It also can be found that WTA for commuting is left-skewed distribution, while, for the shopping and leisure purpose, WTA is right-skewed distribution. This is not consistent with existing findings [17]. 4.2. Effect of Time Savings In theory, VTTS alters

with the change of travel time due to the money budget constraint and it has been validated [11]. This implicates that VTTS and WTA should not keep order Seliciclib constant with change of time saving size. Figure 1 shows the relationship of WTA and the time savings for commuting trips. It is illustrated that WTA decreases with the raising of time saving Δt. The same feature is also presented in WTA for shopping trips (see Figure 2). Figure 1 Variability of WTA with time saving for commuting trips. Figure 2 Variability of WTA with time saving for shopping trips. The value of small time saving is a contentious issue in estimating VTTS [10]. This issue also arises for WTA. Figures ​Figures11 and ​and22 show that WTA is higher than it is expected (for commuting trips, it is higher than 120CNY/hour

and 200CNY/hour for shopping trips) for small time savings (less than 5 minutes). It can be explained that, for the small time savings, other characters such as comfort and level of service are dominated [16] and that some travelers would not give up driving passenger car. 4.3. Effect of Cost Saving Table 2 lists the statistics of the cost savings for the three kind trips (commuting, shopping, and leisure). From the statistics, it is found that although there are differences among the cost savings, the range of the upper and the lower bound for 95% confidence interval of each kind trip cost saving is very small which means that the cost budget constraint plays a role. Therefore, while the time saving size

varies greatly, the cost saving keeps constant (Figures ​(Figures33 and ​and44 illustrate the change of cost saving Δc with the time saving Δt for commuting and shopping trips, resp.). It is reasonable that the small time savings are accompanied with higher WTA and WTA decreases with increase of travel time savings. Figure 3 Change of cost saving for commuting trips. Figure 4 Change of cost saving for shopping trips. Table 2 Summary of cost saving for different trip purposes. 4.4. Discussion of the Results The GSK-3 effects of variables (e.g., individual income, trip length, trip mode, sex, and career) are discussed in some literatures [1, 7–14]. Therefore, these factors are not analyzed in this paper. This does not indicate that the influences of these variables are unimportant. For this paper, the influences of time saving and cost saving are mainly studied due to the fact that they are often ignored. 5. Modelling A linear model is built to describe the relationship of WTA with the influencing variables. In the model, the trip length, saving time, saving cost, allowance, and individual income are considered.