Skip to main content
. 2022 Aug 20;15(10):1917–1928. doi: 10.1002/aur.2794

TABLE 3.

Polynomial regression with response surface parameters predicting autism‐spectrum quotient

Autistic individuals Non‐autistic individuals
Effect Estimate (SE) p value β Estimate (SE) p value β
CE −3.62 (0.46) 5 × 10−15 −0.265*** −3.15 (0.13) 5 × 10−115 −0.19***
EE −0.5 (0.37) 0.17 −0.03 −1.14 (0.14) 3 × 10−16 −0.07***
CE2 0.18 (0.22) 0.425 0.027 1.6 (0.13) 5 × 10−36 0.125***
EE2 0.235 (0.14) 0.1 0.03 0.44 (0.12) 0.0002 0.04***
CE × EE 0.265 (0.22) 0.24 0.03 −0.45 (0.14) 0.0015 −0.03**
Response surface parameters Group comparison (p value)
a1 −4.13 (0.52) 3 × 10−15 −0.295*** −4.29 (0.15) 1 × 10−163 −0.26*** 0.76
a2 0.68 (0.24) 0.036 0.087* 1.59 (0.14) 1 × 10−28 0.135*** 0.001**
a3 −3.11 (0.65) 2 × 10−6 −0.235*** −2 (0.14) 3 × 10−18 −0.12*** 0.11
a4 0.15 (0.44) 0.74 0.027 2.5 (0.29) 8 × 10−18 0.195*** 0.00001***

Notes: Parameters of the polynomial regression with response surface analysis of emotional empathy (EE) and cognitive empathy (CE), predicting Autism‐Spectrum Quotient score in autistic and non‐autistic individuals. The response surface parameters (a1–a4) are shown with a1—the linear effect of total empathy, a2—the curvilinear effect of total empathy, a3—the linear effect of empathic disequilibrium, and a4—the curvilinear effect of empathic disequilibrium. *p < 0.05, **p < 0.005, ***p < 0.0005.