arbitrary selected day care centers without any previous information on
the children's health conditions or their home environment which
reduced sample selection bias. The core questions on children's health
questionnaire have been validated and proved as reliable in epidemiological
studies (Norrby et al., 2003). One limitation of this study was
that since both the information on exposure and childhood health was
collected in the same questionnaire, the reporting bias might occur if
parents with children with asthmatic/allergic symptoms are more
aware of home environment factors (Larsson et al., 2011). We did not
have exact data on the degree of awareness on associations between
indoor factors and health in this area. However, the consistent results
based on the analyses by different regression models indicated that
our results were not seriously biased by selection bias or recall bias.