A Vocabulary Quiz

[Magna]

I'm fairly sure a couple of the questions are wrong . I don't want to post any here or even discuss them in any detail, lest I invalidate the test.

If you think there are some errors, please email them to me at gameswithwords@gmail.com.

Thanks.

I put them into comments at the end - will that work? Otherwise, I can email them.

[Magna]

Math also seems to present a challenge: "Out of every 10 people who did this experiment, you would score better than 10."

What's the issue with the math? Compared with 10 other people, you could score better than anywhere from 0 to 10, right?

Maybe it's really a grammatical error - something akin to a dangling modifier. When a sentence says "Out of [a group], you..." it implies that "you" are being counted among the group, not compared to a group of other people. (For example, I would only say "Of all the men I know, you are the best chess player" to a man - not to a woman.) As it stands, the sentence implies I would outscore myself, which is impossible.

Btw, I didn't get a perfect score.

[gameswithwords]

@Magna -- Thanks for putting them in the comments. That's perfect. I'll take a look.

As far as the modifier ... I'm not convinced that the sentence actually implies you are in the group. I think what may be playing a role is having to have learned to include oneself in the group when interpreting percentile. But I remember how hard it was to learn how to think about percentiles, and many (most?) people don't understand them well, so I don't like using them. But then there is this issue of tripping up people who *are* used to percentiles.

I'll think about whether there is a phrasing that makes things more clear.

And yes, you don't have to get them all right in order to do better than 10 randomly chosen other participants. It is a hard quiz with a high ceiling (though a handful of people have gotten them all correct).

[trainman]

I was annoyed at nebbish. It's not an English word!

Hmm, it's in my (American) English dictionary.

Seems about as English as "kibitz."

[Magna]

@Magna -- Thanks for putting them in the comments. That's perfect. I'll take a look.

As far as the modifier ... I'm not convinced that the sentence actually implies you are in the group. I think what may be playing a role is having to have learned to include oneself in the group when interpreting percentile. But I remember how hard it was to learn how to think about percentiles, and many (most?) people don't understand them well, so I don't like using them. But then there is this issue of tripping up people who *are* used to percentiles.

I'll think about whether there is a phrasing that makes things more clear.

And yes, you don't have to get them all right in order to do better than 10 randomly chosen other participants. It is a hard quiz with a high ceiling (though a handful of people have gotten them all correct).

You could say "You scored as well as or better than ___ out of 10 people who took part in this experiment."

I hope that's helpful!

[PowerofHoodoo]

I got 32/32. I don't know if I'd say any of the answers were wrong, but it gave me pause that some of the words that were "closest" were not all that close.

[gnash]

Math also seems to present a challenge: "Out of every 10 people who did this experiment, you would score better than 10."

What's the issue with the math? Compared with 10 other people, you could score better than anywhere from 0 to 10, right?

Maybe it's really a grammatical error - something akin to a dangling modifier. When a sentence says "Out of [a group], you..." it implies that "you" are being counted among the group, not compared to a group of other people. (For example, I would only say "Of all the men I know, you are the best chess player" to a man - not to a woman.) As it stands, the sentence implies I would outscore myself, which is impossible.

Btw, I didn't get a perfect score.

It definitely does not imply that you would outscore yourself. It may, however, be interpreted as saying that you would outscore everybody else who took the test, and that almost certainly wasn't the intended meaning. It is unclear what the intended meaning is. That you are in the top 10%? That's contradicted by gameswithwords' reply to your post (because then it would be impossible to "score better than 0"). That you would more likely than not outscore 10 randomly drawn participants? That seems to be the only meaning not precluded by obvious contradictions. It would mean you are in the top 5%. But I am not sure if that really is the intended meaning.

Nobody should ever say "out of every [sample size] people..."

[gnash]

@Magna -- Thanks for putting them in the comments. That's perfect. I'll take a look.

As far as the modifier ... I'm not convinced that the sentence actually implies you are in the group. I think what may be playing a role is having to have learned to include oneself in the group when interpreting percentile. But I remember how hard it was to learn how to think about percentiles, and many (most?) people don't understand them well, so I don't like using them. But then there is this issue of tripping up people who *are* used to percentiles.

I'll think about whether there is a phrasing that makes things more clear.

And yes, you don't have to get them all right in order to do better than 10 randomly chosen other participants. It is a hard quiz with a high ceiling (though a handful of people have gotten them all correct).

The way I interpret your answer, "better than 0" means in the bottom 5%, "better than 1" in the bottom 15% (but not in the bottom 5%), etc., and, finally, "better than 10" means in the top 5%. I think it is the only interpretation that is logically consistent with your explanation. But I also think that the vast majority of people would not interpret it that way.

If this is indeed the interpretation, you should express it in ventiles: "You scored better than N out of 20 people who took the test." N could be between 0 and 19, inclusively, but not 20. You'd give the same information to the top and bottom 5%, and a little more information to the rest, and most people would understand you.

[gnash]

I got 32/32. I don't know if I'd say any of the answers were wrong, but it gave me pause that some of the words that were "closest" were not all that close.

"All that close" (or just "close") is not a well-defined concept. "Closest among the 4 alternatives" is.

I missed one, where I simply didn't know the word. (And it is an interesting word, I learned some scientific facts I didn't know.) For all the rest, it was very clear to me why the correct answer was correct, and I don't think I could make the case for any different answers (and I have an infamously keen eye for finding faults with questions).

[bomtr]

Nobody should ever say "out of every [sample size] people..."

Yeah, absolutes are bad...

[PowerofHoodoo]

I got 32/32. I don't know if I'd say any of the answers were wrong, but it gave me pause that some of the words that were "closest" were not all that close.

"All that close" (or just "close") is not a well-defined concept. "Closest among the 4 alternatives" is.

I missed one, where I simply didn't know the word. (And it is an interesting word, I learned some scientific facts I didn't know.) For all the rest, it was very clear to me why the correct answer was correct, and I don't think I could make the case for any different answers (and I have an infamously keen eye for finding faults with questions).

I was mainly addressing what Magna said about some questions being wrong. I went in with more of an SAT mindset of picking a synonym from four fairly closely related choices. The correct choice often had no relation to the others and was not as closely related to the test word as in most vocabulary tests. I got through the test fine, but, for me, it goes against the grain to choose an answer when the connection seems somewhat tenuous.

[Magna]

I got 32/32. I don't know if I'd say any of the answers were wrong, but it gave me pause that some of the words that were "closest" were not all that close.

"All that close" (or just "close") is not a well-defined concept. "Closest among the 4 alternatives" is.

I missed one, where I simply didn't know the word. (And it is an interesting word, I learned some scientific facts I didn't know.) For all the rest, it was very clear to me why the correct answer was correct, and I don't think I could make the case for any different answers (and I have an infamously keen eye for finding faults with questions).

I was mainly addressing what Magna said about some questions being wrong. I went in with more of an SAT mindset of picking a synonym from four fairly closely related choices. The correct choice often had no relation to the others and was not as closely related to the test word as in most vocabulary tests. I got through the test fine, but, for me, it goes against the grain to choose an answer when the connection seems somewhat tenuous.

Good observations. Maybe I should have said some questions seemed invalid, rather than wrong. To the extent the questions are aimed at reasoning out which are the least tenuous connections, it's less a vocabulary test and more a test of abstract or intuitive reasoning. The questions could still be wrong, though, and I'm pretty sure one or two are.

To put some context to this, here's a made-up example:

red
a. distributive
b. abased
c. coral
d. obscure

Everyone knows red is a color, but someone with a good vocabulary would also know it can mean "communist" or "embarrassed." "Obscure" doesn't seem to fit, but then you are left to reason out whether the coral really is a shade of red (rather than orange or pink), and if so, whether it is a closer connection than red=communist=distributive or red=embarrassed=abased. I don't know how closely ability to answer that question would correlate to a good vocabulary, but I suspect not very close. Even to someone who knows all the meanings of those words (i.e., who knows all the vocabulary required to answer this question), it's not completely clear what the answer should be.

[Magna]

As it stands, the sentence implies I would outscore myself, which is impossible.

It definitely does not imply that you would outscore yourself.

I agree it's illogical to say I would outscore myself, but grammatically, that's what the sentence means. It's an example of the common error caused by omitting "other" when comparing one thing to members of the same group.

For example, you could accurately say "Out of all 50 U.S. states, California has the highest population" - but not "California has a higher population than any U.S. state." If the thing being compared is not part of the group, you could use the comparative form (without "other") but you would not use the "out of ..." construction. For example, you could (correctly) say "India has a higher population than any U.S. state" but not "Out of all 50 U.S. states, India has the highest population."

Another way to fix the error is to use the superlative form (e.g., "Out of all 50 states, California has the highest population.") But that option isn't available here, because the website can't accurately say the top or bottom 10% of scorers have the "highest" or "lowest scores" - that's true only of two individual test-takers.

[jpahk]

surprise! it wasn't actually a vocabulary quiz at all:

out of 10 people who took this quiz, you are more pedantic than: 10 of them

[Magna]

surprise! it wasn't actually a vocabulary quiz at all:

Fwiw, it claims it is a linguistics experiment by MIT researchers.

[gnash]

To put some context to this, here's a made-up example:

red
a. distributive
b. abased
c. coral
d. obscure

Everyone knows red is a color, but someone with a good vocabulary would also know it can mean "communist" or "embarrassed." "Obscure" doesn't seem to fit, but then you are left to reason out whether the coral really is a shade of red (rather than orange or pink), and if so, whether it is a closer connection than red=communist=distributive or red=embarrassed=abased. I don't know how closely ability to answer that question would correlate to a good vocabulary, but I suspect not very close. Even to someone who knows all the meanings of those words (i.e., who knows all the vocabulary required to answer this question), it's not completely clear what the answer should be.

That's nice, but your made-up example bears no logical resemblance to the questions in the test.

Actually, I take that back. Even in your example, designed to illustrate your point, the word "coral" is clearly the closest to the word "red". One of its main meanings is nearly synonymous with the main meaning of "red". All other connections are indirect and tenuous. For example, red is now more associated with the Republican Party than with communism, and the connection between "communist" and "distributive" is somewhere between unclear and mistaken. It's like claiming that two words are homophones because you can get from one to the other if you just play telephone long enough.

[gnash]

As it stands, the sentence implies I would outscore myself, which is impossible.

It definitely does not imply that you would outscore yourself.

I agree it's illogical to say I would outscore myself, but grammatically, that's what the sentence means.

It's illogical (rather, nonsensical) to say that a sentence means something grammatically. But anyway, it's clearly not the only possible meaning of the sentence, i.e., the sentence does not imply it.

It's an example of the common error caused by omitting "other" when comparing one thing to members of the same group.

No. The sentence would not be any better if "other" were inserted. It first needs to lose "every" - that's the biggest problem. What needs to be changed then, depends on the intended meaning.

For example, you could accurately say "Out of all 50 U.S. states, California has the highest population" - but not "California has a higher population than any U.S. state." If the thing being compared is not part of the group, you could use the comparative form (without "other") but you would not use the "out of ..." construction. For example, you could (correctly) say "India has a higher population than any U.S. state" but not "Out of all 50 U.S. states, India has the highest population."

Yes, but you are making an unjustified assumption that "you" is necessarily part of the group. It's unjustified in two ways. First, the group being compared to is obviously a sample, unless only 10 people have taken the test. Second, You just took a test now and are being compared to those who took it before you. You are not part of that group.

[Magna]

No. The sentence would not be any better if "other" were inserted. It first needs to lose "every" - that's the biggest problem. What needs to be changed then, depends on the intended meaning.

OK, I agree "every" doesn't need to be there. (I think it's a hangover from TV commercials - "four out of every five dentists surveyed" and the like.) But it's clear to me that the intended meaning is to divide the test-takers into deciles, and tell people which decile they scored into. Saying "you scored as well as or better than ___ out of __ people" would at least fix the logical problem.

Another way to look at this is to point out that if those in the top decile are told they scored better than 10 of 10, the second-highest decile should be told they outscored 9 of 10; third-highest, 8 of 10; and so on, down to the lowest decile, who will be told they outscored 1 of 10. That's strange, because no one scored lower than the lowest decile. Even if someone got zero correct, they would still be told they scored better than 1/10 of the test-takers. Or, working in reverse, if we fix that problem by telling the bottom decile is told they scored better than 0 of 10, then the top decile should be told they scored better than 9 of 10 and not 10 of 10. The only other alternative is that there are eleven deciles.

[dhkendall]

For example, red is now more associated with the Republican Party than with communism

Which has always surprised me (and I'm sure you as well, being originally from overseas). Everywhere else in the world, blue has been used for conservative parties, and red for socialist parties. It wasn't until 2000 (in other words, very recently) that the colours were associated the way they are, and it was mainly done that way by accident (simply most of the news broadcasters just happened to use those colours for those parties that time, serendipitously coming together). So this is at least one that we can't chalk up to the Americans deliberately going against the grain (like their spelling of words or the metric system.)

[Magna]

Actually, I take that back. Even in your example, designed to illustrate your point, the word "coral" is clearly the closest to the word "red". One of its main meanings is nearly synonymous with the main meaning of "red". All other connections are indirect and tenuous. For example, red is now more associated with the Republican Party than with communism, and the connection between "communist" and "distributive" is somewhere between unclear and mistaken. It's like claiming that two words are homophones because you can get from one to the other if you just play telephone long enough.

I made this example up to show that what I was testing was no longer vocabulary. Assuming, for the sake of discussion, that your reasoning is clearly the best and soundest, the kind of reasoning you are using to answer the question is not the same as vocabulary, which is what the experiment seems to be trying to measure.

Fwiw, I deliberately picked an ambiguous color name. In different dictionaries and industries, "coral" can be red, pink, or orange. By way of example, the colors of the Miami Dolphins' uniform are aqua and coral, and I think most people would call that shade of coral an orange. Also, which nuances or definitions predominate depends a lot on factors like what generation you come from, where you live, what you read/listen to/watch, and whom you hang around with. Those types of things can introduce unwanted randomness and decrease a test's validity.