3 ms·

The author is cherry picking studies to try and prove his hypothesis, ignoring that the "studies" he cites show only a possible correlation. First, he's taking

The author is cherry picking studies to try and prove his hypothesis, ignoring that the "studies" he cites show only a possible correlation. First, he's taking the results of a voluntary survey of users of his own site. He'd be better off asking a polling firm to randomly contact people, ask if they regularly read any more well-known sites that are more "intellectual" in nature, and then ask about the person's birth order. Second, he uses anecdotal data about Harvard students, which ignores several potential sources of bias. As one example: Harvard is expensive, and students are more likely to come from families with large incomes. High-income families have fewer children. Another example: younger siblings often specifically don't want to go to the same school as their older siblings, so if an older sibling gets into Harvard, they are less likely to bother applying. There's also a bit about males being more likely to be gay based on having more older brothers. I could probably spend an hour listing reasons you shouldn't take that research seriously.

naasking 9y ago
> The author is cherry picking studies to try and prove his hypothesis, ignoring that the "studies" he cites show only a possible correlation Not really, he starts out explaining that most previous research on birth order effects was well and truly debunked. This doesn't disprove the existence of birth order effects, and he makes a decent case. > First, he's taking the results of a voluntary survey of users of his own site And he thoroughly discusses the self-selection bias. > Another example: younger siblings often specifically don't want to go to the same school as their older siblings, so if an older sibling gets into Harvard, they are less likely to bother applying. If first-borns are more likely to go to Harvard, that's a birth order effect too. If they're not more likely to go to Harvard, then Harvard should have an even mix of birth orders and so aren't subject to the bias you point out.

naasking 9y ago

> The author is cherry picking studies to try and prove his hypothesis, ignoring that the "studies" he cites show only a possible correlation Not really, he starts out explaining that most previous research on birth order effects was well and truly debunked. This doesn't disprove the existence of birth order effects, and he makes a decent case. > First, he's taking the results of a voluntary survey of users of his own site And he thoroughly discusses the self-selection bias. > Another example: younger siblings often specifically don't want to go to the same school as their older siblings, so if an older sibling gets into Harvard, they are less likely to bother applying. If first-borns are more likely to go to Harvard, that's a birth order effect too. If they're not more likely to go to Harvard, then Harvard should have an even mix of birth orders and so aren't subject to the bias you point out.