Table 4.

Parameter Estimates for the 12-Site Location Model

ParameterMNLMNLHMNPMNPHSMNPSMNPH
Distance- 15.2825**- 14.8763**- 3.0370**- 3.0914**- 4.5352**- 4.3512**
(0.32)(0.31)(0.08)(0.07)(0.15)(0.13)
Revenues0.0534**0.0861**0.0121**0.0114**0.0115**0.0090**
(0.01)(0.01)(0.00)(0.00)(0.00)(0.00)
Steller0.1790**- 0.0666*0.0362**0.0438**0.0161**0.0162**
(0.04)(0.04)(0.01)(0.01)(0.01)(0.01)
Miss Dum- 0.8449**- 0.4264**- 0.2316**- 0.1132**- 0.1649**- 0.0986**
(0.09)(0.10)(0.02)(0.02)(0.02)(0.02)
ρ0.6211**0.5547**
(0.03)(0.03)
Embedded Image0.3619**0.3996**
(0.01)(0.02)
κ3,30.5148** (0.03)0.5828** (0.03)
κ4,40.4639** (0.02)0.4814** (0.02)
κ5,50.3113** (0.02)0.3615** (0.02)
κ6,60.2374** (0.01)0.2992** (0.02)
κ7,70.1747** (0.02)0.2246** (0.02)
κ8,80.2520** (0.02)0.2621** (0.02)
κ9,90.4249** (0.03)0.4006** (0.03)
κ10,100.1817** (0.01)0.1921** (0.02)
κ11,110.3453** (0.02)0.3088** (0.02)
κ12,120.3783** (0.02)0.3604** (0.02)
LL value- 2,867.35- 2,702.27- 3,274.30- 3,040.10- 3,087.41- 2,888.12
  • Note: The log-likelihoods for the MNL and MNP models cannot be directly compared because they are based on a different likelihood function. However, the log-likelihoods for all the MNP and SMNP models can be compared to one another. LL, log-likelihood. Model names are defined in Table 2.

  • * Statistically significant at the 90% level;

  • ** statistically significant at the 95% level.