Introduction
Utilising the dataset from a previous challenge (week 43), this week we are going to build a bubble likert chart.
Likerts are usually ordered categorical data, and often it does not make sense to convert categories to a numerical score to find averages. Sometimes, likerts can be treated as interval data. https://www.scribbr.com/methodology/likert-scale/. For the purposes of this exercise we will be calculating average scores of a 5-star rating system : https://www.retently.com/blog/customer-satisfaction-score-csat/#:~:text=Reflect%20on%20how%20the%20data,how%20the%20score%20is%20calculated.
In this example we use a bubble plot to show distribution of ordered data, plus overlay the averages. Average scores between 3-4 are shown to be in the “good” (amber) target areas.
Requirements
- Obtain the data from the previous challenge
- Model Data
- With the Deneb custom visual, layer circle marks sized by survey response count, with a layer of tick marks positioned by average score. Ensure the axis is shared.
- (optional) Concatenate another view showing average score per agent as lollipops
- (optional) concatenate another view creating a marginal histogram of number of ratings.
- (optional) include target area of lollipops,
Dataset
This week’s dataset is downloadable from Github
Share
After you finish your workout, share on Twitter using the hashtags #WOW2024 and #PowerBI, and tag @MMarie, @shan_gsd, @KerryKolosko. Also make sure to fill out the Submission Tracker so that we can count you as a participant this week in order to track our participation throughout the year.
Solution
Via Data Stories Gallery
Coming soon.