Week 4 Lab: The Observer in the Room
This week
You cannot watch your own day without changing it, so this week we measure the size of the change instead of pretending it away.
Before you run this
- When to stop
- Stop if the record turns into a scorecard, if you notice yourself feeling ashamed of a number rather than curious about it, put the log down for the day.
- What this is not
- This week produces no target and no limit. There is no correct amount of phone use, and anyone who tells you there is has skipped the part where they say how they measured it.
- If this one isn't for you
- Log a different observable instead (how often you open the fridge, how often you check the time) or work from a partner's log with their consent.
What you bring to the lab
The lesson below is how you get there. The lab is where you show it.
- your log covers three days of the same observable
- you can point to the moment the logging changed the behaviour
- you estimate how big that change was, and say how you estimated it
Last week the instrument disturbed your sleep. This week the instrument is you.
Reactivity is the name for it: the act of measuring changes the thing measured. You already met a version of this in week 1, when recording your day made you want to improve it. Now we stop treating that as a nuisance and start treating it as a quantity.
The exercise#
Log every time you pick up your phone, by hand, on paper, for three days.
Not the built-in screen-time figure. By hand. The awkwardness is the instrument: writing a tally mark is slow enough that you will sometimes decide not to pick the phone up rather than record it.
That decision is the data. You are not measuring phone use. You are measuring the gap between your phone use and your phone use while being watched by yourself.
Careful with the story#
The effect is usually taught through the Hawthorne Works: a factory outside Chicago where, between 1924 and 1932, researchers changed the lighting and workers’ output went up, and went up again when they changed it back, apparently because the workers knew they were being studied.
It is a good story and the phrase “Hawthorne effect” is now permanent. It is also shakier than it sounds. When economists went back to the surviving illumination data decades later, the pattern that the story rests on was much fainter than the retelling suggests.
So we are in an odd position: the effect is real and you will observe it in yourself this week, but the famous demonstration of it is not strong evidence. Both things are true and you should be able to hold them at once. Getting that distinction right (between this happens and this study showed it happens) is most of what this course is for.
What comes next#
From week 5 you start intervening. Everything you do from here on will be contaminated by this week’s finding, permanently and unavoidably, and the only available response is to know roughly how large it is.
Write that number down. You will want it in week 10.
Evidence on file
- 23 people not blinded
Asked how many times a day they picked up their phone, people's estimates did not correlate with what their phones had logged. Estimates of total duration held up better. Note the sample size before you carry this anywhere.
Andrews S, Ellis DA, Shaw H, Piwek L, Beyond Self-Report, PLOS ONE 2015;10(10):e0139004- History, not measured
The illumination experiments at Western Electric's Hawthorne Works between 1924 and 1932 gave the effect its name, but later reanalysis of the surviving data found the original evidence for it much weaker than the story it produced.
Levitt SD, List JA, Was There Really a Hawthorne Effect at the Hawthorne Plant?, American Economic Journal: Applied Economics 2011
Teaching team
- Tobias Veck, tutor