I’ve been able to spend some of my off-work time exploring data viz and am making my way through several different Tableau programs at the moment. I decided to revamp some of our Student Needs Report data, which was a project I did with two amazing colleagues, Francis and Victoria.
I’ve attached the link to Francis’ SquareSpace so you can see where the journey began, but basically, the Student Needs Report was created during WKU’s 10-Year Strategic Plan process. The University was shaping up for a new direction, and our team felt that student voice should be represented in facilitating that movement.
It consisted of a three-step mission:
- Engage more students in the university’s future-building process
- Collect qualitative and quantitative data on student opinions of needs and wants
- Create a path for the university to better meet student needs
We created student brainstorming teams, discussing ideas and visions for the university. Those ideas were then translated to a survey, which was distributed to the entire school. The online 18-question survey was presented to the Steering Committee, and was a data collection of student voices, representing information about wellness, education, and demographics. Over 3,062+ unique responses were recorded, resulting in the creation of 11 distinct directives for the university.
I was the lead Data Captain, meaning that I worked through all that data (3,000 individuals x 18 questions = 55,116 unique data points) in Excel. Now, just shy of two years later, I wanted to revamp the design of the data and the presentation of the ideas.
The beginning of the PDF report was a demographic survey of the students – where were they coming from, what were they involved in, and how did that impact their experience at WKU?
We had an overwhelming amount of data and had several ways that we could move forward with the story behind it. I often find myself focused on the numbers and forgetting the story.
I wanted to challenge myself to overcome that, or at least begin the journey. I spent time in two intro to Tableau courses, focused on understanding the bare bones of the program. My skills are 1st grade, if that. But the power behind Tableau is mind boggling, and I am glad I took time to explore it.
I found my own weaknesses creeping into my data presentation – the idea that more numbers would automatically mean that things were better. NOT THE CASE. You will see in the deck that I still made that mistake. Overwhelming amounts of data points and not a lot of connection between all of them.
DEMOGRAPHICS
This was the first iteration of the Demographics page of the Student Needs Report. It was really well-designed by Francis. But I wanted to fit more onto one page, and I wanted to expand the content for a larger interaction piece.
This was the piece that I designed on Tableau. I’ve linked the storyboard to Tableau Public, so you can interact with the data, which was the whole point of putting it on Tableau. But I wanted to make the colors more congruent, which I did through a blue/green color scheme.
[STUDENT NEEDS REPORT VS WKU] I added a comparison of the Student Needs Report vs. WKU, looking at the profiles of each respondent and how they relate to each other. Darker the color, more representation, and vice versa.
[GRADE LEVEL] I also create a packed bubbles graph for grade levels, because it was an interesting way to visualize the breakdowns of each of the years, as well as see the “size” of each grade in representation to the other grades.
[STUDENT STATUS] I also wanted to clean up the student status graph. Each demographic is organized into a “set” – four in total. Former residency contains international, Kentuckian, and out of state. Honors Status is Honors, No Comment Honors, and Non-Honors. Location Status is commuter, distance, and live on-campus. Transfer contains transfer. It’s also color coded by set, making it easier to visualize the relationships.
[RACE/ETHNICITY] This was another packed bubbles graph, but more organized. I wanted people to be able to see the overwhelming number of white respondents, without reacting to a percentage point initially. You can roll over each dot to see the percentage point.
[GENDER] Another spot of color. I wanted to keep this rainbow, just because it felt right.
COLONNADE
Next up was breaking down Colonnade, our school’s gen ed program. I was not a fan of the trendline on a ranked graph, nor a big fan of the pie graph. I also wanted a better color scheme.
[COLONNADE SHOULD BE] Here I kept focus again on the color scheme but wanted to depict the story in a bar graph with break out data points. I then filtered that data out by grade level, once again organized by color density and percentages. So most first years think that the process should be more major specific, and most older students think that it should be more flexible.
That makes sense, because when you are trying to graduate, gen eds can sometimes be a huge setback. That shows up in our data.
[BENEFIT OF AVG COLONNADE COURSE] Here I broke out the same data – how people wanted to change Colonnade – but organized it by ranking. So, students that wanted the program to be more major specific were negative about their experience. Students that wanted it to be more flexible were a bit more positive. I then broke it down by grade, color density etc. First years felt the most benefit from the course work, as compared to a less impactful experience recorded by older students.
CLASSROOM
Wanted better viz. Not a fan of the tables of data because if you don’t understand the row/table relationship, its hard to understand the table.
[FEEDBACK IS TAKEN SERIOUSLY] Here I did sort of a diagonal deal and put the Key Takeaways on the side, just to shake things up. I didn’t want the presentation to get overwhelmed by percentages, so I just numbered the Honors vs. non-Honors students. Color density tells the meat of the story for the next graph, comparing grade level versus how students feel as though their voice is being heard.
[PROFESSOR ENGAGEMENT WITH STUDENTS] Color density à bubble graph. Comparing grade level to engagement preferences. Then breaking that engagement down into want the students prefer, assigning numbers without overwhelming the density graph.
REAL WORLD
Same story. Big graphs. Lots of tables. Two decimal points when one would have sufficed.
[PREP FOR REAL WORLD] Once again, sort of a diagonal deal going on here. Prep for real world and the relationship between a simple yes / maybe / no, and grade level. The number of students who feel prepped remains relatively the same as they age, which is promising.
[HONORS ACCESSIBILITY] Most students believe that Honors students have more access to resources, or they are unsure.
[ENGAGEMENT PROGRAM] Pretty positive reviews for 24/7 creative center.
CONCLUSION
This project was fun to work on, and I have more appreciation for data viz. There are beautiful visualizations out there that are actually mind-boggling.
I’ve always struggled with design (one of the perks of being very math-y). But I am a big fan of stretching my mind in this direction, and it was fun to revisit a project that means so much to me.
Excited to explore Python and some more work in R next. 🙂
kylascanlon.com 
Leave a Reply