Verbal Expression (VE) is not a test you sit. The scoring system computes it from your Word Knowledge and Paragraph Comprehension standard scores, roughly 0.655 x WK + 0.406 x PC - 3.03, so a Word Knowledge point moves VE about 60 percent more than a Paragraph Comprehension point. It counts twice because the AFQT is defined as AR + MK + 2VE, which gives your two verbal subtests the same weight as the two math subtests. VE also feeds General Technical, clerical and administrative line scores in the Army, Navy, Marine Corps and Air Force.
On test day you answer Word Knowledge (WK) and Paragraph Comprehension (PC) items. The scoring system turns each subtest into a standard score with a mean of 50 and a standard deviation of 10, then builds VE from those two numbers, the same way it derives Auto and Shop from Auto Information and Shop Information.
The current definition dates from the 1997 score scale. Its report set the weights so that PC kept roughly the smaller share it already had in effect on the 1980 scale, and noted that the smaller weight suits PC's lower reliability. The page on how the AFQT is calculated covers the old 1980 definition and how it changed.
Take about two-thirds of your WK score, add about four-tenths of your PC score, subtract 3, and round to the nearest whole number. Written out, VE is roughly 0.655 x WK + 0.406 x PC - 3.03. The constant is there so that a WK of 50 and a PC of 50 return a VE of exactly 50.
The 1997 report states the formula as equation 2.3, in terms of the underlying ability estimates rather than standard scores. Combining it with the report's Form 04D conversion parameters in Table 2.4 gives the standard-score version used here, with exact constants of 0.654918, 0.405617 and 3.026712. The ratio of the two weights is 0.61934, the figure the report gives for PC relative to WK.
Both columns use the same two scores. Swapping them costs 2 VE points, and the weights account for all of it. With the exact constants the results are 55.24 and 53.24, which round the same way.
A hand calculation can land 1 point away from the VE the scoring system computes. The report builds VE from unrounded WK and PC standard scores and rounds only at the end, while the WK and PC scores you see are already rounded. Job qualification runs on the system's figure, not yours.
VE sits on the subtests' scale, with a mean of 50 and a standard deviation of 10. The 1997 report points out that standard scores on this scale are not cut off at 20 or 80, the way they were on the 1980 scale. Across the practical range of 20 to 80 on both subtests, VE runs from about 18 to about 82.
The weights add up to 1.06, which is why VE spreads slightly wider than its inputs when WK and PC move together. When the two split apart, VE lands between them and closer to WK.
The AFQT composite is AR + MK + 2VE, as defined in section 2.4 of the 1997 report. Arithmetic Reasoning and Mathematics Knowledge each enter once, and VE enters twice, so verbal skill and math skill carry equal total weight. The AFQT page shows how that composite converts to a percentile.
Doubling changes the price of each subtest point. Multiply the VE weights by 2 and you get the value of one standard score point to the composite.
Take AR 52, MK 50, WK 53 and PC 53. VE is 53, the composite is 208 and the AFQT is 56. Raise WK by 3 points and VE becomes 55, the composite 212 and the AFQT 61. Getting the same 4 composite points takes 4 more points of PC, or 4 more points of AR.
The formula sets the price of a point. Your current scores decide how hard each point is to get. If your verbal and math standard scores are close, a WK point buys the most composite and a PC point the least. If math sits 10 points behind, those math points are usually the cheaper ones to collect. Vocabulary also lasts: words you learn in week 2 still count on test day, so start it early. For splitting time across all 4 AFQT subtests, see how to improve your AFQT score.
The Army's own ASVAB page says line scores determine job opportunities, and lists Verbal Expression among the subjects behind 6 of its 10 line scores, including Clerical, General Technical and Skilled Technical. The table below shows the composite definitions in the 1997 score scale report and how far one VE point moves each one. For every composite each Service uses, with or without VE, see ASVAB line scores explained.
The scales differ by Service. The Army rescales GT to a mean of 100 and a standard deviation of 20, as 1.074292 x (AR + VE) - 7.443781, rounded. The line score calculator applies that formula, so its Army GT is the rescaled score.
The Army's other 9 composites use non-integer weights on the same scale. VE carries a weight in all 9, largest in CL and ST, even in the 4 where goarmy.com does not list it. The Marine Corps rescales each raw sum to that same scale. The Navy reports the plain sum, and the official ASVAB site lists the same Navy sums under newer names.
The Air Force reports each composite as a percentile. Appendix C of the 1997 report printed the conversion tables of that time, but neither airforce.com nor the official ASVAB site posts a current one, so the table above shows only the change in the raw index.
Two applicants share AR 52 and MK 50, with every other subtest at 50. Applicant A has WK 62 and PC 44. Applicant B has the mirror image: WK 44 and PC 62. Both have 106 points across the two verbal subtests.
A strong vocabulary covers for weak reading far better than the reverse. Applicant A's PC of 44 has the most room to grow, and every 5 PC points adds about 2 VE points. Applicant B's weak subtest is also the heavier one: each WK point is worth about 60 percent more VE than a PC point, so vocabulary is the obvious lever. Both should put their hours into the weaker subtest. The weights only set how much each point gained returns.
Word Knowledge, in the official description, measures the ability to identify the best synonym for a given word. Synonym questions reward knowing a word well enough to recognize it in a new setting, and roots are the fastest way to do that for words you have never studied. Learn roots and prefixes in families: bene (benefit, benevolent, benign), mal (malice, malign, malady), chron (chronic, chronology, synchronize). Then test yourself on synonyms, not definitions.
To work a family at a time, type a root such as bene into the search box of the ASVAB vocabulary trainer, which lists the practice words that contain it. Short daily sessions beat a weekend cram, because a word has to be recognized quickly to be useful in a 9-minute subtest.
Paragraph Comprehension measures the ability to obtain information from written passages. The habit that pays most is finding the sentence that supports your answer before you choose it. Wrong choices are often true in general, or true of a similar passage, but not stated or implied in this one. Sort each question into stated fact, main idea or inference, and check the passage for the right kind of support.
Work timed sets in reading practice and write down, for every miss, whether you misread the question, skipped the evidence or chose an answer the passage never supported. That error log tells you which of the three to fix next.
The two verbal subtests run at very different speeds. The table gives the official limits without tryout items. Each examinee also gets tryout items in 2 to 4 subtests, and those subtests run on a longer limit.
On the CAT-ASVAB you cannot change an answer once you move to the next question, so settle each item before you submit it. Running out of time is costly: unfinished items are scored as if you had guessed at random, and the penalty grows with the number left. On the paper test, unanswered items count as wrong. Practice at the per-item pace above until you finish with time to spare every time. Then put the slack into reading PC passages carefully instead of racing.
A workable week has 5 short vocabulary sessions of 20 to 30 minutes, 2 or 3 reading sets checked against the evidence, and 1 timed WK and PC check under real limits. Track the subtest standard scores from full practice tests rather than raw percentages, and move hours between WK and PC as the gap between them changes.