Viewer's Choice

OPPORTUNITY ATLAS

By: Aiswarya Baiju

CRITIQUE SITE: OPPORTUNITY ATLAS

INTRODUCTION:

  • This visualization is the answer to the question of in which neighborhood of America will the children have a better chance to rise out of poverty
  • The objective of this visualization is to measure the average outcomes of children who grow up in each neighborhood in America by demographic subgroup
  • The neighborhood where people grew up is given much focus instead of where they live as adults
  • Utilized to trace the roots of today's affluence and poverty back to the neighborhood where people grew up
  • It's an initial release of social data and this data is actually a combination of anonymized data from three sources
  • In this website, it also allows you to visualize where and for whom opportunity has been missing for those users who want to visualize the data to develop local solutions to enable more children to rise out of poverty
  • The users can choose where they want to look, who they want to look at and what outcomes they want to see. The Atlas lets you look at outcomes not just for all kids but for specific demographic groups
  • Result of a collaboration between researchers at the Census Bureau, Harvard University and Brown University
  • Purpose: An interactive tool which can be used by the users to find data on children's outcomes in adulthood for every Census tract in U.S. The Opportunity Atlas aims to show which neighborhoods in America will offer the children the best and the worst opportunities to rise out of poverty. This tool lets the user explore a wide selection of data which shows children's adulthood characteristics depending on where they grew up. It is intended this was so these data can be utilized to make more informed decisions in order to address issues related to poverty and to provide opportunities for a better economic upward mobility.

Data Collection

The data for the visualization was used from anonymous data following 20 million Americans from childhood to their mid-30s

Combined three sources of anonymized data and was collected from:

  • the 2000 and 2010 Decennial Census short form
  • Federal income tax returns for 1989,1994, 1995, 1998-2015
  • the 2000 Decennial Census long form and the 2005-2015 American Community Surveys

Constructed an analysis sample of those Americans born between 1978-1983 and mapped these individuals back to their Census tracts that they lived in through age 23 and then for each 70k tracts, the children's outcomes across a range of measures are estimated



THE DATA

The data used for the visualization purposes were retrieved from the Census Bureau. The estimates of upward mobility are based on the outcomes of children born between 1978-83.

Any opinions and conclusions expressed are those of the authors and they don't necessarily reflect the views of the U.S. Census Bureau.

Predicted outcomes are displayed based on actual outcomes of kids who grew up in each geographic area and the prediction model used for this visualization accounts for the fact that they do not observe children at every single parental income level in every single tract and demographic subgroup.

They do not publish estimates that rely on fewer than 20 children because there aren't large enough samples to yield reliable estimates.

The data presents estimates for children in three core parental income groups in the standard version of the Atlas:

  • Low (25th percentile)
  • Middle (50th percentile)
  • High (75th percentile)

Look at the table below:

The users have several options to alter the data and see the changes and variations.

  • First, the users can pick a location of their choice. It could be a city, state, county or an address as shown below in the image:



  • They can then select the outcome they would like to see from their chosen location. There are different spectrums for every selection in the outcome section so based on what's selected, it will be shown like this (in this case, household income was chosen):



  • The visualization also allows the user to see neighborhood information for the chosen location as shown below:




  • Based on the chosen location, the user can see the place on the map (Idaho was selected for this example)



  • The users can also choose the population they want to visualize for the various options they will be selecting such as the type of parent income, child's race or child's gender:


  • The users can then select the children's outcomes in adulthood by neighborhood where they grew up.


  • The users can also vary and alter the neighborhood characteristics and the characteristics of the current residents in each area.

CHALLENGES FACED

  • The data is very limited since each subgroup and income level is not represented in each Census tract so it's not possible to simply calculate average outcomes based on the child's race, gender and parental income rank (this issue was addressed with a regressing model)
  • They had to account for movements across tracts as children often live in more than one Census tract during their childhood

VISUALIZATION

CRITIQUE

The Opportunity Atlas is a phenomenal visualization creation done by the researchers from various places and it goes above and beyond for what is expected for the user's selection. It is the first dataset that provides such longitudinal information at a detailed neighborhood level.


INTERACTIVE FEATURES:

  • the users can see not just where the rich and poor currently live but also whether the children in a given area tend to grow up to become rich or poor
  • the users can also save map images and download the data as you go future references
  • considering the amount of options one can alter, the filtering tools of this visualization remains until you remove it even when you change your selected outcome measure or toggle between different demographic subgroups
  • whenever a filter is in use, a green tile appears to remind you exactly how the data are filtered and one can also overlay their own geographic dataset
  • The visualization offers more than 50 options to choose from for visualization purposes and they can alter them in hundreds of combination to estimate average outcome for the chosen neighborhood. One disadvantage of this would be how the excess options would cause confusions. There is so much on the screen that it becomes overbearing and overwhelming due to the information that's present on the screen. Because there's so much going on the screen, the users will need some time to get the hang of it and to get familiar and without proper instructions, they would be be a little lost.
  • One of the best features is how the developers also included a feature on the side to avoid any confusion and to show the numbers for a quick preview like the image below:
  • Another feature that's fascinating and extremely useful is their implementation of "GUIDE" feature which tells the first time users how to navigate through the data and effectively find what they are looking for. This definitely increases efficiency since the users can find everything after being given a quick step-by-step walk through.



AESTHETICS:

  • The way they presented the data might have been a little overwhelming at first but once you had a walk through of the website and it was very convenient to get clear takeaways by looking at the data representation because they presented it in a logical manner on the side as well along with its representation on the map.
  • They cleverly used the color scheme and graphs and legends in a way that made it very easy to understand the trends and stats of the design elements.
  • Most of the focus is given on the map and the location to show the stats of the requested place and even though there's a lot of data to choose from, one you decide which to go with and how to filter it, then it's very easy to visualize. The entire screen is filled with parts of the data that's available for the users.
  • The users also have the freedom to collapse tabs and filters on the sides in case they want more room the visualize the data that's on the screen
  • Color scaling and Zoom: Changing the color scaling is a powerful way to uncover local variations in outcomes. Changing the color scaling never changes the underlying data, it only changes the colors used on the map


DATA:

As aforementioned, there are over 50 options from which the users can choose from so there is indeed an immense amount of data since we ere accounting for 20 million people. It could be a little difficult to filter multiple things at once but the initial release of this social data has been organized quite well with different modes so they have even more control of what they are viewing.

Basic Mode: the simplest way to view the data

Advanced Mode: adds the ability to filter the corresponding map regions for specific criteria and allows you to look at two additional parental income levels

Compare Mode: allows you to select any two groups and directly compare their outcomes in each neighborhood

Filters: Very useful way to select which estimates appear on screen based on the children's outcomes or neighborhood characteristics because this lets you hone in on neighborhoods with specified characteristics. 

Improvement:
  • After the users select the location, they have to pick the outcome they want to visualize but it's present twice in the page in different formats and this wastes a lot of space. Having it displayed on two sides in a different way leads to confusion considering the amount of information that's already present. They could have had the second one just for definitions and the first one to show the numbers and percentile of those outcomes
  • The hovering feature changes value very fast so if someone has it on a border of a county and they accidentally move it even a millimeter, it will change the value so if they implement a click and select feature, it would have been beneficial
  • Another important improvement, quite an essential one, is having the "Advanced" and "Compare Outcome" mode displayed on the screen so the users know that it exists as well as have it displayed on the initial pop-up that appears so the users are aware of it. Without this, the users won't know that there exists such two features that will enable them to do multitudes of things

CONCLUSION

  • The Opportunity Atlas is available for those who are interested in measuring the average outcomes of children growing up in American neighborhood based on the demographic subgroup. The objective is to measure the average outcomes of children but this tool also helps answers questions regarding:
  1. neighborhoods where people grow up to see if that has any causal effect
  2. children's outcomes across a range of measures who were mapped back to the Census tracts
  • The users can manipulate the visualization using over more than 50 filters which makes it a very robust tool in analyzing the social data. The users can notice any underlying trends and data of other filters as well on the side in case they are wondering about any correlations. Like any other tool, this one could have also used some improvements to enhance the user's experience. But overall, it's an exceptional visualization tool that incorporates multiple dataset to form a result the user is looking for and it's the first tool to provide longitudinal information.

USERS

  • This visualization will help those who want to see which neighborhood will give children more and better opportunities to rise out of poverty
  • This can also be used by those who are trying to find neighborhood that doesn't provide good opportunities to children so they can develop solutions and help them receive better chances on rising out of poverty
  • The data achieved from tract-level will let these users reach different conclusions in different places
  • Individuals such as researchers and policymakers who are constantly seeking to improve opportunities in neighborhoods in America can use this customized platform
  • Users who are keen on visualizing neighborhood characteristics can use this and can manipulate various demographic combinations to focus on factors that affect the future of children