Jesus Ramos
September 24 2015
Do ratings from users reliably predict ratings from critics?
Estimate Std. Error Pr(>|t|)
(Intercept) 8.74 0.40 0.00
UserRating 0.05 0.05 0.28
PlatformPS4 -1.21 0.56 0.03
PlatformWiiU -3.92 0.60 0.00
PlatformXBoxOne -0.92 0.52 0.08
UserRating:PlatformPS4 0.06 0.07 0.37
UserRating:PlatformWiiU 0.35 0.07 0.00
UserRating:PlatformXBoxOne -0.02 0.07 0.74
According to the figures, every 1-point increment in user ratings brings a 0.28 increase in critic ratings. Though this relationship is very significant at a p-value of 0, it only explains 15.98% of the variance. However, if we control for platform we can account for 60.37% of the variance, but then the only good predictor between ratings is for WiiU.