October 13th; Statistics ii

Character bank: α

There are good slides, but they are not yet available. 

=====

You have data. You have preconceived notions: a null hypothesis (where it is due to chance) and an alternative hypothesis that you want to show. You first put this through the α level, which is the parameter you set (generally .05%) to make sure that it is not due to chance. In a two tailed test, you divide by two. 

Sl. 51: Effect Size

* In cohen's d formula- s: standard deviation(?) ##think so?

 Sl.53:
 
 Distributions are different when compared the influence of coffee intake on the exam time and the exam performance. (Why?)
 
 

Variability_total = Variability_between-groups + Error_within-group

∑/i ∑/g (Y_ig - GM)^2 = w∑(y_y -GM)^2 + ∑/i ∑/g(y_iy - Y-y)