Table 3.
Model Results
| Explanatory Factor | Model I | Model II | Model III | |
|---|---|---|---|---|
| A. Alternative Specific Constants | ||||
 Breskens | - 8.889***(- 3.619) | - 9.066***(- 3.695) | - 9.235***(- 3.770) | |
 Braakman | - 10.015***(- 4.071) | 10.004***(- 4.067) | - 10.107***(- 4.140) | |
 Saeftinghe | - 6.427***(- 2.611) | - 5.396** (- 2.169) | - 5.942** - 2.410)( | |
| B. Price and Income (Pijt and HjYi in Equation [1]) | ||||
 Price (Pijt) | - 0.0178***(- 19.943) | - 0.0178***(- 19.939) | - 0.0176***(- 19.690) | |
 Income (log) (HjYi) | 1.359***(4.144) | 1.349***(4.110) | 1.368***(4.200) | |
| C. Site Attributes (Xijt in Equation [1]) | ||||
 Walking: moderate | 0.376***(5.143) | 0.381***(5.180) | 0.367***(5.014) | |
 Walking: good | 1.006***(15.619) | 1.006***(15.474) | 0.988***(15.221) | |
 Swimming: moderate | 0.527***(4.028) | 0.536***(4.071) | 0.528***(4.002) | |
 Swimming at Breskens: good | 0.578***(6.580) | 0.583***(6.639) | 0.577***(6.614) | |
 Swimming at Braakman: good | 1.406***(11.757) | 1.414***(11.813) | 1.394***(11.563) | |
 Nature: moderate | 0.389***(5.329) | 0.391***(4.110) | 0.397***(5.434) | |
 Nature at Breskens and Braakman: good | 0.860***(11.437) | 0.871***(11.560) | 0.864***(11.509) | |
 Nature at Saeftinghe:good | 1.093***(10.049) | 1.155***(10.305) | 1.162***(10.421) | |
| D. Distance Effects and User Effects (Components of f(Distij,Zij,Userij) in Equation [1]) | ||||
 User of Breskens | 1.021***(9.025) | 0.963***(8.393) | 0.821***(5.767) | |
 (Distance to Breskens)2/ 1,000 | - 0.105***(- 5.495) | - 0.224***(- 7.406) | - 0.147***(- 6.399) | |
 User of Braakman | 1.557***(9.143) | 1.843***(10.054) | 2.058***(11.897) | |
 User of Braakman X Distance to Braakman | - 0.017***(- 3.543) | - 0.026***(- 4.844) | - 0.033***(- 6.129) | |
 log(Distance to Saeftinghe + 1) | - 0.772***(- 11.386) | - 1.030***(- 11.601) | - 0.891***(- 9.736) | |
 User of Saeftinghe X log(Distance to Saeftinghe + 1) | 0.177***(5.975) | 0.178***(5.992) | ||
| Standard deviation of random parameter on user dummy (μu Userij) | 1.257***(11.372) | 1.264***(11.269) | 1.031***(11.716) | |
| E. Substitutes (Components of f(Distij,Zij,Userij) in Equation [1]) | ||||
 Distance to substitutes of Breskens | 0.022***(4.977) | - | ||
 User of Breskens X Distance to substitutes of Breskens | - | 0.041***(2.993) | ||
| F. Directional Effects (Components of f(Distij,Zij,Userij) in Equation [1]) | ||||
 | 0.662***(3.620) | |||
 | - 0.337***(- 3.161) | |||
 X User of Breskens | - 0.052** (- 3.078) | |||
 X User of Saeftinghe | - 0.467***(- 3.643) | |||
| G. Error Component (λijHj in Equation [1]) | ||||
| Standard deviation of error component for the three sites (λijHj) | - 2.819***(13.921) | 2.802***(13.754) | - 2.797***(13.858) | |
| Model Summary Statistics | ||||
| Number of observations | 3,150 | 3,150 | 3,150 | |
| LL | -3,341.7 | -3,331.1 | -3,328.8 | |
| AIC | 2.14 | 2.13 | 2.13 | |
| BIC | 2.18 | 2.17 | 2.18 | |
Note:t-Values are in parentheses. The notation included with the variables refers to the equations in Section III. The models were estimated in NLOGIT 4.0 using 400 Halton draws for the simulation. The abbreviations Bres, Braa, and Saef refer to the sites in set J: Breskens, Braakman, and Saeftinghe, respectively. AIC, Akaike information criterion; BIC, Bayesian information criterion; LL, log-likelihood.
↵** Significance of the coefficients at 5%;
↵*** significance of the coefficients at 1%.