Now that I am halfway through the semester, I am still working to fulfill the initial data visualization priorities we determined in the beginning of the semester. Our first priority was to create a map displaying the distribution of HPS-run schools and include achievement data. Thus far, I have been able to produce a two-layered map with data points of schools that are color-coded based on the CMT proficiency of that school. There are also outlines of the four school zones, so viewers can easily visualize which schools are in which zones. Now that I have figured out how to create the legend, I can move forward with making a map for the CAPT proficiency for HPS-run high schools. In mapping the high schools, I have run across the issue of multiple high school’s at the same address, so when you click on the data point, you only see one school (ex. Engineering, Law & Gov, and Nursing are all coded with the same address). Once this issue is resolved, the map should be fairly easy to complete. Finally, since we now have a workable template for a searchable feature from Derek Eder, I should be able to now include this.
Update by Jack on March 18th: To illustrate what Marissa wrote above, I have pasted the embed code from her post last week
After listening to Alvin Chang discuss his methods of data visualization, I found his suggestion of making sure you understand what your question is very helpful. When I initially began creating the map, I understood what it showed, but I wasn’t entirely clear on the particular story we wanted to show. After speaking with Rob about this, we determined that the main question for the maps is whether one zone has better choice options than another. I feel more confident moving forward now that I have this question in mind.
The second priority involves interactive line charts that display achievement data over the span of five or six years, enabling viewers to see each school’s progression. The first priority for charts is to create one for cohorts. At the moment, Achieve Hartford has static line graphs that shows cohort progression for each school, but with the help of interactive visualization tools, Rob and I would like to have one line chart to compare different schools. Since one line chart with the 30+ schools on it would be way to cluttered, we determined it would be better to have a checkbox feature that enables the viewer to select specific school’s, or two drop down boxes for viewers to select two specific schools. I would like to have the Hartford average permanent on all graphs, so viewers could compare two schools, but also specific schools to Hartford overall. Once I am able to create this chart, it should be relatively easy to create a second interactive chart that better displays the achievement data across time from the maps.
Marissa,
Thanks for sharing your progress memo, and you have successfully created a 2-layer interactive map to display both points and polygons. Here are my responses to issues you raised above and thoughts about next steps:
As you noted, we have a better template now to display 2-layer maps, and I recently updated the code comments to make it more user-friendly. See demo at http://jackdougherty.github.io/FusionTable-Map-2-layers/ and follow link to the code template.
As you discovered, when two or more schools share the same street address (such as the Hartford Public High School academies), they appear on top of one another in a point map, because the geocoded locations are identical. This is one limitation of mapped data. While it’s possible to modify the address (or XY coordinates) to give points more distinct appearances, that would create fictitious data and still would not be visible to users who are zoomed out.
A different approach is to visualize data in two ways, with a combined map and table. See Derek Eder’s GFT to HMTL table (which allows users to sort by column) at http://derekeder.com/fusion-tables-to-html-table/
One easy solution is to paste the HTML table code at the bottom of your current map page, where they would co-exist side by side. An ideal solution would be to make the map and table interactive (meaning that when a user clicks on an address, the pop-up appears on the map, which is what happens in http://smartchoices.trincoll.edu). But we don’t have an easy way to code this . . . yet.
Good reflections on the story behind the data visualization, thanks to Alvin’s presentation.
As for your 2nd visualization concept about an interactive line chart (with the ability to turn on/off lines), I’ll paste here what I wrote in my prior email to you:
hope that helps — a bit
Jack