Skip to main content
. 2016 Jun 21;47(12):3822–3832. doi: 10.1007/s10803-016-2840-3

Table 2.

Ordered logistic and linear regression models investigating group differences in anxious imagery and emotional valence controlling for age, FSIQ and use of non-anxious imagery

ASD high anxiety (N = 29) Non-ASD high anxiety (N = 17) ASD low anxiety (N = 14) Non-ASD low anxiety (N = 18) Statistical test Group differences
Spontaneous anxious imagery condition
 Other images (% none/1–5/>5) 10/57/33 18/76/6 28/36/36 61/49/0 χ 2 (6) = 28.6, p  0.01 ASDanx = nASDanx = ASD > nASD
 Image frequency (%none/monthly-yearly/weekly/hourly-daily) 17/17/21/45 17/17/23/45 29/29/21/21 50/20/10/20 χ 2 (6) = 15.2, p = 0.02 ‡ (ASDanx = nASDanx) > (ASD = nASD)
 Negative emotional valence (1–63) 34.9 (10.1) 41.9 (14.4) 28.1 (9.3) 28.4 (11.8) β = 1.9, p = 0.24, ω2 = 0.03 ASDanx = nASDanx = ASD = nASD
Suggested anxious imagery condition
 Negative emotional valence (1–63) 30.0 (9.7) 29.6 (14.8) 28.4 (16.6) 18.8 (7.4) β = 3.3, p  0.01
ω2 = 0.06
ASDanx = nASDanx = ASD > nASDш

Linear regression;  ordered logistic regression

Significant group differences at p < 0.01 are presented in bold

шASDanx = ASD high anxiety; nASDanx = non-ASD high anxiety; ASD = ASD low anxiety; nASD = non-ASD low anxiety