ASVAB Word Knowledge: The Subtest People Under-Study

Word Knowledge carries more weight in the AFQT than either math subtest, and it is the subtest that receives the least study time. The practice books put their thickest chapters on algebra and give vocabulary a word list. The scoring formula runs the other way.

The mechanism is the AFQT formula, and it is the most useful thing on this page. Word Knowledge is combined with Paragraph Comprehension into one score called Verbal Expression (VE), and the AFQT formula is 2 x VE + AR + MK. Verbal Expression is doubled. Arithmetic Reasoning and Mathematics Knowledge are not. Reading and vocabulary together are worth half of the score that decides whether you can enlist at all, and Word Knowledge is half of that half.

The short version:

• 15 questions, 9 minutes on the computer version. That is 36 seconds each, the fastest pace on the test
• Feeds the AFQT through Verbal Expression, which is double weighted
• Feeds the Administrative and General line scores
• Two question forms: a word given alone, and a word used in a sentence
• Roots, prefixes and suffixes are the highest-yield thing to study, because they work on words you have never seen
• The most common wrong answer is the opposite of the right one, and it is there on purpose

What Word Knowledge actually tests

Nothing but vocabulary. There is no grammar, no spelling, no punctuation and no sentence correction on this subtest. You are shown a word and four candidate meanings, and you pick the closest one.

That makes it unusual on the ASVAB. Most subtests give you something to work out. This one does not. You either know the word or you do not, and the small amount of work available to you is in eliminating choices and in reading the word's parts. Treat it as a recall test, because that is what it is, and plan your time accordingly.

The two forms look like this:

The sentence form carries information the bare-word form does not, and it is routinely left unused. A sentence shows whether the word is doing something positive or negative, whether it is a thing or an action, and often roughly how strong it is. With no recall at all, that is usually enough to eliminate two of four choices.

Where the score goes

Word Knowledge never reaches a composite on its own. It goes through Verbal Expression first:

Two composites and a doubled AFQT contribution. Compare that to the way most people budget their study time:

If you have four weeks and you are spending all of them on fractions, you have mis-allocated. An hour a day on vocabulary for two weeks moves the AFQT further than an hour a day on algebra, for most people, because most people start from a lower base on words and because the words are doubled.

The pace: 36 seconds a question

Nine minutes for fifteen questions is not a lot, and it is deliberately not a lot. The test is not asking you to reason your way to a meaning. It is asking whether the word is already in your head.

That pace has one practical consequence: deliberation returns nothing here. Sixty seconds on a word you do not know produces the same answer as ten seconds on it. Eliminate what you can, choose, move on. There is no penalty for a wrong answer on the ASVAB, so a guess is strictly better than a blank.

Roots, prefixes and suffixes: the highest-yield study

A memorized word list only helps if the test shows you a word from your list. Word parts work on words nobody could have predicted, which is why this is where to start. This is the table to memorize:

The parts assemble: benevolent is good plus will, wishing well. Circumvent is around plus come, get around. Verbose is word plus full of, wordy. The method will not decode every word, but it decodes enough to convert blind guesses into two-way guesses, and across fifteen questions that is worth real points.

The rest of what to study, in order

1. Word parts. The table above. Two or three sittings and you own them, and they never stop paying.
2. A word list, in short daily sessions. Twenty minutes a day for three weeks beats seven hours the weekend before, and it is not close. Vocabulary is stored by repetition over time, which means cramming is close to the worst possible method for this one subtest.
3. Use the word in your own sentence. Not a definition you copy. A sentence about your own life, out of your own head. Writing "the recruiter was candid about the wait" fixes candid in a way that "candid: frank, honest" does not.
4. Read things that are not written for you. Any nonfiction written for adults: a news site, a long magazine feature, the enlistment documents you are about to sign. Reading widely is the slow method and the only one that also lifts Paragraph Comprehension, which is the other half of the same doubled score.
5. Learn synonym clusters, not single words. Gregarious, affable, convivial, garrulous live near each other. Learning them together builds the sense of closeness the test is actually measuring.

Notice what is not on the list: grammar drills, spelling practice, and learning to pronounce anything. None of it is tested.

The five places people lose points

Every one of these is a decision error, not a knowledge gap, which means every one of them is free to fix.

Choosing the antonym. The test writers reliably put the opposite of the correct answer among the four choices, because a half-remembered word carries a sense of the subject without the sense of the direction. Laud is clearly about how something is spoken of, which makes criticize plausible when the answer is praise. Before you commit, check whether your choice is the meaning or the mirror of it.

Choosing the word that sounds similar. Prompt: ingenuous. One of the choices will point at ingenious. Prompt: affect. A choice will point at effect. Sound-alike distractors are among the most common on the subtest, and they work on anyone moving fast, which on a 36-second clock is everybody. One deliberate half-second on the letters, not the sound, defeats them.

Picking the most impressive-looking option. The correct meaning of a hard word is frequently a plain one. Abate most nearly means lessen. Nothing exotic. An option selected because it is unfamiliar is a guess, not a deduction, and the difficulty of the prompt word says nothing about the difficulty of the answer.

Overturning a correct first answer. This subtest punishes second-guessing more than any other, because the first reaction is recall and the second is rarely more than doubt. A word known on sight should be answered and left. Change an answer only when you can say what specifically was wrong with the first one.

Expecting an exact match. "Most nearly means" asks for the closest of four, not a definition out of a dictionary. The correct answer is regularly an approximate fit, and an approximate fit is not evidence of a wrong choice. It is what "most nearly" means, and applicants talk themselves out of correct answers over it every test day.

How to work one under time pressure

Thirty-six seconds, spent in this order:

1. Read the word, and answer it in your head before you look down. If a meaning arrives, check that one choice matches it and go. Five seconds, and that is most of the questions you actually know.
2. If nothing arrives, look for the parts. Prefix, root, suffix. They often give the direction even when they do not give the meaning: good or bad, toward or away, more or less.
3. If there is a sentence, use it. Decide whether the word is approving or disapproving, and whether it is a noun, a verb or an adjective. Cut any choice that is the wrong part of speech or the wrong tone.
4. Eliminate, do not select. On a word you do not know, the useful question is which choices are impossible, not which one is right. Two eliminations turn a 25 percent shot into a 50 percent one.
5. Choose and leave. Answer everything, and move on immediately, because on the computer version there is no going back anyway.

The habit to build in practice is the leaving. Steps one through four are already within most applicants' reach. What costs them the subtest is ninety seconds spent on question four and then a rushed run from eleven to fifteen.

Question counts and timing, both versions

Nearly everyone now takes the computer-adaptive test at MEPS. The paper version still runs at some Mobile Examining Team sites and in high school testing.

The computer version adapts: answer correctly and the next word gets harder. A stretch of difficult words is therefore an indication that you have been answering correctly. Do not read rising difficulty as failure, and do not let a hard run late in the subtest change how you work.

You cannot skip a question, and you cannot return to one. Every question is answered when you meet it, which again argues for the eliminate-choose-move discipline rather than any plan to come back.

What counts as a good score

Word Knowledge is reported as a standard score where 50 is average and roughly two thirds of test takers land between 40 and 60. On its own that number does nothing. It matters once it is combined with Paragraph Comprehension into Verbal Expression, and Verbal Expression is what your Administrative and General line scores and your AFQT are built from.

So the target is not a Word Knowledge score. It is the Administrative or General minimum for the job you want, and those run from the high 30s into the 60s depending on the career field. The table further down this page lists every enlisted Air Force career field with an Administrative or General requirement, with the number it demands. Find your job, find your number, and work back from there.