Introduction
The other day, I saw a graph that I thought was really informative. This graph included a 10-year period of daily happiness values as determined using data from Twitter. I wanted to know more about it, so I did my own study. What I learned taught me a few things that I would like to share.
At first, I didn’t know who created the work and the chart in particular. It took me some time to find the project source. Once I found it, I spent a couple of hours looking at the data.
This article is a result of my study, although I didn’t plan on doing this work. The primary reason I am writing this article has to do with something I found in my evaluation. I want to see if others can spot this obvious anomaly and to think about what it means.
Background
I’ve been thinking about happiness for a long time. In my pre-puberty years, I sometimes found myself being a sulker. I’d walk around feeling sorry for myself. I am not sure why I did this, but I know it happened because of things I did like the one I am about to tell you.
One night, I decided to sit outside our house in a short-sleeve shirt in the middle of a very cold Chicago winter. I wanted to see if I could freeze myself in place. Meanwhile, my Mom was laying on the couch reading a newspaper, oblivious to the self-pity game I was playing.
While my core temperature dropped, I realized that I could die during this ridiculous activity and that it was pointless. I think I was about 12 or 13 years old, and I probably had some serious hormonal imbalances going on. All I can say is that kids do stupid things sometimes.
A few months later, I saw myself in a mirror and noticed something I did not like. I looked at the sullen look on my face and I decided at that minute to always be a happy person. It was like I ran an analysis of myself and I couldn’t find any reason to be unhappy. I realized that I was blessed/gifted in so many ways that I had nothing to complain about. From that day forward, I have been a happy person. I flipped the personal happy switch. I am an optimist, and I love life!
The Happiness Score
Figure 1 contains information about the happiness score according to the creators of this project. It is an interesting topic, so I recommend that you take some time to think about how this works.
Figure 1 – The Definition of happiness.
The Happiness Graph
With that as a background, I want to show you the happiness chart that initially caught my attention. It is shown in Figure 2.
Figure 2 – I have selected Saturdays because they seem to be the happiest day of the week!
As beautiful, functional, engaging and informative as this chart is, I wanted to know the answer to two fundamental questions, along with a couple of other questions I had. The primary questions are: (1) What makes people happy, and (2) What makes people unhappy?
To answer these questions, I downloaded the data from the project site. Once I had it, I used this data in Tableau to answer the questions.
Happy and Unhappy Days
I think this data set is a great candidate for Makeover Monday activities. If anyone is actually reading this, please make this recommendation for me to Andy Kriebel and/or Eva Murray.
To answer my questions, I created a quick dashboard. I wanted to accomplish three things:
(1) What are the happiest / unhappiest days of the week? I thought I had a good idea of the answers to this question.
(2) What makes people happy?
(3) What makes people unhappy?
Figure 3 is my quick dashboard. For work like this, I don’t go deep in artistic design, and I try to minimize my time on the project because I work on a lot of different topics. In the dashboard, the basic color scheme is this: green is happy, blue is unhappy (if you are feeling blue, you are feeling sad).
Figure 3 – The happiness dashboard.
(1) What Are The Happiest and Unhappiest Days of the Week?
When I first looked at the time series graphs by day of the week, I could see a couple of things. First, happiness seems to have a bit of a wave pattern to it. There are periods of overall higher happiness followed by declining happiness. I wonder if this could be correlated to economic conditions or some other phenomenon. I would really like to know what causes this type of happiness momentum.
Secondly, the average happiness scores by day of the week across these 10 years confirmed the answers my first question: Saturday is the happiest day and Tuesday is the least happy day.
There are obvious reasons for this, so I think people should contemplate why our happiness varies by day of the week. Do you feel this way during the week? For me, I’m equally happy every day of the week because I love my life, my job, and my family. I am equally happy to be given each day sent my way.
(2) What makes people happy?
To find the answer to my second question (what makes people happy), I plotted the top 10 happiest days (upper right quadrant). When I went back to look at Figure 2, I could only guess at the reason for the happiest days because the pattern was not clear. However, the upper right panel of my dashboard confirmed it. Christmas is the happiest day for everyone – year after year.
I wondered if this is true because of religious reasons, or is there some other reason? I wonder if we consider Christmas as the happiest day of the year because we exchange gifts and we have the potential to receive something that we want.
I certainly hope that our materialism isn’t the reason Christmas is the happiest day of the year! I hope it is because we spend our time together as families and reflect on the miraculous nature of life itself.
For me, I receive more happiness giving gifts than I do in receiving them. However, I will admit that my wife has brought me to tears with some of the gifts she has given me through the years.
(3) What makes people unhappy?
To answer this question, I was tempted to get geeky in Tableau and build a dashboard action that went out to the internet and pulled back the major events that occurred on the days that rate lowest on the happiness score. To create this type of action, I normally do a few test cases in Tableau to verify the concept. What I learned in the process of doing that blew me away.
The bottom 10 happiness scores are shown in the bottom right quadrant of Figure 3. It is easy to see that 80% of these events have happened in the past 2 years. When I saw these dates, I thought it would be easy to identify the cause of the unhappiness. I anticipated that a simple Google search would pull back the events, highlighting the events that lead to the unhappiness. With the events being so recent, I figured that this would be a “piece of cake” to accomplish.
In doing this work, I learned that what I expected to happen didn’t happen. In fact, I had to diligently search to find the events. As shown in Figure 4, the internet sources I had to use to identify the primary unhappiness events were quite varied. It was almost as though these events were being intentionally wiped from existence. I couldn’t believe it.
Figure 4 – The days that represent the bottom 10 happiness scores.
There were eight primary events in the bottom ten happiness scores. Nine of the ten most unhappy days were related to mass shooting events and violence in general. The Las Vegas mass shooting and the Charlottesville violence events appeared twice.
The heinous Las Vegas event, however, is in a class by itself according to its happy score (5.774). This score is significantly below all other events due to the severity of the event and the widespread coverage it received. Overall, the happiness score varies from a maximum of 6.357 to a low of 5.774, which creates a tight range of 0.583. The lowest score is about 90% of the maximum score.
To see how much unhappiness momentum these events created, I did some experimental work using moving average happiness scores. That technique showed a lot of promise in identifying events that produced longer-lasting unhappiness, but I’m not going to bog this article down with those details for one particular reason.
If you took a look at Figure 4, something should have jumped out at you and slapped you upside the head. If you didn’t see it, you were not paying attention. There is an anomaly in that table that floored me when I saw it. Can you spot it? Look at the column titled “Major Event”.
The Anomaly
When I first saw the anomaly, I started to think about what I should call it. Was it a counter-intuitive result, was it an outlier, was it an anomaly, or was it a reflection of the truth? Well, item number 4 is the anomaly – the day that Trump was elected president.
Why is this an anomaly? First, all the other bottom 10 unhappiest events are related to violent acts of mass killing. This makes item #4 a categorical anomaly, even though it might not be an anomaly based solely on the magnitude of the happy score. The confirmation of a President cannot be seen equivalent to a mass killing event, at least not in my mind. In fact, I was very surprised that this had a lower happiness score than the Sandy Hook school shooting. That seemed implausible to me because the Sandy Hook shooting and the Las Vegas shooting were the first and second worst events to ever happen in my life, in my opinion. Those events linger in my mind so much so that I have been thinking of ways to use data science to combat gun violence.
I remember the celebrations that occurred when Barak Obama was initially elected on Tuesday 11/4/2008, or six weeks after this dataset started. There were many joyful people. I wonder if his election was the reason for the first happiness wave that lasted for about 2.5 years through mid-2011.
Second, if Trump were elected President by a fair and impartial vote (without Russian tampering/influences), wouldn’t that day have been celebrated by his contingent? That question helps frame the context of this anomaly. Where were all the happy voters, and why weren’t they celebrating on Twitter like the President is so fond of doing?
Since I am not much of a political person, I haven’t been paying too much attention to the shenanigans being conducted by President Trump and his administration. The continuous barrage of small lies, medium lies, and damn lies have become commonplace. Firings, new hires, arrests, subpoenas, hidden agendas, unethical behavior, and all kinds of investigations seem to be the primary characteristics of Trump’s time in office. Trying to identify something truthful in this administration has become the challenge, and there has not been too much national pride established over the past two years.
With this being said, I wondered how his election ended up as item number 4. All I can surmise is that many people that were as shocked as I was when he was elected. These unhappy people used Twitter to express their disbelief, their anger, and dissatisfaction with the election process. These people voiced their opinion louder than almost any mass casualty event. That result, by itself, should be considered very significant and it possibly could be a predictor of future national unhappiness.
To test that idea, I looked at the time series histories of happiness by day of the week. For every day of the week, the happiness scores have been below average since Trump was elected president. This result indicates that people have spoken, the data has spoken, and I bet Trump would be very happy to blow some smoke at us to tell us how wrong we are about his performance.
I almost never write about politics. However, when I saw item number 4, it caused me to think about what millions of people must have said on Twitter in response to Trump being elected President. The negativity far outweighed the positivity on that election night.
If Robert Mueller is able to complete his investigation, I think we will learn the truth (Figure 5) about the events leading up to that election and why item #4 was created as one of the most unhappy days in the past 10 years.
Figure 5 – In case you do not realize, these headlines are not something to be proud of Mr. President. I wish we could tell you: “You’re Fired!”
Final Thoughts
I instinctively knew in seventh grade that I would never be a politician. This is one reason I rarely write articles that include that topic. The other reason is that I just write the truth as I understand it to be.
The unhappiness that occurred during Trump election night should be considered a significant event. This finding should be remembered for what it really represents. We have learned that money can buy power, which can lead to corruption and an erosion of our political system and national identity. This sequence of events does not make me happy.
Fascinating assessment, especially of the “anomaly!” Rings true – sadly.