Table 3. Performance on mathematical tasks by patients.
| Test | S.A. | S.O. | P.R. |
|---|---|---|---|
| Estimation test (maximum 20) | 20 | 19 | 20 |
| Calculation tests (maximum 20) | |||
| Addition | 19 | 16 | 20 |
| Subtraction | 19 | 19 | 19 |
| Multiplication | 19 | 13 | 17 |
| Division | 19 | 11 | 16 |
| Adding and subtracting fractions (maximum 30) | 27 | 27 | 20 |
| Multiplication (maximum 36) | |||
| Easy known tables (time, sec) | 36 (115) | 36 (158) | 36 (74) |
| Hard known tables (time, sec) | 35 (208) | 23 (537) | 31 (127) |
| Novel tables (time, sec) | 36 (508) | 32 (967) | 33 (313) |
| Reversibility (maximum 40) | |||
| Subtraction | 40 | 35 | 37 |
| Division | 37 | 34 | 38 |
| Number infinity (maximum 30) | 30 | 29 | 19 |
| Bracket expressions | |||
| Calculation accuracy | 45/64 | 52/64 | 43/64 |
| Serial order errors | 4 | 1 | 2 |
| Bracket generation and calculation | 4/5 | 4/5 | 2/5 |
The estimation test assessed nonsymbolic quantity estimation and involved marking a given value on a vertical line. The calculation tests examined the integrity of all basic arithmetic functions using whole-number problems. The fractions test assessed the ability to add and subtract fractions. The multiplication test examined the patient's ability to perform known and novel multiplications. The known problems were drawn from the 1–12 multiplication tables and were subdivided into “easy” (× 2, 5, and 10) and “difficult” (× 7, 8, and 9) tables. The novel multiplication problems were tables that had not been rote learned (× 13, 15, and 18). Responses were scored for accuracy (out of a maximum possible of 36) and timed in seconds. The reversibility tests involved 40 paired subtraction and division problems, with the larger integer appearing in first position in half of the problems and in the second position in the remaining half (i.e., 59–13 and 13–59). The number infinity tasks required the participant to generate 10 numbers that were bigger than n but smaller than n + 1 (two trials) or smaller than n but bigger than n - 1 (one trial). The bracket expressions involved two tasks. First, the patient was required to calculate the result of 64 expressions containing syntactic brackets [e.g., 36 ÷ (3 × 2)]. If the participant failed to understand the embedded nature of the expression and calculated the result from a left-to-right serial order strategy [i.e., 36 ÷ (3 × 2) = 24], this was recorded as a serial order error. Second, the participant was presented with five identical unbracketed numerical expressions (e.g., 7 + 4 × 3 + 17 =). The participant was asked to mark up expressions with brackets and to calculate the result. If the participant was able to produce at least two different results for the string, he was credited with passing the block. There were five blocks in this test.