Gifted Bulletin Board

Welcome to the Gifted Issues Discussion Forum.

We invite you to share your experiences and to post information about advocacy, research and other gifted education issues on this free public discussion forum.
CLICK HERE to Log In. Click here for the Board Rules.

Links


Learn about Davidson Academy Online - for profoundly gifted students living anywhere in the U.S. & Canada.

The Davidson Institute is a national nonprofit dedicated to supporting profoundly gifted students through the following programs:

  • Fellows Scholarship
  • Young Scholars
  • Davidson Academy
  • THINK Summer Institute

  • Subscribe to the Davidson Institute's eNews-Update Newsletter >

    Free Gifted Resources & Guides >

    Who's Online Now
    0 members (), 600 guests, and 19 robots.
    Key: Admin, Global Mod, Mod
    Newest Members
    parentologyco, Smartlady60, petercgeelan, eterpstra, Valib90
    11,410 Registered Users
    March
    S M T W T F S
    1 2
    3 4 5 6 7 8 9
    10 11 12 13 14 15 16
    17 18 19 20 21 22 23
    24 25 26 27 28 29 30
    31
    Previous Thread
    Next Thread
    Print Thread
    Joined: Feb 2011
    Posts: 1,432
    Q
    Member
    OP Offline
    Member
    Q
    Joined: Feb 2011
    Posts: 1,432
    Okay - I clearly have too much time on my hands! Reviewing the recent concordance tables between the new SAT vs. the old SAT vs. the ACT lead me to the concordance tables between the old SAT and the ACT. Anyhow, concordance tables are meant to establish an accurate comparison between SAT scores and ACT scores. Just out of curiosity, I compared the minimum SAT and ACT scores set by DYS and noticed that applicants need much higher SAT scores than ACT scores to qualify. I am sure that DYS very carefully set their minimum scores so I can only assume that talent search applicants must somehow score lower on the ACT than on the SAT. Of course, the SAT (at least the old one for sure) has always been closer to an abilities test while the ACT is closer to an achievement test so it may makes sense that the curves are more disparate on the ACT than on the SAT for talent search applicants (mostly age 11 to 14) versus high school juniors/seniors.

    Any thoughts? I think I will check how NUMATS award levels are for SAT compared to ACT for last year.

    Joined: Feb 2010
    Posts: 2,639
    B
    Member
    Offline
    Member
    B
    Joined: Feb 2010
    Posts: 2,639
    There is some controversy over the concordance table released by the College Board:

    What’s a college test score worth? An ACT-vs.-SAT dispute.

    Joined: Feb 2011
    Posts: 1,432
    Q
    Member
    OP Offline
    Member
    Q
    Joined: Feb 2011
    Posts: 1,432
    Yeah, I read that article and a few others as well a couple of weeks ago. There is no agreement. In one of the articles, there was a mention of at least one university (I think it was in Texas), that contrary to College Board's concordance tables, supposedly concorded the old/new SAT in the other direction - meaning that a 730 on the new SAT was considered higher than a 730 on the old SAT. Per College Board's concordance tables, a 730 on the new SAT math was equivalent to a 700 on the old SAT math. I am not sure what the concordance should be - both my kids found the new SAT to be unequivocally more difficult as far as advanced concepts and problem solving skills but at the same time, the curve is more (too?) generous in that you can miss quite a few more problems to get the same scaled scores.

    Joined: Feb 2011
    Posts: 1,432
    Q
    Member
    OP Offline
    Member
    Q
    Joined: Feb 2011
    Posts: 1,432
    Originally Posted by Quantum2003
    Okay - I clearly have too much time on my hands! Reviewing the recent concordance tables between the new SAT vs. the old SAT vs. the ACT lead me to the concordance tables between the old SAT and the ACT. Anyhow, concordance tables are meant to establish an accurate comparison between SAT scores and ACT scores. Just out of curiosity, I compared the minimum SAT and ACT scores set by DYS and noticed that applicants need much higher SAT scores than ACT scores to qualify. I am sure that DYS very carefully set their minimum scores so I can only assume that talent search applicants must somehow score lower on the ACT than on the SAT. Of course, the SAT (at least the old one for sure) has always been closer to an abilities test while the ACT is closer to an achievement test so it may makes sense that the curves are more disparate on the ACT than on the SAT for talent search applicants (mostly age 11 to 14) versus high school juniors/seniors.

    Any thoughts? I think I will check how NUMATS award levels are for SAT compared to ACT for last year.

    Okay, I am going to have to correct at least one assumption in my original post after reviewing the statistics from NUMAT, CTY and Duke. Davidson's significant disparity between SAT and ACT scores does not appeared to be based on Talent Search results (at least NUMATS, CTY or Duke). For example, Davidson's 7th grade minimum SAT composite (Math + CR) of 1300 is 96.4 percentile for NUMAT while its ACT composite of 24 is only 84.4 percentile for NUMATS.

    NUMATS has the best and most comprehensive statistics but it must be proprietary while CTY and Duke have charts that can be accessed easily online. Anyhow, it was interesting comparing the results of the three talent searches. I would have expected all three Talent Searches to have similar scaled scores to percentile conversions even if their Grand Ceremony qualifications varied significantly. That was not the case at all as the actual disparities were huge. For example, to score in the 99th percentile for SAT Math as a 7th grader, you need 690 for Duke, 770 for NUMATS and 800 (males)/780 (females) for CTY. For 99th percentile for SAT CR (verbal) as a 7th grader, you need 640 for Duke, 680 for NUMATS and 710 (males)/720 (females) for CTY.

    As arbitrary points of comparison, I looked at the DYS minimum scores for 7th graders on the SAT Math (660) and CR (640) as well as on the ACT Math (23) and Reading (28). The minimum SAT Math score of 660 equals 98 percentile for Duke, 92.9 percentile for NUMATS, and about 85.9 percentile [80.8 percentile (males) and 91 percentile (female)] for CTY. The minimum SAT CR score of 640 equals 99 percentile for Duke, 97.3 percentile for NUMATS, and about 94.6 percentile [94.7 percentile (males) and 94.5 percentile (females)] for CTY. The minimum ACT Math score of 23 equals 91 percentile for Duke, 78.8 percentile for NUMATS, and about 72.8 percentile [67.4 percentile (male) and 78.1 percentile (females)] for CTY. The minimum ACT Reading score of 28 equals 94 percentile for Duke, 88.4 percentile for NUMATS, and about 84.9 percentile [85.3 percentile (males) and 84.4 percentile (female)]for CTY.

    And if your 7th grader wants to attend a Grand Ceremony, the best bet is Duke (around 600 CR, 620 Math, 30 Reading and 25 Math) compared to NUMATS (670 CR, 740 Math, 33 Reading and 31 Math) and CTY (700 CR, 700 Math, 28 Reading and 31 Math). CTY cut-offs are constant but NUMATS' and Duke's may change a bit from year to year.

    Anyhow, what I find the most interesting is the scaled score to percentile differences for the three talent searches. It can't just be explained by Duke perhaps being more inclusive and casting a broader net. There are actually a lot more extremely high scorers (near or at 800) testing through NUMATS and CTY than through Duke.

    I can't wait for College Board to come out with the study on old vs. new SAT for Talent Search kids. There was a really old study (a couple of decades?) on the College Board site that compared the curve (score distribution) for these younger kids to the curve for Juniors/Seniors.

    Last edited by Quantum2003; 05/30/16 08:33 PM.
    Joined: Feb 2010
    Posts: 2,639
    B
    Member
    Offline
    Member
    B
    Joined: Feb 2010
    Posts: 2,639
    Originally Posted by Quantum2003
    I am not sure what the concordance should be - both my kids found the new SAT to be unequivocally more difficult as far as advanced concepts and problem solving skills but at the same time, the curve is more (too?) generous in that you can miss quite a few more problems to get the same scaled scores.
    Is a table showing how raw scores are mapped to scale scores available for any recent administration of the SAT. The mapping will of course vary over time.


    "To see what is in front of one's nose needs a constant struggle." - George Orwell
    Joined: Feb 2011
    Posts: 1,432
    Q
    Member
    OP Offline
    Member
    Q
    Joined: Feb 2011
    Posts: 1,432
    I don't believe that any raw score to scaled score tables for recent administration of the SAT are available to the public. However, my conclusion is based on the four tables for the four practice tests prepared by College Board as well as actual data points provided by my two children who took the March SAT. Not as reliable individually but probably accurate in the aggregate, College Confidential has a couple of threads where posters provided raw scores to scaled scores for a number of versions of the SAT that were taken in March as well as school administrations of the test in March and April.

    Just to reiterate, only the new Math SAT is consistently more difficult for sure according to my kids who tried two of the practice tests and took the March administration. In reviewing old SAT tables and new SAT tables as well as some data points for the March adminstration, the curves are always more generous on the new SAT than on the old SAT. This is partly what puzzles me - it would seem simple enough to establish slightly harsher curves on the new SAT to avoid the need for a concordance table, at least for the Math section. The new Reading section is actually easier and thus has a harsher curve - missing only a couple to a few knocks your score down to the middle 700's (after converting from the 40 subscale to the 800 scale) whereas on the old SAT you would be around the high 700's to 800.

    As you stated, the mapping will vary over time. In fact, the mapping does vary even among the four practice tests as well as among the different versions of the March and April SAT. As you already know, different versions of the old SAT have always corresponded to different curves - missing a single question on the old Math section can result in a score of 770, 780, 790 or 800 although most commonly a 780 or 790. From what I could infer, at least some "school" administrations of the March and April SAT were markedly easier (as reported in the media) than the version my kids took during the "regular" March administration but it would appear that those curves were harsher per data points provided on the College Confidential threads.


    Moderated by  M-Moderator 

    Link Copied to Clipboard
    Recent Posts
    Testing with accommodations
    by aeh - 03/27/24 01:58 PM
    Quotations that resonate with gifted people
    by indigo - 03/27/24 12:38 PM
    For those interested in astronomy, eclipses...
    by indigo - 03/23/24 06:11 PM
    California Tries to Close the Gap in Math
    by thx1138 - 03/22/24 03:43 AM
    Gifted kids in Illinois. Recommendations?
    by indigo - 03/20/24 05:41 AM
    Powered by UBB.threads™ PHP Forum Software 7.7.5