Xiggi's SAT prep advice

<p>In actuality, to estimate your score simply pick the middle of the range. See, when ETS gave a field test in March 03 to a bunch of HS kids, it used a prototype scale. Theyr'e using that scale in this book, but what they did is add and subtract the Standard Error of Measurement for each of the three sections (that's the estimate of the range your score will stay in if you retake the exam) to create the range. I say "estimate" because the score grid (raw to scaled score conversion) is different for each actually administered exam. The blue book consists of Frankentests...old Q&A booklets with Writing and modified Math IC questions (all of which I've seen on the SAT II's) stuck in. So all of these questions appeared on administered exams, but some on the Reasoning test, and some on the subject tests...in other words, not in this same form. However, ETS did make the book conform to its own design specs in terms of distribution of topics and difficulty. It's just that only a few of the "new" math topics actually appear in this book. Better bet is to get the March 05, October 05, and January 06 (which is the same as May O5 except for the essays, since those are released) QAS from your friends, and take those tests full length.</p>

<p>I'd expect a new blue book in 2008 at the earliest...as soon as ETS has given 8 QAS administrations, it will time for them to make even more money off those exams. If you want more blue book style exams, there are 6 in the CB online course. One of those is the March 03 field test, though.</p>

<p>If you want to maximize your vocab and start early, use lots of mnemonics, flash cards, and especially roots, so that you are learning pieces of many words, not just one at a time. Then get Cracking the SAT (PR book). The Hit Parade in there is based on 12 years of released tests, with easy words eliminated, and the most frequently occuring words (with a weighting system based on right/wrong answer, SC vs. CR) grouped into themes. After you learn that list, expand: get a hold of every Real SAT book (and any other QAS booklets you can) and learn all of those. ETS="Every Test Sucks because we are lazy; why invent when you can simply recycle".
By the way, folks, the "New" SAT is not New. For example, the March SAT in 2005 was an old SAT from 11/03 with some Math IC questions on the math. I had both real and experimental grammar, but recognized all of them from SAT-II Writing exams which were administered (and then repeated many times) as far back as 2000-2002. Some of them appeared on the Test of Standard Written English for many many years, back in the Middle Ages.</p>

<p>I've personally designed and calibrated hundreds of tests (and thousands of individual questions) for SAT Reasoning and Subject Tests and the GMAT. The process of generating questions, choosing distractor answers, generating top-down organization, difficulty, biserial correlations, and calibrating the test is exceedingly complicated. Most companies and individual tutors don't understand the nature and real aim of these questions, and so can't do anything but review content which most students already know. And the books by ARCO, Barrons, REA, and McGraw Hill prove that there is no effort to align the exams with the real deal. Do your research before you start selling a product that claims to include "SAT Practice Exams", for real!! There are a few tutors who know the way the test is designed, and so can turn their knowledge of the test into strategies to help their students, but it is hard to know whether a tutor knows what they are talking about. Unfortunately, many tutors simply pass on advice that they were told by someone else or read in some book or other, and which they assume to be true without first researching the test, and so pass on strategies which are not valid.</p>

<p>The PR book came out a full year before the CB book, and had a lot of typos (and was based on some guesswork, obviously). The 2006 edition was a lot better, and almost all the questions are spiraled from real examples or bunches of real examples, so they are not "unauthentic" as you claim.</p>

<p>Boy, I could write a book on this. Oh wait, I have. Several. There's no "one size fits all" magic pill for CR. You might try to survey the questions first so you know what to look for, then annotate the passage as you read it. Don't answer the questions as you read though, or you lose the flow of the passage. This works well on very factual, neutral tone passages that ask you to simply retrieve details or figure out word definitions in context.
Or you might skim the passage first and take notes...read the italics up top, a topic sentence or two in each paragraph (until you know what that paragraph is about), and the beginning and end. This is crucial for anything that smacks of being a "position paper". But these new darn Literary Fiction passages lend themselves to neither strategy. You really have to read it, and start making a chart of each character, their characteristics, point of view, and feelings. If you can keep that stuff straight, you'll be able to eliminate loads of trap answers. As far as duals (short and long), well, that's a long story, ha ha. As far as question strategy...hey, do I have to give you all the answers?</p>

<p>Based on the posts I've read so far, you seem to be the best informed of anyone on this forum. For example, you are correct that Sentence Completions test vocabulary and the logic of sentences (not merely the latter, as some have claimed). Each question tests those two factors to varying degrees. And your "counterintuitive" advice is pretty close to the mark. See, if the last question on a section (#5/8/6 as the case may be) is really simple in its structure, then you should always pick a word that most people have never heard of. But if the logic of the sentence is complicated, then the question is already hard, so you should cross off any impossible word (along with any super-easy ones), and pick the next hardest word, usually of medium-hard difficulty.</p>

<p>Taking the two factors above into account, anyone care to guess some of the types of distractors that ETS includes on SC questions? Whoever comes up with the most right answers wins 50 points.</p>

<p>On hard questions, the solution is simple: do not depend on your ear. How do you think the test writers make a question hard? Awkward-sounding answers contain no actual grammar errors. Fine-sounding ones contain an error you don't notice. Know the rules instead.</p>

<p>was repeated in January 06. Get a copy from a friend who took that test, and analyze it. You may figure out what you did wrong, or what was tough.</p>

<p>This is hysterically funny, probably because it is true for so many.</p>

<p>In fact, the College Board book only contains one test of each type, and most are 9-12 years old. The best thing to do is find a tutor who takes the exam regularly and knows what is really on it. Apart from that, the two best books are "Cracking the SAT Math Subject Tests" and "SAT II Math Workout" both by The Princeton Review. The authors of those books (Graham Sultan; Morgan Chase) and their collaborators based the content on 45 administrations of the SAT-II for which they sat.</p>

<p>You are correct that the answer to this type of question is the one slightly lower than the average, and that this is a quick method of solution (this same question has been appearing for decades on various ETS exams). However, it's important to explain why: Since it takes you longer to cover the distance at 8 than at 12, your average speed will be closer to the 8 than to the 12. That's "quantitative reasoning", which is what the test is designed to measure.
You are also correct in pointing out the importance of speed and going over answers you got right on a section. See, on many questions (of all levels of difficulty) there are 2-4 different methods of solution, and those who see the elegant ones will have more time to work on other questions. ETS also includes a question or two per section which is really a stumper, i.e. one that, instead of leading many students into confidently picking #1-#3 distractors (wrong answers) is just far too non-linear for the average test-taker to even start. Most folks skip these...and you need to have really analyzed many SATs to know what's really going on with such questions.</p>

<p>Much has been said about vocabulary. However, the facts are as follows:
Sentence Completion is a mix of vocabulary and logic. Without a command of the former, you simply cannot get a lot of questions right.
Math tests a great deal of vocabulary. Not only do the Number and Operations questions rely on it, but also your understanding of various terms is essential to understanding other questions correctly.
Though analogies are gone, the level of vocabulary on the SAT is increasing.
Furthermore, the passages now contain a much higher level of vocabulary. If you can't understand the specific claims of the passage, you cannot answer questions on them. And don't forget, that along with other literal comprehension questions, definitions in context (i.e. what does "heralds" mean in line xx) are a full 1/4 of the CR passage-based questions.
On grammar, the occasional diction question, some of the verb idioms, and especially redundancy errors that I've seen on recent SAT's, all rely on knowing specific words in the sentence.
So, as comforting as it is to say that vocab is not as important as other things, it's a bit misleading. There's only so far your "reasoning" will take you. On CR alone, if increasing your vocabulary allows you to convert only 8 questions from ones you got wrong (-0.25 raw score points) to ones you got right (+1 raw score point), that's an increase of 8*1.25=10 raw score points, which could mean as much as 120 points on your CR score, depending on where you are on the bell curve. But it could mean a lot more, depending on where your vocabulary is right now. It's literally a "no-brainer" to learn droves of words that have appeared on past SAT's. There's nothing complicated or subtle about learning vocab--no weird strategy to learn. Just do it!</p>

<p>Xiggi is dead on...you can get all the editions of the New PSAT (back to 1997) from ETS. If you can't find both the Form S and T (and W, in recent years), start talking to your friends' older siblings....everyone gets their booklet back in December/January. But studying for the SAT is simpler....remember, all the Math and CR questions on the PSAT are from old SAT's. The grammar is from SAT-II Writing tests...they just dump those questions little by little onto the PSAT, and now they're dumping them on the SAT. Certainly, the PSAT doesn't test the high end of the Algebra 2 topics, but it's easier to just prepare for both, take them in October (you'll have to prepare for the essay on the SAT of course), and be done, than study twice.</p>

<p>There are a lot of good ideas about these subjects on this forum.
MYTH: If I start my son/daughter studying for the P/SAT too early, it will be bad for their health and whip them into a testing frenzy. It's better to call a tutor the week before the November SAT in their senior year of high school.
FACT: Longer prep is better. Sure, I can cite dozens of examples of students I've met for two tutoring sessions a week before the exam who saw incredible increases, but these students were not only crazy-smart (not just valedictory/4.0GPA at a tough high school, but already doing research, published, etc.), but supremely motivated. Plus, they probably had a good testing day. Also, I tend to work well with high-scorers, which is not the norm in the testing business. Many large companies will say that it is harder to move a high-scorer than a low-scorer. That's because, as so many have already said, that very few tutors know enough about the test to teach a high-scoring student much of anything. However, longer prep is better. If you want to move your CR score, you better start learning vocab and reading 19th century British novels (along with every other type of writing you can get a hold of) NOW! Certainly, you can learn the little actual math and grammar content in less time, and it won't take terribly long to learn what Pearson is looking for in scoring the essay, but you need to practice and analyze both what you've done, and the test itself, to achieve the maximum score gain. Plus, working long-term familiarizes you with all the ins and outs of the exam, and taking full-length practice exams builds up your endurance and desensitizes you to test anxiety. This is true whether you work on your own or with a tutor/class/online course. And if you've got lots of other homework, you'd better start really early. Plus, a National Merit (or College/Corporate Merit) scholarship may only pay for a couple of courses at today's high-priced colleges, but it may help get you into them, and it's one more thing in your life that you can feel proud of.
TUTORING/SELF-STUDY
Reasons to self-study:
You are organized, self-aware, disciplined, have plenty of time to prepare, and don't have the means to pay for any outside help. You have the ability to analyze your own processes, and the design of the test, in extreme detail, then can form a plan to change your understanding and approach to the test.
Reasons to sign up for a class/tutor:
You know you need the structure and accountability that these provide. You want someone to do the above analysis (of the test and you) for you. You realize that a company that has had dozens of people studying the SAT for decades might pass on perhaps a few insights than all those discovered by your 16-year-old brain. Your parents are willing to spend the money. The fact is, college is really expensive today, whether you go to an Ivy, any-size liberal arts, large state college, whatever. You and your parents will be spending a great deal on college. Why not try to get into a more selective one that fits your interests? Weigh the costs of preparation vs. 4 years of tuition (and more, if you plan to continue your education)
Reasons to sign up for a tutor:
You work better in a one-on-one setting. Your score is really low or really high on the test (even retail courses grouped by score can't usually cater to anyone with an outlying score). You have studied on your own for a while or with a course, have actually completed all the work, and still aren't getting anywhere. Your parents have the money to pay for one. You and your clients have spoken to former clients of this tutor and they rave about him/her. Your parents' friends, unprompted, recommended "this great tutor" to you based on results (e.g. Johnny's score went up 900 points from one SAT to the next). See, lots of people think that they're being smart by hiring the cheapest tutor, or buying a classroom course with a "guarantee". The fact is, you get what you pay for. If you get the cheapest tutor, and then meet with them for a zillion hours, and your score doesn't go up, you've actually spent more money with no result (dumb!) than a dozen hours with a really good tutor. And there is no "guarantee". If you take a course, and your score doesn't go up, that company either lets you take another course (gee, thanks), or refunds your money. In other words, they borrowed your money for a couple months at 0% interest, and then gave it back to you. Advertising is deceptive. And it's hard to know who's good at tutoring. There is so much misinformation, or merely basic information. Bizarrely enough, anecdotal evidence is your best bet. If several kids you know (older ones, perhaps) were tutored by someone, and their scores increased a great deal, then that's not a coincidence. Arun Alagapan charges $700 for a 50-minute hour in New York City. He makes clients sign a legal agreement that they will not sue him if the score does not increase. But come on, you don't think anyone would pay that kind of money unless he achieved stellar results with his tutees, do you? Those parents are not paying so much just because it's super-chic to be tutored by Arun.
Other questions to ask a tutor/classroom teacher or their company (this would only apply if you are interviewing a very top-level tutor):
How many times has the teacher taken the actual exam?
Have they been involved in test development or course development (for a company)?
Have they written any books?
Do these books sell?
Have they ever trained anyone in their methods?
Have they ever trained folks who have trained others in their methods?
What is the typical score improvement for a student who does all the assignments and meets over a four-month period for about 20 hours of tutoring? (this way, you have a specific, average benchmark with which to compare prospects). Is it at least 500 points? How many students have seen score improvements of over 700 points?
What does the SAT really measure? (this is to see if the tutor is at least superficially aware of the intent and design of SAT questions)
If someone can give satisfactory answers to all of these questions, then they are worth every penny, as long as you have the pennies. It's perhaps silly, but the fact is, that this will be the most important test you take in your life.
My best friend from elementary school went to Yale undergrad (he turned the Yale Daily News from a half-baked rag into a serious newspaper), Harvard Business School (he did that crazy 16-months-straight degree when they had it), and has worked in a variety of high-powered jobs. When we were 14, we used to be the analytical department for the most successful day-trader of all time. Anyway, when he was applying for a position at a venture capital firm, they not only asked for his resume, academic credentials, and GMAT score (that's the test for admissions to business school), but they asked for......his SAT score. So, in spite of all his various accomplishments since, they still were extremely focussed on that number. Sorry for the long post, but I hope you find it interesting.</p>

<p>Certainly, those who are familiar with a lot of advanced math will be advantaged on a couple questions. But the key to so many questions like this one is to simply rearrange the information, and find the pattern. This is one of the few types of questions on any standardized test to assess inductive, rather than deductive, reasoning. So B=(2+4+6...+100), and A=(1+3+5+...+99). Both series have the same number of terms (50), so just subtract each one. Notice that 2-1= 1, and 4-3=1, etc. 50*1=50, your answer. Like so many SAT questions, computation is always the longest way to go, and in some questions, it's impossible, given the time limit. Simply set up your work and look for the pattern. Less "math"=quicker solution=higher score.</p>

<p>You bring up one more point I forgot to mention about a tutor. If a student is serious about long-term prep, you want a tutor that has all the QAS and PSAT booklets back to 1997, at least. Real materials, and lots of them, is what gets the top students (and the best authors) to a real understanding of the test. Of course, that is necessary, but not sufficient. You need to do a lot more research than just looking at tests to truly understand the psychometrics game.</p>

<p>I just discovered this board yesterday, and am posting because it is so interesting. There is no way I could transmit all of my knowledge to this forum, even if I sat down and wrote for the next two years for free. There are so many things that an expert about many related subfields "knows" that there is no way for them to make a list of it and somehow account for it all. Although I've certainly spent many years trying, with very limited success, to make others in my image who can do some of these tasks to a decent degree, I believe that there is simply no substitute for the variety of experiences I've undergone in regards to testing. Many of our areas of knowledge are cognitive skills that only appear in the situation in which they are needed, and vanish just as quickly. You can't spill out the mind as if it were a vat. It's not that kind of "container", and "container" is a terrible metaphor anyway.</p>

<p>Check schools and public libraries (even via inter-library loan).
The prior edition of REAL Subject Tests (same book as REAL SAT II's) came out in 1996, and has a red/orange/blue/gold cover.
Also, you can get the white with orange, or white with blue College Board Achievement Tests books...they are really not that different, and with the few sources available, what choice is there? These prior editions are the only ones they have ever released. If you know a tutor with a photographic memory who took the Math I, Math II, Writing (and other) tests over 45 times and typed up their notes with all the questions and answer choices verbatim, that would help, too. Perhaps they wrote a book based on their experiences....ETS reuses many of these exams (in fact, there are only a couple of test forms for some of the less popular exams), rotating them from the less popular administrations to the popular ones, before ripping 80% of the questions from two different old exams, adding a few new variations, and putting that "new" test form back into play. It's funny when you take the SAT-II's and notice that the copyright dates on the front of the test booklet go back to 1997;)</p>

<p>You're right that <em>most</em> of what appears in the ETS books are well-edited, and they are certainly better than the tests in any prep book.
Bu ETS routinely makes mistakes in logic, formatting, math notation, and especially their knowledge of correct usage, vocabulary, and grammar rules. Many. These folks aren't content experts in any of those areas, see. So, you shouldn't feel so frightened of the big SAT, just get to know it. And we know that they don't bother to run the scan-tron bubble sheets through their Scan-Tron 8200 more than once to double check the scoring (that would take about 1/2 second for each one, oh my!), or they wouldn't have had all those messed-up scores last year. Unforgivable, for the $700 million a year they pull in for all of this.
I've had several of my students write in to ETS to complain about math errors that occurred on their exams that I've noticed. ETS always denies that there was an error (and with a non-disclosed test form, there is no way to argue), but the next time they re-use that test, the question has been FIXED! So, I guess the emperor is actually wearing clothes if they say so?</p>