A correlation between typing precision and stability?

Some statistics

My typing performance varies a fair amount day to day.

For instance, two days ago, on April 26, my average speed was 66.13, and my average accuracy was 97.18%. Yesterday, my average speed was 60.14, and my average accuracy was 66.68%. (All these numbers are from speed practice on typeracer.)

That may not seem like a very large difference, but I think that on no other day has my average accuracy been more than one or two wpm above my long term average.

More rigorously, my typing speed (on this test) is approximately normally distributed…

…with a mean of 60.57 (for now) and a population standard deviation of 6.531.

So an average typing speed of 66.13 wpm is just slightly less than a standard deviation from the mean (technically, a z-score of 0.8513).

I think that this gives me a p value of 0.1973. Which is not significant at the p value of “>= 0.05”, long established by statistical custom.

(But I notice I’m confused. “Almost a standard deviation from the mean” sounds moderately impressive, but my p value is not anywhere near significant. And why didn’t the p value ask for my sample size. It seem like a small effect can still be known to be reliable, if you have a large sample size. Naively, if I had a sample of 1000, which had an average that deviated from the mean by 0.8513 standard deviations, I would feel pretty confident that there was some some actual effect in play (because of the central limit theorem?), unless I had taken “many” samples of that size, and was cheery-picking this one.

Note to future self: learn what p values actually are, how they work, and why they are defined that way.)

Also the sample size for that day is only 15, which is less than 30, which is supposed to be the threshold of a rough rule of thumb that I might be fooling myself by seeing patterns in noise.

So…currently my epistemic state is…

• I observed an effect that was big enough to be distinctly noticeable without any statistical analysis.
• I don’t actually know if that effect is just noise that I’m falsely perceiving as a real effect.

[Edit 2020-05-04: The problem was that I was confusing standard error with standard deviation. If I take the standard error, with n = 15 (the size of my unusual sample), I get 1.791. Using that standard error to calculate a z-score of 2.883, I end up with p = 0.001964 < 0.05 (one tailed).

I used a slightly different standard deviation to calculate the standard error, than the one that I cite above, because I’m making use of the data that I’ve collected since I originally wrote this post.

It looks like I was successfully confused by something that didn’t make sense.]

Some speculation

Anyway…I came here to posit a correlation, and I might as well do that, although I’m now more uncertain if there is an effect to explain.

I posit that my typing speed (and my typing precision in particular) is correlated with my stability.

When my stability is high, my fingers are more precise, and my typing is smoother, and therefor faster. When my stability is low, especially when I have a low grade background panic, either my precision is lower (I hit the wrong keys) or I tend to either make more transposition errors, and I am less accurate, and therefore slower.

This does jive with my personal experience. My subjectively-assessed stability was very low yesterday (at least compared to the rest of the month), and the sort of agitation that often comes with low stability feels like the experience of jerkily, and repeatedly making typing errors as I get in my own way. They both have a character of compulsive panicked reactivity, and different “considerations” (in this case global priorities, as well very local keystrokes) competing for bandwidth and yanking me around.

This would also be useful if true, because then I could use typing accuracy as an objective measure of stability. Unfortunately, I don’t know how to easily test if it’s true, because I don’t have an objective measure of stability. Maybe I could do subjective surveys along with typing tests? This seems ripe for placebo / confirmation bias, because my expectation of low stability might influence my typing effort.

On that note, I wonder if EEG is a good measure of the phenomenological state that I’m called “stability.” Naively, they seem like they would correlate, at least, and that measure seems harder to game on accident. (Although I think I’m falling prey to a conceptual trap here, and that I’m failing to clearly distinguish agitation levels from stability levels. Obviously, they trend together, but I think it is possible to have high agitation, and to be on top of it / be taking it as object.

Reviewing my day, April 3, 2020: morning compulsiveness

[This post focuses on my own idiosyncratic (I assume) bug, followed my stream-of-consciousness work trying to solve it.]

What happened

Today, I had a pretty unproductive day, for a very clear reason. I spent the whole morning, from 7:26 AM (actually, probably earlier than that, 7:26 is when I started time-tracking) until 12:54 PM, neurotically trying to get a good price on some graphic novels.

[toggl data goes here]

I had an early signal of something. When I woke up, I felt distinctly low energy, in an emotional way, rather than a “poorly rested way”. I felt sort of “ughhh” and didn’t feel interested in getting up for the day. (I distinctly remember noting the feeling, because I had the fleeting thought that I should record it in my log, but didn’t this time.)

I started with random web browsing, and then got a hankering for some specific out of print graphic novels. Being out of print (and having had limited print runs), they were pretty expensive, up in the $60 to $300 range. I wasn’t going to pay that much, so I started searching used bookstores for better prices. Which, overall, was fine.

But eventually, I got to the point of finding a few ebay sellers who had what I wanted, and then obsessively trying to figure out which ones to buy from which sellers. This was not entirely straightforward because…

1. if I bought multiple items from a single seller, I could get a discount on shipping.
2. some of the items were “buy it now” but some were auctions. Auction prices tended to be lower than “buy it now”, but I would have to account for the risk of the price rising higher and overshooting the “buy it now” price, or of missing out on the item entirely.
3. there were different editions, and some (overall dis-endorsed) part of me was concerned about the sets matching, and in particular, my not ending up with a book that said “volume 2” on the back, if that wouldn’t match the other ones I was getting.

So I was neurotic and compulsive about this for about 5 hours(!). I even finally made a decision, and then immediately tried to cancel the order, because I had a moment of compulsive not-quite-panic, and thought of a “better plan.”

I was mildly self-conscious of how maladaptive this was, while in the process. In my log (spelling errors in the original):

• 11:12 – I just spent multiple hours carefully trying to get the best possible price on some comics that I was buying. Even now, I feel the urge to cancel some of my purchases, to buy them less expensively a different way (I just did that).
  • This is crazy, given that I bought and sold thousands of dollars, and most of my net worth, more-or-less based on impulse over the past few weeks. I should have spend hours carefully considering _those_ transactions, and not worrying about these ones, which at worst would cost me an extra $30 or so.
  • Can I make that switch somehow?
  • This is crazy
    • current plan:
      1. Buy Knights of the Old Republic + Freedon Nad rebelion
      2. Bid on the lowest “Dark Lords of the Sith”
      3. Buy a copy of “Sith Wars
  • I think this counts as #compulsive #compulsiveness

This kind of thing happens to me sometimes. I occasionally get obsessed with trying to complete some task. Its as if I have an itch and I just have to scratch it. I once spent ten hours trying make a text-to-speech audio-book of a book that I had bought, so that I could listen to it as I was walking. (Despite many set backs, I eventually succeeded.)

I speculate that one important part of the structure is something like feeling, at every moment, that success is imminent. (This also accords well with my model of human motivation, which I’ll write up one day.)

I call this state “compulsiveness”. It might also be described at “neurotic”, and is related to, but distinct from, my term of art “urgey”. (When I’m urgey, there isn’t a single obsessive focus. I’m just sort of compulsive in a general way.)

[Also, I did recover, and then had a pretty excellent second half of my day.]

What should I do about it?

[this part is pretty stream-of consciousness]

So, what should I do so that I never have this problem again?

Meditation is supposed to help with this sort of thing, by “extending the space between stimulus and response.” But that’s a long run solution, and a very fragile one: the months when I am under the most strain and pressure, are the months when my meditation habit is most likely to break down.

(I also predict that this doesn’t happen when I have strength trained in the past 2, maybe 3, days.)

Similarly, one thing that would have forestalled this situation, in this case at least, is if I had an iron habit regarding what I do when I wake up, and it isn’t look at my computer. But that solution runs into the same problem: it is most likely to fail when I am most likely to need to rely on it.

One thought is to try Frame-by-Frame debugging: step through the moments in sequence, noticing the places where an intervention would have been effective.

Doing that…

• One possible intervention point was the moment that I noticed that I felt kind of “ughhh” / unenthused.
• Another point of intervention is upon noticing that I’m compulsive (though at that point most counter actions would be pushing against a strong psychological force.
  • One solution might be to go exercises, immediately, as soon as noticing the compulsion, but then allow myself to decide to go back into into it after exercising. (Exercise clears the “physiological short term memory”, and seems to have a major centering and stabilizing effect, for me at least. When I booted up today, after this whole thing, the first thing that I did was deadlifts.)
    • Maybe I should greet the compulsiveness first?
    • In order for this to work, I need to have a really robust strategy for exercise. I don’t want to be held up because of some trivial inconvenience, like that I would need to change my pants.
  • Or maybe my TAP should be to just become more strategic about serving the needs of the compulsiveness. If I had spent some time doing meta, clarifying what my goals were and which thought processes were most likely to meet them, I think I would have done much better.)

One problem that I have will all of these is that I’m unlikely to encounter this situation again for a while, so I’m apt to forget any policies I come up with now.

…which suggests that I need to induce this state somehow, under controlled conditions, so that I can practice the responses I want?

(In many cases, I would not want to have any set response at all: I would just want to become conscious and reason about what’s best in the moment. The problem is that some states are particularly low on consciousness, and I have to have a solution that doesn’t depend on me magically being sensible or strategic in the moment.)

I need something like a smooth off-ramp. What actions could I take while in the midst of compulsiveness that aren’t in resistance to the compulsiveness?

(This skill might be the same skill that I need for falling asleep when my mind is churning. )

Maybe I want to go with the flow of the compulsiveness, at least at first? The “strategically serving the neurotic need” seems to do that. Maybe start, by listing what count as the win conditions?

[Pete would tell me about “leaning out”. But I don’t know how to do that.]

I want to have a robust solution here, and then murphyjitsu it, but I don’t have enough…data(?) for that?

What’s blocking me from having this bug solved forever?

Well, all solutions depend on either 1) having a lot of consciousness available to work with or 2) having something like a detailed enough sense of what the thing is that I can work with it “from here”, where I’m not in direct feedback-y contact with it.

So maybe I should aim to get more feedback-y contact, while I have some consciousness? So I should be looking for small versions of this problem, that are not overwhelming to my capacity.

What might those include?

• When I’m wanting to sleep but my mind is churning?
• When I feel antsy
• When I feel urgy.
• When I have a literal itch?
• When I feel antsy during a meditation?
• When there’s something that I want, and I want it right now.
  • (When I put it this way, it feels like the obvious solution is to take pleasure in waiting, at being stronger than the urge.)
• When I’m horny?
• When I’m doing some kind of endurance training, and I want to get up from my wall-sit.

[That’s 8. I should force myself to get to at least 10.]

• When I’m flipping through Tinder / Bumble uselessly (I don’t do this much any more)
• When I’m watching, anything, on youtube? (Not anything. I think when I check, most of the time when I’m on youtube, I won’t have much of this phenomenology.
• When I feel like I need to sneeze?

Ok. So what are my next steps? To look out for when something like this is happening, and observe it? I feel like that’s not going to actually happen. I could try and arrange for a situation in which one of these things will be dominant, but I don’t really feel like it’s worth an hour of planning and an hour of execution right now.

I know that I will have the urge to masturbate sometime in the next two weeks. Maybe I should study that?

TAP: notice “motion towards” masturbation -> . . .

Then what, study it?

What’s the specific response?

TAP: notice “motion towards” masturbation -> Freeze and be alert…then note it in my log.

Is that going to fire? Maybe, but I’m out of steam right now, and want to go to sleep. We’ll see if any of this percolates anything in me.