Hacker News new | ask | show | jobs
by Pyramus 1080 days ago
Real life is rarely about evidence or not evidence, which is a categorical decision (in contrast to evidence in court, which can be circumstantial).

Some more context in addition to the sibling comment: The classical example is this. You want to test the effectiveness of two drugs A and B in some sort of clinical trial. You gather a lot of data and learn that drug A has an average effectiveness of 4.0, while drug B has 4.9 (more is better). You apply a statistical test and it tells you B is not significantly different from A.

Can you conclude that B is not better than A? You can't. The absence of evidence is not evidence of absence. There could be any number of reasons why we don't have significant evidence, e.g. because the experiment was done with a sample too small. B might be better than A, but there is no evidence it is.

The NTSB gathers a lot of data regarding the crash. The board looks at all the factors and deliberates whether they contributed to the crash. It determines that there is no significant evidence that the mental state of the pilot was relevant.

Can you conclude that the mental state was irrelevant? You can't. It's likely the pilot experienced stress but it is impossible to determine whether this stress was significant. It might well be, but there is no evidence it was.

Parent seems to think it was not a factor. But by relying on a flawed argument we will not find out why he actually thinks that way, ultimately leading to an unproductive discussion.

1 comments

Can you conclude that the mental state was irrelevant? You can't.

No, but you can’t claim it was relevant either. So either scenario is likely, so we rely on the experts at NTSB to make a call.

Saying “we cant prove its not relevant” is not actionable information, so it’s not really worth raising it.