Hacker News Reader

Comment by cgriswald | Hacker News Reader

> The study was very upfront about the limitations.

That's an overly generous interpretation of their comments. See deogeo's post with some questions about why many of these things are even mentioned in the paper. Where they were upfront has already been acknowledged--by me. My criticisms were where they implied, but did not support a comparison to US-based finance companies generally. Here they were 'upfront' but they shrank what should be several papers worth of arguments into a single word: 'similar'.

> A better analogy would be someone trying to debunk the theory of relativity by claiming it doesn't cover perpetual motion machines.

If you think this is a better analogy, you've missed the point entirely.

View on HN

You're beginning to sound like a conspiracy nut. The paper is very upfront that more studies should be done in other countries:

"the data necessary for the analysis, such as paygrades, demographics and manager assignments, are probably available for most large organizations. We hope this methodology will be applied to other firms, countries, and industries, which will help to generalize the findings and identify where these biases are more or less problematic"

> You're beginning to sound like a conspiracy nut.

If I sound that way to you, it's probably because of your own biases.

> The paper is very upfront that more studies should be done in other countries:

You aren't even addressing the part of the paper I criticized. You haven't even acknowledged it exists, but you expect to, what, convince me or others that I'm wrong by throwing random quotes my way and calling me names?

The quote I provided shows that the authors want other people to study the US or other countries to see if the results apply in a general manner. The paper in no way attempts to generalize the US other than that large companies typically track the sorts of data which would make similar studies possible.

> The paper in no way attempts to generalize the US other than that large companies typically track the sorts of data which would make similar studies possible.

In the section 3.1 Institutional Context, the paper explicitly compares the bank and the country to firms in the United States and the United States respectively over a handful of values. There's no explanation given for why the United States specifically was compared against nor is rationale given for the selection or exclusion of points of comparison. A reader has to conclude that they felt these were the relevant points of comparison because these were the things likely to affect their results.

This necessarily implies generalization. Any other interpretation can only make sense if we disbelieve their results or find their comparison criteria between firms and countries lacking.

Also from that section "we do not want to claim that the evidence is representative of the world"

I acknowledged that disclaimer in my first post and in some of my subsequent posts. I really think you're arguing in bad faith here.

Reply on news.ycombinator.com