Crunching Data into a Dashboard

https://dexterrules.github.io/tcheckr/index.html

How can we create a dashboard which crunches all of the data at one place, with easy to follow insights?

The Problem

For months, I was working on a data project on Facebook Trolling. I wanted to learn more about how are people trolled, which demographic gets affected the most, and which country/location has the most amount of troll incidents. This required me to gather data on-

• Demographics
• Geography
• Type of Trolling

Step 1: Gathering Data

The first step was to gather data through simple survey responses. I created a simple survey to collect responses on the data points mentioned above-

Online Survey Software | Qualtrics Survey Solutions

Step 2: Deploying a second survey

My second survey was more exploratory with Likert scale questions to measure intensity and impact of trolling. This helped to me gain some insights on Male vs Female trolling and the relevant impact.

Online Survey | Built with Qualtrics Experience Management™

Step 3: Analyzing responses: RStudio

After I collected data for about a month or so, I was able to get 53 responses for my first survey and 58 responses for my second survey. Now, it was time to crunch some data and dig some insights. I used RStudio to conduct some simple tests like t.test and generate some graphs like Parallel coordinate plots. I was seeking some answers to questions like-

• Which gender gets more amount of trolling?
• Which gender gets more amount of community troll vs individual troll?
• Which gender uses social media more than the other?

Here is a summary of results from both surveys-

Tcheckr

This is what it looks like-

Parallel plot for average total time spent between Male-Female

Another Parallel plot to show correlation between average total number of times Fb was opened, and total time spent.

Step 4: Going a step further: Tableau

While Rstudio is great in terms of crunching data, formulating graphs and finding out results, it has limitations in terms of data sharing and how data is visualized. Also, R is a very user-unfriendly language, and even for generating simple bar graphs, it requires significant amount of code prep. The error messages are very cryptic and requires lot of guesswork to fix.

I decided to link my data to Tableau through a simple Qualtrics API. The integration works like a charm and you can extract live data from survey responses. Once I integrated my Qualtrics with Tableau, I started creating demographical and geographical charts.

Creating a Workbook on Tableau for charts

Final step: Creating a dashboard

The final step to integrate everything into a dashboard. I chose to use Tabler UI kit to quickly create a dashboard. Here is a simple IA map for the entire dashboard.

Then, I started creating Tableau embeds to be integrated into the Dashboard.

Want to experience the dashboard yourself?

TCheckr

Final thoughts

I learnt a lot of things while working on this project, such as working with Data Visualization, creating Tableau workbooks using Qualtrics API, conducting surveys and ultimately putting them together on a Dashboard. This entire project required a lot of hard skills, from R to front-end development, but more than that, it required UX thinking and how information is being presented. I will keep working on this project, integrating different types of charts and graphs, and maybe creating an API which fetches live data from CSV files as they get updated. If you are interested in working on this project, let me know.

Liked this case study? Visit my portfolio http://sanchitsoni.co/