Pretty sorry execution of the usability tests I'd say. Isn't quality of the sign one of the most obvious factors to consider? I wonder if they only tested in perfect weather conditions, too.

I guess I get a decent example for confounding variables for our intro material out of this so there's that.