape scenes with regard to scenic beauty. The Scenic Beauty Estimation Method was used to derive interval scale beauty values (SBEs). For each landscape image, areal measures of vegetation in each distance class and for each vertical section were taken and used as predictors. Presence of haze, clouds and human impacts were also recorded. Among the most important contributors to scenic beauty were amount of center middleground vegetation, and center background vegetation. Left foreground vegetation and right foreground vegetation were found to have significant and opposing regression weight signs—negative for the left and positive for the right. Study two was conducted to determine whether these opposing regression weight signs for foreground vegetation were due to a perceptual right-left bias or to some specific content in the image itself. In the second study, the photographic slides used to present landscape scenes to subjects were reversed so that the content which was previously on the right was now left, and vice versa. Thirty-nine students rated the reversed slides with regard to their scenic beauty. The signs of the regression weights in study two shifted such that left foreground was now positively valued and right foreground was negatively valued. This finding suggests that viewers are sensitive to foreground content and its placement in the image, and not simply to one side of the field of vision.

Manipulation of dimensionality in landscape preference judgments: A quantitative validation∗

A Paired Comparison Methodology was used to scale preferences for landscapes depicted in 35mm color slides. Several S groups, varying in their familiarity with forestry, evaluated a series of photographically controlled forest scenes with various levels of insect damage. Some Ss were told a priori that insect damage was present (experimental), others were not (control). The objective was to see if experimental Ss would cue on a single “damage dimension” in their preference judgments, thus providing reliable interval measures of their preferences and estimates of esthetic impacts resulting from insect damage. Results of statistical tests indicate that dimensionality can be manipulated and that esthetic impacts can be measured as a result of changes in a specific landscape dimension.

Distance and Scenic Beauty A Nonmonotonic Relationship

Distance to major topographic features was found to be highly predictive of perceived scenic beauty of natural scenes. In addition, the hypothesis of a nonmonotonic functional relationship was supported. The implications of nonmonotonicity were addressed, including such things as the implied presence of maximal and minimal levels of impact. These implications were contrasted with Buhyoff and Wellmans (1980) suggestion of a monotonic logarithmic function and were compared with similar findings in experimental aesthetics (Berlyne 1974).

Interval scaling of landscape preference by direct- and indirect-measurement methods

Abstract Buhyoff G.J., Arndt L.K. and Propst D.B., 1981. Interval scaling of landscape preference by direct- and indirect-measurement methods. Landscape plann. , 8: 257–267. While interval-level scaling of landscape preferences may be desirable, the psychometric techniques which can be used to derive such measures are quite complex. These methods require intensive subject involvement and extensive data manipulations. This paper investigates the relationship between landscape preference metrics derived under an indirect method through the application of the Law of Comparative Judgment and a more direct, less data-intensive magnitude-estimation method. Results for two landscape stimulus sets indicate that extremely similar interval scales emerged from both techniques.

Seasonality bias in landscape preference research

Little attention has been paid to potential seasonality biases in landscape preference research using photographic simulations. This paper presents findings that substantiate the hypothesis that there is an interaction between the season in which photographs are taken and the season in which they are evaluated. This interaction may partially invalidate research in which it occurs. Using nine unspectacular landscapes, four of which had fall coloration, in a paired comparison routine, the preferences of similar subject groups in two different seasons were compared. The preferences of the two groups were not significantly related, and a strong shift in the preferences for landscapes according to their coloration was found.

Individual and group reliability of landscape assessments

Abstract The reliability of landscape quality assessments made by individual observers as well as their group consensus was examined using a test—retest procedure. After the elapse of more than a year, the individual observer reliability estimates averaged nearly 0.80. Group consensus values were derived using the Law of Comparative Judgment psychophysical scaling procedure. These derived group consensus values proved very reliable ( r > 0.95) after the elapse of a year. Individual rater reliability was good, but group reliability was excellent and should probably be used whenever possible, especially when used as an aid to long-term decision making.

Microcomputer-resident comprehensive statistical analysis

Hardware and Firmware SPS programs were written for a Tandy Radio Shack Model I computer that utilizes the l80 microprocessor and Microsoft BASIC. System configuration must be a minimum 32K RAM, at least one disk drive, and the TRSDOS or NEWDOS operating system. Hard-copy output can be enabled via operating system utilities or insertion of program statements to change PRINT to LPRINT (Buhyoff & Hull, 1980). Either serial or parallel printers may be used.

Statistical microcomputing: A critical look

Abstract With the introduction of the microcomputer, sophisticated statistical computing can now be done at a persons desk. However, since the microcomputing environment is a subset of the mainframe environment, there exist limitations to statistical microcomputing, such as precision and processing efficiency. However, the microcomputer does offer advantages for statistical computing if the statistical software for microcomputers exhibits certain characteristics. These software characteristics and the advantages and disadvantages of statistical microcomputing are discussed.

A Note on Impacts of Adjacent Land Use on Blue Ridge Parkway Operating Costs

The National Park Service has long been interested in effects of adjacent land uses upon costs of operating national parks. Although studies have described such effects, little work has been done to quantify them [1-4]. Objectives of this study were to determine how factors internal and external to the Blue Ridge Parkway affect the cost per visit and total cost of operating the Parkway. Parkway administrators were seeking information which could be useful in budgeting and land use planning. Only costs of administration and maintenance were considered, not construction or non-market costs such as scenic damage or congestion. PROCEDURES The Parkway, which meanders 470 miles along the crest of the Blue Ridge Mountains in Virginia and North Carolina, was divided into twelve segments with varying degrees of development in adjoining counties and with different concentrations of Parkway facilities. Segments averaged about forty miles long * Work was supported by National Park Service grant number CX500081113. 157