Autism and Types of Empathy

 fThis is another SPSS post. For this one, the data was gotten from a thread on the Wrong Planet forum, about this test (Word document) assessing four different kinds of empathy - Fantasy, Perspective Taking, Empathetic Concern and Personal Distress.

Twenty eight people posted their scores on the forum (including myself). One was excluded from this analysis because her profile described her as the family member of an autistic person instead of describing her as autistic. The remaining 27 were 11 males and 16 females (anecdotal observation suggests a more even gender ratio of autistic posters on forums than is found in epidemiological studies, probably suggesting that women are more likely to seek out an autistic community). According to their profiles, 8 were diagnosed with Asperger Syndrome, 4 were diagnosed with another autism spectrum condition, 8 were self-diagnosed autistics and 7 were unsure if they were autistic or not.

Firstly, the accuracy of self-diagnosis has been questioned, and the unsure individuals could turn out to have conditions other than autism as well. So I compared diagnosed to undiagnosed on all four subtests, and found no significant differences (fantasy p=.836, perspective-taking p=.740, empathic concern p=.188 and personal distress p=.699). In addition, a one-way ANOVA of all four diagnostic groups found no significant differences, though the 'other ASD' group was very small so this has limited accuracy.

Average scores for the whole group were (comparison scores in brackets taken from the thread):

• Fantasy - 16.19+/-7.67 (average scores: 15.73 for men; 18.75 for women)
• Perspective Taking - 9.85+/-7.06 (average scores: 16.78 for men; 17.96 for women)
• Empathic Concern - 17.19+/-6.63 (average scores: 19.04 for men; 21.67 for women)
• Personal Distress - 16.22+/-6.17 (average scores: 9.46 for men; 12.28 for women)

As the comparison averages indicate, men and women tend to differ on multiple subtests. So I ran a T-test comparing male and female autistics next. Only Fantasy showed a significant difference at p=.002, with women scoring higher (19.69 vs 11.09). This is consistent with Lord et al (1982)'s finding that autistic boys showed more stereotypic play than autistic girls.

Then I ran a one-sample T-test to compare the participants' scores to the previously-reported averages. For Fantasy, since it showed gender differences, I compared males and females separately to their own gender's means. For the others, I took the average of the two gender means for comparison. The results were:

• Fantasy - male p=.068 (almost significant), female p=.778 (not significant)
• Perspective Taking - p>.001 (highly significant)
• Empathic Concern - p=.020 (significant)
• Personal Distress - p>.001 (highly significant)

So, autistics score lower on Perspective Taking, somewhat lower on Empathic Concern and higher on Personal Distress. Perspective Taking deficit is described as a central feature of autism spectrum conditions, so that result is not surprising. While all other score correlations were non-significant, Empathic Concern and Perspective Taking scores had a correlation of .591 (two-tailed p>.001), which explains the lowered Empathic Concern scores. Higher Personal Distress is in contrast with the findings of Corona et al (1998), who found that autistic children showed less reaction to an experimenter banging her knee and acting distressed as a result. However, this could be due to differences between self-report and observation. Autistics often state that they often are unaware of another person's distress, but if they are aware of it, they react emotionally. So perhaps the autistic children in Corona et al's study did not realize the experimenter was distressed, while the respondents on this test were reporting on situations in which they knew there was a problem.

Oh, and incidentally, my own scores were Fantasy 25, Empathic Concern 19, Perspective Taking 12, and Personal Distress 25.