Trying harder works, sometimes

post by Kaveh-Sedghi · 2022-02-03T05:04:16.903Z · LW · GW · 1 comments

Contents

  A Story of Habit Change
  Reflections
None
1 comment

It seems to be commonly accepted that pressing the "Try Harder" button does not work. [LW · GW]While I agree with this in general, I would argue that there are edge cases where trying harder is a valid option.

To illustrate this, is a story: of science, and how I used it to change a particularly pernicious habit.

A Story of Habit Change

I used to have the bad habit of lying on my bed and using my computer, rather than utilizing my desk. This was bad for several reasons:

1. Reduced productivity, as doing work on paper becomes practically impossible.

2. Net-negative impact on long-term health due to bad posture.

Naturally, I wanted to get rid of this habit, and I turned to one of the most effective tools I know of - the Scientific Method. I hypothesized that the reason I was lying on my bed rather than using my desk was due to comfort - the bed was the more appealing option.

I resolved, then, to reduce the comfort appeal of my bed by relocating my blanket, along with my pillows, to an unused room. I reasoned that this would make the friction of attempting to get back the comfort high, helping me break the habit.

For a time, this strategy worked. My productivity increased, and my posture became qualitatively better. But after about 12 days, my strategy failed.

The strategy was sound, and the hypothesis seemed to correspond to observations (while being intuitively correct as well). Yet, it did not work in the long run.

As a rationalist, and as a knower, I should have recognized that perhaps my hypothesis was flawed, and tried a different approach. Instead, I opted to try the same strategy again - I repeated the experiment. I did not attempt to change anything, and simply did what I had done earlier. But this time, I tried harder, and resolved to be more committed.

It has been 3 months, and so far, this seems to be working.

Reflections

My experience reminded me of something I had read, in Philosophy of Science: A Very Short Introduction. Okasha, the author, points out that there are problems associated with Popper's demarcation of science from pseudo-science. Popper criticized ideas such as Freudian psychoanalysis because he thought that they could explain any observation - essentially, he thought that they couldn't be falsified.

However, there are instances where sticking to a theory and explaining away failures and new observations is helpful. The example given in the book is of the apparent failure of Newton's theory of gravitation to predict the orbit of Uranus. The observations contradicted the predictions, yet two scientists working independently stuck with the theory and postulated that there was another planet that was affected the orbit of Uranus. Ultimately, they turned out to be right - the planet was Neptune.

Relating back to my example, I stuck with my strategy, and just tried harder; it payed dividends.

I could just as easily have failed by doing this, however, so the interesting question becomes: when do you stick with a theory, even though reality appears to contradict it, and when do you abandon it in favour of a new one?

1 comments

Comments sorted by top scores.

comment by Pattern · 2022-02-08T19:11:39.884Z · LW(p) · GW(p)
It seems to be commonly accepted that pressing the "Try Harder" button does not work.
[From that linked post, link broke when text was copied:]
But just noticing yourself pushing the Try Harder button isn’t enough - you need to do something stronger to change this. You need to find strategies that actually work.

There's a distinction that could be made, that your approach wasn't just 'try harder to work at the desk'. It was 'change the environment and try harder'. (also see the next section below)


As a rationalist, and as a knower, I should have recognized that perhaps my hypothesis was flawed, and tried a different approach. Instead, I opted to try the same strategy again - I repeated the experiment. I did not attempt to change anything, and simply did what I had done earlier. But this time, I tried harder, and resolved to be more committed.

Sometimes it seems like there's a mistake that happens around 'reproducibility'. It goes something like:

  • 1 (One**) study is enough (this is usually implicit)
  • If it's replicated then it's good.

And, uh...maybe it's just a problem with articles about science, and not the approach people doing the experiments have, but that's missing something. Fire is reproducible with matches - reliably. And while there are conditions that might mess with that, we know what they are - wind, rain, soggy matches, etc. Not everything has to be at that point, but fire is not something that has only been reproduced once. It's reliably, systematically, reproducible, and fairly easily so.

Experiments (on something) don't necessarily start out like fire. (Also, new skills, don't necessarily just happen instantly. If they do - that's fantastic, and I'm really interested in your methodology.)

(This seems related to 'try harder' except it's usually expressed as 'something something patience'.)


I could just as easily have failed by doing this, however, so the interesting question becomes: when do you stick with a theory, even though reality appears to contradict it, and when do you abandon it in favour of a new one?

Having more theories can help.


While reading this post, I thought, 'maybe I'd use my desk more if:

  • I put it in a different spot.'
  • The chair was more comfortable, or more suitable for working at that desk.'

Personally, I'm in a different spot where using my bed like that isn't comfortable*, so I don't do it a lot. But if I changed that and moved my desk to be next to my bed, would I be able to use paper on it? Or is there a height difference that makes that not work?


*Your hypothesis/approach is simpler than mine above and more general. It was immediately clear how it might (and might not) be relevant to me. But there's not enough info about your setup for other people to make a lot of inferences like: 'could the position of the desk relative to solar illumination be relevant?' Are the desk and the bed in the same room?' So it's easier for me to make guesses like 'this is how my setup can be improved'.

Offhand this would suggest that scientific knowledge may grow slowly around 'how to make working at your desk more productive' versus 'which airport has the most comfortable chairs', because there's more shared space. But overall at home knowledge might advance faster because/when more time is spent there. It seems like this is especially because it's the same place. And yet, trying out a bunch of different environments and seeing what works best might be faster (if they were productive), even if it doesn't have the same time concentrated in one location (aside from, say, 'your airport', and ones you go through a lot).


**The way the I's (capital i) look exactly like 1's (ones) in this font look the same, can create ambiguity.