The statistics tag has been used more and more to refer to any question that has a quantitative element to it.
I think the tag should be used only about questions that are fundamentally about statistics (or probability, or sampling). Like:
- Can we determine the percentage of statistics made up out of thin air?
- Is a coin toss fair?
- Did the National Rifle Association (NRA) block research into statistics related to gun control?
- How accurate are the d20 dice used in role playing games?
- Does Gallup assume people answer honestly in their creationist vs. evolution polls?
These are some (what I would call) misuses of the tag:
- Are youths that watch violent X-rated material 6 times more likely to force others sexually?
- Does it rain more often on weekends than weekdays?
- Are Fairtrade farmers paid worse than other farmers in the same regions?
- Does temperature drive CO₂?
Just because statistics would be useful/necessary to answer the question, that doesn't inform us at all about what the question is about. It seems that it is being used as a meta-tag.
If we are fine with those fours question (any many others) being tagged statistics, why not:
- Do infants have fussy periods at well-established periods after birth?
- Do headers in football (soccer) cause brain damage?
- Does drinking water after eating watermelon cause cholera?
- Does nuclear radiation increase the risk of heart attacks?
- Do young black males have on average higher testosterone than other races?
My point is that a too liberal use of statistics makes it either meaningless or a meta-tag.