| Message |
tirerack simply didn't have as large of a selection, so that is a possible explanation for the difference in sample means (I could've included the two /40 tires that had tread width data as well, but a sample of two would certainly be a poor representation of the population). But regardless the reason, in actuality, the means of all three tires could be the exact same... it's just in this sample that the sample mean of the /45 tire was 0.2" larger. And as I've shown, that's statistically insignificant. (The statistics actually suggest that there's about a 1 in 5 likelihood of getting those results if all the means are, in fact, truly equal...) By the way, my sample is not "weak"... I'm an applied statistician by profession and am well aware of how to interpret data (for samples both small and large). If I had used Z-intervals rather than t-intervals, you'd have a point, but I know what I'm doing (not to mention, ANOVA cares not about sample size). I accept that you're admitting to being wrong with your previous statement, in part attributed to my calculations, but you're in no position to question my statistics. No offense, at all... but it'd be like telling a professional baseball player that he could improve on his swing.
____________________________________________________________ I love elitists who aren't really elite in any way... they amuse me greatly. |
 |
