Question

Design a stated preference survey for modeling the mode choice of commuters. There are three choices of modes: bus, rail, and ferry. Please address three alternative specific attributes in the hypothetical scenarios and three decision maker specific attributes in the survey questions.

Answer 1

Stated preference are an efficient method to analyze consumers’ evaluation of mode choice alternatives when there are hypothetical choice alternatives and new attributes. Let us develop a stated preference questionnaire for three modes of travel i.e bus, rail and ferry.

The most important thing while setting up a stated preference questionnaire is to understand the different attributes that contribute to a user choosing that mode. The three specific attributes I would keep in the survey questionnaire are

a) Travel Cost

b ) Travel time

c) No of changes/transfers

d) Chance of getting a seat

So we can design our survey questionnaire like this

1) Do you give more importance to travel cost or travel time while you commute to work.

2) Suppose you were forced to take public transit - how important is it to get a seat while travelling. Three choices - important, very important, not important.

3) When taking public transit to work - how much importance to you give to the "number of transfers that you make". What is the maximum number of transfers that you are willing to make.

=======================================================================================================================================================================================

Lets assign the "sign and values to these coefficients.

a) Travel Cost - assume it costs more to travel in ferry and the lest cost is by bus, while rail cost are in between busy and ferry. As cost increases, the utility functions should decrease, so the sign of he attribute parameter are -ve. So assume that coefficient for Travel Cost is -0.1

b) Travel time - clearly as travel time increases, the mode is less preferred, so again the sign of the coefficient is -ve. Typically ferries are present in areas which have a river access and can significantly reduce travel time. So assign a steeper coefficient to this parameter. Let coefficient for Travel time = -0.2. This basically says users give twice the emphasis on travel time as compared to travel cost

c) No of changes/transfers. - In road trips you dont have any transfers, you go in one automotbile trip from your origination to destination. As you have more transfers the mode becomes less attractive. So assign a coefficient of -0.05 for # of changes/transfers.

d) Chance of getting a seat - This is an important parameter for many travellers. It often influences the decision makign process and whether a particular transit mode is preferred. For cars we always get a seat, For rail and ferry we may or may not. So lets have the coefficient for chance of getting a seat defined as follows.

The sign of this attribute will be +. As chances of getting as eat increase, the utility function value should increase.

Assume this attribute can take the following values

a) .1 if the probability of getting a seat is equal to 1

b) 08 if the probability of getting a seat is greater than 0.5 and less than 1

c) .04 if the probability of getting a seat is between 0.25 and 0.5

cd 0.01 if the probability of getting a seat is less than 0.5

So now we can set up the utility function as below

Ui = 0.1* Travel Cost -0.2 Travel time -.05 * number of transfers + Coefficient for Seat Availibility

so Ubus = -01* Travel costbus - 0.2*Travel timebus - (0.05 * number of transfer) +.1

Urail = -01* Travel costrail - 0.2*Travel timerail- (0.05 * number of transfers) +0 .08

Uferry = -01* Travel Costferry- 0.2*Travel timeferry- (0.05 * number of transfers) +0 .04