Introduction to Excel for Data Visualization

online course for a small social enterprise

Purpose & Role

This course aims to teach international development and monitoring and evaluation professionals from all backgrounds how to create both static and dynamic data visualizations in Excel. Beyond Excel skills, this course also explores key data visualization principles, case studies, and two custom scenarios. As the lead instructional designer for this course, I researched content, ideated storyboards, screencast and edited demos using Premiere Pro, Audition, and AfterEffects, designed and developed all course interactions using Articulate Storyline, completed quality assurance testing and editing, inserted custom javascript triggers to track learner engagement, and analyzed the data using Keen.io. Other team members illustrated the backgrounds and provided editing and quality assurance support.

Process

The project began with a focus on R&D to identify in-demand Excel skills and how this course could differentiate itself from other courses available online. After identifying the ability to perform basic data cleaning, complete key formulas such as index-match and vlookup, and create interactive visualizations using PivotTables and slicers as learning outcomes, I decided upon brevity and interactivity as the missing components of other courses on the market.

A screencast that I recorded in Articulate Replay, edited in Premiere and Audition, and animated in AfterEffects.

As such, I designed each screencast on data cleaning, formulas, or interactive visualizations to cover a single topic within a few minutes. For interactivity, I built engaging presentations and case studies. I also created two custom scenarios in which the learner chooses the appropriate Excel function or visualization technique for a given set of criteria. By actively adjust elements of the chart and learning from mistakes made, the learner demonstrates understanding of the visualization techniques.

In this interaction, the learner clicks parts of the chart that should be improved. In this screenshot, the learner is hovering over the legend.
After clicking a correct part of the chart, the chart dynamically changes to reflect the change, which in this case involves integrating the legend with the bars.
In addition to the chart dynamically changing, the learner recieves an explanation for why their action was correct.
The exercise is complete when all three numbered circles are filled with green checkmarks.

After building the course, I worked with a team of developers and other instructional designers to insert custom javascript triggers to record student activity through our custom Learning Records Store. As this was the first course to use the system, the process involved a lot of deliberation regarding what we should track through the system and how. We ended up deciding to track completion, navigation, and interaction.

Within one month of launching the course, over two hundred people enrolled in the course. One of the most significant insights was that only 25% of learners got past the first module. However, once past the first module, 65% completed the second module and an even larger portion went on to complete the entire course. The second, third, and fourth modules, while much shorter, were also much more interactive. The chart interaction shown above, for example, is in the fourth module. After digging deeper into the data to see exactly which slides learners failed to complete in the first module, I designed a series of interactive exercises to preempt such slides in order to increase overall completion of module one.

The example below shows a section of module one that, in the first iteration, was solely screencast videos. To make the content more engaging, I created exercises like the one shown below to help learners review content.

After watching a screencast video about the index match formula, learners apply this knowledge to an exercise that immediately follows the video. In this screenshot, the learner is hovering over the first field.
If the learner drags the field onto an incorrect part of the function, the field becomes red and they receive feedback to try again.
If the learner drags the field onto the correct part of the function, the field becomes green and they receive "well done" feedback.
After the learner drags all fields onto the correct part of the function, they receive feedback to click next to continue onto the next screencast.

As of December 2016, over 600 people have enrolled in the course. We continue to analyze our Learning Records Store data to see how improvements could be made. We also ask learners to complete a qualitative course evaluation. Between these evaluations and the data we receive, this course is in a constant state of iteration in order to deliver the best possible learning experience for students.

Below is a demo video of the course in its initial iteration, before I created additional interactions to add to module 1.