I think, more to the point, is that more data really is not better data. Oftentimes it's worse -- the larger the volume of data you collect, the harder (and mor

I think, more to the point, is that more data really is not better data. Oftentimes it's worse -- the larger the volume of data you collect, the harder (and more expensive) it can be to manage sources of systematic bias in your data collection. One of the more extreme examples of this situation is Big Data, where everything is a convenience sample, and therefore nothing is particularly trustworthy. Combine it with standard statistical hypothesis testing, whose p-values, being designed for the size of experiment that one could realistically manage in a lab setting circa 1910, are as much a reflection of your sample size as of anything else, and you're at great risk of leading yourself into a tar pit.