Dataset
The data we will be working with will be US census data from 2020
([Link] We will be considering 2-3 demographic attributes, like
ethnicity, educational attainments and income at the state level.
Goal
The goal of this exercise will be to compare the efficiency of three different types of maps, which
display 2 or 3 demographic characteristics at the state level for US census data. The goal is to explore
which type of map or combination of maps is efficient for understanding the relationship between 2
or 3 variables, and also to extract the exact values of the variables based on the other variables.
Different types of designs are available, including one single map with a number of layers, where the
details can pop up by a tooltip; the layers being available for filtration; pie charts or bubbles overlaid
on the layer; a split screen showing the variables separately; bar charts that pop up in a separate
window once you click a state etc. The goal is to understand which combination is best for different
tasks.
Tasks
The tasks pertaining to the map are as follows:
(1) Understanding the spatial relationship of variables and the relationship of one variable to the
other
(2) Noting down the values of the different variables.
The details are given below:
(1) Understanding the spatial relationship of variables and the relationship of one variable to
the other
Why is a task pursued? (goal): This is to understand the spatial pattern of a variable to gain overall
understanding about the variable. After observing this for more than one variable, the relationship
between both the variables is also understood.
How is a task conducted? (means): By visual observation and comparison between maps showing the
different variables.
What does a task seek to learn about the data? (characteristics): The variation in values across space,
the relationship between different variables.
Where does the task operate? (target data): On the demographic variables for census data.
When is the task performed? (workflow): After the preferred variables are chosen from a filter/or are
displayed
Who is executing the task? (roles): User.
(2) Noting down the values of the different variables.
Why is a task pursued? (goal): This is to note down specific values of certain demographic variables.
How is a task conducted? (means): By visual observation and noting down.
What does a task seek to learn about the data? (characteristics): The exact values of certain
demographic variables.
Where does the task operate? (target data): On the demographic variables for census data.
When is the task performed? (workflow): After the preferred variables are chosen from a filter/or are
displayed
Who is executing the task? (roles): User.
Screenshots
Scheme 1:
Scheme 2:
Summary of key elements
Scheme 1:
Drop downs for ethnicity and switch between income/education
Can view texts as labels
Color variation for ethnicity
Bubble size variation for income/education
Pop up with the value for education/income/percentage of ethnicity
Summary of key elements
Scheme 2:
Drop downs for ethnicity and switch between income/education
Color variation for ethnicity
Bubble size variation for income/education
Pop up with the value for education/income/percentage of ethnicity
Two windows with two pairs of filters
No labels – only pop ups
Plan for Evaluation
The target question you want to answer: which visualization is better for comparison? Which
is better for noting down texts?
The people you would recruit to answer that question: 3 of my colleagues
The kinds of measures you would use to answer your data (e.g., insight depth, use cases,
accuracy) and what these measures would tell you about the core question: Time taken to
copy down values of income and percentage of educated; time taken to understand and
enunciate relationships between different variables and different parts of the country.
The approach you will use to answer that question (e.g., a journaling study, a formal
experiment, etc.): Thinkalouds
How you would instantiate those methods (i.e., what would your participants do?)
What criteria would you use to indicate that your visualization was successful
Synthesis of my Findings
Scheme 1: - good for Task 2
(1) Better for noting down information - faster
(2) Provides a single and clean view of the entire country
(3) Can zoom in to read labels directly in all parts of the country
Scheme 2: - good for Task 1
(1) Good for exploration: comparing different parts of the country
(2) Good for comparison between different variables
(3) Good for creating interesting maps to draw attention