We were fortunate to be able to participate in Spurprojects “How is the World Feeling?” datahack. For one week in October 2016, 10,144 people from 104 countries simultaneously participated in the world’s largest survey into how we are feeling. They were asked to rate how they felt at different points across the week and note the activity they were participating in at that time. It is the world’s largest open data source of emotions with 56,992 responses.
About the data
When registering, participants recorded whether they considered themselves to be mostly happy, sad, anxious, peaceful, angry or powerful. Throughout the one week stocktake, participants recorded which one of these emotions they were feeling at different times as well as what they were doing (relaxing, nothing in particular, errands, exercise, work, socialising, commuting or studying). When registering, respondents also noted the their primary form of employment (we will look at this data below). There are a bunch of other fields recorded as well which you can explore via Spurprojects.
It is encouraging to see from Figure 1 that about 64% of the 56,992 responses were positive emotions (happy, peaceful, and powerful). We explore some of these relationships below – and reveal emotions by country and time of day in an upcoming post.
Figure 1: Total emotional responses from the emotional stocktake
There were some wonderful insights to be had and really interesting data sets have emerged.
What stories did we find in the data?
This data is fascinating. On sign up the participants were asked to select an emotion that they felt was most consistent with how they perceived themselves. We were particularly interested in relating how people perceive themselves to how they tracked their emotions across the week – did people who described themselves as mostly angry also consistently log feeling angry during the week and if so what were they doing at that time.
We looked firstly at the breakdown of the different emotions logged by people who perceive themselves as predominately angry. Even though they considered themselves as a mostly angry person, the emotion of anger (red in Figure 2) wasn’t logged that frequently (just a bit more than those who identified as mostly sad or mostly powerful – see figures below).
Figure 2: Those who perceive themselves as mostly angry and the emotions that they feel while doing different things (click on the legend to turn series on and off)
The group of people who identified with sad as their predominant emotive state, reported consistent moments of time in the week where they were experiencing sadness (Figure 3). Commuting seems to be a particularly difficult time (about 40% of logged emotions while commuting was sadness) for people who identify as ‘sad’.
Figure 3: Those who perceive themselves as mostly sad and the emotions that they feel while doing different things (click on the legend to turn series on and off)
In our discussion we pondered when would people feel powerful as an emotion. The data shows that by and large the activity most related to the emotional state of powerfullness is exercise. Figure 4 shows where people who perceive themselves as powerful are most likely to have this emotional state across the week. For further analysis of each of the emotional types and the subsequent reported data of emotions experienced by employment type and by activity have a look at Truii’s Emotional Summaries Dashboard.
Figure 4: Those who perceive themselves as mostly powerful and the emotions that they feel while doing different things (click on the legend to turn series on and off)
Positive vs negative emotions
To simplify the analysis we have lumped all the positive emotional responses (happy, peaceful, powerful) and the negative emotional responses (sad, anxious, angry). We divided the respondents into either mostly positive or mostly negative (as defined by them when they signed up).
Positive and negative emotions and dominant employment type
It is then interesting to see if there is any relationship between employment type and how respondents perceived themselves. Figure 5 shows the percentage of positive responses across the survey from those who perceive themselves as mostly positive or negative. Those who associated mostly negative emotions at the start of the study consistently reported less positive emotions (blue bars are all lower than green bars). Interestingly those that were self employed had the highest overall positive response rate (positive self employed felt positive emotions nearly 80% of the time). At the other end of the spectrum retirees who described their dominant emotion as mostly negative recorded a positive emotion only 44% of the time.
Figure 5 shows that regardless of employment status the amount of positive emotions felt is relatively stable and positive emotions are the predominate type of emotion that someone who perceives themselves as generally positive in their outlook of life experiences across the week. The type of people based on employment type most likely to experience a greater amount of negative emotional states (anxiety, sadness and anger) are the unemployed and looking group.
Considering the group of participants who perceive themselves as generally experiencing negative emotions such as sadness, anger and anxiety (blue bars in Figure 5) we can see that there are differences within the group depending upon employment type (blue bars in Figure 5). The retired, holiday makers feel the least amount of positive emotions across the week followed closely by the unemployed and looking group. The self employed group reported the most amount of positive emotions across the week (66%).
Figure 5: Percent of responses that were positive based on employment type (click on the legend to turn series on and off)
Positive and negative emotions and activity
Taking into consideration what activity people are doing and how that influences their emotional state we can look at Figure 6. This shows people who perceive themselves as positive (green bars) and the influence of activity on their emotional state. In no activity type does the amount of positive emotions fall under 50% however, studying comes close. If you are a positive person and studying this is the activity most likely to induce negative emotions for you according to this data. There is something about ‘doing nothing’ and ‘running errands’ that also makes the world’s more positive types feel negative emotional states.
Similar to Figure 5, Figure 6 shows less overall positive responses for those respondents that identified with a mostly negative emotion at the start of the study. From Figure 6 those respondents who were relaxing, socialising and exercising at the time of responding had higher overall positive responses. Those engaged in study at the time of response logged a positive emotion only half the time. Interesting, those who said they where doing ‘nothing in particular’ generally logged less positive emotions (48% of responses were negative).
Figure 6: Percent of responses that were positive based on activity (click on the legend to turn series on and off)
Type of activity considerably influences the emotions felt by the group of people who identified as having predominately a negative emotional state (blue bars). Figure 6 shows this range of positive compared to negative emotions experienced across the week. Doing nothing does not seem to be a great thing for either group to participate in if they want to stay in a more positive emotional state. For the group who identified as mostly negative in their emotional type found doing nothing to produce negative feelings 61% of the time followed closely by studying at 58% of the time.
A striking picture that this data shares is that there is clearly a difference between doing nothing and relaxing. For all people regardless of whether you perceive yourself as mostly positive or mostly negative the activity of ‘doing nothing’ brings on an increased chance of experiencing a negative emotional state. Whereas, the activity of relaxing is more likely to enable you to feel positive emotions.
Studying is also an activity that regardless of how you perceive your predominant emotional state leads to increased reporting of feelings of negative emotions. The group that considers themselves as mostly positive actually felt the most negative emotional types when studying (Figure 6). Also, it seems that activities such as socialisation, exercising and relaxing are all activities that support a more positive emotional state. I hope that institutions such as Universities and High Schools offer students plenty of opportunities to engage in these types of activities to support their overall health and well being. It seems they are right – Go for a run or talk to a friend!
About the data visualisations
The visualisations in this post are contained within Truii’s Emotional Summaries Dashboard. You can link to the Dashboard, or use the shortcode from each viz to embed or link to specific Viz’s in another web page. You can simply ‘like’ any of the visualisations to post them on your social media feed.
The viz’s are easy to make. All you need to do is create a free Truii account to create and publish your own data visualizations.
Don’t forget to sign up to Truii’s news and posts (form on the right).