Search

Search Results

• 1. [Article] Evaluating the Use of Crowdsourcing as a Data Collection Method for Bicycle Performance Measures and Identification of Facility Improvement Needs

  This research developed a smartphone application called ORcycle to collect cyclists’ routes, users, and comfort levels. ORcycle combines GPS revealed route data collection with new questionnaires that ...

  Citation × Citation Citation Abstract Copy Title: Evaluating the Use of Crowdsourcing as a Data Collection Method for Bicycle Performance Measures and Identification of Facility Improvement Needs Author: Figliozzi, Miguel A., Blanc, Bryan Philip Year: 2015 This research developed a smartphone application called ORcycle to collect cyclists’ routes, users, and comfort levels. ORcycle combines GPS revealed route data collection with new questionnaires that try to elicit cyclists’ attitudes as well as comfort levels and factors that influence their perceived comfort and route choice. The new questionnaires were developed to better understand how cyclists’ comfort levels are affected by route characteristics, route stressors, safety reports, cyclists’ demographics, and cyclists’ cycling attitude. Preliminary results show that many trip characteristics, route choice factors, route stressors and demographic variables are correlated with comfort levels. ORcycle is the first statewide deployment of a smartphone application to collect bicycle specific safety and crash data in addition to travel and comfort data. Potential applications that can take advantage of the ORcycle data include: comfort and route models, prioritization of network improvements, and crash and injury risk models. Title: Evaluating the Use of Crowdsourcing as a Data Collection Method for Bicycle Performance Measures and Identification of Facility Improvement Needs Author: Figliozzi, Miguel A., Blanc, Bryan Philip Year: 2015 This research developed a smartphone application called ORcycle to collect cyclists’ routes, users, and comfort levels. ORcycle combines GPS revealed route data collection with new questionnaires that try to elicit cyclists’ attitudes as well as comfort levels and factors that influence their perceived comfort and route choice. The new questionnaires were developed to better understand how cyclists’ comfort levels are affected by route characteristics, route stressors, safety reports, cyclists’ demographics, and cyclists’ cycling attitude. Preliminary results show that many trip characteristics, route choice factors, route stressors and demographic variables are correlated with comfort levels. ORcycle is the first statewide deployment of a smartphone application to collect bicycle specific safety and crash data in addition to travel and comfort data. Potential applications that can take advantage of the ORcycle data include: comfort and route models, prioritization of network improvements, and crash and injury risk models. Close

• 2. [Article] How to Estimate Pedestrian Demand

  There is growing support to improve the quality of the walking environment and make investments to promote pedestrian travel. Such efforts often require analytical non-motorized planning tools to estimate ...

  Citation × Citation Citation Abstract Copy Title: How to Estimate Pedestrian Demand Author: Clifton, Kelly, Singleton, Patrick Allen, Muhs, Christopher D., Schneider, Robert J. Year: 2015 There is growing support to improve the quality of the walking environment and make investments to promote pedestrian travel. Such efforts often require analytical non-motorized planning tools to estimate levels of pedestrian demand that are sensitive to environmental and demographic factors at an appropriate scale. Despite this interest and need, current forecasting tools, particularly regional travel demand models, often fall short. To address this gap, Oregon Metro and NITC researcher Kelly Clifton worked together to develop a pedestrian demand estimation tool. For generations, planners have been using statistical models to forecast travel demand, but these models have traditionally been auto-centered. The new tool will allow planners to allocate infrastructure based on pedestrian demand in the Portland, Oregon metropolitan area. The tool is also designed to be replicable, so that other metropolitan areas can adapt the model to begin estimating pedestrian demand in their cities. Title: How to Estimate Pedestrian Demand Author: Clifton, Kelly, Singleton, Patrick Allen, Muhs, Christopher D., Schneider, Robert J. Year: 2015 There is growing support to improve the quality of the walking environment and make investments to promote pedestrian travel. Such efforts often require analytical non-motorized planning tools to estimate levels of pedestrian demand that are sensitive to environmental and demographic factors at an appropriate scale. Despite this interest and need, current forecasting tools, particularly regional travel demand models, often fall short. To address this gap, Oregon Metro and NITC researcher Kelly Clifton worked together to develop a pedestrian demand estimation tool. For generations, planners have been using statistical models to forecast travel demand, but these models have traditionally been auto-centered. The new tool will allow planners to allocate infrastructure based on pedestrian demand in the Portland, Oregon metropolitan area. The tool is also designed to be replicable, so that other metropolitan areas can adapt the model to begin estimating pedestrian demand in their cities. Close

• 3. [Article] Examining Consumer Behavior and Travel Choices

  This study represents a first attempt to answer a few of the questions that have arisen concerning multimodal transportation investments and the impacts of mode shifts on the business community. This research ...

  Citation × Citation Citation Abstract Copy Title: Examining Consumer Behavior and Travel Choices Author: Clifton, Kelly J., Muhs, Christopher Devlin, Morrissey, Sara, Morrissey, Tomás, Currans, Kristina Marie, Ritter, Chloe Year: 2013 This study represents a first attempt to answer a few of the questions that have arisen concerning multimodal transportation investments and the impacts of mode shifts on the business community. This research aims to merge the long history of scholarly work that examines the impacts of the built environment on non-work travel with the relatively new interest in consumer spending by mode of travel. This empirical study of travel choices and consumer spending across 89 businesses in the Portland metropolitan area shows there are important differences between the amounts customers spend on average at various businesses by their mode of travel. However, these differences become less pronounced when we control for demographics of the customer and other attributes of the trip. This study of consumer spending and travel choices has some compelling findings that suggest some key spending and frequency differences by mode of travel that will likely invigorate the discussion of the economic impacts of these modes. Key findings are: 1) Bicyclists, pedestrians and transit riders are competitive consumers: When demographics and socioeconomics are controlled for, mode choice does not have a statistically significant impact on consumer spending at convenience stores, drinking establishments and restaurants. When trip frequency is accounted for, the average monthly expenditures by customer modes of travel reveal that bicyclists, transit users and pedestrians are competitive consumers and, for all businesses except supermarkets, spend more on average than those who drive and 2) The built environment matters: We support previous literature and find that residential and employment density, the proximity to rail transit, and the amount of automobile and bicycle parking are all important in explaining the use of non-automobile modes. In particular, provision of bike parking and bike corrals are significant predictors of bike mode share at the establishment level. Other findings lend more insight into the relationship between consumer behavior and travel choices. For the non-work destinations studied, the automobile remains the dominant mode of travel. Patrons are largely arriving by private vehicle to most of the destinations in this study, particularly to grocery stores where larger quantities of goods tend to be purchased. But, high non-automobile mode shares and short travel distances exist in areas of concentrated urban activity. In sum, this study provides some empirical evidence to answer the questions of business owners about how mode shifts might impact their market shares and revenues. More work is needed to better understand the implications of future changes and to provide a robust assessment of the returns on these investments and their economic impacts. Title: Examining Consumer Behavior and Travel Choices Author: Clifton, Kelly J., Muhs, Christopher Devlin, Morrissey, Sara, Morrissey, Tomás, Currans, Kristina Marie, Ritter, Chloe Year: 2013 This study represents a first attempt to answer a few of the questions that have arisen concerning multimodal transportation investments and the impacts of mode shifts on the business community. This research aims to merge the long history of scholarly work that examines the impacts of the built environment on non-work travel with the relatively new interest in consumer spending by mode of travel. This empirical study of travel choices and consumer spending across 89 businesses in the Portland metropolitan area shows there are important differences between the amounts customers spend on average at various businesses by their mode of travel. However, these differences become less pronounced when we control for demographics of the customer and other attributes of the trip. This study of consumer spending and travel choices has some compelling findings that suggest some key spending and frequency differences by mode of travel that will likely invigorate the discussion of the economic impacts of these modes. Key findings are: 1) Bicyclists, pedestrians and transit riders are competitive consumers: When demographics and socioeconomics are controlled for, mode choice does not have a statistically significant impact on consumer spending at convenience stores, drinking establishments and restaurants. When trip frequency is accounted for, the average monthly expenditures by customer modes of travel reveal that bicyclists, transit users and pedestrians are competitive consumers and, for all businesses except supermarkets, spend more on average than those who drive and 2) The built environment matters: We support previous literature and find that residential and employment density, the proximity to rail transit, and the amount of automobile and bicycle parking are all important in explaining the use of non-automobile modes. In particular, provision of bike parking and bike corrals are significant predictors of bike mode share at the establishment level. Other findings lend more insight into the relationship between consumer behavior and travel choices. For the non-work destinations studied, the automobile remains the dominant mode of travel. Patrons are largely arriving by private vehicle to most of the destinations in this study, particularly to grocery stores where larger quantities of goods tend to be purchased. But, high non-automobile mode shares and short travel distances exist in areas of concentrated urban activity. In sum, this study provides some empirical evidence to answer the questions of business owners about how mode shifts might impact their market shares and revenues. More work is needed to better understand the implications of future changes and to provide a robust assessment of the returns on these investments and their economic impacts. Close

• 4. [Article] Understanding Residential Location Choices for Climate Change and Transportation Decision Making: Phase 2 Report

  This research builds on the related Phase 1 project. In this second phase, we continue to study neighborhood and housing preferences that shape the residential location decision process. An online experimental ...

  Citation × Citation Citation Abstract Copy Title: Understanding Residential Location Choices for Climate Change and Transportation Decision Making: Phase 2 Report Author: Clifton, Kelly J., Gehrke, Steven R., Currans, Kristina Marie Year: 2015 This research builds on the related Phase 1 project. In this second phase, we continue to study neighborhood and housing preferences that shape the residential location decision process. An online experimental survey tool is developed to investigate lifestyle preferences and tradeoffs that households make in their location decisions. This computer-aided experimental survey draws upon stated preference methods to engage participants in questions about residential location and transportation options. The survey infrastructure was extensively piloted (6-10% response rate). The 10-minute survey can be deployed for future investigations. This infrastructure is a contribution for the integration of visualized neighborhood typologies, or concepts, which were objectively defined using data from 25 of the most populous metropolitan regions from around the United States. The construct of neighborhoods is based upon national data to account for potential options not currently available in Oregon. These visualizations help ground the survey respondents in the same reality and were carefully crafted to convey various attributes of the built and transportation environment. The initial analysis of the preference data collected in this survey (N=1,035) indicates that the preferences for neighborhood, housing, and transportation characteristics have a greater influence on the preferred neighborhood concept than the more typically used socio-economic characteristics (income, household size, age). Another interesting preliminary finding is that 27% of respondents would prefer to live in a more urban neighborhood than they currently reside. These “urban seeking” respondents had no particular demographic trend, providing little evidence that specific socioeconomic markets had specific preferences for the built environment. Title: Understanding Residential Location Choices for Climate Change and Transportation Decision Making: Phase 2 Report Author: Clifton, Kelly J., Gehrke, Steven R., Currans, Kristina Marie Year: 2015 This research builds on the related Phase 1 project. In this second phase, we continue to study neighborhood and housing preferences that shape the residential location decision process. An online experimental survey tool is developed to investigate lifestyle preferences and tradeoffs that households make in their location decisions. This computer-aided experimental survey draws upon stated preference methods to engage participants in questions about residential location and transportation options. The survey infrastructure was extensively piloted (6-10% response rate). The 10-minute survey can be deployed for future investigations. This infrastructure is a contribution for the integration of visualized neighborhood typologies, or concepts, which were objectively defined using data from 25 of the most populous metropolitan regions from around the United States. The construct of neighborhoods is based upon national data to account for potential options not currently available in Oregon. These visualizations help ground the survey respondents in the same reality and were carefully crafted to convey various attributes of the built and transportation environment. The initial analysis of the preference data collected in this survey (N=1,035) indicates that the preferences for neighborhood, housing, and transportation characteristics have a greater influence on the preferred neighborhood concept than the more typically used socio-economic characteristics (income, household size, age). Another interesting preliminary finding is that 27% of respondents would prefer to live in a more urban neighborhood than they currently reside. These “urban seeking” respondents had no particular demographic trend, providing little evidence that specific socioeconomic markets had specific preferences for the built environment. Close

• 5. [Article] Can Protected Bike Lanes Help Close the Gender Gap in Cycling? Lessons from Five Cities

  Even in areas with increased levels of bicycling, there remains a significant “gender gap” in bicycling in the United States, in contrast to many other countries with high rates of bicycling. The primary ...

  Citation × Citation Citation Abstract Copy Title: Can Protected Bike Lanes Help Close the Gender Gap in Cycling? Lessons from Five Cities Author: Dill, Jennifer, Goddard, Tara, Monsere, Christopher, McNeil, Nathan Year: 2014 Even in areas with increased levels of bicycling, there remains a significant “gender gap” in bicycling in the United States, in contrast to many other countries with high rates of bicycling. The primary objective of this paper was to explore whether protected bike lanes could help reduce the gender gap. To do so, the authors used survey data from a comprehensive evaluation of protected bike lanes in five large U.S. cities (Austin, TX, Chicago, IL, Portland, OR, San Francisco, CA, and Washington, DC) that included survey responses of 1,111 intercepted bicyclists and 2,283 residents. Both men and women overwhelmingly felt that the lanes increased their safety while riding in them. On several measures of safety and comfort, women bicyclists using the lanes did have significantly more positive associations with the protected lanes than men. A survey of residents, which included people who do and do not bicycle, revealed that women generally feel less comfortable than men bicycling on roadways, though the addition of some physical separation to a striped bike lane, does increase stated levels of comfort to a level that might increase rates of bicycling. Overall, women were also more likely to indicate that the new protected lane had increased their overall levels of bicycling. Analysis of the demographics of the intercepted bicyclists revealed differences between men and women that point to the challenge of increasing women’s rates of bicycling, primarily that the women intercepted bicycling were much less likely than men to have children. Title: Can Protected Bike Lanes Help Close the Gender Gap in Cycling? Lessons from Five Cities Author: Dill, Jennifer, Goddard, Tara, Monsere, Christopher, McNeil, Nathan Year: 2014 Even in areas with increased levels of bicycling, there remains a significant “gender gap” in bicycling in the United States, in contrast to many other countries with high rates of bicycling. The primary objective of this paper was to explore whether protected bike lanes could help reduce the gender gap. To do so, the authors used survey data from a comprehensive evaluation of protected bike lanes in five large U.S. cities (Austin, TX, Chicago, IL, Portland, OR, San Francisco, CA, and Washington, DC) that included survey responses of 1,111 intercepted bicyclists and 2,283 residents. Both men and women overwhelmingly felt that the lanes increased their safety while riding in them. On several measures of safety and comfort, women bicyclists using the lanes did have significantly more positive associations with the protected lanes than men. A survey of residents, which included people who do and do not bicycle, revealed that women generally feel less comfortable than men bicycling on roadways, though the addition of some physical separation to a striped bike lane, does increase stated levels of comfort to a level that might increase rates of bicycling. Overall, women were also more likely to indicate that the new protected lane had increased their overall levels of bicycling. Analysis of the demographics of the intercepted bicyclists revealed differences between men and women that point to the challenge of increasing women’s rates of bicycling, primarily that the women intercepted bicycling were much less likely than men to have children. Close

• 6. [Article] Development of a Pedestrian Demand Estimation Tool: a Destination Choice Model

  There is growing support for improvements to the quality of the walking environment, including more investments to promote pedestrian travel. Planners, engineers, and others seek improved tools to estimate ...

  Citation × Citation Citation Abstract Copy Title: Development of a Pedestrian Demand Estimation Tool: a Destination Choice Model Author: Muhs, Christopher D., Clifton, Kelly, Singleton, Patrick Allen, Schneider, Robert J. Year: 2015 There is growing support for improvements to the quality of the walking environment, including more investments to promote pedestrian travel. Planners, engineers, and others seek improved tools to estimate pedestrian demand that are sensitive to environmental and demographic factors at the appropriate scale in order to aid policy-relevant issues like air quality, public health, and smart allocation of infrastructure and other resources. Further, in the travel demand forecasting realm, tools of this kind are difficult to implement due to the use of spatial scales of analysis that are oriented towards motorized modes, vast data requirements, and computer processing limitations. To address these issues, a two-phase project between Portland State University and Oregon Metro is underway to develop a robust pedestrian planning method for use in regional travel demand models. The first phase, completed in 2013, utilizes a tool that predicts the number of walking trips generated with spatial acuity, based on a new measure of the pedestrian environment and a micro-level unit of analysis. Currently, phase two is building upon this tool to predict the distribution of walking trips, connecting the origins predicted in phase one to destinations. This presentation will focus on phase two, which is one of the first studies to focus on destination choices among pedestrians separately from other modes. The approach can be extended to identify the spatial extent of potential pedestrian paths to these destinations. Ultimately, the products developed from the research can estimate various aspects of pedestrian demand – trip generation, trip distribution, and areas of potential pedestrian activity. These tools will add to the analytical methods available for transportation modeling, pedestrian and safety analysis, health assessments, and other pedestrian planning applications. Title: Development of a Pedestrian Demand Estimation Tool: a Destination Choice Model Author: Muhs, Christopher D., Clifton, Kelly, Singleton, Patrick Allen, Schneider, Robert J. Year: 2015 There is growing support for improvements to the quality of the walking environment, including more investments to promote pedestrian travel. Planners, engineers, and others seek improved tools to estimate pedestrian demand that are sensitive to environmental and demographic factors at the appropriate scale in order to aid policy-relevant issues like air quality, public health, and smart allocation of infrastructure and other resources. Further, in the travel demand forecasting realm, tools of this kind are difficult to implement due to the use of spatial scales of analysis that are oriented towards motorized modes, vast data requirements, and computer processing limitations. To address these issues, a two-phase project between Portland State University and Oregon Metro is underway to develop a robust pedestrian planning method for use in regional travel demand models. The first phase, completed in 2013, utilizes a tool that predicts the number of walking trips generated with spatial acuity, based on a new measure of the pedestrian environment and a micro-level unit of analysis. Currently, phase two is building upon this tool to predict the distribution of walking trips, connecting the origins predicted in phase one to destinations. This presentation will focus on phase two, which is one of the first studies to focus on destination choices among pedestrians separately from other modes. The approach can be extended to identify the spatial extent of potential pedestrian paths to these destinations. Ultimately, the products developed from the research can estimate various aspects of pedestrian demand – trip generation, trip distribution, and areas of potential pedestrian activity. These tools will add to the analytical methods available for transportation modeling, pedestrian and safety analysis, health assessments, and other pedestrian planning applications. Close

• 7. [Article] A Strategy for Reducing the Impact of Driving Under Influence of Intoxicants in Portland, Oregon

  Driving under the influence of intoxicants (DUII) applies to operating a vehicle under the influence of any substance that can impair driving performance. Nationally and locally there is a strong desire ...

  Citation × Citation Citation Abstract Copy Title: A Strategy for Reducing the Impact of Driving Under Influence of Intoxicants in Portland, Oregon Author: Monsere, Christopher, Chi, Delia Year: 2008 Driving under the influence of intoxicants (DUII) applies to operating a vehicle under the influence of any substance that can impair driving performance. Nationally and locally there is a strong desire to reduce the societal impacts of this significant social problem. The trend in U.S. alcohol-related motor vehicle crashes was generally decreasing in the mid to late 1980’s but has remained fairly flat since the early 1990s. In 1982, approximately 60% of all US traffic fatalities were alcohol-related. By 1994, this had decreased to 43% but in the decade that followed it has changed little. In 2005, nearly 39% of fatal crashes still involved alcohol. This contrasts sharply with the trends in many other peer industrialized countries such as Australia which has seen a 60% drop in the total number of fatal alcohol-involved crashes since 1980. In most U.S. jurisdictions, the trend for alcohol-involved crashes mirrors the disappointing national trends and in some has even been slightly increasing. The reasons for these trends are many but can be found in changing demographics, reduced resources for enforcement, prosecution, and treatment as well as changing public behaviors and attitudes. The objective of this study was to identify tools, techniques and strategies that could help reverse this trend for the Portland metropolitan area. It is clear that the system to prevent, enforce, adjudicate, and treat alcohol related problems is complex and will require the cooperation and energy of many different stakeholders. Title: A Strategy for Reducing the Impact of Driving Under Influence of Intoxicants in Portland, Oregon Author: Monsere, Christopher, Chi, Delia Year: 2008 Driving under the influence of intoxicants (DUII) applies to operating a vehicle under the influence of any substance that can impair driving performance. Nationally and locally there is a strong desire to reduce the societal impacts of this significant social problem. The trend in U.S. alcohol-related motor vehicle crashes was generally decreasing in the mid to late 1980’s but has remained fairly flat since the early 1990s. In 1982, approximately 60% of all US traffic fatalities were alcohol-related. By 1994, this had decreased to 43% but in the decade that followed it has changed little. In 2005, nearly 39% of fatal crashes still involved alcohol. This contrasts sharply with the trends in many other peer industrialized countries such as Australia which has seen a 60% drop in the total number of fatal alcohol-involved crashes since 1980. In most U.S. jurisdictions, the trend for alcohol-involved crashes mirrors the disappointing national trends and in some has even been slightly increasing. The reasons for these trends are many but can be found in changing demographics, reduced resources for enforcement, prosecution, and treatment as well as changing public behaviors and attitudes. The objective of this study was to identify tools, techniques and strategies that could help reverse this trend for the Portland metropolitan area. It is clear that the system to prevent, enforce, adjudicate, and treat alcohol related problems is complex and will require the cooperation and energy of many different stakeholders. Close

• 8. [Article] Understanding Travel Modes to Non-work Destinations: Analysis of an Establishment Survey in Portland, Oregon

  During the past three decades, research in travel behavior has generally proceeded from broad-level, aggregate analysis of mode share--the proportions of walking, bicycling, transit, and vehicle travel ...

  Citation × Citation Citation Abstract Copy Title: Understanding Travel Modes to Non-work Destinations: Analysis of an Establishment Survey in Portland, Oregon Author: Muhs, Christopher D. Year: 2013 During the past three decades, research in travel behavior has generally proceeded from broad-level, aggregate analysis of mode share--the proportions of walking, bicycling, transit, and vehicle travel occurring in traffic analysis zones, census tracts, neighborhood, or other geographical units--to fine-grained, disaggregate analysis of mode choices and other trip-making attributes at the individual level. One potential issue is whether there are differences in the types of conclusions drawn from results of analyses performed at these different levels, as these results directly inform transportation planning and policy. This thesis aims in part to confirm whether the types of conclusions drawn from different levels of analysis are different, and to what extent. We also examine the relationships between the built environment and non-work travel choices from a unique analysis perspective. To do this, we use data from a 2011 travel intercept survey in the Portland, Oregon metropolitan region that was administered at convenience store, bar, and restaurant establishments. We estimate, for each of the travel modes--walk, bicycle, and automobile--two analysis models: one binary logistic regression model for mode choice of the individual traveler going to the establishment and one multiple linear regression model for mode share of shoppers at the establishment. Both models control for socio-demographics, trip characteristics, and built environment measures of travelers. For the binary logistic regression models, the data are disaggregate and particular to the individual traveler. These models also controlled for attitudes and preference towards travel modes. For the multiple regression models, data are aggregated to the establishment. The built environment data in each model represent characteristics of urban form surrounding the establishment. The data being oriented to the destination-end of the trip, as well as providing controls on land use make this analysis unique in the literature, as most non-work travel studies use residential-based data. Results suggest that analyses performed at the two different levels provide policy-relevant but somewhat different conclusions. In general, characteristics of the individual and the trip have stronger associations with mode choices of individuals than when aggregated to the establishment and analyzed against the mode share patterns of shoppers. Instead, mode shares have stronger relationships with characteristics of the built environment. The built environment surrounding the destination has a much more pronounced association with mode shares at the establishment than with mode choices of individuals. The results highlight the usefulness of simple aggregate analysis, when appropriate. We also find large differences between modes in which characteristics are important for mode choice and mode share. Walking and automobile models behave somewhat similarly but in opposite directions, while bicycling behaves quite differently. These differences suggest on their own a move away from non-motorized travel to be considered as equivalent or assessed as one item in research and in practice. Title: Understanding Travel Modes to Non-work Destinations: Analysis of an Establishment Survey in Portland, Oregon Author: Muhs, Christopher D. Year: 2013 During the past three decades, research in travel behavior has generally proceeded from broad-level, aggregate analysis of mode share--the proportions of walking, bicycling, transit, and vehicle travel occurring in traffic analysis zones, census tracts, neighborhood, or other geographical units--to fine-grained, disaggregate analysis of mode choices and other trip-making attributes at the individual level. One potential issue is whether there are differences in the types of conclusions drawn from results of analyses performed at these different levels, as these results directly inform transportation planning and policy. This thesis aims in part to confirm whether the types of conclusions drawn from different levels of analysis are different, and to what extent. We also examine the relationships between the built environment and non-work travel choices from a unique analysis perspective. To do this, we use data from a 2011 travel intercept survey in the Portland, Oregon metropolitan region that was administered at convenience store, bar, and restaurant establishments. We estimate, for each of the travel modes--walk, bicycle, and automobile--two analysis models: one binary logistic regression model for mode choice of the individual traveler going to the establishment and one multiple linear regression model for mode share of shoppers at the establishment. Both models control for socio-demographics, trip characteristics, and built environment measures of travelers. For the binary logistic regression models, the data are disaggregate and particular to the individual traveler. These models also controlled for attitudes and preference towards travel modes. For the multiple regression models, data are aggregated to the establishment. The built environment data in each model represent characteristics of urban form surrounding the establishment. The data being oriented to the destination-end of the trip, as well as providing controls on land use make this analysis unique in the literature, as most non-work travel studies use residential-based data. Results suggest that analyses performed at the two different levels provide policy-relevant but somewhat different conclusions. In general, characteristics of the individual and the trip have stronger associations with mode choices of individuals than when aggregated to the establishment and analyzed against the mode share patterns of shoppers. Instead, mode shares have stronger relationships with characteristics of the built environment. The built environment surrounding the destination has a much more pronounced association with mode shares at the establishment than with mode choices of individuals. The results highlight the usefulness of simple aggregate analysis, when appropriate. We also find large differences between modes in which characteristics are important for mode choice and mode share. Walking and automobile models behave somewhat similarly but in opposite directions, while bicycling behaves quite differently. These differences suggest on their own a move away from non-motorized travel to be considered as equivalent or assessed as one item in research and in practice. Close