I'm a behavioral scientist and I find both are useful. If you never look at a histogram it's surprisingly easy to fool yourself about what exactly the ecdf is t

I'm a behavioral scientist and I find both are useful. If you never look at a histogram it's surprisingly easy to fool yourself about what exactly the ecdf is telling you in certain situations, particularly when comparing distributions.