11–20 of 90 posts
> Sprinters don’t accelerate to the top by acquiring the perfect pedigree. They do it by making bold career moves over the course of their career that catapult them to the top. This article looks like yet another one that screams for a survivorship bias disclaimer: https://xkcd.com/1827/
> Sprinters don’t accelerate to the top by acquiring the perfect pedigree. They do it by making bold career moves over the course of their career that catapult them to the top. This article looks like yet another one that screams for a survivorship bias disclaimer: https://xkcd.com/1827/
Haha yeah, the ones how end up on top always think they made bold moves, while the others didn't. It is exactly as you said, survivor bias and nothing else.
What type of bias is it when you just invent an explanation for something out of thin air?
Something that has always surprised me a bit is how many developers want to run their own large companies. Wanting to run a small company is largely about personal freedom that comes for not having a boss, but wanting to be the head of a large business is a whole different ballgame. It's a role that encompasses strategy, leadership, and management - 3 things that can only really be done by talking to people and under…
Think they want to run their own large companies.
> Sprinters don’t accelerate to the top by acquiring the perfect pedigree. They do it by making bold career moves over the course of their career that catapult them to the top. This article looks like yet another one that screams for a survivorship bias disclaimer: https://xkcd.com/1827/
Well, it promises the fastest way to become a CEO, not safest once.
It’s a shame, they have the data and could have asked the question.
| # sprinters | # non-sprinters |
--------+----------------+-----------------+
CEO | | |
--------+----------------|-----------------|
non-CEO | | |
--------+----------------+-----------------+
I don't see anything much that's non-obvious here. - Go small to go big = It's easier to be CEO in a smaller company. Kind of obvious. - Inherit a Big Mess = It's easier to be CEO in a situation where nobody wants to be CEO. Kind of obvious, too. - Make a big leap = Take risky bets, in which case you're not represented in the sample of CEOs when they don't play out.
Something that has always surprised me a bit is how many developers want to run their own large companies. Wanting to run a small company is largely about personal freedom that comes for not having a boss, but wanting to be the head of a large business is a whole different ballgame. It's a role that encompasses strategy, leadership, and management - 3 things that can only really be done by talking to people and under…
Earlier quoted context omitted.
Haha yeah, the ones how end up on top always think they made bold moves, while the others didn't. It is exactly as you said, survivor bias and nothing else.
Any study of successful people (or really successful anything, according to any criteria you choose for “success”) can be dismissed as a “survivorship bias”. But it’s quite baseless to presume that no insight can be gained by studying them. What type of bias is it when you just invent an explanation for something out of thin air?
If you select a sample of observations matching your criteria, and then do a factors analysis of them, you're effectively committing the Texas Sharpshooter Fallacy (another way of describing survivorship bias).
If instead you were to randomly select from a cohort of individuals and track their career trajectories, then you'd be looking at what factors contribute most to success.
I suspect the latter might be rather more strongly weighted to the generally-recommended career advice of good school + business degree + management-track positions early in career, though of course, that's precisely the hypothesis that you're trying to establish.
Point being that there are a lot more people who graduate from the School of Hard Knocks and advance through smaller firms (of which there are far more than there are top-school graduates).
Then you get to address questions such as whether or not school actually is determinative in careers, or is merely selecting for those destined to succeed regardless. (There was an interesting incidental study done based on NYC magnet school examination cutoffs, which are quite arbitrary, and for which there's significant evidence that those at least immediately bordering the selection criteria (and possibly more broadly) have roughly equivalent career / earnings trajectories).
Among the few interesting elements in the HBS "study" is the possible role of high-risk gambles in career advancement. That also ultimately points to survivorship bias (it excludes those who took such gambles and lost), though there's the prospect of multiple-iteration games as well. So long as the upside gain (CEO-track) is high, and the downside loss (failed that advancement attempt, but no penalty) is low, there's little harm in making such attempts.
If, however, failed attempts are read as a flag against future consideration, the calculus changes remarkably. This is typical of many "up or out" advancement policies.
> Sprinters don’t accelerate to the top by acquiring the perfect pedigree. They do it by making bold career moves over the course of their career that catapult them to the top. This article looks like yet another one that screams for a survivorship bias disclaimer: https://xkcd.com/1827/
Yes, this.