Arboriculture & Urban Forestry 35(1): January 2009 utility lines) were avoided. The No Vegetation district scenes contain no trees or shrubs, showing uninterrupted rows of store- fronts. The Mature Vegetation scenario depicts a similar street scene with street trees of mature height and associated shrubs. No conflicts of trees with structures or infrastructure are apparent. Respondents were asked to project their shopping behav- ior. Each respondent saw both scenarios but responded to one hypothetical place. Three sets of variables measured consumer response. Respondents provided ratings on perceptual descrip- tions of the district and reported their likely behavior on five patronage variables. A final section presented a contingent val- uation method exercise, asking respondents to indicate their willingness-to-pay values for a list of goods and services. Con- tingent valuation is an economic analysis tool typically used to assess values for nonmarket, environmental public goods (such as wildlife conservation, clean air, and environmental protection). In this study, the tool was applied to an urban forest environment. Respondent Sampling Residents of three major cities in the Pacific Northwest region of the United States were selected for the survey sample frame: Seattle, Washington; Tacoma, Washington; and Portland, Oregon. These cities were chosen because they are the three largest popu- lation centers in the geographic region and have extensive strip mall development. It was assumed that the selected respondent populations and the strip mall settings of these cities do not differ distinctly from other temperate climate urban centers. A random list of mailing addresses was purchased from a commer- cial list broker generated within zip codes selected to represent typi- cal urban areas having strip malls. After pretesting, self-administered surveys were mailed to a sample of 1,200 households. Mailing pro- cedures (including reminders) generated 165 reasonably complete responses, and given that 63 were nondeliverable or not completed, the response rate was 14.5%. A 20% to 25% return was expected of urban mail surveys decades ago (Sommer et al. 1990), but response rates have fallen (Elmendorf and Luloff 2001). The questionnaire ended with several demographic and back- ground items. Thirty-three percent of respondents were in their 20s and 30s, 44% were in their 40s and 50s, and 24% were in their 60s or older. The majority of respondents were male (60%) and white (85%). Twelve percent of responding households contained a single person, whereas 36% were two-person households with 46% claiming to be composed of three or more persons. Forty- seven percent of households had persons 18 years or younger. Stated annual household income varied considerably. Twenty- one percent claimed $35,000 or less per year, 49% selected the $35,000 to 50,000 range, whereas 30% earn greater than $75,000 annually. Meanwhile, U.S. median household earnings for the study year were $38,885 and regionally $39,067 for Oregon and $47,421 for Washington states. There are issues of generalizability resulting from survey response. The low response rate raises possibilities of nonresponse and/or self-selection bias. Resources were not available to test for nonresponse issues. Also, the respondent profile indicates poten- tial inference limitations with regard to gender and ethnicity. ANALYSIS AND RESULTS Data analysis involved several approaches. Descriptive statistics, followed by factor analysis, revealed categories of response in 35 the preference ratings. For scenarios, individual response items were first tallied, then combined using data reduction methods to look for underlying categories, and then compared for differ- ences in response between conditions of streets having and not having trees. Visual Preference Prompt images depicted a gradient of vegetation treatments. Does the presence of vegetation affect visual quality of small mall retail settings? Considering the range of mean preference ratings, scenes having only lawn in the foreground came in at approximately 1.5 and display uninterrupted views of mall build- ings. Images having trees, alone or in association with shrubs, registered the highest means of approximately 3 to 3.5. Principal axis factor analysis with Varimax rotation was used to extract underlying common dimensions based on observed covariation of individual images. Several decision rules were used to determine inclusion of any item within a specific cat- egory (Kaplan and Kaplan 1989). Four categories emerged from the analytic procedure, accounting for 66% of the total variable variance, and included 25 images. Each group was assigned an assemblage epithet describing principal content. After dimensional analysis, new variables were constructed by aggregating means across all category items ( Figure 2 ). Factor analysis categories typically reflect multiple perceptual influences such as vegetation content, spatial configuration, and built to nature balance. The images of Mixed Screen display a combination of trees and shrubs at the property edge. The vegeta- tion composition is varied, depicting mixed species and irregu- lar placement of plants. Foreground landscape obscures views of midground buildings. Ordered Trees images contain large trees planted in regular order at the property edge with the ground plane planted in grass. Category 3, Shrub Edge, depicts border plantings of both sheared and naturalistically maintained low- growing shrubs. Unlike the two other categories, background buildings are in full view behind the vegetated edge. The base images, containing No Vegetation in the foreground, formed a fourth category with uninterrupted views of a parking area and buildings. The lowest rated category contains views of strip mall build- ings visually unmitigated by landscape vegetation. From the most barren scenes of buildings and parking lots, category preference means increase with increased quantities and structure of vegeta- tion. Preference ratings take a major step up with the addition of shrub and hedge plant materials. Another preference increment is added when trees are planted with the highest visual quality values being assigned to tree plantings having associated shrub materials. Scenario Analysis Place Perceptions While viewing two scenarios, respondents were asked to rate their level of agreement with a series of descriptive statements about one of the locations using Likert scales (1 = “strongly disagree” to 7 = “strongly agree”). Categories were statistically derived and labeled using the analytic approaches described for visual pref- erence data. Four categories contained 16 of 25 original items ©2009 International Society of Arboriculture
January 2009
| Title Name |
Pages |
Delete |
Url |
| Empty |
Search Text Block
Page #page_num
#doc_title
Hi $receivername|$receiveremail,
$sendername|$senderemail wrote these comments for you:
$message
$sendername|$senderemail would like for you to view the following digital edition.
Please click on the page below to be directed to the digital edition:
$thumbnail$pagenum
$link$pagenum
Your form submission was a success. You will be contacted by Washington Gas with follow-up information regarding your request.
This process might take longer please wait
