A few weeks ago, Gazelle International completed its three year assessment report for CLICK, our system for virtual exchange. We compiled and interpreted survey data for 334 students, 29 teachers, and 14 CLICK projects. As the Data Analyst intern, my task was to produce clear charts from the CLICK survey results. With fourteen questions, closed and open responses and a plethora of student variables -- “pre” and “post” completion of a CLICK project, US vs. Abroad students, Year 1 vs. Year 2 vs. Year 3 -- how could I show the full story of our program’s results? How can I visualize our CLICK survey results? Challenge “on”!
Take one survey question. It asked students whether they believed the CLICK experience positioned them for success in the global workforce, a key goal of the program. Each year saw steady improvement as we re-phrased the question for years 2 and 3 and redesigned training materials for our teachers. The standard bar chart in Excel, sadly, did little to make it clear that we even had a “story” behind the graph, much less one that showed improvement.
How can we convey our story better? Because it included all of our key variables, the bar chart was much too busy and dense. See what you think of my new graph. Does it communicate our important results? and in a more visually appealing and accessible manner?
*note: sample size varies from pre to post; post surveys have fewer respondents.
To simplify the bar chart, I combined “Strongly Agree/Disagree” and “Agree/Disagree” into a broader “Agree” and “Disagree” categories. I subtracted the “Pre” results from the “Post” results to obtain a percentage difference. Negative results indicate a lower percentage of students in the “Post” survey answered, say, “Agree” than in the “Pre” survey, and vice versa for positive results. The new chart captures the dramatic improvement in student sentiment around workforce readiness over the past couple of years.
Another goal of the surveys was to get a sense of our students’ attitude towards cross-cultural interaction and global awareness. Take, for example, the question that asked about the importance of learning and communicating in another language. The basic bar chart, again, was a weak visual to reveal the differences I saw between US and Abroad students’ responses.
US students and Abroad students generally seemed to agree but their level of agreement differed quite significantly. Turning to Stephanie Evergreen’s blog on data visualization for help, I created a back-to-back bar chart that conveyed the magnitude of difference between US and Abroad students.
This back-to-back bar chart visually and quickly reveals a stark difference between US and Abroad Students. US Students expressed less of a need to learn another language to thrive in a global workforce. Since English has been the default language of communication in all but two of our CLICK classrooms, it was no surprise that the students abroad showed a greater recognition of the importance of learning other languages. They had to work harder to use English, often their second or third language.
Effective data visualization makes it possible for us to identify trends and patterns, highlight successes of teachers and students, and target specific routes for improving our training or suggesting gaps for teachers to address. For the overall program, these high-impact visuals are critical in providing concrete evidence of CLICK successes so far, and how the data-led tweaks and refinements we’ve made over the years have borne fruit.
Are you already involved with virtual exchange? Let us know if we could support your evaluation and assessment efforts. Or are you interested in getting started with virtual exchange at the institutional or individual level? Register for a workshop or contact us today!
August 14, 2020
Post by @jsphwei