VE score: Verbal Expression from Word Knowledge and Paragraph Comprehension
Type your WK and PC standard scores. The calculator rebuilds VE with the published formula of the reference form, within about one point, and shows the AFQT it gives.
Estimated VE standard score
51
Within about one point of the VE on your sheet
| Weight of PC versus WK | 0.62 |
| AFQT with this VE | 50 |
| AR + MK + 2VE | 202 |
Estimate: formula of the reference form 04D (Segall 2004, section 2.3). Other forms use slightly different constants, so the printed VE can differ by a point.
Prepared by Radif Partners · Method · official sources read on October 22, 2026 · published formulas and tables
Verbal Expression, printed as VE, is not the average of Word Knowledge and Paragraph Comprehension. The 1997 score scale report defines it as an optimally weighted combination of the unrounded WK and PC standard scores, in which PC carries 0.619 times the weight of WK, then places it on the same scale as the subtests, mean 50 and standard deviation 10. With the constants published for the reference form 04D, WK 50 and PC 50 give a VE of 50, WK 60 and PC 50 give 57, and WK 50 and PC 60 give 54: a point of WK moves VE more than a point of PC. Since VE counts twice in the AFQT sum, each point of VE is worth two points of sum. Other test forms use slightly different constants, so treat the result as an estimate within about one point of the VE on your sheet.
The formula
The 1997 score scale report describes VE in section 2.3. On the older scale, VE was the sum of number-right scores on WK and PC. On the 1997 scale it is built from the two ability estimates of the item response model, weighted to predict best, and then rescaled. For the reference form 04D, the report gives the standard score of WK as 11.03 times its ability estimate plus 47.81, the standard score of PC as 12.35 times its estimate plus 45.89, and VE as 7.23 times the WK estimate plus 5.01 times the PC estimate plus 46.90. The calculator inverts the first two lines to recover each ability estimate from your standard scores, then applies the third and rounds.
| PC 40 | PC 45 | PC 50 | PC 55 | PC 60 | |
|---|---|---|---|---|---|
| WK 40 | 39 | 41 | 43 | 45 | 48 |
| WK 45 | 43 | 45 | 47 | 49 | 51 |
| WK 50 | 46 | 48 | 50 | 52 | 54 |
| WK 55 | 49 | 51 | 53 | 55 | 57 |
| WK 60 | 52 | 55 | 57 | 59 | 61 |
Why it is an estimate
Each form of the test has its own transformation constants, equated to the reference form, and VE uses the unrounded ability estimates, while your sheet only shows rounded standard scores. These two effects can move VE by about one point. That is enough to change the AFQT by one or two percentiles near a threshold, so if your goal sits within a point or two of a branch floor, use the VE printed on your sheet rather than the estimate.
Now put that VE in the AFQT
AFQT percentile
50
| Composite AR + MK + 2VE | 202 |
| VE counts twice | 2 × 51 = 102 |
| Points to reach 65 | 14 on the sum (7 on VE alone) |
How much VE weighs
The AFQT sum is AR + MK + 2 × VE, so VE is half of it. The Army GT is built from AR and VE, the Air Force General composite also uses AR and VE, its Administrative composite uses MK and VE, and its Mechanical composite counts VE twice. The Marine Corps uses VE in its GT and Clerical composites. A weak VE therefore drags down more numbers than any other score on the sheet, and a strong one lifts most of them. The line score calculator shows the effect on every composite at once.
Working on the two parts
Word Knowledge asks for the closest meaning of a word, alone or in a sentence. It rewards a wide vocabulary and the habit of using roots, prefixes and context; the Word Knowledge practice page has questions written for this site. Paragraph Comprehension gives a short passage and asks for its main idea, a detail or an inference; the Paragraph Comprehension practice page trains that skill. Both are timed: on the computer version, Word Knowledge allows about 36 seconds per question and Paragraph Comprehension about 162.
A worked example
A candidate has WK 47 and PC 55 and wonders which to work on. The estimate gives VE 50. Raising WK by four points, to 51, gives 53; raising PC by four points, to 59, gives 52. With AR 50 and MK 50, the AFQT moves from 48 to 54 in the first case and to 52 in the second. The weaker subtest often has more room to grow, and Word Knowledge pays more per point, so here it is the better bet.
VE and the Auto and Shop composite
VE is not the only composite built this way. On the computer test, Auto Information and Shop Information are separate subtests, and the report combines them into AS with equal weights before rescaling to mean 50 and standard deviation 10. The paper version gives a single Auto and Shop subtest. Either way, the AS on your sheet is the number to use in the Air Force Mechanical composite and in the Marine Corps Mechanical Maintenance composite. Like VE, it is a rounded composite: two people with the same AI and SI can see AS differ by a point when their unrounded scores fall on either side of a half.