Tracking Everyday Situations
In this post, I explain how I measure and log everyday situations and environment, and share some interesting patterns that I found in my data so far.
How Do Scientists Measure Situations?
There is plenty of research on measurement of situations and environments in psychology and sociology. As a self-tracker, I prefer a self-centric approach (measuring situations as related to me) and found the trait psychology approach to rating situations to be the most suitable for my purposes.
After thorough review of research literature, I ended up using four metrics to describe everyday situations and environments. The first three dimensions (passive-active, negative-positive, and personal-social) are adopted from Magnusson’s taxonomy of situations. The fourth dimension (unfamiliar-familiar) was adopted from Endler, Hunt and Rosenstein:
- Passive-Active: situation was active or passive in terms of my involvement
- Unfavorable-Favorable: situational context was positive or negative for me
- Solo-Social: situation involved communicating or otherwise engaging with other people
- Atypical-Typical: situational context was familiar or unfamiliar to me
Tracking Situations in My Daily Log
I track all four metrics in my rTracker log using simple binary flags. For example, biking to work and having my tire go flat would appear in my log as “Active, Negative, Solo, Atypical”:
I do allow for some ambiguity (e.g., situation can be both personal and social, etc) when logging certain situations, but in analytics I focus mainly on the extremes.
Some Interesting Patterns
This data allows me to look at my life and life metrics in the situational context. For example, the chart below shows probabilistic profile of my situations at any given moment across all days of the week in 2019:
As you can see, my Fridays tend to be the most active, and my Sundays tend to be the most unusual. Sundays are also happen to be most favorable. My Saturdays tend to involve more social interaction than any other days of the week.
I can also speculate that situational context can affect my life indicators. For example, by fitting a simple linear regression, I can see how situations affect my happiness:
# effects of situations on happiness hapsit<-PubHourlyLog %>% select(happiness,SitAct,SitPas,SitPos,SitNeg,SitSoc,SitSol,SitAty,SitTyp) %>% rename(active=SitAct,passive=SitPas,favorable=SitPos,unfavorable=SitNeg,social=SitSoc,solo=SitSol,atypical=SitAty,typical=SitTyp) mod<-lm(happiness~.,data=hapsit) tidy(mod) d <- tidy(mod) %>% mutate( low = estimate - 1.96*std.error, high = estimate + 1.96*std.error ) %>% rename(situation=term,happiness=estimate) ggplot(d, aes(happiness, situation, xmin = low, xmax = high, height = 0)) + geom_point(size=2) + geom_errorbarh()+ theme_minimal()+ ggtitle("Impact of Situational Context on Happiness")
The resulting plot of regression coefficients suggests that I tend to be more happy in more active and favorable situations, but passive situations also have a positive effect. Unfavorable situations, situations that involve social interaction, and atypical situations, on the other hand, have a negative effect on my happiness.
All four metrics are available in my public logs.