Laserfiche WebLink
<br />The validity of these results is reinforced by <br />multicollinearity analysis (Appendix I). For an equa- <br />tion with II terms and 17 social science variables, <br />Ihe inlercorrelation between the Independent vari- <br />ables is quite low. Of the 55 bivariate relationships, <br />only six are significant at the .20 level and most are <br />not significant at .10. Even the multiple r2 's from <br />linear multiple regressions, using all the other tenus <br />except the dependent variable as predictors, are com- <br />paratively low with the majority between .05 and <br />_12 and the highest three between .36 and .38.22 <br />This low multicollinearity is imporlant because co- <br />efficients are so unreliable when multicollinearily <br />is severe that even changes in signs of important vari- <br />ables may occur (Gordon, 1968; Darlington, 1968; <br />Schoenberg, 1971; Johnston, 1972; Wonnacott and <br />Wonnacott, 1972; Duncan and Goldberger, 1973). <br />Low multicollinearity is a necessary, although not <br />sufficient, condition of valid coefficients. <br /> <br />Figure 5.3 illustrates the formulation of the <br />terms used in Equation 5.24. This chart can be com- <br />pared to Figure 5.2 in order to see how a general con. <br />ceptualization was applied to obtain a working equa- <br /> <br />2brhese three were the aesthetics, recreation, and eeo. <br />logy terms and did cause some difficulties. <br /> <br />FLOOD CONTROL <br />PROPOSAL <br /> <br />Table 5. 7. Terms of the population evaluation equa- <br />tion (Equation V). <br /> <br />Te!.!!!. <br /> <br />(1) Number of Types of Groups With Whom Discussed <br />Flooding Problems. (KTYPE) <br /> <br />(2) Proximity of Flood Experience to Present Resi- <br />dence. (KLOSF) <br /> <br />(3) Proportion of Single Unit Structures in Block <br />(KSINUN) <br /> <br />(4) Willingness to Follow Government Agencies <br />(AGENL-4) <br /> <br />(5) Knowledge of Local Flood Control Proposals <br />(KNOWL) <br /> <br />(6) Perception of Need for Improved Flood Control <br />in Area (CONCL-8) by Perceived Effectiveness of <br />a Proposal (IREFF) <br /> <br />(7) Attitude Toward the Effect of Man~rnade Objects <br />upon the Beauty of Nature (MANL4) by Perceived <br />Aesthetics Effect of Proposal (IRAES) <br /> <br />(8) Outdoor Recreation Orientation (RECL-6) by Per- <br />ceived Recreation Effect of Proposal (lRREC) <br /> <br />(9) Ecological Orientation CECOL-B) by Perceived <br />Ecological Effect of Proposal (IRECQ) <br /> <br />(10) Willingness to Pay for Government Expenditures <br />(PA YL-6) by Perceived Low Cost of Proposal <br />(IRCOS) <br /> <br />(11) Willingness to Follow Experts (EXPTL-6) by Evalu- <br />ation by Government of Proposal (IDAGEN) <br /> <br />Xu' x13' X14. XIS' Xl6 <br /> <br />GOVERNMENT <br />.DECISION <br />AGENCY <br /> <br /> <br />X12 <br /> <br />X13 <br /> <br />X'4 <br /> <br />XIS Xl6 <br /> <br />(X6XI2) (X7XI3) (X8XI4) (~Xl 5) ex. WX16) <br /> <br /> <br />x, <br /> <br />x, <br /> <br />x, <br /> <br />XIO <br /> <br />x. <br /> <br />POPULATION <br /> <br />X, <br />X, <br />x, <br />x., <br />X, <br />x, <br /> <br /> <br />POPULATION <br />EQUATION (5.24) <br /> <br />Figure 5.3. Diagram of conceptualization of processes relUlting in the equation for predicting public evalua- <br />tion <br /> <br />73 <br />