Why Care About Social Networks in Travel Demand Forecasting? Testing the Predictive Power of Social Attributes in Modeling Discretionary Trip Frequencies

In recent years, social networks are gaining increasingly more attention in explaining personal travel behavior. A fundamental question in this regard is what value does social network data add to explain and predict travel patterns? Although several studies have documented the empirical significance, a test of the added predictive value of an individual’s social networks has not yet been conducted. In this paper, the authors test whether the inclusion of social network variables, alongside the conventional personal socio-demographics, improves the predictive power of trip frequency models. The empirical analysis is undertaken in the context of a 2011 Dutch survey that retrospectively collected information about the evolution of social networks and travel patterns. The findings offer valuable insight into how different social network characteristics vary in their predictive power both within and across activity purposes. First, not all social network variables are statistically significant for all travel purposes, e.g. the number of strong ties is not significant for shopping trips. Second, even if a parameter is statistically significant in estimation, prediction results vary, e.g., number of geographically close alters asserts statistical significance in the estimation model of active recreation trips yet adds no value to the prediction outcome. This is an important finding and puts the prediction efforts into perspective in that fitting a model to the data does not necessarily establish value in prediction and forecasting. Finally, social network variables show none to phenomenal improvement in prediction across activity types. In summary, the relationship length and geographical spread of social networks are found to be significant across all travel purposes. However, the size of social networks contributes to better predictions.

• Supplemental Notes:
  • This paper was sponsored by TRB committee ADB10 Standing Committee on Traveler Behavior and Values.
• Corporate Authors:

  Transportation Research Board

  500 Fifth Street, NW
  Washington, DC  United States  20001
• Authors:
  • Sharmeen, Fariya
  • Sivakumar, Aruna
• Conference:
  • Transportation Research Board 96th Annual Meeting
  • Location: Washington DC, United States
  • Date: 2017-1-8 to 2017-1-12
• Date: 2017

Language

• English

Media Info

• Media Type: Digital/other
• Features: References; Tables;
• Pagination: 19p
• Monograph Title: TRB 96th Annual Meeting Compendium of Papers

Subject/Index Terms

• TRT Terms: Demographics; Mathematical prediction; Recreational trips; Traffic forecasting; Travel behavior; Travel demand
• Candidate Terms: Social networking
• Geographic Terms: Netherlands
• Subject Areas: Highways; Planning and Forecasting; Safety and Human Factors;

Filing Info

• Accession Number: 01632421
• Record Type: Publication
• Report/Paper Numbers: 17-04092
• Files: TRIS, TRB, ATRI
• Created Date: Apr 19 2017 9:24AM