FLOOD09697 - Laserfiche WebLink

Home Browse Search

ent to the senses. Similar evidence to that described for the relationships of the recreation and aesthetics tenns was manifested for the ecological and aesthetics tenns; the calculated correlalion coefficient in this case was not quite as high, 0.2345. Despite any atten- uation or weakening caused by correlation with other variables, this lerm is shown by Equation 5.25 as being one of the more important. Certainly, the interest in ecology warrants inclusion of such a term in analyses of the merits of flood control proposals. The results of analysis of the term interrelaling the "Willingness to Pay for Government Expenditures Scale," PAYL, and lIPerceived Cost of a Projece' indi- cated Term 10 to be Ihe least important. The cost ac- ceptance function had the lowest standardized coeffi- cient ofany of the interaction terms in Equation 5.25. The scale score P A YL for willingness to pay had the second highest reliability of any of the scales (Appen- dix B). This would indicate that concern about flood control cost is not as great a concern as other aspects of a project. The general public may usually assume a proposal is affordable if it is proposed by an offi- cial agency and costs are not direct; they may also ex- pect officials who oversee the flood control decision to be responsible for keeping costs down. If this is true, monetary considerations will be very important to an agency, especially the decision agency, in deter~ mination of acceptable flood control proposals. Such a result is su~ested by Equation III, but requires fur- ther testing. 3 In any case, a cosl acceptance func- tion term should be included in all studies, whether it be an agency or public group being analyzed, until furlher testing indicales specifically when and where it is unimportant. The correlation between the "Ecological Orienta- tion Scale," ECOL, and MANL was computed as 0.3745. It is interesting also Ihat use of leisure orien- tation, LEIL, in place of MANL in this acceptance function term (Term 7 of Equation 5.21) results in al- most exactly the same coefficients and correlation as in Equation 5.24. This point is illustrated by replac- ing (MANIA) by (LEIL-6) (see Appendix D) in Equa- tion 5.24. The resulting equation in unstandardized farm is: (I) (2) (3) (4) POPEVE = .381-.124XI + .124X2 - .480X3+.049X4 (5) (6) (7) + .063X5 + .023X6X12 + .032XI 33Equation IV and VI do not indicate strong impor- tance of cost factors because the effect of monetary consid- erations is largely absorbed by the Type IV term for the influ- ence ofprcviousagency evaluations. The effect of the cost factors is probably underrated even in Equation III because of the fact that it appears in both Terms 2 and 3 (see Gordon, 1968). (8) (9) + .0090X8X14 + .0079~XI5 (10) (11) + .0084XlOX16 + .0065XII Xl7 (5.26) In standardized fonn is: (I) (2) (3) (4) POPEVE = -.I72XI +.1J4X2".107X3+.108X4 (5) (6) (7) + .087X5 + .339X6X12 + .196X7X13 (8) (9) (10) + .1 11 X8X 14 + .124~XI5 + .081XIOX16 (II) + .1l9XIlX17 . .. . (5.27) The r2 of Equation 5.26 or 5.27 is 0.490. Equation 5.26 was not used inslead of Equation 5.24 because of the slightly smaller value of the co- efficient in Term 7 and slight reduction in r2. The results would have been easier to explain had this lasl equation been used on the hypothesis that leisure ori- entation and outdoor aesthetics orientation are corre- lated. All that can be expressed with certainty at this stage of development is that the aesthetics acceptance function is important and must be included for realis- tic analysis of public flood control proposal evalua- tions. What the actual value of the coefficient will be in this context with satisfactory measurements of the variables remains for further study. The last term in Equation 5.24, the only Type IV term in the equation, accounts for the influence of the County Flood Control Division agency. It is the predicted value of that agency's revised evalua- . tion which is entered inlo this lerm in the model. Th~ companion variable is measured by the "Willingness to Follow the Advice of Experls," EXPfL The scale was slightly more effective in this interaction than the "Willingness to Follow Government Agencies Scale, II AGENL, Term IV; this could partly be because of the higher reliabilitl and longer length of this scale (see AppendIX C).3 This type of relationship should al- ways be included in any attempt to predict actual re- action to proposals by the public. It may be useful to add more acceptance functions of this kind with each one representing the effect upon the public's evalua- lion of some significant group. In summary, all the acceptance functions in- cluded in this population analysis seem reasonable 34rhe fact that a variable is used as a Type I variable did not prevent its selection for a Type IV term. The selec- tion of the variables for the interaction terms was done be- fore the addition of other variables to an equation. 77