Fallacies of Certain Studies
When creating a survey or an experiment, there are a number of errors that must be minimized. Obviously it is important that the "experiment" must actually test what you are wanting to look for. Height of person vs IQ would be problematic because babies tend to be quite short and do poorly on both written and verbal tests. It is equally important that bias needs to be controlled for as much as possible.
While we can discount those illustrations above as symptomatic of a time, this error persists even today. Just ask your doctor the last time they conducted an experiment. I am rare to the extent that my family physicians are monitoring my blood samples and meeting with me every 3 months while I am on a Keto diet and intermittent fasting. This means that my doctors are participating in an "experiment" despite their best efforts.