Source(s): https://flowingdata.com/2015/12/15/a-day-in-the-life-of-americans/ & https://flowingdata.com/2017/05/17/american-workday/ (By Nathan Yau)
Organization: Flowing Data
Assignment Instructions: Find an interactive visualization that provides complex analysis and is well done.
Investigate:
Is it interactive - can the user meaningfully change the presentation of the data, or the visual encodings, or otherwise steer what is presented to them? In other words, no infographics or static visualizations for this design critique.
Is it a data visualization - we aren't looking for user interfaces, but rather for visual presentation of data in an interactive format.
It is interesting or complex enough to talk about for a design critique - probably a simple bar chart won't be quite interesting enough to discuss in a presentation. Look for visualizations with interesting visual or interaction design choices or novel approaches to presenting data, or novel data that is challenging to visually represent.
Simulation of 1,000 people's daily routine from ATUS (American Time Use Survey). The author extracts the data from the ATUS Extract Builder.
ATUS stores the data in databases of historical tables. The author took a sample of the timeline and then simulated the user's hourly "status" of primary activity and then analyzed the path of the simulation. Averages per day and by category and gender, allow the organization to develop annual trends every year.
Data Collection & Extraction Process
The data was taken from ATUS Extract builder where the analyst can set the conditions/filters in which they need from the entire database. Once you click "retrieve data" it populated an excel sheet of the tabular data from the Bureau of Labor Statistics. Example below. (https://www.bls.gov/tus/database.htm)
This visualization has minimum interactions. The user can change the pace of the simulation: slow, medium, and fast.
Interface Component Observations:
Color of the dot changes as it transitions to another state
Text goes below the time when a trend is highlighted (major times in the day, ie: getting ready, eating lunch, traveling back from work, etc.)
Long-duration of activities are placed in a circle and traveling is in the middle since it's an "intermediate step" in someone's day.
Percentage by the primary activity is updated as the dots travel through the different categories.
Waking Up
Lunch Hour
Getting off Work
Winding Down
Color Coded by category
Color transitions based on current activity
Creating a “Travelling” category and placing it in the center
Graph shape being circular
Simulation pace is fixed in 3 modes – user choice based on what they are looking for
Key Times/Phases of the day – the subtitle text below the time provides an “analyzed” trend for that period
User Actions are restrictive to simulation speed
Visualizations are 1 view at a time (cannot see the data overlaid/overall)
Lack of filtering to understand data trends more than just activities (gender, age, occupation)
Consolidating all visualization
Put all visualizations in 1 view (dashboard-like)
Apply filters to all visuals at once
Have a "tracking" mode to see the trends like seen in the findings section
While looking at the webpage and interactive visualization. I found the author provided a supplementary simulation, analyzing work days of jobs that are "9-5" structured. Overlapping this data with the average day in the life helped provide more insight to some of the "anomalies" seen in the day schedule.
Blue collar/Union Related jobs tend to work in "spurts" and as a result, they work later in the night
Business professions stick to the 9-5 schedule but often times stay in the "somewhere else" category and then return to workplace. This shows that individuals in that category tend to work more post their typical work day
Production shows the most irregularities which align with the demands of the manufacturing industry