Research Article
Development of the Short-Form Yin Deficiency Scale Using Three Item Reduction Approaches
Table 2
Category characteristics of the 7-, 6-, and 5-point Likert scales.
| Scale | Category (response) | Observed count (%) | Average measure | Expected measure | Infit MnSq | Outfit MnSq | Step calibration | Category measure |
| 7-point Likert | 1 (disagree very strongly) | 992 (22) | −0.89 | −0.83 | 0.93 | 1.00 | None | (−2.18) | 2 (disagree strongly) | 1141 (25) | −0.59 | −0.63 | 1.00 | 0.97 | −0.87 | −0.94 | 3 (disagree) | 428 (9) | −0.32 | −0.44 | 1.05 | 0.89 | 0.45 | −0.43 | 4 (neither disagree nor agree) | 667 (15) | −0.25 | −0.26 | 0.97 | 0.97 | −0.79 | −0.08 | 5 (agree) | 567 (12) | −0.12 | −0.08 | 1.10 | 1.16 | 0.00 | 0.31 | 6 (agree strongly) | 554 (12) | 0.09 | 0.12 | 1.08 | 1.17 | 0.04 | 0.98 | 7 (agree very strongly) | 214 (5) | 0.31 | 0.31 | 1.07 | 1.09 | 1.17 | (2.48) |
| 6-point Likert | 1 (disagree very strongly) | 992 (22) | −1.19 | −1.12 | 0.96 | 0.99 | None | (−2.50) | 2 (disagree strongly) | 1141 (25) | −0.74 | −0.77 | 0.92 | 0.87 | −1.08 | −1.10 | 3 (disagree) | 1095 (24) | −0.38 | −0.45 | 0.95 | 0.89 | −0.57 | −0.31 | 4 (agree) | 567 (12) | −0.17 | −0.15 | 1.06 | 1.10 | 0.36 | 0.30 | 5 (agree strongly) | 554 (12) | 0.12 | 0.16 | 1.10 | 1.18 | 0.03 | 1.07 | 6 (agree very strongly) | 214 (5) | 0.38 | 0.46 | 1.16 | 1.21 | 1.26 | (2.56) |
| 5-point Likert | 1 (disagree very strongly) | 992 (22) | −1.37 | −1.31 | 0.96 | 1.01 | None | (−2.61) | 2 (disagree strongly) | 1141 (25) | −0.81 | −0.85 | 0.92 | 0.87 | −1.22 | −1.19 | 3 (disagree) | 1095 (24) | −0.34 | −0.40 | 0.89 | 0.88 | −0.58 | −0.21 | 4 (agree strongly) | 1121 (25) | 0.04 | 0.08 | 1.10 | 1.16 | −0.19 | 1.08 | 5 (agree very strongly) | 214 (5) | 0.52 | 0.59 | 1.13 | 1.10 | 1.99 | (3.17) |
|
|
Disordering response category, where step calibration value became lower than that of the previous category. MnSq, mean square. |