Sujan Sikder

Copula-Based Method for Addressing Endogeneity in Models of Severity of Traffic Crash Injuries: Application to Two-Vehicle Crashes

An emerging copula-based methodology is used to address endogeneity in models of crash injury severity. Specifically, two important sources of endogeneity are addressed: (a) endogeneity due to correlations between the injury severity of the two drivers involved in two-vehicle crashes and (b) endogeneity of collision type and injury severity outcomes. Two sets of copula-based joint model systems are formulated and estimated by using data on two-vehicle crashes from the 2007 Generalized Estimates System: a copula-based joint ordered logit–ordered logit model of injury severities of the two drivers involved in two-vehicle crashes and a copula-based joint multinomial logit–ordered logit model of collision type and injury severity outcomes of two-vehicle crashes. The model estimation results and elasticity estimates underscore the importance of accommodating endogeneity in models of crash injury severity. The results shed new light on the determinants of injury severity in two-vehicle accidents.

Spatial Transferability of Person-Level Daily Activity Generation and Time Use Models: Empirical Assessment

An empirical assessment is made of the spatial transferability of person-level daily activity generation and time use models in regions in Florida and between Florida and California. The empirical models are for unemployed adults and are based on the multiple discrete-continuous extreme value structure. The prediction properties of the model are examined first. The results shed new light on the prediction properties of the multiple discrete-continuous extreme value model that have implications for transferability and that provide insight into how the model structure could be improved. Two approaches to transferring models are evaluated—naïve transfer and updating model constants—with measures such as log likelihood–based metrics, aggregate predictive ability, and model sensitivity to changes in demographic characteristics. Results suggest that accurate prediction of aggregate observed patterns is not an adequate yardstick with which to assess transferability; emphasis should be placed on model sensitivity to changes in explanatory variables. Updating of constants improves a transferred models aggregate prediction ability, but not necessarily its policy sensitivity. The extent of transferability between regions within a state is greater than that across states. Within Florida, there is greater transferability between urban regions (especially between Southeast Florida and Central Florida) than between urban and rural regions.

Spatial Transferability of Tour-Based Time-of-Day Choice Models: Empirical Assessment

An empirical assessment of the transferability of tour-based time-of-day (TOD) choice models across different counties in the San Francisco Bay Area of California is presented. Transferability was assessed with two approaches: (a) an application-based approach that tests the transferability of a model as a whole and (b) an estimation-based approach that allows the analyst to test which specific parameters in the model are transferable. Also tested was the hypothesis that pooling data from multiple geographic contexts helps in developing models with better transferability than those estimated from a single context. The estimation-based approach yielded encouraging results in favor of transferability of the TOD choice model, with a majority of parameter estimates in a pooled model found to be transferable. Pooling data from multiple geographic contexts appears to help in developing better transferable models with better transferability. However, attention is needed in selecting the geographic contexts from which to pool data. The pooled data should exhibit the same demographic characteristics and travel level-of-service conditions as in the application context.

Is Usual Share of Commuting Mode Always Greater Than Its Actual Share

With data from the 2001 National Household Travel Survey (NHTS), recent research showed that transits usual share was greater than its actual share for workers in the United States in a variety of commuter markets. A modes usual share is the percentage of workers who state that they usually use that mode for commuting in a week, whereas the actual share of a mode is the percentage of work trips by that mode by the same workers on a typical work day. This study explores whether this relative relationship between a modes usual and actual shares holds true for common modes other than transit for the United States. Mathematically, it is determined that this relative relationship cannot hold true for all modes; in other words, the usual share has to be smaller than the actual share for one or more modes other than transit. Empirically, the same 2001 NHTS is used to test this relative relationship for three common modes—the privately owned vehicle (POV), walking, and biking—and for a variety of commuter markets. The empirical results confirm the mathematical conclusion that the relative relationship holds true for biking but not for POV and walking. In addition, the relationship between usual and actual shares is determined not solely by the mode but also by individual commuter markets. Finally, the deviation between usual and actual shares in percentage terms is large for transit and walking, but small for privately operated vehicles and bikes. One direction of future research would be to determine the reasons for these differences in the usual–actual relationship across modes and commuter markets.