If the pedestrian decides to continue shopping, the next decision assumed is to select a walking direction. For each direction, three factors are considered relevant for a pedestrian's decision. The first factor is whether the direction is the same as the one that the pedestrian just came from, represented by a dummy variable dY (Y=E (East), S (South), W (West), N (North)). Because there is a natural tendency that pedestrians follow previous directions and try to minimize the number of back-turns, a positive influence is expected from this factor. The second factor is the total retail floorspace in the direction, qY. The variable represents the pedestrians' estimate about the attractiveness of retail activities based on his/her perception of the environment. The third factor, lY, is the length of the part of the street that is pedestrianized in the direction, representing the amenity of walking. Tailored to the situation of ENR, the last factor is the location of The Bund. Because The Bund is a special landmark in this area and just located at the eastern end of ENR, it is common among pedestrians to use it for orientation, especially for tourists. It is represented by a dummy variable, bY.

According to the HHM framework, the value function of each direction is defined as:

(5.32)

where Wx and Δx are vectors of state values and factor thresholds, similarly defined as in the go-home model; wd and wb are scalar state values for the two dummy variables. It is assumed that the pedestrian pairwise compares the values of alternative directions under a certain discriminant threshold, λR, and the direction with the highest value rank is selected. If no choice can be made on this basis, random choice is assumed.

Found a mistake? Please highlight the word and press Shift + Enter