Supporters of stereotypes often take refuge in facts and statistics. For example, people might support the stereotype that children of Asian immigrants in America are high academic achievers. These supporters will dig up data from spelling bees, olympiads and standardized test score to establish their arguments. Often, little fault can be found with their data.
The problem lies with the inference. In emphasizing ethnicity in analysing academic achievement, the implication always is that ethnicity is the cause of said achievement. These kids are good at spelling bees and olympiads and standardized tests because they are Asian immigrants. Correlation is confused with causation. (If not explicitly then at least implicitly.) And that is problematic.