Writing the new SAT writing section (slog, slog, slog)

Writing the new SAT writing section (slog, slog, slog)

So I’m in the middle of rewriting the workbook to The Ultimate Guide to SAT Grammar. After the finishing the new SAT grammar and reading books, I somehow thought that this one would be easier to manage. Annoying, yes, but straightforward, mechanical, and requiring nowhere near the same intensity of focus that the grammar and reading books required. Besides, I no longer have 800 pages worth of revisions hanging over me — that alone makes things easier. 

However, having managed to get about halfway through, I have to say that I’ve never had so much trouble concentrating on what by this point should be a fairly rote exercise. Even writing three or four questions a day feels like pulling teeth. (So of course I’m procrastinating by posting here.) 

In part, this is because I have nothing to build on. With the other two book, I was revising and/or incorporating material I’d already written elsewhere; this one I have to do from scratch. I’m also just plain sick and tired of rewriting material that I already poured so much into the first time around. 

The problem goes beyond that, though. 

I’ve recently come to realize is that the tests I’ve written in the past were comparatively easy to construct because the test they were based on were themselves well-written. Sure, the current workbook was a drag (and don’t even get me started on the trauma of having Word spontaneously decide to create multiple pages 45s — I still flinch whenever I see that number), but writing for the new SAT has made me realize just how much I took for granted. I’ve never actually had to mimic a poorly written test.

Although focusing on The Skills That Matter Most sounds like a lovely idea in principle (who would want students tested on skills that don’t matter?), it’s a little less lovely in practice — or at least in the sort of practice that the writers of the new test have cooked up.

As I’ve written about before, one of the most salient features of the new test is that the majority of the questions focus primarily on a “core” group of skills or topics, with a much smaller number of questions testing a much wider range of topics. The major problem with this setup is that students who are aiming for high scores will need to spend a significant amount of time studying topics that stand a relatively small chance of actually appearing on the test.

To use the type of rhetoric normally applied to SAT vocabulary, students will now be required to devote lots of time “memorizing obscure rules that they are unlikely to ever be tested on.”

I knew this in theory before, but I didn’t fully realize how it would play out on a practical level until I started writing full-length sections.

As has been extensively remarked upon by now, the new SAT multiple choice writing section is essentially a cheap ripoff of the ACT English section — kind of like a fake designer handbag. At first glance, it seems pretty much the same, but when you start to look more closely, you can see that the quality just isn’t as good. At some level, I get the sense that even the people writing the test think it’s bullshit. There’s a carelessness to it that I’ve never felt on another exam. (Cat portraits? Seriously?)

Another big issue involves length. Given the fact that all of the reading is now lumped together in a single, seemingly interminable 65-minute section (nearly twice as long as the ACT reading action) and the writing section will immediately follow the reading section, it is understandable that the writing section would be made somewhat shorter. (As a side note, having to do two verbal sections consecutively is going to be a major turnoff for all but the strongest readers). The problem is that with only 44 questions, 12 of which must involve inserting/deleting information and another 1-3 of which must involves graphics of some sort, there simply isn’t room to test the full range of concepts. As a result, there’s a constant tradeoff: some concepts that I would argue are extraordinarily important — comma splices, for example, or verb tense — are given very short shrift.

Furthermore, with room for only one or two questions in a given category per test, there’s very little way to test concepts at various levels of difficulty. To take just one example, the current SAT writing section focuses heavily on subject-verb agreement, with at least three and as many as five questions of varying difficulty levels appearing throughout the test. Most of these questions test that concept in fairly predictable ways: students can be reasonably confident, for example, that they will encounter sentences that place non-essential clauses and prepositional phrases between subjects and verbs, as well as sentences with compound subjects (two singular nouns connected by “and”). Not coincidentally, these are the structures that students are most likely to employ in their own writing. Because of the frequency with which these structures are tested, students have very good reason to spend time mastering them.

Based on the material released by the College Board, however, the new test appears to test subject-verb agreement no more than two times per test, and sometimes not at all. Furthermore, there is no way to tell what forms those questions will take when they do appear: they could be  as pedestrian as subject-prepositional-phrase verb, or they could be as out-there as “that/whether/what” as subjects — a construction that few people except professional writers (or pretentious Harvard Crimson reporters) use.

But, you say, isn’t that a good thing? Won’t that make the test harder to “beat?” Well, yes, it will make the test less predictable, but predictability is not always a negative. Here’s the thing: the basics of subject-verb agreement are really, really important. Even if a student struggles with some of the more complex variations, they should at least understand how to avoid the most common errors involving it by the end of high school.

Now, given that these common constructions may or may not be tested, students will have far less incentive to study them to the point of mastery. The fact that something is straightforward to learn does not mean that it is isn’t important, or that it shouldn’t be tested. A test that truly focused on “the skills that mattered most,” or one that genuinely gave students a clear blueprint of what to study, simply would not operate this way. 

The ACT English test, in contrast, has 75 questions — that’s plenty of room to test a much fuller range of topics in various ways, from a variety of standpoints, and with varying degrees of difficulty. There are numerous rhetoric questions, but they feel far more balanced with the grammar questions. The writers of that test clearly understand that both are important. Looking back, when I’ve written mock ACT English section, I’ve never felt as if I had to pick and choose; there always seemed to be ample opportunity to include all of the question types I wanted to include. The same for creating mock SAT Writing tests: there was plenty of room to incorporate everything, and I never felt as if I was having to make constant judgment calls about what to test.

Now, though, I find myself writing and rewriting, shifting things around, trying to figure out what’s important enough to include. Decisions that seem like they should be simple and straightforward are instead tedious and drawn-out. The problem is certainly compounded by test-writing fatigue, and by my tendency toward perfectionism, but still, it seems like this shouldn’t be so hard. 

Furthermore, trying to capture the ACT’s earnest, down-home, politically correct hokeyness may be an eye-rolling experience, but at least it’s never made me feel like a shill for the tech industry.

The College Board’s prevailing philosophy seems to be that making students read vapid passages that basically scream “LOOK, WE’RE BEING RELEVANT!” and that present tired clichés about technology and the new economy, with an occasional bone tossed condescendingly to the humanities just so no one can accuse them of being total Philistines (look, it’s ok to take some philosophy course because they can help you make money!), will somehow make students “college and career ready.”

To be perfectly blunt, that is one of the stupidest ideas I’ve ever encountered. Only people who live in an echo chamber completely divorced from reality could buy into — or think that students would buy into — something so patently condescending.

It’s also such a blatantly desperate ploy to recapture market share that it comes off as more than a little pathetic. (Not that I’m feeling sorry for anyone — you ruin the test, you suffer the consequences). You can just picture the strategy sessions at which it was decided that “Relevant!” would be the official buzzword of the redesigned SAT, to be repeated ad nauseum lest anyone continue to think that the good people at the College Board were somehow out of touch. 

Unfortunately, there doesn’t seem to be a way of creating accurate materials that don’t reflect this particular ideological bent. In order to resolve this moral quandary and assuage my conscience, I’m thinking that I may need to put a disclaimer in the book.

Notice: the views expressed this work are intended to represent those held by the technology-happy testing companies and education “reformers” responsible for the installation of Bill Gates’s lackey as the head of the College Board and the redesign of the SAT. In no way should they be taken to reflect my personal opinions. 

I’m only half-kidding.

What parts of speech can be subjects?


Nouns are the most common type of subjects. They include people, places, and things and can be concrete (book, chair, house) or abstract (belief, notion, theory).

Example:  Bats are able to hang upside down for long periods because they possess specialized tendons in their feet.



Pronouns are words that replace nouns. Common pronouns include she, he, it, one, you, this, that, and there.

Less common pronouns include what, how, whether, and that, all of which are singular. They are typically used as part of a much longer complete subject (underlined in the second example below).

Example: They are able to hang upside down for long periods because they possess specialized tendons in their feet.

Example: How bats hang upside down for long periods was a mystery until it was discovered that they possess specialized tendons in their feet.



Gerunds are formed by adding -ING to the ends of verbs (e.g. read – reading; talk – talking). Although gerunds look like verbs, they act like nouns. They are always singular and take singular verbs.

Example: Hanging upside down for long periods is a skill that both bats and sloths possess.



The infinitive is the “to” form of a verb. Infinitives are always singular when they are used as subjects. They are most commonly used to create the parallel structure “To do x is to do y.”

Example: To hang upside down for a long period of time is to experience the world as a bat or sloth does.


No semicolon before “and” or “but”

Shortcut: semicolon + and/but = wrong 

If you see an answer choice on either the SAT or the ACT that places a semicolon before the word and or but, cross out that answer immediately and move on.

Why? Because a semicolon is grammatically identical to a period, and you shouldn’t start a sentence with and or but.

The slightly longer explanation: In real life, semicolon usage is a little more flexible, and the choice to use when can sometimes be more a matter of clarity/style than one of grammar. It is generally considered acceptable to place a semicolon before and or but in order to break up a very long sentence, especially when there are already multiple commas/clauses.

For example:

Pamela Meyer, a certified fraud examiner, author, and entrepreneur, became interested in the science of deception at at business school workshop during which a professor detailed his findings on behaviors associated with lying; and she subsequently worked with a team of researchers to survey and analyze existing research on deception from academics, experts, law enforcement, the military, espionage and psychology.

In the above sentence, either a comma or a semicolon could be used before and. In this case, however, the sentence is so long and contains so many different parts that the semicolon is a logical choice to create stronger break between the parts.

Why not just use a period? Well, because a semicolon implies a stronger connection between the clauses than a period would; it keeps the sentence going rather than marking a full break between thoughts. Again, this is a matter of style, not grammar.

The SAT and the ACT, however, are not interested in these details. Rather, their goal is to check whether you understand the most common version of the rule. Anything beyond that would simply be too ambiguous.

Because, since, and commas

Because, since, and commas

So after months of living under a rock in a desperate attempt to finish my revised SAT books in time for the summer stampede, I’m finally taking a few tentative steps into the sunlight.

Several of you have written to me lamenting my lack of acerbic commentary on the June SAT scoring scandal, the new SAT, and other educational embarrassments, and I promise you that I am well aware of that shortcoming. Rest assured that I do in fact have many, many things to say; I’ve merely had to devote every last brain cell to rewriting my books for the last half-year, leaving me precious little room for other thoughts, acerbic or otherwise. I’m hoping to begin posting some commentary here in the next weeks.

For the moment, though, I want to address a topic that a couple of people have written to me about recently, namely the “no comma before ‘because’ or ‘since’ rule.”

Actually, it’s not really a “rule” per se. It’s more of a guideline. In the universe of SAT grammar errors, it’s a pretty minor one, and for that reason it’s pretty certain that an answer’s correctness/incorrectness will never depend exclusively on it. It’s merely a secondary consideration.

Let me back up and explain.

There are two main types of conjunction: coordinating and subordinating.

Coordinating conjunctions (for, and, nor, but, or, yet, so) are also known by the acronym FANBOYS. They can be used to join two independent clauses (complete sentences) and must always follow a comma when employed in this way.

For example: The city’s government has curtailed spending on all non-essential services, so the new theater company has been forced to suspend its productions for lack of funding.

A FANBOYS conjunction can only be used at the start of the second clause; it cannot be used at the start of the first. Even if you don’t know the grammatical rule, you should be able to recognize that this usage does not make any sense.

Incorrect: So the new theater company has been forced to suspend its productions for lack of funding, the city’s government has curtailed spending on all non-essential services.

Subordinating conjunctions on the other hand, can be used to start either the first or the second clause. Common subordinating conjunctions include becausewhilesincewhenuntil, and unless.

Now, here’s the rub. If you are being absolutely, technically correct, a comma should be used to separate two clauses when the clause begun by the subordinating conjunction comes first.

Correct: Because/Since the city’s government has curtailed spending on all non-essential services COMMA the new theater company has been forced to suspend its productions for lack of funding.

When the clause begun by the subordinating conjunction comes second, no comma should be used.

Correct: The new theater company has been forced to suspend its productions for lack of funding NO COMMA because/since the city’s government has curtailed spending on all non-essential services.

Incorrect: The new theater company has been forced to suspend its productions for lack of funding COMMA because/since the city’s government has curtailed spending on all non-essential services. 

That’s the strict version of the rule. The more complicated truth is that it depends on context; if a sentence is very long, or a comma before because is necessary to make the meaning of the sentence clear, then it’s acceptable to that construction.

From what I’ve observed, the College Board/ETS/whoever it is setting the rules tends to adhere to the strict version. That is, answers that contain the construction “comma + because/since” are pretty much always wrong. That said, wrong answers with that construction almost always contain an additional problem that makes them absolutely, incontrovertibly wrong. So maybe it’s more correlation, not causation.

A while back, someone contacted me about a question in the online program that violated that rule; an answer with the construction “comma + because” was correct. That did make me rethink things, but only up to a point. I’ve found slight inconsistencies between the online tests and real tests in the past, and I’ve never seen a question from an administered test that broke this “rule” (although if anyone reading this post has seen one, please let me know!).

My suggestion, therefore, is to work from the standpoint that any answer with the construction “comma + because/since” is probably wrong, and only reevaluate that assumption if there’s a clear grammatical problem with every other answer. To be fair, the SAT does break its own “rules” sometimes, and this one isn’t particularly hard and fast. But still, you’re better off playing the odds and working from there.

Pre-publication copies of “The Ultimate Guide to SAT® Grammar,” 3rd Ed., are now available through CreateSpace

For those of you who are just so eager to get started studying for the new PSAT Writing section that you can’t wait another few weeks, I am offering pre-publication copies of the new edition of The Ultimate Guide to SAT® Grammar through CreateSpace, for the reduced price of $19.95 (nearly 40% percent off the Amazon list price).

To purchase, go to https://www.createspace.com/5464532.

Hard questions, easy answers (or: when you hear hoofbeats, think horses, not zebras)

Perhaps you’ve heard the saying “when you hear hoofbeats, think horses, not zebras.” If you’re not familiar with the expression, it means that when searching for an explanation, you should always consider obvious possibilities before thinking about more unlikely options. Whenever I tutor the Writing section of the SAT, I find myself uttering these words with inordinate frequency.

I’ve worked with a number of students trying to pull their Writing scores from the mid-600s to the 750+ range. Most have done well on practice tests but then unexpectedly seen their scores drop on the actual test. Unsurprisingly, they were puzzled by their performance on the real thing; they just couldn’t figure out what they had done differently. And at first glance, they did seem to know what they were doing. When I worked very carefully through a section with them, however, some cracks inevitably emerged. Not a lot, mind you, but just enough to consistently pull them down. They would be sailing along, identifying errors like there was no tomorrow, when all of the sudden they would hit a question whose error (if any) they simply could not identify.

When that happened, they would stop and read the sentence again. And when they couldn’t hear anything wrong, they would read the sentence again, slowly, trying to hear whether something was wrong (mistake #1). Then, if they really didn’t want to choose “No error” but weren’t sure whether something was truly wrong, they would start searching for an explanation, usually a somewhat convoluted one, for why a perfectly acceptable construction was ambiguous or awkward or otherwise wrong (mistake #2). Almost always they did so when the actual answers — answers based on concepts they understood perfectly well — were staring them right in the face.

One of things that it’s easy to forget — or, in the case of many natural high-scorers who haven’t needed to study the framework of the test, to never realize — is that “hard” questions are not necessarily hard because they test hard concepts. Most often, they are hard either because they test (relatively) simple concepts in hard ways or because they combine concepts in unexpected ways.

Hard questions can — and often do — have “easy” answers. That does not mean that the answer is the option that sounds weird (that’s the distractor answer). It does, however, mean that the answer is likely to be an extremely simple word like “is” or “are” or “it.” It also means that the answer probably involves an extremely common error, like subject verb agreement or pronoun agreement, not some obscure rule you’ve never heard of.

The challenge is figuring out which concept is being tested, not understanding the concept itself. So when people who can usually hear the error come across a question whose answer they don’t instantly hear, their instinctive reaction is to look for something outlandish to be wrong with it, not to think systematically about what the most common errors are and check to see whether the question contains them. In others words, they hear hoofbeats and imagine that a herd of zebras is about to come racing around the corner.

For example, consider the following question, which a very high-scoring student of mine recently missed:


(A) Thanks to the strength (B) of the bonds between (C) its

constituent carbon atoms, a diamond has exceptional

physical properties (D) that makes it useful in a wide

variety of industrial applications. (E) No error


If you spotted the error immediately, great, but bear with me for illustrative purposes. The sentence itself is rather challenging: it discusses a topic (chemistry) that many students are unlikely to have unpleasant associations with, and it also contains the word “constituent,” which many weaker readers will have difficulty decoding, and whose meaning many slightly stronger readers will not know or be able to figure out. So right there we have two big stumbling blocks likely to distract from the grammar of the sentence. Many test-takers are also likely to think that “Thanks to” sounds too casual and would be considered wrong on a serious test like the SAT. Many other test-takers are likely to just not hear any error.

In that case, the most effective approach is to consider the structure of the test. The most common error is subject-verb agreement, and when in doubt, it’s the error you should always check first. There is exactly one underlined verb in the sentence: “makes.” It singular (remember: singular verbs ends in “-s”), which means that it’s subject must be singular as well.

But what is the subject? “Physical properties,” which is plural, so there’s a disagreement. The answer is therefore (D). The sentence should read “…physical properties that make it useful.”

The moral of the story is that if you don’t spot an error immediately, whatever you do, don’t fall into the loop of endlessly rereading the sentence and trying to figure out whether something sounds funny. Instead, check systematically for the top five or so errors: subject-verb agreement, pronoun agreement (check “it” and “they”), verb tense (pay attention to dates and “time” words), adjectives vs. adverbs (easy to overlook), and, if you’re at the end of a section, faulty comparisons.

If all of those things check out, the sentence is probably fine.