Research Article

Random-Parameters Behavioral Models to Investigate Determinants of Perceived Safety in Railway Stations

Table 4

Comparison between Binomial Logit and Mixed Logit estimated models.

MODEL 1MODEL 4MODEL 3MODEL 5

SpecificationBinomial LogitMixed LogitBinomial LogitMixed Logit

DescriptionAll attributesAll attributes with random parametersConsistent attributes in sign only and interactions with SE characteristics and travel habitsConsistent attributes in sign only with random parameters and interactions with SE characteristics and travel habits

#observations1208120812081208
Log-likelihood function (constants)−835.5−835.5−835.5−835.5
Log-likelihood function (fitted)−648.1−574.5−627.9−562.3
Pseudo-R20.2240.3140.2490.328
Akaike information criterion1316.31189.11285.91162.6

VariableCoeff.t-ratioCoeff.t-ratioStd. dev.t-ratioCoeff.t-ratioCoeff.t-ratioStd. dev.t-ratio
Surveillance cameras0.466.251.213.770.940.800.456.010.905.291.843.84
Security personnel0.8911.812.584.101.781.401.1011.222.366.763.887.73
Security personnel, employed−0.47−3.34−1.12−2.84
Commercial activities0.081.210.291.702.362.570.081.180.201.60
Road crossings1.126.493.273.421.350.920.784.371.663.422.783.44
Road crossings, freq. car high0.492.090.951.61
Tactile path and signage−0.09−0.72−0.01−0.035.013.99
Greenery0.252.180.792.360.780.650.563.211.164.04
Greenery, female−0.54−2.32−1.14−3.04
Intermodal infrastructure0.453.691.342.565.433.660.271.840.391.03
Intermodal infrastructure, freq. PT high0.672.701.722.52
Artificial lighting0.090.730.261.081.461.780.090.740.191.07
Crowding0.343.810.782.174.093.200.100.840.040.11
Crowding, female0.462.620.821.73
Decorum and maintenance1.118.883.443.724.952.441.007.362.264.854.335.74
Decorum and maintenance, freq. rail low0.722.161.722.13

The symbols , , and indicate significance at 1%, 5%, and 10% level.