The Ratio and Root Tests always fail when x is an endpoint of the interval of convergence, so the endpoints must be checked with some other test. Why do you think The Ratio and Root Tests always fail when x is an endpoint of the interval of convergence

For a science project, a high school research team conducted a survey of local air temperatures. based on the results of the survey, the found that the average temperatures were around 10 degrees higher than expected. this result was wrong. the trouble with the survey was that most of the locations were exposed to direct sunlight and located over asphalt or sand, which resulted in higher temperatures than normal. this is a classic example of an error in which phase of inferential statistics?