Table 2.
Summary statistics for the four AQ rating scale categories
| Category | Frequency of use (%) | Average AQ person measure (logits) | Threshold measure | Category infit mean square | Category outfit mean square | ||||||
|---|---|---|---|---|---|---|---|---|---|---|---|
| ASD | Non-ASD | ASD | Non-ASD | ASD | Non-ASD | ASD | Non-ASD | ASD | Non-ASD | ||
| 1 | Definitely agree | 1507 (23) | 3750 (34) | −1.77 | −2.17 | None | None | 1.01 | 1.01 | 1.04 | 1.02 |
| 2 | Slightly agree | 1539 (24) | 3850 (35) | −0.49 | −0.59 | −0.25 | −0.84 | 0.93 | 0.94 | 0.90 | 0.90 |
| 3 | Slightly disagree | 1790 (28) | 2333 (21) | 0.47 | 0.61 | −0.11 | 0.08 | 0.89 | 0.98 | 0.86 | 0.97 |
| 4 | Definitely disagree | 1662 (26) | 1017 (9) | 1.80 | 2.14 | 0.36 | 0.76 | 1.05 | 1.05 | 1.08 | 1.09 |
Frequency of use’: the number of persons rated in that category. ‘Average AQ person measure’: observed mean person measure (in logits) in each rating category. ‘Threshold measure’: the difficulty measure between every two adjacent categories. Infit mean square and outfit mean square examine the consistency of use of each rating category; this should not exceed 2.0. (AQ = Autism-Spectrum Quotient, ASD = autism spectrum disorder.)