Visualizing time spent on my research 2018 to 2021

Intro

With the application for the Irish Research Council Postgraduate Scholarship, I submitted a Gantt chart about how I intended to spend my time. It was an idealistic plan, but so are most project proposals. Now that the thesis is submitted, I wanted to create an actual visualization of the time spend on each part of the project.

Data Source

I use Project Hamster, a convenient desktop time tracking software. You simply choose activity, start and end time and the software keeps track of how you spend you time. Records can be exported as tables. Activities are connected to projects, so for example my research activities would be linked to the "Uni" project like so : teaching@Uni, Case Study 1@Uni

Data Cleaning

  1. Since the time tracker software exports all entries for a given time period, and I occasionally track time for non-project related activities (such as job applications), the exported data had to be filtered to only show activities relating to the Uni project.
  2. Some items were grouped to reduce the range of entries. Many different admin items were grouped into a general "Admin" group. Different classes taught were grouped into "Teaching".
  3. Spelling and capitalization had to be checked and corrected.

Data Processing

  1. The time tracking software tracks activities down to the second. This is nice, but for a visualization I was looking to group data by month. The Datetime elements for starttime were reduced to a MM/YYYY format.
  2. With the reduced Datetime elements, it was possible to convert the data into a pivot table. The columns are the Datetime elements, the row is the individual activities. The field values are the total time in minutes spend on each activity.

This table looked like the example below:

Start |Activity 1|Activity 2|Activity 3
01/18|60 min |0 min |0 min
02/18|0 min |30 min |0 min
03|18|20 min |60 min |0 min

Visualization

I used RAWGraphs to create a streamgraph of the table. This gave me the option to quickly try out different settings to find a good way to visualize this big data set.

And here it is:

viz
Click here for high-res file

So what are we looking at?

  • The bigger the stream, the more minutes spend that month
  • Literature review and methodology took up much of the first half of the project, longer than anticipated
  • Conferences (both attending and organizing) took up time throughout the project
  • The three-and-a-half case studies (one eventually became a sort of epilogue) were written as first drafts within half a year, quicker than expected.
  • You can see the waves of teaching as the semesters begin and end.
  • The last half year was filled with revision, intro and conclusion only pop up shortly.
  • Augusts were my least productive months, no regrets.

Limitations

Obviously the data from time tracking had to be reduced to allow for a reasonable visualization of three years' work. This meant to group data and reduce complexity. I can also not guarantee that every single entry over these three years is correct in both duration and category.

Conclusion

This project for me was an honest and liberating way of putting three year's work into one graph. I hope it can both be an interesting visualization example as well as an encouragement to early career researchers.