I’m jumping on the bandwagon this year and making a New Year’s #resolution, even though I think they’re kind of silly.
Change isn’t easy. Often when we try to change, we just end up not changing, and instead feel guilty. Not fun. Not productive.
Then again, bad habits can be unproductive too.
So this year, I’m gonna resolve to be a little less perfectionist.
The 80/20 rule says that roughly 80% of the output is created from 20% of the input 📉
In our case, let’s use time worked as our input unit (x-axis in our plot below), and research output as our output unit (y-axis).
Let’s hypothesize that the relationship between time worked and research output follows a power law. Gotta love power distributions, they’re everywhere!
And let’s use writing a paper as the research task.
If we follow the Pareto principle, the relationship between output and input is 80/20: 80% of our paper is generated by 20% of the total work we do. Thus the remaining 80% of the time we spend on our paper only contributes to the final output by 20%.
Depending on the actual shape parameter, we will hit the ideal threshold at some point. After that, large amounts of time spent will only contribute to tiny improvements.
A perfectionist keeps working beyond the ideal threshold towards the perfectionist’s threshold, aaaaaaall the way at the right end of the graph – achieving only very small gains for large amounts of time spent working. That sounds very inefficient!
The idea is to adjust your threshold so it’s closer to the ideal threshold and much earlier than the perfectionist’s threshold. Stop fiddling with the paper after a certain point.
But where’s the ideal threshold? ⌛
The problem is that a certain amount of thoroughness is needed. You can’t send your paper out with a ton of typos, poor structure, or missing references.
Where do we draw the line between being productively detail-oriented and being unproductively perfectionist?
This post sums it up nicely. Basically, healthy perfectionism is motivated by positive reasons (e.g. achieving your goals). But unhealthy perfectionism stems from fear and self-doubt and gets in the way of achievement.
Practically, I interpret this as follows:
Healthy perfectionist activities
- Not skimping on the research itself – fully explore your data with descriptive statistics, spend time doing a good literature review, read the most relevant references, test out various parameters, and document your work well
- Writing iteratively – create an outline, dump your ideas on paper, rework them until they tell a good story, clean up typos and funny sentences
- Paying attention to good design principles in communicating your research
Unhealthy perfectionist activities
- Obsessively trying to generate research that’s 100% perfect – send out the paper and give the community a chance to comment on your work
- Obsessively trying to find every last paper written on a particular topic – take a break and come back to it later
- Agonizing over the perfect sentence – learn to let go; perhaps give your co-authors or reviewers a chance to give you suggestions
- Agonizing over the perfect structure – yeah that’s me; same advice as above
- Agonizing over the perfect graph – yeah, that’s me too; same advice as above
- Getting side-tracked with fun tangents – find the right time and right place to explore those interesting topics, perhaps after your big deadline
Figure out what’s good enough
Otherwise, you risk outputting NOTHING as you fiddle that perfect sentence into oblivion.
I was particularly drawn to this explanation from prof. Matt Might:
The metric academics need to hit is “good enough,” and after that, “better than good enough,” if time permits. Forget that the word perfect exists. Otherwise, one can sink endless amounts of time into a project long after the scientific mission was accomplished. One good-enough paper that got submitted is worth an infinite number of perfect papers that don’t exist.
That’s some sage advice.
Practice what you preach
Since this is a New Year’s Resolution, this post is going out now. I’m not going to edit this post to death (too late, ha!) and obsess over whether my graph above is the right representation.
This is my chance to be a little less perfectionist and let go of being afraid to make (public!) mistakes.
Here’s your chance to comment and tell me where I’m wrong!
🎉 Happy New Year! 🎉
AbnormalData 