Sunday, July 8, 2007

More lying with statistics - a primer

Welcome to TVNewser readers, and kudos to Brian for posting our rebuttal to the Chi-Trib nonsense; we did, after all, use his own stats to prepare it, so he no doubt felt obliged.

Here's yet another way people who want to massage the data can make figures say whatever they want; ignore trends and use inappropriate periods to compare against.

Say you have the following two shows, rated over uniform periods, in arbitrary quantities:

Show A: 100 ... 80 ... 60 ... 40
Show B: 30 ... 40 ... 50 ... 60

Which show is the better bet, ratings-wise? Proponents of Show A could easily say it was superior - after all, its average rating over the period is 70, versus a mere 45 for Show B. Never mind that Show B clearly is growing an audience and Show A is losing theirs.

Remember to look behind the figures and question the motives of people promulgating bogus factoids like the Chi-Trib canard.