How literal are the % from the DAT?

[white fang]

---

Members don't see this ad.

I am curious about the meaning of the percentages on the DAT so I created this unnecessary thread out of a pure curiosity.

It seems unlikely to me that a 23AA, which my score said was 98.7%, is actually 98.7%. That would mean 1.3% x 8000 (avg amount of people who take it in a given year I believe) scored higher than me or anyone else with a 98.7%.

That is only 104 people.

It seems that every other breakdown on this website has someone that got a 23+ on the DAT. It makes me think a lot more people get a higher score than the percentages indicate and that the %'s are actually meaningless.

Thoughts?

 

[Double Bonded]

It's the percentile for your particular version of the test (there are multiple versions of the DAT given because it can be administered any day of the year). Also SDN is a skewed sample because (1) people with higher scores are the ones who will actually post them and (2) being on sdn allows people to see great resources that increase their score (resources that non-sdn users don't know about)

 

[white fang]

It's the percentile for your particular version of the test (there are multiple versions of the DAT given because it can be administered any day of the year) so the number of people who scored better than you for that particular test is probably way less than 104.

But the yearly amount seems like there are probably far more when adding together all versions of the tests. I think since I have been reading the site closely last fall I see 15 people a month post on this site alone with a higher score than that. I am thinking the % isn't a real number of people or anything like that.

 

[white fang]

It's the percentile for your particular version of the test (there are multiple versions of the DAT given because it can be administered any day of the year). Also SDN is a skewed sample because (1) people with higher scores are the ones who will actually post them and (2) being on sdn allows people to see great resources that increase their score (resources that non-sdn users don't know about)

This is probably very true and I didn't think of it. Without reading SDN I would have never known to use anything besides Kaplan.

 

[Dgeorg6]

[/B]


This is probably very true and I didn't think of it. Without reading SDN I would have never known to use anything besides Kaplan.

This is so true and amazing.. I know about all these materials yet everyone else in my pre-dental society take kaplan courses and that and MAYBE know about crack dat pat and dat destroyer and thats it

 

[uva09]

But the yearly amount seems like there are probably far more when adding together all versions of the tests. I think since I have been reading the site closely last fall I see 15 people a month post on this site alone with a higher score than that. I am thinking the % isn't a real number of people or anything like that.

Hmm, so you question the validity of the percentiles on your unofficial copy of the DAT but you do not question the validity of DAT scores posted by members on SDN, PreDents, etc.? 😕

The official versions do not include the percentiles so it's a moot point anyway.

 

[Tootyfruity]

Hmm, so you question the validity of the percentiles on your unofficial copy of the DAT but you do not question the validity of DAT scores posted by members on SDN, PreDents, etc.? 😕

The official versions do not include the percentiles so it's a moot point anyway.

mm the %iles were on teh back of my official cDATs

 

[white fang]

Hmm, so you question the validity of the percentiles on your unofficial copy of the DAT but you do not question the validity of DAT scores posted by members on SDN, PreDents, etc.? 😕

The official versions do not include the percentiles so it's a moot point anyway.

I question that maybe we don't understand correctly what the percentiles are. I wonder if they are actually based on outdated information where if you get a certain amount right or wrong you get 'x' percentile and not actually indicative of how test takers have performed on a given test.

Speaking of 'official versions', how do you get that? Was I supposed to receive that in the mail or will only the schools get that?

 

[kster]

I question that maybe we don't understand correctly what the percentiles are. I wonder if they are actually based on outdated information where if you get a certain amount right or wrong you get 'x' percentile and not actually indicative of how test takers have performed on a given test.

Speaking of 'official versions', how do you get that? Was I supposed to receive that in the mail or will only the schools get that?

I've searched for information on the percentiles, and it's basically just a lot of guesses. I also had someone say it had to do with the type of questions you got right compared to everyone else who answered that question (e.g. getting one right that most people missed, or vice versa).

At any rate, my AA 23 was a 99.7%ile. I'd much rather have an AA 24 that was a 98.9%ile, since as far as the schools are concerned, percentile means nothing (afaik, they don't even see it).