During the planning of the study little attention was paid to the possibility of contamination through other sources of information. The potential for crossover of information from the mothers in the test group to those in the control group was
minimal, because of the geographical distances among the residences of mothers. Residences were distributed over 151 postcodes across Adelaide and its suburbs. In addition, any contamination from other information sources does not mitigate our results, as it would not have differentially favoured either the test or control group