Launch, grow, and unlock your career in data

January 3, 2016

Dear Data Two | Week 38: Negative Thoughts

3 comments's supposed to be the most wonderful time of the year, at least that's what the song says. Coincidentally, for week 38 (Dec 21-17), the topic was negative thoughts. Given the level of stress that I felt all week, I'm actually happy this was the topic because it helped me understand what and who was "causing" my negative thoughts.

I tracked the date, what was on my mind, the general topic, where I was, who I was with, the stress level and what I was doing.

I tend to always look at what Giorgia and Stefanie produce each week as a way of looking for inspiration. This week, I really liked how Stefanie created piles for her negative thoughts.

Stefanie included a classification of inward/outward for each negative thought then split them up so you could see how much more there was on one side vs. the other. So I went back to my data and added a similar classification. More on this in a minute.

Here are some of the highlights from the data analysis I performed on my data:

  1. My inward negative thoughts more than doubled my outward negative thoughts. This means that I was keeping my negativity to myself more than expressing it to others. For me, this is good because I have always tended to just say whatever is on my mind. Does this mean I am learning to control myself a bit more?
  2. Most of the negative thoughts were me by myself. Otherwise, Beth was second most involved. This really isn't surprising given it was Christmas week, we went shopping together, and every year I get stressed about how much we spend on Christmas.
  3. 75% of my negative thoughts were low stress...phew!!

So back to the postcard. I thought Stefanie missed an opportunity to think about the "weight" of her negative thoughts. How could I represent my data on what looked like a scale? How could I incorporate more of the metadata? I came up with this Tableau viz:

What I tried to do was make the inward negative thoughts appear "heavier" than the outward negative thoughts by adjusting the y-axis. I was also able to include who I was with (inner circle) and the topic (outer circle).

For my postcard, I was able to make it look even more like a scale and I also was able to include where I was. This turned out to be a good example of how you can do way more by hand than you can with any software.

Explore the entire week in the Tableau story below, including my thought process for how I decided which data to put in the final viz.


  1. This one is cool, Andy. The stack of circles on the scales is definitely eye-catching, which I know has been a hot topic on Twitter lately. :-)

    One question: instead of scaling the Y axis to show the Outward thoughts as being "lighter" than the Inward thoughts, why not instead just scale the Stress measure down so that the size of the Outward circles ends up smaller? The current layout doesn't necessarily suggest that the Outward thoughts are lighter, unless by having them on a scale higher up suggests they're floating. :-) But having the Inward and Outward thoughts share the same Y scale, with the Outward thoughts being smaller, would convey that the Outward thoughts weigh less. At least that works for me. I made this simple adjustment and posted it here:!/vizhome/DearDataTwoWeek38Modified/Week38

    Let me know what you think.

    Happy New Year.


    1. Thanks for the feedback Michael. The idea for the scale was for the total weight of all of the data, not each individual record. The outer circle is specifically scaled so that high is 3x low and 2x medium, so the size of the outer circle is more meaningful. I like your idea though!!

  2. Yeah, I did worry that the encoding of the outer circle got compromised by my redesign. It really is tough sometimes to effectively convey multiple measures inside of a single viz.