Final Project - Effectively Communicating 2014 Mean SAT Scores and Participation Rates in the USA through a Bivariate Map

Every year, thousands of students across the nation prepare for one of the most important test of their lives: the SAT. This standardized test plays a major role for students wanting to attend a university or college of their choice. More students than ever are participating in the SAT exam. In fact, in 2014, 1,672,395 college-bound seniors took the test. The College Board produces various datasets depicting program results for the year's results.

For our Final Project, we were placed in the seat of a GIS Analyst for the U.S. Department of Education National Center for Education Statistics, to create a bivariate map (that is, 1 map with 2 themes over one geographical area) for the Washington Post. We were provided data from the College Board of 2014's mean SAT scores (Critical Reading, Math, and Writing separately) and participation percentages rates by U.S. States, plus District of Columbia. We were also tasked to a write a 3-page minimum paper justifying how we effectively communicated the spatial statistical data in our map based on all the cartographic skills and map design principles we've learned throughout the semester. Below is my analysis.

Thematic Mapping Methods

After combining the mean scores of each SAT subject area, to derive a composite mean score, I chose to use a choropleth thematic mapping method to visualize the data. To represent the participation rates, I used the graduated symbol thematic mapping method. I did not need to standardize the data since it was already standardized. The reason I used graduated symbols for percentage rates and composite scores for choropleth mapping is because when I tried it the other way around, it didn’t do as great of a job at differentiating the data. This has to do with the fact that the SAT composite scores dataset has a larger spread of number possibilities, not just in the dataset I used, but in real life as well. For example, for each subject matter (Math, Reading, and Writing), a student can earn a maximum score of 800, for a grand total of 2400.

Data Classification

I manually classified both thematic datasets to even-like number increments. I placed myself in the shoes of my audience (students and colleges) and quickly realized that the other default classification methods provided by ArcMap would not make it easy or effective for my audience to ascertain the scores and percentage dynamics of the nation’s results. For example, all of the classification methods (such as Natural Breaks, Quantile, Equal Intervals, etc. ) provided ranges like 1309 to 1387, 1388 to 1485, and so on for the mean composite SAT scores. Most universities advertise minimum SAT scores in even increments, like Fordham University at 1200, University of Miami 1270, Duke University 1540, and so on. It only made sense to classify the data as so. Similarly with the percentage rates, the ArcMap classification methods broke the data at odd increments. A good marker I used to class my data was to make sure that it would break at least at exactly 50%. This is an effective and universal way of viewing data.

For both data classification methods, I classed my data using 5 classes. This is an ideal class number for easy visual processing. Too many classes can make it difficult for viewers to comprehend the data; yet too few classes can inaccurately represent and corrupt the information.

Layout and Designs

My map utilizes a functional map layout maximizing the use of space (also known as map real estate) for a well-balanced finish. For example, I used a landscape orientation which fits the shape of the contiguous United States. Also, the map elements (such as legends, scale bar, north arrow, and source data) are non-scattered, but instead uniformly aligned to all other elements, allowing for quick data acquisition and map navigation.

For the choropleth mapping, I chose a sequential color scheme called ‘Brown Light to Dark’. This color ramp works great for people who are color blind. I verified this accessibility design at www.Vischeck.com where it allowed me to simulate what people with color deficits (specifically Deuteranope and Propotanope) would see it as, and it was the same.

Additionally, I purposely chose a dark blue as my title background and graduated symbols because in color psychology, blue is associated with reliability; and personally I think it’s a good color for academia topics as well.

With the use of stylistic effects in Adobe Illustrator, and by applying Gestalt’s design principles, I achieved visual hierarchy in my map by emphasizing thematic symbols with the use of drop shadows. There is also evidence of adequate contrast by using different (yet complimentary) colors (i.e. browns and blues) to distinguish polygon features and datasets’ symbologies from one another. Also, I established a figure-ground relationship by accentuating the United States polygons in contrast to the blue ocean background. Finally, I incorporated visual balance in my map by aligning map elements in a clean, organized, and symmetrical display.

All in all, my map layout and designs do not interrupt the flow of information displayed, nor distract the reader, and are inclusive to people with disabilities.

Results and Conclusions

The bivariate map I created, creatively and spatially visualizes two themes that would otherwise be cumbersome to decipher by just looking at an Excel table. By gleaming over the mapping data, one can quickly ascertain that high participation rates does not necessarily mean high SAT scores by state. In fact, there is a higher correlation of it being the complete opposite. From a quick glance, we learn that the mid-northwestern states received the highest mean SAT composite scores in 2014, yet their participation rate was only at a range of 2 to 10%. On the east coast, the SAT scores are lower, but participation rates are well over 50%. It might be that the east coast high school seniors are more confident and less prepared, and that mid-northwestern high school seniors aren’t confident (or maybe they aren’t being encouraged to attend college by their school counselors or parents), but the ones that are participating, sure are scoring very high!

As you can see, a lot of questions can arise from this two-themed map. With the use of geographic information systems and science (GIS), and with the application of appropriate cartographic skills and design principles, data visualizations like these can change the landscape of a geographic region by deriving new information to find better solutions.