Randomized studies of productivity

I believe in the rough benchmarks of ~4 hours of work a day.

I also believe that people are widely scattered around that mark.

I've been doing an experiment where I print out my work day in 5 minute increments, with three tick boxes next to each time-period - "Hard" work (Coding), "Soft" work (reading e-mail), and being distracted. As the day goes along, I tick off the increments as I remember to do so, and I get a pretty interesting picture of my day: - I rarely do "hard" work for more than 45 minutes at once - I do about 1.5 - 5 hours of "hard" work / day, depending on meetings, sleep, drinking the night before, all sorts of things - I'm improving my ability to focus on "hard" work over time.

I definitely recommend this!

Despite what he says, I'm sure it would be possible to design a study that gives an idea of the amount of 'hard' mental work is optimal per day. It might be necessary to do different studies for different tasks, though. I imagine the brains capacity for learning the violin might be at a different rate to learning Python, for example.

Even if the study's conclusion is that "it depends on the person", that would still be a useful result.