Page 120 - Proceeding of Atrans Young Researcher's Forum 2019_Neat

“Transportation for A Better Life:
Smart Mobility for Now and Then”

23 August 2019, Bangkok, Thailand

In the stated adaptation survey, about 100 Table 5. Estimated binary logit models of mode
respondents were interviewed. Each respondent was choice in adverse weathers
given four hypothetical mode choice scenarios Car/Taxi vs Other MC/MC Taxi vs Bus vs Other
corresponding to four trip purposes. In each Model Other
hypothesized choice scenario, the respondent was variable Est. Std z- P(> Est. Std z- P(> Est. Std z- P(>
.E. val |z|)
.E. val |z|)
.E. val |z|)
asked to choose between two available modes (i.e., Trip cost - - - - - -
Mode i and Mode j). In total, there were 400 diff. 0.00 0.00 2.97 0.003 2 0.01 0.00 6.09 1.1e- 0.00 0.00 1.25 0.21
4
9
4
2
7
2
7
3
9
-
-
-
-
hypothesized choices made by the respondents Trip time - 0.05 0.02 2.63 - 0.04 0.01 2.57 0.01 0.02 0.02 1.12
7
0
(Table 4). Based on the hypothetical choice data, diff. 2 0.57 0.32 1.80 0.008 1 6 3 0 5 2 2 0.26
three binary logit models were estimated using the R Gender 8 1 1 0.071 - -
software. Results of model estimation are presented 0.12 0.07 1.75
in Table 5. Income 5 1 5 0.07

Career dummy

Junior 2.63 1.13 2.31
Table 4. The number of choice samples by choice engineer 0 8 0 0.021
set Skilled 1.39 0.79 1.75 0.08
worker
8
8
2
% % 1.24 0.41 2.97 0.00
Total Unskilled labor 1 7 5 2
Choice set (Mode i vs Mode j) choosing choosing
samples 1.70 1.11 1.52
Mode i Mode j Retired 6 7 7 0.13
Car/Taxi vs Other(Motorcycle, 238 33% 67% Vehicle Ownership dummy
Motorcycle Taxi, Bicycle, Ebike, Bus) 1.92 1.07 1.80

MC/MC taxi vs Other (Car, Taxi, Bus, 236 69% 31% MC 9 2 0 0.07 1.81 0.83 2.18 0.02
Bicycle, Ebike) Bus ticket 0 0 1 9
Bus vs Other (Taxi, Motorcycle, 158 53% 47% Trip Purpose dummy
Motorcycle Taxi, Bicycle, Ebike) - -
Commut. 2 3.02 1.15 2.62 0.009
3
1
Key findings from the model estimation can - 2.94 1.14 2.57
-
be summarized as follows. Firstly, “travel time” and Shopping 0 - 2 4 0.010
-
“travel cost” are statistically negative factors to Leisure 2.43 1.13 2.14
mode choice, meaning that people tend to choose a 3 - 5 2 0.032
-
faster and cheaper mode. People choosing Other 2.59 1.14 2.27 0.022
1
8
9
motorcycle/motorcycle taxi and car/taxi are Weather Condition dummy
-
sensitive to both cost and time factors while people Heavy rain before departure - 1.00 0.38 2.60 0.00 0.80 0.45 1.74 0.08
8
7
choosing bus are not sensitive to both travel time and time 4 6 3 9 1 0.60 0.45 1.34 0
travel cost. The insignificance of travel time and Flood during the trip 6 2 3 0.18
travel cost to the choice of bus is because travel time Summary statistics 236 158
N°ofObs. 238
by bus is always much longer than by car and Null dev. 329. 327.
motorcycle and travel cost by bus is usually fixed. Res. 9 269. 2 271. 219
Secondly, males tend to choose car or taxi than Dev. 4 7 193.
6
females. This tendency isn’t observed in case of Res.Df 228 233 150
193.
271.
269.
choosing motorcycle and bus. Thirdly, personal Log-like. 4 7 6
income has a negative effect on choosing bus, AIC 289. 277. 209.
7
4
6
meaning that lower income people tend to choose C&S R 2 0.23 0.18 0.14
9
7
6
bus. This tendency isn’t observed in case of choosing Na. R 2 0.31 0.24 0.19
6
8
8
car and motorcycle. It suggests that free ticket of Over. % 70.6 70.3 69.0
public transport might be considered in adverse Signif. codes: 0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1
weathers to facilitate the “shift” traffic management
solutions. Fourthly, several career groups have
positive effects on choosing car/taxi, particularly
young engineer, skilled workers, and artisans.
Meanwhile, unskilled labors have a tendency of
choosing bus. This finding might suggest that trip
reduction through flexible work schedules could be
considered for specific job groups to influence their
mode choice. Fifthly, motorcycle ownership has a
positive effect on choosing car, meaning that
95