This weblog submit incorporates one small analogy for explaining family-wise error to your college students.
I used to be making French toast for dinner the opposite night time. Whereas I used to be measuring out cinnamon, I noticed utilizing one tablespoon as a substitute of three teaspoons to keep away from measuring errors is form of like utilizing a one-way ANOVA with three ranges as a substitute of doing three t exams to keep away from Kind I error.
Follow me right here. If I have been to make use of three teaspoons to measure out an ingredient, there’s a probability I might make a mistake 3 times. Three alternatives for air pockets. Three alternatives to not completely degree out my ingredient. In the meantime, if I simply use one tablespoon, I’ll solely threat the error related to utilizing a measuring spoon as soon as.
Equally, each time we use NHST, we settle for 5% Kind I error (properly, if you’re a psychologist and utilizing the 5% gold normal, however I digress). Utilizing three exams (t exams) after we might use one (ANOVA) will enhance the danger of a false constructive.
I do not find out about you, however I by no means know which rationalization for a statisitcal phenomena will stick in my college students’ brains. You possibly can discuss, like, measuring no matter you wish to use this analogy, however I like that the three to 1 ratio that applies teaspoons and tablespoons as properly three t exams versus an one-way ANOVA and three t exams.
