Laserfiche WebLink
<br />1792 <br /> <br />000078 <br /> <br />BROWN AND DANIEL: LANDSCAPE AESTHETICS OF RIPARIAN ENVIRONMENTS <br /> <br />TABLE 3. Regression Models of Poudre River Scenic Beauty <br /> <br />Variable- <br /> <br />Tucson <br />Ratings <br />(SBE) <br /> <br />-64.58 <br />(-1.3I)t <br />0.09 <br />(2.92) <br />-4.12 x 10-' <br />(-3.63) <br />-7.40 <br />(-5.82) <br />-3.97 <br />(-4.80) <br /> <br />Unit <br /> <br />Constant term <br /> <br />FLOW <br />(FLOW)' <br />SOIL <br /> <br />cf, <br /> <br />(cf,)' <br /> <br />percent <br /> <br />SKY <br />TREES <br />GRASS <br />BED <br /> <br />percent <br /> <br />percent <br /> <br />percent <br /> <br />-5.34 <br />(-3.26) <br />-5.35 <br />(-3.36) <br />25,40 <br />(2.74) <br /> <br />percent <br /> <br />VIEW <br /> <br />1-5 rating <br /> <br />LIGHT <br />WCOLOR <br />PHOTOL <br /> <br />1-3 rating <br /> <br />I-~ rating <br /> <br />16.13 <br />(2.89) <br />37.11 <br />(3.42) <br />-49,55 <br />(-5,52) <br />0.46 <br />19.32 <br />79,91 <br />216 <br />1092 <br /> <br />1-5 rating <br /> <br />PHOTOC <br /> <br />1-5 rating <br /> <br />R:!.~ <br />F <br />Standard error <br />N <br />Flow at maximum <br />scenic beauty <br /> <br />cf, <br /> <br />Experiment <br /> <br />Fort Collins <br />Ratings <br />(SBE) <br /> <br />-88.50 <br />(-2,59) <br />0.13 <br />(4.56) <br />-5.06 x 10-5 <br />(-4.70) <br />-)1.03 <br />(-9,17) <br />-3.91 <br />(-4.59) <br />1.40 <br />(3.07) <br /> <br />38.02 <br />(3.97) <br />-40.13 <br />(-5.52) <br /> <br />0.47 <br />28.36 <br />81.38 <br />216 <br />1285 <br /> <br />Paired <br />Comparison <br />(PCE) <br /> <br />-16.19 <br />(-1.33) <br />0.09 <br />(4.17) <br />-3.03 x 10-' <br />(-3.86) <br /> <br />-14.96 <br />(-2.22) <br /> <br />0.35 <br />6.49 <br />25.03 <br />32 <br />1485 <br /> <br />SSE. scenic beauty estimate, peE. paired-comparison estimate. <br />.Variables not listed did not enler the stepwise regression model for any experiment. Entries left <br />blank represent variables thai did not enter the stepwise model. <br />tNumbers in parentheses are Student's 1 ratios. <br />.+Adjusted R~. using the .adjustment procedure in the statistical package for the social sciences. <br /> <br />flow levels. Although the instructions to respondents again <br />avoided mention of specific scene features, and 40% of the <br />pairs compared completely different views. we expected that <br />forcing a choice between many pairs that differed only in <br /> <br /> 30 <br /> 20 <br />W W <br />" <br />a. 0 <br />0 <br />"' .W <br />'" <br />!'l <br />~ .20 <br />3 <br />. <br />. .30 <br />CD <br />u <br />. .40 <br />. <br />.x .50 <br /> .60 <br /> .70 <br /> wo '00 <br /> <br /> <br />~ ForI COllins 'al1ngs <br />_Tucsonrilhngs <br />_ Tucson. paul!c camp <br /> <br />700 1000 1300 1600 1900 2200 2500 <br /> <br />Flow (CISI <br /> <br />Fig. I. Scenic beauty versu!> flow from regression models for lhe <br />Poudre River. evaluated given mean values for non flow variables. <br /> <br />flow level would lead respondents to more heavily empha- <br />size flow than they did in the rating format. <br />The imponance of flow was indicated by the R2 of <br />regression models developed from the sets of scenic beauty <br />judgments obtained from the two response formats. For each <br />format. scenic beauty was regressed on just flow variables <br />(FLOW and FLOW') and then on the full ,et of scene <br />features (Table 2) using a stepwise procedure. With the <br />paired-comparison experiment the adjusted R2 was 0.26 for <br />the flow-only model and 0.35 for the full stepwise model <br />(Table 4). For comparison with these paired-comparison. <br />models. rating models were developed using the same 32 <br />cases upon which the paired-comparison model was based. <br />The adjusted R2 values of the flow-only models were 0.08 for <br />the Tucson sample and 0.18 for the Fort Collins sample but <br />rose to 0.62 and 0.42. respectively, for the full stepwise <br />model. <br />Flow variables alone account for considerably more of the <br />explained variance in the paired-comparison scenic beauty <br />metric than in the rating-based metric. and the availability of <br />nonflow variables led to much greater increases in R2 in the <br />rating.based models than in the paired-comparison model. <br />Of course. the significance of the differences in R2 cannot be <br /> <br /> <br />tested <br />the r" <br />diffen <br />sugge <br />comp <br />Impac <br />exten <br />focus <br />We <br />entire <br />River <br />ship I <br />sen[s <br />river. <br />accou <br />range <br />ex pee <br />am on; <br />woulc <br />At tht <br />only <br />repeti <br />only t <br />such i <br />accou <br />Th, <br />set of <br />one I. <br />trips 1 <br />recre, <br /> <br />1 <br /> <br />i <br />l JUS! fll <br />,. full m <br /> <br />