## Participants

The study collected data from 237 observations on; normal TV watching, less TV watching, normal outdoor activities, more outdoor activities, autism level, the theory of mind measure.

## Measures

The analysis was developed using a regression model in which the level of watching TV, outdoor activities and interaction of TV and outdoor activities was used (1) and then theory of mind measure was added in the second (2) model:

*Autism = β _{0 }+ β_{1}TV_{i}+ β_{2}Outdoor_{i}+ β_{3}(TV*Outdoor)_{i }+ ε_{i}*

*Autism = β _{0 }+ β_{1}TV_{i}+ β_{2}Outdoor_{i}+ β_{3}(TV*Outdoor)_{i }+ β_{4Theory}_{i }*

*+ ε*

_{i}Where *β _{0, }*

*β*

_{1, }*β*

_{3, }*β*are constants

_{4 }*ε*

_{i }*is the error term associated with observation i.*

Finally, a regression model of the independent variables was made on the theory of the mind using the equation 3 below,

*Theory = β _{0 }+ β_{1}TV_{i}+ β_{2}Outdoor_{i}+ β_{3}(TV*Outdoor)_{i }+ ε_{i}*

Where *β _{0, }*

*β*

_{1 }*and β*are constants

_{3 }*ε*

_{i }*is the error term associated with observation i.*

## Results

The linear regression analysis of the independent variables: TV watching, outdoor activities, interaction between the TV watching and outdoor activities and theory of mind measure on the dependent variable autism symptoms was better compared with the model where the theory of mind was removed as given by the adjusted R-square value which is higher in the first case. In fact the model without the variable had higher error of estimation or 1.432 while when the variable is included the error reduces to 1.420 which shows that the full model was more accurate.

The ANOVA results of the model are both the full model (*F*_{3,234 }= 9.489, *P* < 0.01) and the reduced model (*F*_{4,233}=10.853, *P* < 0.01) which are significant even at 1% level of significance. The results indicate that both the reduced and full models are significantly important and can be used to predict the autism symptoms. The last section of the regression analysis gives the regression coefficients for both models.

For the reduced regression model, the coefficients for the constant (*β*_{0 }= 9.052, *t *= 57.771, *P *< 0.001), TV watching (*β*_{1 }= – 0.714, *t *= -3.708, *P* < 0.001), outdoor activities (*β*_{2 }= 0.481, *t *= 2.574, *P *= 0.011) and the interaction of TV and outdoor activities (*β*_{3 }= 1.211, *t *= 3.146, *P *= 0.002) are all significant at 5% level of significance. The reduced model can thus be represented using the equation:

*Autism = 9.052 – 0.714TV+ 0.481 Outdoor+ 1.211 (TV*Outdoor)*

On the other hand, from the full model, the constant (*β*_{0 }= 8.796, *t *= 45.355, *P *< 0.001), TV interaction (*β*_{1 }= – 0.63, *t *= -3.238, *P *= 0.001), outdoor activities (*β*_{2 }= 0.495, *t *= 2.669, *P *= 0.008), interaction between TV and outdoor activities (*β*_{3 }= 1.256, *t *= 3.284, *P *= 0.001) and the theory of the mind (*β*_{4 }= 0.110, *t *= 2.205, *P *= 0.028) are significant at 5% level of significance. These results can be represented using the equation:

*Autism = 8.*796 – 0.630 *TV+ 0.495 Outdoor+ 1.256 (TV*Outdoor) + 0.11 Theory*

The results reveal that all the independent variables are important in the model in which the TV watching negatively affects the autism symptoms whereas all the other variables affect the autism symptoms positively.

According to the mean results, without considering the nature of TV watching nor the outdoor activities autism symptoms has an average of 9.03, whereas autism has a mean of 9.32 under normal TV watching and 8.55 under less TV watching which reveals that watching TV tend to increase the prevalence of autism.

With inclusion of more outdoor activities, the average autism symptoms are 9.30 whereas at normal outdoor activities the autism is at 8.74 which reveal that the increased outdoor activity increases the chances of autism. The largest value of the analysis (9.33), which is the mean of the autism symptoms under normal TV levels and more outdoor activities are as expected from the results above. Thus, increased TV watching coupled with more outdoor activities tends to increase the autism symptoms. The vice versa is accordingly true (8.00) for less TV watching and normal outdoor activities.

The results above indicate that watching TV and outdoor activities affect the prevalence of getting autism. However, the results are not in the same direction based on the regression coefficient which revealed a negative effect of watching TV on autism and positive effect of outdoor activities on watching TV.

Carrying out the regression of the other independent variables on the theory of the mind produces unreliable important model due to its very small coefficient of determination (adjusted *R-Square*=0.028) with large standard error (1.862). However, the model is significant and shows that at least one of the independent variables used in the model affects the theory of the mind (*F=*3.302, *P*=0.021).

The analysis of the coefficients of each independent variables reveals that the constant (*β*_{0 }= 2.327, *t* = 11.419, *P *< 0.001) and TV watching (*β*_{1 }= – 0.760, *t = –*3.034*, P *= 0.003) significantly affect the theory of the mind while outdoor activities (*β*_{3 }= 0.128, *t = *– 0.525, *P *= 0.600) and the interaction between TV watching and outdoor activities (*β*_{4 }= – 0.406, *t = *– 0.810, *P *= 0.419) do not affect the theory of the mind.

The model for the theory of mind can therefore be estimated using the equation:

*Theory of mind *= 2.327 – 0.760 *TV watching*

Assuming the level of watching TV and outdoor activities are the same, the mean for the theory of mind measure is 5.97, whereas for the normal level of TV watching the mean is 6.26 while it is 5.51 with less TV watching. This shows that the theory of mind increased with increased level of TV watching. Under outdoor activities, the mean theory of mind measure is 6.00 for normal outdoor activities and 5.95 for more outdoor activities revealing that the theory of mind measure decreases with increased outdoor activities which is the same with the regression coefficient although this is not significant.

## Conclusion

The results revealed that autism is affected by the level of watching TV and the outdoor activities whereby increased “watching TV” and increased “outdoor activities” is likely to increase the chances of a child developing autism in the future life. The exposure of TV to children under the age of 2 years will make them not develop theory of their minds. Due to the above results parents should be encouraged to reduce the exposure of their children under the age of 2 years to television and outdoor activities.