18 Comments

Utilizing imperfect information properly also requires a certain precision of thought. One has to be aware of all the contextual elements which go into relying on that data. While it varies amongst the population, everyone has a maximum capacity of variables they can simultaneously hold in their mind before the complexity falls apart into chaos. When people hit their personal limit, it is always interesting (and sometimes disappointing) to see who defaults to, "This is too complex for me to properly understand" from "This cannot be understood." My general preference is to avoid the latter.

Another element I find is that, in their quest for certainty, people seem to heavily discount how much fun imperfect information is. Combining a variety of reasons into a hunch and finding out whether or not one got it right is way more entertaining than adding two-plus-two together and always getting four.

Or the fact that the more one practices making and refining decisions with imperfect information, the more one hones their instincts to navigate its blind spots. People who scoff at thinking about a situation in probabilistic terms and updating those probabilities based on new information are usually more likely to dismiss someone as simply lucky or blessed with good fortune (source: anecdotal inkling).

Which is partly true. An intuited probability model still involves chance. But, for similar reasons you raised in your post, it unfairly discounts the fact that someone intentionally (if imperfectly) positioned themselves to heighten their odds. While simplistic, overtime a small edge derived from imperfect information can compound into a completely different life experience compared to those who simply shrug and say, "It can't be known".

Expand full comment

Good post. You are very good at survey methods and you will get multivariate regression and factor analysis in due time. Really nothing to critique in your essay, here is a man observation: if the critique is so broad and general that “anyone can say the sample is to [large / small / restrictive / biased / … ]” and the critic is not providing details on why the critique is relevant to the specific analysis or interpretation you are working on, then put little weight on that feedback. When the critique is targeted, backed up with citation or thoughts experiment that fits the situation, and paired with recommendations on how to improve your game, weigh that feedback more. Anyone who has taken one course in research methods can come up with 15 different “rival hypothses”, the good critic is one who is discerning and helpful and who critiques the most relevant issue with tact, like a good teacher would.

Expand full comment

Good essay.

I have read similar essays by credentialed academic researchers whose results reached a wider audience and thus elicited methodological blow-back from other scientists whose results (I snarkily assume) did not reach as wide an audience. Explaining post-publication why your methodology is valid is part of the process. Congratulations on hitting this milestone 😀

Expand full comment

> And my polls regularly reflect real phenomenon!

Typo, should be "phenomena".

Expand full comment

(not a dig on Aella, just pulling on a thread)

I’ve noticed this happening a lot (mostly in podcasts) over the past 9-12 months: people flipping “phenomenon” and “phenomena.” More often than not, using the plural as the singular. It’s happened often enough where I’m wondering if this is a genuine linguistic fad.

Expand full comment

It seems to me people often believe that imperfect data is *worse* than no data.

Expand full comment

I'm reminded of Copenhagen ethics.

Expand full comment

Makes you wonder how many people would actually say this to your face. The internet makes it very easy for people to just be little pieces of shit. Because regardless if male or female, being rude in public sometimes gets you punched... It's true violence is not the answer but sometimes it's the solution.

Expand full comment

Great post, here is a late question (this was written back in 2022 and now it is 2023!) coming from someone just starting in the area of surveying and data science and all this. Could you suggest good books on stats and python for analysing data and for understanding stats? Thanks once more for the post. It was an eye opener for me in the sense that it got me thinking about stats!!!

Expand full comment

My guess is you are deeply Bayesian but you don't say it often!

Expand full comment

> I also sometimes wonder how much of the disproportionate

> criticism is leveled due to me being an open sex worker on the internet

Maybe, but as a monogamous homebody I'm also quite familiar with these attitudes. I think it's rather that if a little skepticism shows that you're sophisticated and open minded without being naive, then, well gosh, a lot of skepticism must be even better. After all, the most obvious relationships are linear:

MOAR SKEPTICISM = MOAR BETTER

Expand full comment

Undergraduate showed me how little I should trust research by desperate grads trying to finish their degree.

Expand full comment

Hi Aella, thank you for your interesting surveys.

May I ask, are you familiar with The Ricky Gervais Show, at one point the world’s most downloaded podcast? May I ask, do you think there is potential for making a Substack about listening to that show and making surveys from its audience? 

Expand full comment

Aella, You will get pushback as you are using an unpopular thought process called "logic."

Also, best I can tell, we live in a country in which low self-esteem is endemic. Some/most people with low self-esteem look for opportunities to boost their ego by putting someone else down. For guys, this means being a Bully, either in real life or on line. For women, this can manifest as being a Mean Girl. Hang tough girl. The vast majority of your followers see the value in the work you're doing.

Expand full comment

This clown prefers anal sex (9/10 times)

Expand full comment