Your example made me laugh, and I agree that an actual survey would be very difficult to carry out (difficult, not impossible - any PhDs in stats around here who would like to volunteer to design a survey??)
I'm gonna go ahead and guess that there is a correlation, though, between severity of spankings in childhood and higher tolerance for severity in stories. This guess is based on other research, nothing to do with spanking, that shows that jury members in trials tend to judge defendants more harshly if they themselves can personally relate to the defendant. For instance, men judge other men accused of rape more harshly than women ("If I can restrain myself from raping a woman, this other guy could have too."), and women judge other women accused of harming their children more harshly than men ("If I can keep from stabbing my child, this woman could have too.").
So based on that - totally unscientific conclusion coming right up! - my guess is that people who were spanked harshly will be more likely to poo-poo milder spankings in stories ("If I could put up with worse, this character could too.") Right?