Data Visualisation #2
My week of reading
Last week, I decided to track my reading habits. Before I started collecting any data, I decided to review one of the digital apps I use to track my daily activity, workouts, steps, and sleep. I wanted to see if it would give me any ideas for my next visualisation and what data I would collect. I have used the Healthmate (Withings) and Nokia smartwatch for the last year. And I have never logged in to the backend to view all the data available.
On reflection, I really only review this data in the ‘now’, for example, my BPM during a workout, or at a push, I might review my sleep over a few days if I have noticed that my sleeping patterns have been disturbed or I’m overly tired. Looking at all the data available, I did start to wonder why I collect and store such a wealth of information on my activity and sleep habits when I never use it?
In particular, I was drawn to the stat below.
My night-time heart rate on the 31st January was 74bpm, well above my average night-time heart rate of 55bpm. While this stat flags a potential issue, it does not provide context and the cause for this increase. Was it stress-related? Were the cats particularly active that night? Or did I have one too many margaritas? The data only gives us part of the story and I think it is important to always have that in the back of our minds when looking at data. How we fill in the blanks and who fills in the blanks are critical questions to have at the forefront of our minds.
So, with this in mind, I wanted to track a bit more of that context and touch on some data that my Healthmate app can’t track, namely my feelings and emotions. I decided to track my reading activity as I knew my work activity over the next few weeks has a particular research focus.
I tracked my activity in a notebook where I recorded details on the reading content, duration, the device I used, and the predominant feeling I had after the reading activity. In particular, I was interested in seeing whether my feelings and emotions impacted my learning as a whole and whether certain emotions would snowball and flow from one to the other.
After collating my data, it was clear that reading ‘actual’ books is now a thing of the past for me. We don’t even have a bookcase. So I decided to draw my data on a bookcase, with each day represented by a shelf. Monday is the first shelf and so on until Friday.
Here are some reflections of what I learnt about my reading activity:
- I have clear bouts of inspiration and motivation, flowing from one activity to the next. However, this is also reflective when some more negative emotions take hold.
- I think it would have been useful to track my emotions from the start of each activity to the end, to see if there was any change and whether the reading activity itself manifested these feelings or was it something external?
- Going forward, I should probably avoid reading the news during the day, as it potentially impacts the flow of the day and subsequent activities.
- Overall, I have positive feelings following work or university related readings, which suggests that I am engaging well with the material.
- I am completing my university reading either in the early morning or the late evening, which at times contributes to my feelings of tiredness and could impact the quality of my reading and understanding. Perhaps there is scope to break up the day with a lunchtime read when I am a little more refreshed!
- I still read for pleasure! I was delighted that I managed to squeeze in two hours of reading actual books.
- I break up my reading time into fairly manageable chunks of time, with only two reading activities breaking the one hour mark. However, perhaps this reflects my ability to focus on one reading activity at a time or maybe other distractions force me to move to a different task and shorten my reading windows.
- What I didn’t include in this exercise was capturing my focus levels or any distractions. I think this would have provided a more well-rounded picture of my productivity and I imagine I would have captured more instances of random scrolling.
Very inventive visualization. Though the fact you represented everything visually as books caused me a bit of cognitive dissonance – hard to think of these as representations of digitally-mediated activity. I guess that is something significant about data visualizations in general though – they can shape your interpretation, or the way you apprehend a certain phenomenon, through the graphical display itself as much as the data depicted by the display.
An additional question for me is about how you have broken up ’emotions’ into distinctive types of ‘feelings’, e.g. angry, inspired, energized etc. Are these your own categories, or specific ‘scientific’ ways of categorizing emotion? In the branch of learning analytics that focus on student emotional responses to learning experiences (emotion learning analytics), often a taxonomy of ‘Basic Emotions’ from 1970s psychology is used. But Basic Emotions is a highly contested way of measuring human affect – it seems to assume that everyone experiences equivalent emotions in the same ways across cultural contexts, and that these can be measured through things like pulse rate, skin innervation, and facial expression analysis. Do you think such forms of analytics would be able to capture your own feelings this week, as you have recorded them?
I can definitely see why you experienced a bit of cognitive dissonance but perhaps that serves to emphasise my own discomfort at how my reading habits have evolved over the years, from paper to digital form.
My emotions were my own categories and an ‘in the moment’ reflection and perhaps this is something I need to consider going forward. Further thought and research could add a layer of depth to these visualisations over the coming weeks. I think that the analytics that you mention could supplement and support the data that I captured during the week. However, I am unsure whether it could accurately and reliably capture my emotions in their entirety. Given my own experience with tracking my heart rate for exercise, I know that the data has to be taken with a pinch of salt and is not always reliable or accurate. Also, given the irrationality of human emotions can they accurately modelled? Are these systems designed using data sets that are modelled on the average user and therefore are we applying predefined reactions regardless of context, gender, culture etc…? While this data is useful, I think it needs to be considered with a critical eye and I’m not sure it could capture in-depth specifics and a full range of emotions. However, reflecting on this has sent me down a rabbit hole looking at emotion AI and the potential of machine learning in this area, so I may change my mind on this!