User Guide
- In your web browser, go to the following link:
http://dataexplorer.psrc.org/household-travel-survey/
- Along the top, tabs allow you to select whether you would like to select a summarizing a single data item, in the One-way table tab, or two data items together, Two-way table tab. For example, you may wish to summarize trip travel modes - a one-way table- or trip travel modes by trip destination purpose - a two-way table. The figure below shows the location of the One-way or Two-way table selection.
- If the user selects a One-way table, then to obtain the summary table, they first select the category of the data variable, and then the variable itself.
-
Next the user presses the "Create Table" button to show the table in their web-browser.
-
The user can filter to only show households who live in the City of Seattle.
-
The data can be downloaded to an Excel file by pressing the download data button.
- If the user selects a two-way table, the following screen will be created.
-
To obtain the summary table, the user can select the category of both data variables, and then the two variables. In this example, the user has selected travel mode and travel distance.
-
Next the user presses the "Create Table" button to show the table in their web-browser.
-
The user can select either a data table or a visual chart.
-
Different options for summarization can be selected from the radio buttons:
-
The households used for summarization can be selected with a filter button to select only Seattle households.
- Finally the data can be downloaded by selecting, the Download Data Button.
Travel surveys contain a wealth of rich information that could be used broadly across transportation and land use planning. However, travel survey data has been used mainly for travel model estimation without wider use in planning arenas. One reason many planners do not use survey data is that they lack the skills to weight data, merge tables, and understand margins of errors.
To make travel survey data accessible to a broader audience, the Puget Sound Regional Council (PSRC) data science team has developed an open-source web application called the Travel Survey Data Explorer. The travel survey data explorer is partially inspired by similar Census web applications that allow for the selection and creation of tables that query Census datasets .
The Travel Survey Data Explorer can be found at http://dataexplorer.psrc.org/household-travel-survey/. The app is open source, and its code is on Github at: https://github.com/psrc/travel-study-stories
The app reveals underlying relationships between variables about people and travel. PSRC has created a series of data stories that explore such topics as mode choices, autonomous vehicles, and older people’s travel . These types of stories can be created by any data savvy planner by using the travel data explorer app. We also hope to expand the app to allow for multiple years of data so that it is easy to analyze travel trends over time.
The travel survey data explorer app allows for the generic specification of variables and their values, and thus could be extended to other travel surveys. PSRC would like to collaborate with other travel survey partners to spur collaborative travel survey tool development across the industry. We would like to collaborate on broad array of tools for travel surveys, such as this data explorer but also algorithms and techniques to clean travel survey data.
The app was originally developed with on-going discussions with a group of data savvy planners to ensure that it met their needs. The app is linked on the PSRC travel survey study page so it can be found by any PSRC website visitor. The development team chose to use RShiny as the backend because it is an open source environment that allows for easy data manipulation and clean-looking data visualizations.
The app allows for the creation of a one-way (single dimension table) or a two-way table with a tab selection.
If a one-way table is selected, the user can choose from a category of data and then a specific variable to analyze. The user can download the data to an Excel file, by selecting the download data button.
The app guides the user to consider the sample sizes and margins of error for all the estimates. Many planners have expressed confusion over when it is appropriate to report the data, and when the data should be suppressed. The app is designed to alleviate some of the confusion by linking them to information about margins of errors and graying out any cells in which the number of samples is less than 30.
The two-way table tab allows the user to select two variables to create a crosstab to show the data across two dimensions, as shown in Figure 3. Then the user can select to either see the data as a bar chart or a table. As with the one-dimensional table, information about margins of error is provided and the data can be downloaded to an Excel file.
Simple data tables from travel surveys can be powerfully illuminating on various topics. PSRC has written data stories about older adults, autonomous vehicles, employer transportation benefits, home choices, and mode choices. The following are examples of simple data from the travel survey app that is valuable to planning:
- A large majority of workers (82%) reported having free parking at work.
- Older adults are in cars on about 85% of their trips.
- Households making less than $25,000 a year are three times as likely to use transit on each trip, and twice as likely to walk than households making $25,000-$100,000 a year.
The travel survey data explorer app is in beta form and PSRC will continue to improve it over time. Some of the improvements we hope to include are: handling of multiple years of data, continuous incorporation of feedback from planners, functionality for grouping and filtering data. We greatly hope that other travel survey researchers will join with us to build this app over time.