Essentialism in Research

I’m currently reading about essentialism, an approach for “doing less but better” in work and life.

One of the most useful ideas from the book, which seems obvious in retrospect but probably not widely done, is to write down criteria for deciding whether to work on a project.

There will be “minimum criteria”, without which it makes no sense to consider the project. But there are also “extreme criteria”, which are high bar criteria that will mean the project will be a really great use of time. The core to essentialism is to only accept those projects that meet all the minimum criteria AND at least a few extreme criteria.

In thinking what my criteria list would be for deciding to work on research projects, here is what I came up with:


minimum criteria

[must meet all criteria]

1. Is this a machine learning project?

2. Is it research: (i) is there a scientific explanation hypothesis (ii) that has not been answered before?

3. Is there something intriguing about the project? Could I see myself giving an elevator pitch to another researcher with enthusiasm?

4. Is it possible to prove or disprove the hypothesis given current knowledge and capabilities?

5. Do I have the resources (data, computational power) to complete the project successfully?

extreme criteria

[must meet at least two criteria, answer can also take the form “yes, if…” or “no, if…”]

1. Does it require me to learn more about a topic that I really want to learn about?

2. Is it a contrarian idea? (Does it contradict prevailing wisdom or methodology?)

3. Could this be a high impact paper? High citation potential? Could significantly change the way things are done?

4. Am I uniquely qualified to work on this idea?

James McInerney