The article discusses how researchers use different methods to figure out if certain things cause specific outcomes. They talk about how randomized trials are the best way to do this, but sometimes they can't be done. In those cases, researchers use observational studies, which can have biases. To deal with these biases, researchers use something called propensity score approaches. These approaches help to make sure that the results are more accurate. The researchers used these methods to look at how breastfeeding and complementary feeding affect the health of infants and children in Malawi and Zambia.