Image by Gordon Johnson from Pixabay

Introduction

Our coach was kind enough to give us a day off dashboard week yesterday to prepare for our upcoming Tableau exams. I very much appreciated the chance to recover somewhat from the last few days of intense dashboarding.

The dataset used today was survey data collected on participants’ music taste, frequency of listening, mental health indicators and whether listening to music improved their mood. It was motivated by a desire to understand the therapeutic effects of music.

Data Investigation

The data was clean. I didn’t use Alteryx but I grouped some genres and listening times together in Tableau for simplicity. I excluded Latin and Gospel because they were only a handful of respondents’ favourite genres.

Respondents were asked how much music they listened to per day and whether listening to music improved their mental health conditions. So, the treatment (listening to music), and the outcome (effect on mental health), were reported in very general terms, across an individual’s recent lifetime. A better way of conducting the survey would be to narrow down the questions to a specific event e.g., the last time the respondent listened to music.

In addition, there were a lot of columns compared to the number of rows. So it was necessary to choose a few to focus on to avoid sparse categories.

Nonetheless, this is an interesting topic, and to be honest it was a relief to rule out complex analyses early and be able to start dashboarding. I often get bogged down in the details of the data and feel rushed by the time I finally start putting a dashboard together.

The Dashboard

I was inspired by some of my cohorts’ long-form dashboards and wanted to give that a go. I started with some simple charts summarising the listening habits of the participants, and whether music improved their mood. I then drilled down a little to look at genre and if that changed the outcome. I think the most interesting finding was that exploratory listeners who seek out new artists and genres were more likely to report benefits from listening to music.

My choice of charts was motivated by simplicity and clarity. Really the point of the dashboard was to highlight a few interesting correlations. There was less need for interactivity than in most business style dashboards. I enjoyed the chance to make a long form dashboard that can be read like a story. I broke up sections with blue lines. A useful tip is to use PowerPoint to create lines and save them as image files to use in dashboards like this.

You can see the dashboard here

The Data School
Author: The Data School