Page 118 - 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 route chosen. For shopping trips, trip transport mode and route at higher level of rate when
cancellations are much more common for shopping they are informed during trips.
trips, about 30%-60% respondents canceled their 4.2 Influential factors to the behavioral
shopping trips. Trip postponements are also very changes
common. These behaviors are different from In the previous section, changes in people’s
commuting trips and the differences could be travel behavior in response to adverse weather
explained by fact that it is much easier to postpone conditions are dependent on the context and the trip
or cancel shopping activities than work or study purpose. Furthermore, it is important to understand
activities. how such changes might be influenced by factors
related to socio-demographic characteristics of the
Table 1. Percentages of travel behavior changes by travelers, vehicle ownership status, and trip
adverse weather condition and trip purpose characteristics. Based on the interview data, Pearson
Heavy rain Heavy rain Flood Flood chi-square independence tests were deployed to
Weather
Condition before the during the before the during the determine significantly influential factors.
departure trip departure trip The results of the analysis have confirmed
Trip purpose C S L C S L C S L C S L the dependency of behavioral changes on three
Trip 4.2 39. 45 1.6 11. 0.0 7.1 59. 59. 1.4 18. 18. groups of factors, including individual characterizes
cancelation % 0% % % 8% % % 0% 8% % 2% 2% (e.g., gender, age, vehicle ownership, personal
Delayed income, multiple job status, and the main job), the
33. 30. 29 22. 10. 16.
departure n.a n.a n.a n.a n.a n.a context of adverse weather condition (i.e., heavy
5% 1% % 2% 5% 3%
time rain/flood before or during the trip), and trip features
Waiting for 45. 53. 36. 39. 37. 37. (e.g., trip purpose, mode used, trip length, and trip
resuming of n.a n.a n.a 1% 8% 1% n.a n.a n.a 9% 6% 6% frequency). The null hypothesis of independence is
the trip rejected for all behavioral changes with a p-value
Destination 1.1 3.4 4.0 0.9 3.7 8.5 0.1 8.7 4.7 0.4 9.1 9.1 smaller than 0.005. Cramer’s contingency
change % % % % % % % % % % % % coefficient (referred to as Cramer’s V) was
Route 7.6 0.1 2.1 2.4 2.4 8.2 44. 5.6 10. 27. 13. 13. calculated for determining the strength of the
change % % % % % % 2% % 2% 5% 6% 6% association between the factor under consideration
2.5 0.5 2.6 1.4 0.0 0.7 4.9 0.0 2.6 0.1 0.1 0.1
Mode change and the behavioral change. The results of Pearson
% % % % % % % % % % % % chi-square independence tests show that gender and
55. 26. 17. 48. 28. 46. 21. 16. 6.4 30. 21. 21.
None age are highly influential to the behavioral changes
3% 9% 6% 6% 3% 5% 5% 2% % 7% 4% 4% under heavy rain and flood conditions for
100 100 100 100 100 100 100 100 100 100 100 100
Total commuting and shopping trips because Cramer’s V
% % % % % % % % % % % %
Notes: “C” = commuting trip, “S” = shopping trip, and “L” = leisure trip, “n.a” = is about 0.4-0.5. Female and older people tend to
not applied make trip cancelation, time change and mode change
much more than male and younger people. Similarly,
Similar to the case of shopping trips, trip car and motorcycle ownership may have stronger
postponements and cancellations are the most influence to the behavioral changes as Cramer’s V is
common for leisure trips. About 25%-60% of the about 0.3-0.4. The motorcycle and bicycle users tend
respondents postponed their leisure trips and 15%- to switch to car in case of heavy rain. As shown in
35% canceled their leisure trips. Destination Table 2, trip length and trip frequency have the
changes, routes changes and mode changes are more strongest influences as the Cramer’s V is about 0.6-
frequent for leisure trips than shopping trips. 0.7. The longer and more frequent the trip the higher
However, nobody postponed their leisure trips in likelihood of change.
case of heavy rain during the trip. There seems to be
behavioral homogeneity between leisure trips and Table 2. Dependence of travel behavioral changes
shopping trips because both activities are non- on trip length and trip frequency
obligatory and more flexible than study and work
activities. Wea. Beh. Trip distance Trip frequency
When different contexts are compared, it is cond. change χ2 df Sig. Cra χ2 df Sig. Cra
found that if travelers have information on weather Heav Trip cancel. 3070 10 .000* 0.691 3537 12 .000* 0.742
y rain Time chan. 3104 10 .000* 0.695 3150 12 .000* 0.700
and traffic conditions before trips, they reschedule befor Des. change 3124 10 .000* 0.697 3192 12 .000* 0.705
their journey in terms of cancellation, destination, e the Route chan. 3144 10 .000* 0.700 3211 12 .000* 0.707
dep. Mode chan. 3126 10 .000* 0.698 3191 12 .000* 0.705

93