A public health researcher in Delhi is investigating the association between air pollution (PM2.5 levels) and respiratory symptoms in school children. The study finds a strong positive correlation. However, the researcher notices that children from lower socioeconomic status (SES) schools are overrepresented in the high-exposure group, and lower-SES children also have higher rates of malnutrition and indoor cooking smoke exposure. What is the most appropriate next step to address this methodological concern?