Skip to main content
. 2024 Aug 21;24:900. doi: 10.1186/s12909-024-05886-1

Table 4.

Rasch reliability and fit statistics for the full PTMQ, the two subgroups identified from multidimensionality assessment of the PTMQ, and each of the six individual subdomains in the PTMQ survey

PTMQ Subgroups PTMQ Subdomains
Parameters Rasch Model Expectations PTMQ Altruisticb Careerc Intrinsic Identified Introjected Career External Amotivation
Misfitting Items Fit residual ± 2.5 logits[a] [19] None - - None None None None None None
Targeting Item Location (logits) Difference in person and item means < 1 [36] 0.00 ± 1.05 0.00 ± 1.21 0.00 ± 0.94 0.00 ± 1.06 0.00 ± 1.54 0.00 ± 1.93 0.00 ± 0.77 0.00 ± 1.60 0.00 ± 1.53
Person Location (logits) -1.46 ± 0.81 -2.57 ± 1.35 -1.07 ± 1.22 -3.17 ± 2.11 -5.47 ± 2.17 -2.34 ± 2.64 -0.50 ± 1.56 -2.79 ± 2.40 -2.60 ± 1.97
Item-trait Chi-square Interaction χ2 - 80.45 67.41 35.59 12.48 3.94 2.46 8.49 3.13 7.44
df - 36 20 16 8 6 4 6 6 6
p p > 0.05a < 0.0001 < 0.0001 0.003 0.13 0.69 0.65 0.19 0.79 0.28
PSI > 0.7 [37] 0.82 0.78 0.83 0.74 0.57 0.72 0.72 0.68 0.61
Item fit residual (logits) 0 ± 1 [19] 0.23 ± 1.26 0.02 ± 2.23 0.20 ± 1.28 -0.15 ± 1.10 0.37 ± 0.51 0.24 ± 0.38 0.21 ± 0.71 0.23 ± 0.23 0.18 ± 0.57
Person fit residual (logits) 0 ± 1 [19] -0.20 ± 1.26 -0.18 ± 0.97 -0.25 ± 1.11 -0.40 ± 0.85 -0.20 ± 0.61 -0.34 ± 0.53 -0.40 ± 0.95 -0.23 ± 0.71 -0.36 ± 0.97
PCA (Eigenvalue for 1st Contrast) < 2.5 [2] [36] 4.93 2.31 2.53 1.84 1.64 1.95 1.66 1.60 1.95
Variance by the first factor > 50% [36] 27.40% 23.10% 31.70% 45.97% 57.75% 97.31% 55.30% 53.32% 64.93%
Differential Item Functioning (DIF) Course - - - - None None None None None None
Sex - - - - None None None None Item 15 (F = 8.16, p = 0.0054) None
Teaching Years - - - - None None None None None None

a. Non-significant Chi-square values using a Bonferroni correction at the p = 0.01 level

b. Intrinsic, Identified, and Amotivation subdomains based on individual item loadings on PC1 from the PTMQ analysis

c. Introjected, Career, and External subdomains based on individual item loadings on PC1 from the PTMQ analysis