@Speedy2019 how is one chosen? Randomly selected? Do you know if these field testers from all over the country??
@likestowrite #1973 I think this looks reasonable. One of the problems, Iâm sensing, is that there might be a lot clustering that makes adjusting the percentiles hard. This bound to happen with compression. I sense a big cluster in the 214-218 range and another in the 205-209. So where you have 205-209 at 98%, this might really be 97%. I have nothing hard to back this up â just my own DDâs case (209 SI concording to 201-206, which is mostly the 2014 97% range) but thatâs my sense of it FWIWâŠ
@Pickmen what state are you?
I just asked my son. He said that NONE of the passages on his test (10/14) were on the field test. Maybe they were on the 10/28 test. AND, CB did not give them the booklets, questions, or answers after the test was over. He took the test in January 2015.
Thank you Thank you for looking @Pickmen! I have been wanting an opinion. My premise, which may be wrong and I donât mind someone proving how it is, is that the %'s I put on my adjusted chart have nothing to do with my guessing, but instead simply involve changing the %ile numbers back to the old definition. The only place where this is not the case is the beginning of the 99+s. At that point, one has to guess because there is no 99++ to tell you where to stop.
The more I think about concording, the tougher I think it is since we donât have an exact 2015SI% into a 2014SI% concordance chart. If I am right about the SI chart, there seems to be less trouble in interpreting that (up to the fuzzy line that breaks the 99s from the 99+s).
@SLparent, no idea how they were selected. Youâd think CB would select high schools that met certain requirements around the country, racial makeup, gender makeup, geographic regions, economic makeup to match segments of the nation as a whole. I doubt if it was random. If you randomly select a bunch of very poor schools, the results might be too low or too high for affluent areas.
The more I think about it, I seriously doubt the field testers would have been given answers/explanations after completing the test. The students donât need a score. CB would have collected and graded the tests themselves.
@websensation - I feel confident in stating wrt my post #2032, if my daughter was given a chance in 8 months time, with same passages (I donât know how many were repeated), with different questions, she would max out her ERW. Now that she has had a chance to review the score and correct answers with explanations, her SI would jump from 215 to appx. 222, solely based on that single passage (I believe it had 3 questions associated with it). True enough, if field test scores werenât provided with correct answers, this would not matter as much.
@micgeaux, so maybe there were multiple versions of tests during the field test.
Field test scores were NOT provided. No answers were provided. No questions were provided.
Yes, I think there were. Because I think my twins had different versions.
I donât want to start a conspiracy theory. But, you realize CB works with Khan. Who do you think is writing the material for that? So, any student studying Khan had an unfair advantage???
There are probably sample passages with various Lexile Reading levels. Then, there are probably sample questions that are various levels of critical and inferential thinking.
@SLparent â CT
@Pickmen â That is a question I have too. For schools where the PSAT is not a real focus but which have one or 2 NMSFs perhaps a year or every other year, how did they do? Certainly seems a number of high scoring students in schools that typically produce a good number of NMSFs did very well. It might turn out that a number of the states predicted to go down wont r not by as much as predicted by Test Masters.
@OHToCollege â I tend to agree - I am not sure home many of the juniors who took the test had exposure to the reading or other questions through a field study but that certainly bothers me. what resource are we getting the facts from - same passages but different questions? And was it a small sample in a number of states or are there concentrations anywhere?
@LivinProof , #2031
âif a kid can remember a reading passage they spent 5 minutes on 8-10 months earlier, then they should go straight to nmsfâŠdo not pass goâŠdo not collect 200 SI.â
I took SAT on 01/23/2016. I believe it was very similar to one on June 2014. Not sure 100%. But why I remember? Since it was First SAT ever as a freshmen. Later on I found out on CC threads, many kids say the same thing.
@likestowrite You are welcome. Thanks for the work. I appreciate it. Also all the work done by @DoyleB and @thshadow
In response to #2037. I think you well articulated what I have been thinking. This was more of an achievement test and less of an aptitude than ever before. As an achievement test, kids from better schools will do even better relative to worse schools. By better schools, I mean schools with more competition and more expectations from their students, not necessarily âbrighterâ kids on the whole. I suspect kids from these better schools do relatively better on the ACT than the SAT. That is the case at our private school. Although at the top level, those few students do great no matter what test you give them. I think the effect will be that there will be even more clustering of NMSF at particular schools and that might partially explain some of these anecdotes. We will have to see if this is the case. That would really be too bad. The effect would be even greater over time as students/tutors become even more familiar with the new SAT. I donât believe you would need an expensive tutor, but more than ever NMSF would be the most studious and careful from the families with resources and emphasis on education. However, one could argue these standardized tests have never been that great at measuring aptitude but this redesign will certainly not improve that.
On second thought Iâll change my opinion. I think the field test kids were at a disadvantage. Having read the passage 10 months earlier, some of them were likely to remember they had read it and unintentionally relied on their memory rather than reading it carefully during the real test. Memories being what they are, careless errors followed.
@micgeaux, thanks for the info. I think I understand some of this better. Let me explain.
About 9 or 10 years ago, I worked for Pearson (textbooks, education, etc). Back then, Texas had the TAKS tests. High school students needed to pass the TAKS tests in order to graduate. The TAKS tests were replaced a number of years ago with the STAAR tests. Fellow Texans will understand this.
Anyway, one part of the TAKS English Language Arts test, were Short Answer Response (SAR) questions. Students had to read a passage, maybe couple of pages, and then answer a prompt. I was involved in the SAR prompts. Scores ranged from 0-3, with 3 being Exemplary. Some schools were selected to participate in field tests of SAR prompts, several different passages and prompts. The studentâs SAR answers were graded.
The score distribution would then be examined to find the ârightâ distribution. Those prompts would then make there way into real TAKS tests at later test dates. Some prompts produced bad score distributions. This could be for various reasons: passage too hard to understand, prompt wasnât specific enough, etc.
Relating this back to PSAT, it looks like multiple PSAT versions were administered during the field testing by CB (confirmed by @micgeaux). They ultimately used the PSAT with the Douglas passages for the real October, 2015 test. That version of the test produced the best results for them. Hope this is all clear.
@SLparent This is the only info given to me by the GC. Will see how much more I can find out. The GC reported that âmany moreâ juniors scored in the â99%â than in previous years. Will report back next week.
@Speedy2019 I just want to reiterate. My son said that NONE of the passages on his field test were on the 10/14 test.