SAT concordance table - compare old and new SAT scores

Here is the concordance table from the CB website: https://collegereadiness.collegeboard.org/educators/higher-ed/scoring-changes/concordance

https://collegereadiness.collegeboard.org/xls/concordance-tables-old-sat-scores-new-sat-scores.xls

The endpoints are the same, but the shape of the distribution is very different.

@GMTplus7 The College Board played numerous tricks with PSAT percentiles this year. I don’t think percentiles coming from the College Board would be at all transparent.

See the following for details on the games they played, presumably to get more kids who took the 2015 PSAT to thing that the new SAT would be a great test for them.

http://241nkd2yuifs2cob542atybb.wpengine.netdna-cdn.com/wp-content/uploads/2016/01/Problems-with-the-New-PSAT-Compass-1-27-16.pdf

How can we cut our nation’s kids free from the control of the College Board?

So this table documents that the new sat is easier / inflated compared to the old one.
https://collegereadiness.collegeboard.org/pdf/higher-ed-brief-sat-concordance.pdf

For example, a 700 math on the new one is equivalent to a 670 on the old one.

Seems odd that they would inflate the scores, and further compress the high end…

Yes, the 700-800 range on the old SAT Math compresses to 730-800. And, except at the endpoints, you always add 10 to 40 points to the old Math score to get the new Math score. I agree, why do that?

^To make the students taking the New SAT feel smarter. And suckered into taking more SATs.

If CB was waiting for the 2nd round of testing (on Saturday), why not release the scores on Sunday or today? This waiting is ridiculous. We’ll then have 1 day to decide what to take on June 4th, if anything at all.

I doubt CB was waiting for this past Saturday’s scores to release the Concordance tables today. It’s impossible for CB to have received the Scantrons from this past Saturday’s test in time to process the scores on Mother’s Day Sunday.

There’s definitely a different reason for waiting until today to release the Concordance tables.

On the “App”, after you put in the three subscores from the Old SAT (CR, M, W), it gives you a ERW and Math subscore and a total composite score. It also then has an option to “see an estimate based on Critical Reading and Math only.” When you click on this it gives an estimated composite score on the new test, based solely on the CR and M sections of the old test. How is this really possible? Isn’t it necessary to know the score on writing component of the old test in order to get a correct correlation on the new test which has writing questions built into the reading section of the new test? The concordance table doesn’t support this kind of calculation either from what I can see.

It looks like these concordance tables are based on studies, just like the PSAT ones were, not on real test scores: “Concordance tables for the redesigned SAT will be based on two concordance studies. The first was
conducted in December 2014 and focused primarily on concordance tables for the PSAT™10 and
the PSAT/NMSQT. The second study is scheduled for December 2015 and focuses on concordance tables
for the redesigned SAT. More information about these studies is included in Appendix A.” From page marked 7 in https://collegereadiness.collegeboard.org/pdf/college-board-guide-implementing-redesigned-sat-concordance-installment-3.pdf

For the Math section, I’ve matched up the Old SAT percentile ranks to Old SAT scores.  Both New & Old SAT test scales have the same 200-800 endpoints, but the New SAT has a bulge in score increase that is pronounced at the lower end of the percentile ranking (i.e. it makes the low scorers on the new SAT seem a lot “smarter”).  You can copy and paste this table in Excel to graph it and play with it.
 
There’s a similar trend with the Verbal sections.
 
 
 
%tile                      Old                      New                    difference

99%        |              800         |              800         |              +             0

99%        |              790         |              800         |              +             10
99%        |              780         |              790         |

99%        |              770         |              780         |                             

98%        |              760         |              780         |              +             20
98%        |              750         |              770         |

97%        |              740         |              760         |                             

97%        |              730         |              760         |              +             30
96%        |              720         |              750         |                             
95%        |              710         |              740         |                             
93%        |              700         |              730         |                             
92%        |              690         |              720         |                             
91%        |              680         |              710         |                             
89%        |              670         |              700         |

88%        |              660         |              690         |                             

86%        |              650         |              670         |              +             20
83%        |              640         |              660         |                             
81%        |              630         |              650         |                             
79%        |              620         |              640         |                             
76%        |              610         |              630         |                             
74%        |              600         |              620         |                             
71%        |              590         |              610         |                             
68%        |              580         |              600         |                             
66%        |              570         |              590         |                             
63%        |              560         |              580         |

60%        |              550         |              570         |                             

56%        |              540         |              570         |              +             30
53%        |              530         |              560         |                             
50%        |              520         |              550         |                             
47%        |              510         |              540         |                             
43%        |              500         |              530         |                             
40%        |              490         |              520         |

36%        |              480         |              510         |                             

33%        |              470         |              510         |              +             40
30%        |              460         |              500         |                             
27%        |              450         |              490         |                             
25%        |              440         |              480         |                             
22%        |              430         |              470         |                             
19%        |              420         |              460         |                             
16%        |              410         |              450         |                             
14%        |              400         |              440         |                             
12%        |              390         |              430         |                             
11%        |              380         |              420         |                             
9%          |              370         |              410         |                             
7%          |              360         |              400         |                             
6%          |              350         |              390         |                             
5%          |              340         |              380         |                             
4%          |              330         |              370         |

4%          |              320         |              360         |                             

3%          |              310         |              360         |              +             50
3%          |              300         |              350         |                             
2%          |              290         |              340         |

2%          |              280         |              330         |                             

1%          |              270         |              310         |              +             40

1%          |              260         |              300         |              +            

1%          |              250         |              280         |              +             30
1%          |              240         |              260         |              +             20
1%          |              230         |              250         |              +             20
1%          |              220         |              230         |              +             10

I don’t understand this rediculous scaled scoring. Is the collegeboard attempting to make students believe they’re scoring higher in hingsight than they actually are? For example, on the concordance table it states a 25 ACT is equivalent to a 1220 on the new sat. Where a 25 places a student in the 79% on the ACT, one would assume that the 1220 would hold a similar metric in statistical value and be around the 84-85% as that was the percentile on the old SAT. How are admissions officers and others looking at these scores going to use this rediculous scaling with percentiles inflated much or less ± (5-6)% SMH class of 2017 gets the short stick yet again.

The SSAT test (used for prepschool admission) is also reported on a crazy endpoint scale. Therefore, people just ignore the test score and use percentiles. With percentiles, you immediately know where you stand.

Perhaps middle school kids applying for prepschool have less fragile self-esteem than HS students applying to college.

When will percentiles be released for the SAT? Will it be when the scores are released shortly (this month)?

From the CB’s newly released tables:
 
On the 1600 scale, here is the concordance between the New SAT vs. the Old SAT.  Note the scores I highlighted in red.   Look at how there is NOT a one-for-one correspondence between New and Old scores, because for the New 1600 score there are 3 different component sub-scores  (M+CR+W), while for the Old 1600 score it was just 2 different component sub-scores  (M+CR).  
 
Does this make your head ache?  It sure makes mine ache…
 
 
** New | Old | difference **
 1600 | 1600 | + 0
   1590  | 1580 | + 0
-------------------------------     
1590  | 1590 | + 10
 1580 | 1570 | + 10
 1570 | 1560 | + 10
 1560 | 1540 | + 10
-------------------------------     
1560  | 1550 | + 20
 1550 | 1530 | + 20
1540  | 1510 | + 20
-------------------------------     
1540  | 1520 | + 30
 1530 | 1500 | + 30
 1520 | 1490 | + 30
1510  | 1470 | + 30
-------------------------------     
 1510  | 1480 | + 40
 1500 | 1460 | + 40
 1490 | 1450 | + 40
1480  | 1430 | + 40
-------------------------------     
1480  | 1440 | + 50
 1470 | 1420 | + 50
 1460 | 1410 | + 50
 1450 | 1400 | + 50
 1440 | 1390 | + 50
 1430 | 1380 | + 50
1420  | 1360 | + 50
-------------------------------     
1420  | 1370 | + 60
 1410 | 1350 | + 60
 1400 | 1340 | + 60
 1390 | 1330 | + 60
 1380 | 1320 | + 60
 1370 | 1310 | + 60
 1360 | 1300 | + 60
 1350 | 1290 | + 60
 1340 | 1280 | + 60
 1330 | 1270 | + 60
 1320 | 1260 | + 60
 1310  | 1240 | + 60
-------------------------------     
 1310  | 1250 | + 70
1300 | 1230 | + 70
 1290 | 1220 | + 70
 1280 | 1210 | + 70
 1270 | 1200 | + 70
 1260 | 1190 | + 70
 1250 | 1180 | + 70
 1240 | 1170 | + 70
 1230 | 1160 | + 70
 1220 | 1150 | + 70
 1210 | 1140 | + 70
 1200 | 1130 | + 70
 1190 | 1120 | + 70
 1180 | 1110 | + 70
 1170 | 1100 | + 70
 1160 | 1090 | + 70
 1150 | 1080 | + 70
 1140 | 1070 | + 70
 1130  | 1050 | + 70
-------------------------------     
 1130  | 1060 | + 80
 1120 | 1040 | + 80
 1110 | 1030 | + 80
 1100 | 1020 | + 80
 1090 | 1010 | + 80
 1080 | 1000 | + 80
 1070 | 990 | + 80
 1060 | 980 | + 80
 1050 | 970 | + 80
 1040 | 960 | + 80
 1030 | 950 | + 80
 1020 | 940 | + 80
 1010 | 930 | + 80
 1000 | 920 | + 80
 990 | 910 | + 80
 980 | 900 | + 80
 970 | 890 | + 80
 960 | 880 | + 80
 950 | 870 | + 80
 940 | 860 | + 80
 930 | 850 | + 80
 920 | 840 | + 80
 910 | 830 | + 80
 900 | 820 | + 80
 890 | 810 | + 80
 880 | 800 | + 80
 870 | 790 | + 80
 860  | 770 | + 80
-------------------------------     
 860  | 780 | + 90
 850 | 760 | + 90
 840 | 750 | + 90
 830 | 740 | + 90
 820 | 730 | + 90
 810 | 720 | + 90
 800 | 710 | + 90
 790 | 700 | + 90
 780 | 690 | + 90
 770 | 680 | + 90
 760 | 670 | + 90
 750  | 650 | + 90
-------------------------------     
 750  | 660 | + 100
 740 | 640 | + 100
 730 | 630 | + 100
 720 | 620 | + 100
 710 | 610 | + 100
 700 | 600 | + 100

What is most bizarre and RIDICULOUS is that a perfect 800 on CR/writing on the new SAT = a 34 English Writing score on the ACT. That is unreal.

It’s certainly making my head spin. What is the benefit of inflating the score? This really messes up Naviance and other data sheets. On Spykid’s PSAT (No SAT score yet) his reading/writing score was 99 percentile, yet it wasn’t the top of the scale.

Anyone know what the CB does not equate the CR score on the new SAT to the CR score on the ACT? They don’t do the math either. Why not?

I think it is clear what the motive is: inflating scores makes it easier for colleges to admit “special category” students without risking lawsuits.
This is nothing new.
Scores were similarly inflated in 1995 and again in 2006. Wake up people. Colleges are not interested in merit. They are interested in money and diversity politics.

Here is a link to score converters: https://collegereadiness.collegeboard.org/sat/scores/understanding-scores/sat-score-converter

Confusing, eh? I sure wish I’d saved my old Dick Tracy Decoder Ring!

The most important thing to know is that, typically, the new SAT scores are going to seem higher than the old test scores. So don’t go dancing around the dining room just yet thinking that you’re a shoo-in at Skidmore or Syracuse (or Stanford???). The published medians that you’ll see on the College Board Web site (etc.) are based on the old results, which is why the conversion is imperative if you’re trying to determine whether your scores on the March test will put you in the ballpark at your current target colleges.