Hacker News new | ask | show | jobs
by yboris 2151 days ago
My memory of the claims I came across during my time in Educational Psychology Ph.D.: tests in the distant past (early 1900s) were often culturally biased (even if unintentionally). I couldn't quickly find examples, but the gist is language or customs more available to rich white kids than otherwise (think playing golf, polo, etc).

One quick post I came across when looking just now is about "oarsman" and "regatta", words more-likely to be familiar to rich white kids than otherwise:

https://www.clearchoiceprep.com/sat-act-prep-blog/the-most-i...

1 comments

The complaint about "oarsman" and "regatta" is nonsense.

Supposedly it was racist. Supposedly the white kids are all out doing yacht races, so they alone would know the vocabulary.

That is ridiculous. The portion of the population of white kids that go yachting is minuscule. Practically all white kids don't even know a person who does that. Practically all white kids have never even been on a yacht.

The question was fair. It, and others like it, are a test of reading. If you read well and have read a large variety of books, you'll know the words. The whole point of the test is to see if a student has that sort of college-capable ability (Can you read very well?) so of course the question belongs on the test.

The same goes for polo. Nobody does that. The same is nearly true for golf, and I think the desire to watch Tiger Woods might flip any advantage toward the black kids.

Really, it's about reading.

Don't focus on whether it's racist or not - that's not the point. The point is that it is easy for test makers who are not cognizant of culture to slip in questions that to some degree are easier to answer if you have relevant knowledg ef a particular culture.

In my Ph.D. studies in Educational Psychology we were made aware of cultural bias. More than a hundred years ago, when people made tests to assess "intelligence" they may have not been so careful as we try to be now.

Are you sincerely willing to claim that it's impossible to create a test that doesn't make it easier for some subset of the population to outperform another?

It's desirable to create a test with cultural bias. College has a culture. Remember, the test is about the ability to do well in college. Awareness of the cultural expectations will increase the chance of success in college.

If a student couldn't be bothered to learn about the culture, he isn't a good fit.

That said, the questions really are about reading. If you read lots of books, not counting the ones for little kids, you'll have no problem with the questions.

Do you think it's desirable to maintain a college culture that is biased towards the children of people that have yachts?
Those people practically don't exist, so who cares? They are a rounding error, if that.

It is desirable to maintain a college culture that is biased towards people who read about lots of different things, including yachts. Reading is good.

The choice of reading material matters. I don't count "The Cat In The Hat" and "Teen Magazine" the same as "The Grapes of Wrath", "Hamlet", "The Scarlet Letter", and "As I Lay Dying".