Assignment 2

Below is a Tableau Story that illustrates various trends in Biker Ridership on Campus. The data was collected by the Central Lane Metropolitan Planning Organization from bike counters located through Lane County. The visualizations below focus on eight bike counters around the University of Oregon campus. Understand biking patterns on campus can lead to infrastructure improvements, better policy and planning decisions, and ultimately improve the biking experience by making it safer, more efficient, and more enjoyable.

2. There are some rather obvious patterns in the data I chose to look at. In the first panel, there is an obvious correlation between various weather variables (I looked at rain and temperature); the nicer the weather the more people rode. My second panel shows that more people ride to and from campus during the week, which is also intuitive, but I would have expected an even more drastic difference. Lastly, I looked at which counters around campus saw the most traffic. While I would have expected the 13th West of Alder counter to see the most traffic, I was surprised by exactly how much more traffic it saw compared to the other counters.

3. I chose different temporal resolutions based on my data. For the panel dealing with weather I felt that a resolution down to the month was sufficient. While weather is always changing, day to day (minute to minute in Eugene~), I wanted to look at larger overall trends, and breaking the data down by month seemed like the best way to do that. For my other panels I chose to looks at days of the week. I wanted to see how ridership changed throughout the week, while also looking for trends based on whether or not school was in session.

4. Using the filter gave fun insights into the data. Although I only included the campus locations in my Story, I viewed all of my panels with and without the campus location filter. The panel that changed the most between the two was the ridership by days of the week. Without the filter there was very little difference in the rider numbers between the weekends and weekdays (in fact, depending on other variables, weekends saw more riders), but as soon as I selected only the campus locations there was an obvious trend of more ridership during the week and less on the weekends.

5. There are many additions that could make this a more meaningful and ultimately more useful data set. While these raw numbers are a good start, they are only the beginning to understanding deeper patterns and trends in ridership. For starters, better attribute data could be collected for the various locations of the counters. Attributes such as slope of the road, if a bike lane is present, is it on a designated “bike route”, speed limit, and proximity to intersection (and whether there is a stop sign or stop light at the intersection). All of these factors are weighed by bikers to various extents and could prove to be useful data to have. Another type of data that would be interesting and rather hard to collect would be some qualitative assessment of who the various riders are. Are they casual riders just out for a stroll, serious commuters who ride the same routes on a daily basis, or a transient riding along 6th Avenue. Collecting this data would basically require constant surveillance with monitors assigning various codes to each rider and is quite unrealistic (at least to my mind), but nonetheless I think it would shed light on various patterns and trends.

6. A major limitation of the CLMPO data set is its completeness. Missing data and gaps in the data can cause interesting artifacts in various types of analysis. Also, my understanding is that these counters are periodically moved around the county, but I couldn’t find information on when a given counter was at a given location and for how long. Lastly, like with any sensor and data collection, accuracy and precision can always be improved. While playing around with various visualizations I came across some numbers that had to be errors or miscounts. On the whole the data seemed pretty reliable, but it can always be improved.