>"Why is "trend" relevant? Your clarified statement is that modern wage disparity is not due to discrimination. That means that even a handful of counter-examples would suffice to show that that statement is incorrect."
That does not logically follow. If you're offering an explanation for a large dataset which is depicting a trend, the value of that explanation is not undermined by a handful of statistical outliers. If we say that the explanation for women living longer than men is a mixture of biology and cultural factors, a single story about a woman murdering her husband does not speak to the accuracy of a broader explanation at all.
You would need to demonstrate that the counter-examples and the counter-explanation they suggest are statistically significant.
Edit: I'm not engaging in an ad hominem attack here, but I would like to draw attention to previous comments made by this poster to indicate that this is not the first time they've had difficulty parsing how statistics are generally understood to work: https://news.ycombinator.com/threads?id=dalke
I disagreed with the statement 'Wage disparity is caused by gender roles, not by discrimination'.
It's clear from history, as my example of Lilly Ledbetter, that Wage disparity is also caused by discrimination.
I don't see the need for "counter-examples and the counter-explanation" when documents like http://www.eeoc.gov/eeoc/newsroom/release/4-28-11a.cfm ("Gender-Based Wage Gap Persists, Experts Agree at EEOC Forum" (2011)) say things like:
> Gender-based wage discrimination remains a problem today and a percentage of the wage discrepancy cannot be explained by non-discriminatory factors, said government and private experts ...
Two possible objections are that 2011 isn't modern enough, and that the EEOC cannot be believed. Without knowing what is considered valid evidence, I don't see the need to dig up examples when they can be dismissed with a one line response.
That does not logically follow. If you're offering an explanation for a large dataset which is depicting a trend, the value of that explanation is not undermined by a handful of statistical outliers. If we say that the explanation for women living longer than men is a mixture of biology and cultural factors, a single story about a woman murdering her husband does not speak to the accuracy of a broader explanation at all.
You would need to demonstrate that the counter-examples and the counter-explanation they suggest are statistically significant.
Edit: I'm not engaging in an ad hominem attack here, but I would like to draw attention to previous comments made by this poster to indicate that this is not the first time they've had difficulty parsing how statistics are generally understood to work: https://news.ycombinator.com/threads?id=dalke