Fig. 1

Estimated marginal means of ratings on the four scales by sex and diagnosis. Presented above are data from each of the four insight scales, sex means (boy—gray; and girl—white) and diagnosis means (no autism diagnosis and autism diagnosis) are shown. For the figure above, *p < .05, **p < .01, and ***p < .005. A Results from an ANOVA show that girls were significantly rated higher on insight measured by the five levels of emotional awareness scale, p = .023. Differences between diagnosis, p = .281 and an interaction between sex and diagnosis, p = .877, were not significant. B Results from an ANOVA show that non-autistic individuals were rated significantly higher than individuals with an autism diagnosis, p = .013. Differences based on sex, p = .058, and an interaction between sex and diagnosis, p = .798, were not significant. Additionally, differences in girls were found p = .046. C Results from an ANOVA show that non-autistic individuals were rated significantly higher than individuals with an autism diagnosis, p = .028, and sex, p = .001. In the non-autistic group, girls were rated significantly higher than boys, p = .025. Similarly, in the autism diagnosis group, girls were rated significantly higher than boys p = .020. Differences in an interaction between sex and diagnosis, p = .587, were not significant. Additionally, differences in girls were found p = .044. D Results from an ANOVA show that non-autistic individuals were rated significantly higher than individuals with an autism diagnosis, p = .040, and sex, p = .004. In the non-autistic group, girls were rated significantly higher than boys, p = .038. Similarly, in the autism diagnosis group, girls were rated significantly higher than boys p = .041. Differences in an interaction between sex and diagnosis, p = .462, were not significant