That confounds the environment of the "womb" with genetics, by the way.

In addition, twin studies have a problem in that twins are not like other children. Twins tend to have a number of problems more often than non-twins (various cognitive problems are more common).
As an example of how this could easily confound a twin study, suppose everyone has 100 IQ, except for people with a particular condition. 5% of the general population has this condition, as well as 10% of twins. If one twin has it, both twins have it. It is caused by insufficient resources during fetal development. The effect of this condition is that it drops your IQ to 80.
Despite genetics having
zero impact on IQ in this model, you'll find a strong correlation between twin IQ raised apart.