Communication Design Studio — Project 3
Visualizing Patterns
We’ve all seen data visualizations — some of them effective; many of them simply pleasing graphics. We’re going to dive into the process of conceiving and crafting visuals that communicate information in ways that are useful, usable, and desirable. To do so, we need to create or find data to use for the project.
Week 3–9.17.15 — Week 5–10.1.15
Gathering Data for the project…A few potential ideas:
Idea #1
how does caloric intake vary with sleep hours and exercise and is there any correlation between this and workload vs. social activities?
some data points to consider over a 20 day period for this idea are:
- measuring hours of sleep/24 hours
- caloric intake between 12am-12pm and at what times
- caloric intake between 12pm-12am and at what times
- hours of exercise per day and what type of exercise
- overall mood/happiness measure (not sure how I would measure this though)
Idea#2
coastal vulnerability to sea level rise looking at:
- measured monthly water levels from a 10 year range
- coastal change analysis
- tides changes over a 10 year range
- habitat changes over a 10 year range (look at bird migrations or populations possibly)
- temperature maps over that same 10 year range
- sea grass data over that same 10 year range
- fish/marine habitat protection areas that have been set up over that 10 year range — not sure about this one yet
- coastal flooding data
http://catalog.data.gov/dataset?groups=ocean9585&tags=seagrass
Idea #3
Bicycling and walking rates by country as compared to obesity rates for commuter — either compared by country or by region of USA
http://www.ncbi.nlm.nih.gov/pubmed/19164816
https://www.census.gov/prod/2014pubs/acs-25.pdf
factors to compare:
- male vs. female
- country
- exercise
- age
- obesity prevalence
- region
Week 6–10.6.15–10.8.15
After discussing the above ideas with Stacie I decided that the best area with the most interesting data for me to look at would be idea #2 looking at how coastal vulnerability is affected by sea level rise. I had a lot of data sources for different things to compare for this so I needed to take some time to sift through all of the data and figure out which points would be the best for comparison, as well as if I needed to limit my data points to one area.
In trying to figure out the best points for comparison I decided to enlist some more experienced knowledge in the area from my brother-in-law who is a coastal engineer working out of Boston on projects in NYC and New England. He suggested that I would find the best data comparisons by focusing on a few key areas as is detailed on my slack channel:
For project #3 I am planning to look at the coastal vulnerability to sea level rise in the Northeast US. The types of data that I’m planning to compare are monthly averages over the last 10 years. I will be looking at water levels, tides and currents, windspeeds, migratory bird pattern changes, precipitation averages, and temperature averages, if I can find data on dune erosion I will also include that. Most of my data is coming from the us geological survey USGS, national oceanic and atmospheric association NOAA , and the National data buoy center NDBC. This information is useful to many groups from environmentalists concerned with coastal ecosystems to residents living on the coast who are concerned with the safety of their homes and businesses in regards to land erosion over time as well as preparations for the effects of potential superstorms.
Week 6–10.13.15–10.15.15
This week I realized that government websites are actually very hard to extract data from, so I’ve decided to switch course a bit and instead focus my data on weather for Boston for each month for the past 10 years. So far I have all of the average monthly high temperatures, average monthly low temperatures, total monthly precipitation, total monthly snowfall, average ocean temperatures for each month, and average water levels for each month. I want to look at the data to see how the air and water temperatures and sea level rise (or drop) correlate to precipitation and snow fall in regards to increased storm intensity in the region. I want to explore the idea that larger storms and more snowfall in the region are actually a sign proving the existence of global warming not refuting it. My new sources that I can actually read the data from are listed below:
https://www.nodc.noaa.gov/dsdt/cwtg/all_meanT.html
Week 7–10.20.15–10.22.15
Data Gathering/refinement
Week 8–10.27.15–10.29.15
Data Gathering/refinement
Week 9–11.03.15–11.05.15
We worked a bit on gathering more data and refining what we have. I realized that it would be very helpful to look into the historical records of hurricanes and tropical storms/severe weather in Boston to supplement the other weather data that I have. I am also trying to find some flooding data if I can get it for the correct years/ timescale that I am working with.
Here is the organized data set that I’m working with:
Week 10–11.10.15–11.12.15
Beginning to visualize categories of data:
Week 11–11.17.15
More explorations on the elements of visual form:
Some Initial explorations of data visualization using Excel:
Still working through the idea of what the scale of comparison will be.
Week 11–11.19.15
Things to consider:
- Macro vs. Micro scale
- How to compare different variables
- starting point that you move into the visualization through
A first attempt at using shapes to symbolize my data proved to be a very bad idea. This made me think I should probably be avoiding literal interpretations of data and find more tailored, succinct, and creative ways to visualize things.
In the above graph I was starting to look at the data from a 3D perspective to make sense of it, but still struggling with how to overlay the scales.
Two different charts looking at water temperature averages in Tableau. This program helped to look at some of the data in new ways and how it could potentially be represented through different variables.
In the above chart I began to try and map out an interactive flow. This began to frustrate me because there were so many different combinations and paths that could see to take with my data. I then started to just jump straight to the end product because looking at my data piece by piece and building it up just didn’t seem to be working for me. The way that made the most sense to overlay my initial breakdowns of the categories was through envisioning them in a more literal way in an actual landscape diagram of a beach (because this is normal terrain in the Boston area).
precipitation= sound
water temp = gradient
air temp = dot density
water level = position
snowfall = shape/volume
After beginning to wrap my mind around how to best visualize this data through this very literal interpretation it finally occurred to me that I have actually done this type of weather data visualization before by using Ecotect software and also mapping sun path diagrams, and wind roses etc.
These diagrams overlay several types of information: date, month, position, time, and angles to measure the suns path by using it’s altitude and azimuth relative to the geographic location that is being analyzed. This isn’t exactly the best way to visualize the data that I’m working on currently, but thinking of the data in a more architectural context such as this made me realize that I already know how to read and understand these charts and that multiple layers of data such as this can be overlaid to create an appropriate comparison.
This revelation also made me start to think of the possibility that my data visualization doesn’t have to be linear like a timeline just because I’m talking about time. It could also be seen in a more radial diagram that cycles around like the months do. I then began to explore this possibility in addition to the more literal option that I had been considering:
Week 12–11.24.15
I began working on a test chart of one year to see where my interactions could occur within the data visualization and to make sure that my variables and scales were mapping appropriately.
My first exploration in mapping the variables to a coordinate plane system on illustrator did not work out how I had planned. There was a lot of data even in the abbreviated single year that I was trying to map and it was hard to accurately note amounts… this was definitely not the look I was going for.
During our work session this week I further reviewed my data and realized that I was still having issues noting the different types of information that I collected on the same chart.
When I took a step back to consider the main point that I wanted to get across in the end (that increased temperatures of the water equate to increased snowfall and precipitation) I realized that the variable of water level was probably not as important a point of comparison to illustrate my point. So, I decided to cut out this section of data from the data set that I’m working with.
Checklist for next week:
- Data for visualization — see above chart, final comparison will be of average monthly water temperature, air temperature, precipitation, and snowfall for the city of Boston for the past 15 years.
- The coordinate system for the visualization — originally I was thinking this would be linear because of the years, but after the explorations into precedent inspirations noted above, I have decided to use a radial coordinate system based off of the monthly cycle of the data.
- Scales for each piece of data: both water and air temperatures will be on a 5 degree scale ranging from 0–80 degrees F, while precipitation will be mapped to sound decibel ranges (the louder the sound, the greater the precipitation), and snowfall will be mapped to a 5 inch increment scale.
- Sketches for how data types map to visual/aural/temporal variables — see sketches above: precipitation will be represented through sound with varying decibels to signify amounts, air temperature will be represented by gradient, water temperature will be represented by position, and snowfall will be represented by scale.
- Rough plan for the walk through of the piece: standard radial chart to display monthly information with a year ledger on the side bar. The user can select one or multiple years to see/compare monthly information per each year in an overlay on the graph.
Week 13–12.01.15
A more developed digital iteration of the flows for the interaction with the data visualization…. an evolution of visual variables…
Things I have realized:
- color can be a bit of a monster… and confuse things quite a bit
- Also, just because a variable maps well to what it is representing does not mean that it will map well to the coordinate
- sometimes you need to go back to basic black and white to understand your concepts and slowly reintroduce color as a variable and not a backdrop that will distract.
Week 14–12.08.15
Evolution of weather diagram and visual variables:
Thoughts: black is too dark, blue is too purple, clouds are nice way to show air
Thoughts: grey is a difficult background color, be careful not to go too light. It’s also nice to have a lighter contrast rather than a darker one.
The evolution of the weather diagram, clicking water temperatures on and off and looking at all variables for years 2000, 2005, 2008, 2010, and 2015:
Things I realized I need to do after this exploration:
- move the lines to the top of the diagram
- lower the opacity of the air
- unfeather the air opacity
- turn ring sections on one at a time, inner vs. outer
- split the key in half so that the scale numbers do not intersect it
- turn on the key words only at rollover
- the blue is still too purple
- update the explanation and the font
- center the lines on the data separations so that you can tell which is which
- switch the canvas size to pixels for a screen
Week 14–12.10.15
First full presentation of data visualization:
Overall feedback I received and things to reconsider before the final presentation next week:
- numbers and words are still not legible
- Grid lines are still to prominent
- introduce data by individual quadrant instead of by full years. Build up the data.
- make sure there’s not too much text.
- differentiate between the years of the blue sections
- talk more slowly and allow for more viewer exploration
- have a slower introduction into the data instead of everything all at once
- zoom in to number scales so that the audience can read them more
- maybe align month text labels to circle more
Week 15–12.16.15
Final presentation of data visualization:
Reflection:
Did you learn any new skills and/or approaches from doing this project that you think you may find useful in the future?
I think I learned a lot about how to plan for data classification and mapping visual variables. I found it pretty straightforward to map different individual variables to different data types and it was interesting to explore various ways of doing this. It got much harder when trying to overlay more than one data category on the same coordinate system though, especially when working with four or five different categories and corresponding variables. Coordinate systems that were perfect for some, didn’t work well at all for others. Sorting through all of this and learning how to adapt my visual variable type and pare down my data sets to work together more cohesively was an interesting challenge. I definitely learned to think about how things will work together much earlier in the process, so that later on I don’t run into these misfit issues. I also learned about data gathering and how to make sure you have a very distinct question and goal of what you want to show before you start deciding on data categories to compare, so that you can choose the categories that will really be the best to make your point. It’s more interesting when things have a point. I also learned about visual representations and how the same variable shouldn’t represent different data categories, but variations of it can and will help the different data types to be compared in a more easily understandable way.
How do you feel about the outcome of this project? Are there things you would change? If so, what are they, what would you do, and why are they important?
I’m ok with the outcome of the project. I was really happy with my process and how I came to the coordinate structure system that I ended up using because it was based off of a strong precedent that relates and that I’ve worked with before. I do wish that I had explored a few more options of data types for the project though. When I initially thought of data I kind of relegated myself to something with hard numbers. After seeing some of the much more creative and interesting topics that my classmates were able to pull data about I really wish I had chosen to take data from subject that I found more interesting. I think that it would have helped me to be more excited about the visual form of the project in the end and I would have had a lot more inspiration to work with, even if it wasn’t a topic with serious amounts of stock numbers of data, like weather is. I also think I originally approached the project more from a “just find data” mentality rather than questioning mentality. I think it might have worked better if I had started with “find a question” rather than “find data” because when it got down to comparisons much later in the project I felt like I really struggled with justifying why this data mattered or should be interesting or what the question it really answered was? I also think that I went overkill on data gathering. It was nice to have chosen a topic with many advanced archives and resource availability, but I ended up wasting a lot of time finding all kinds of different data points that in the end I really didn’t need and didn’t relate to each other in the data visualization as well as I would have liked them too.
How do you feel about the movement through this project? Were there sessions that did/didn’t resonate with you? Any parting words as we bring the semester to a close?
I feel like the movement through the project worked well. I’m glad we started earlier with data gathering because I wasn’t stressed out by finding my data and I had plenty of time to build and refine the data set. One drawback to this was that I feel like starting the data at the same time as the earlier projects made me just pick the first thing that I could find data on easily, so that I could focus on projects 1 and 2 and not worry about the data collection. In the end of the data project I wish I had been more focused on only choosing the data topic for one of these earlier sessions. I think that maybe I would have then been able to dig a little deeper and find a topic that was more interesting to me and had a bit more staying power since we looked at this project over the longest period of time through the semester. Overall everything was really great this semester and I feel like I’ve learned a lot. I’m also finding myself to be more conscientious of the choices that I make when representing information in all of my projects for other classes as well, which has been a really nice outcome for me.
