Ninety-five children completed the study. The overall dropout rate was 9.52% and the reasons for attrition included withdrawal of consent by parents (N = 4), uninformed change of residence (N = 4) as well as on new information regarding past medical history (N = 2). We stratified those 95 children in to three groups of sub average (N = 33), average (N = 31) and above average (N = 31) intelligence. These groups when compared significantly differed on six socio demographic variables namely, chronological age, type of school, monthly income of the family, father's education, mother's education and mother's age. We considered these variables as potential confounders for our analyses. The socio-demographic, participant and family characteristics of completers are presented in Table .
Socio demographic profile of children and their families (N = 95)
Psychopathology assessment demonstrated low needs on hyper activity, conduct problems, emotional symptoms, peer problems and total difficulties domains. Participants also scored favorably on pro social domain with mean (sd) of 9.14(0.86) and thus we ruled out any major psychiatric morbidity among them. In addition, the principal investigator, a qualified psychiatrist, ruled out any ICD-10 [28
] based psychiatric morbidity with face-to-face clinical interviews. We present the psychological characteristics of the participants in Table .
Psychological profile of children who participated in this study (N = 95)
The mechanical (rho = 0.74, P = 0.001), social script (rho = 0.65, P = 0.001), general sequencing ability (rho = 0.77, P = 0.001) and false belief scores (rho = 0.70, P = 0.001) of PST were significantly correlated with IQ. The selective accuracy Theory of Mind score was calculated by subtracting the general sequencing ability scores from the mean score of false belief sequences.
The linear correlation between ToM and IQ was not significant (rho = -0.14, P = 0.17; Prep = 0.83). Among the sub domains of intelligence, only visuomotor (rho = -0.26, P = 0.02) and social intelligence (rho = -0.29, P = 0.003) were significantly related to ToM. We present the bivariate Pearson correlation matrix between ToM and various domains of intelligence in Table . The total score of Unexpected Contents Theory of Mind task did not correlate with the Intelligence Quotient (rho = -0.15; P = 0.14). ToM did not have significant relationship with IQ within below average (rho = -0.26, P = 0.2; P rep = 0.82), average (rho = -0.27, P = 0.2; P rep = 0.82) and above average (rho = 0.05, P = 0.8; P rep = 0. 57) intelligence groups. Though, ToM significantly differed between the three intelligence groups (F = 7.86, df = 2; P = 0.001), children with above average intelligence [mean (sd) = -0.35(0.79)] had better ToM than those with below average intelligence [mean (sd) = -0.61(1.51)] who in turn had better ToM than those with average intelligence [mean (sd) = -1.50 (1.17)].
Correlationa matrix between Theory of Mind ability and intelligence domains of participants
The multiple linear regression analysis with ToM as the dependent variable also demonstrated the lack of a significant relationship between ToM and IQ when the above stated six socio-demographic confounders were controlled [β(SE) = -0.24 (0.01), t = -1.59, P = 0.12; Prep = 0.86]. The social intelligence domain [β(SE) = -0.55 (0.06), t = -3.75, P = 0.001; Prep = 0.99] and the visuomotor domain [β(SE) = -0.32 (0.09), t = -2.43, P = 0.02; Prep = 0.95] continued to exhibit significant relationship with ToM. When we further controlled for the effects of VABS adaptive behavior composite age equivalent, the lack of a significant relationship between ToM and IQ [β(SE) = 0.02 (0.01), t = 0.05, P = 0.96; Prep = 0.51] and significant relationship of the social intelligence domain [β(SE) = -0.91 (0.11), t = -3.75, P = 0.001; Prep = 0.99] and the visuomotor domain [β(SE) = -0.49 (0.11), t = -2.57, P = 0.012; Prep = 0.96] remained. Other domains of intelligence did not have significant relationship with ToM.