Skip to main content
. 2019 Oct 26;50(10):3561–3574. doi: 10.1007/s10803-019-04260-1

Table 3.

Comparison of measures of affect, repetitive behaviour and activity/impulsivity between participants who did and did not engage in self-harm

Measure Subscale Median scores for participants U Score P value Effect size r interpreted through Cohen’s d
Self-harm
N = 20
No self-harm
N = 63
MIPQ-S
 Mood 17.00 20.00 278.00 < .001  .41
 Interest and pleasure 11.500 14.00 428.500 .032  .24
 MIPQ total score 29.00 20.00 344.500 .002  .33
RBQ
 Stereotyped behaviour 5.00 3.00 453.500 .106  .18
 Compulsive behaviour 8.00 3.00 300.500 .001  .36
 Insistence on sameness 4.00 3.00 322.000 .002  .34
 Restricted preferences 7.00 4.00 395.500 .050  .22
 Repetitive language 6.00 4.00 411.500 .074  .2
 RBQ total score 32.00 18.00 313.500 .002  .35
TAQ
 Overactivity 20.5650 10.00 275.500 < .001  .41
 Impulsivity 18.500 11.00 288.500 < .001  .40
 Impulsive speech 9.00 5.00 322.500 .002  .34
 TAQ total score 47.00 25.00 254.500 < .001  .44

Median scores and Mann–Whitney U statistics are reported. Significant differences (p < .01) are highlighted in bold. Where data violated parametric assumptions effect size r was used as an alternative to standard difference statistics and then interpreted with Cohen’s d (Rosenthal et al. 1994; Fritz et al. 2012). For determining the strength of effect size r as interpreted by Cohen’s d, arbitrary cut-offs were assigned as followed: .1 = small, .3 = medium, .5 = large