Kraft foods wants to see whether a coupon will affect sales for macaroni and cheese. Kraft picks a single grocery store in Philadelphia in which to test two coupon types: ?buy one, get one free? and ?$.50 off.? They also want to test whether it matters on which day of the week the coupon gets delivered: Wednesday or Friday. They choose randomly assigned customers to one of the four conditions. They measure success by total sales of Kraft Macaroni and Cheese at the store.

a) What type of experimental design is this?

b) Do you have any concerns about the design?

c) Does this experiment satisfy the necessary conditions for a properly designed experiment?

d) The analyst computes the means for each cell, which appear in the table below. Based on the data, the analyst claims that it does not make a difference whether the coupon is delivered on Wednesday or Friday. Is the analyst correct? Why or why not? Does it help to create the interaction plot?

Transcribed Image Text:

Coupon Type Buy-One-Get-One $.50 off Mean Wednesday 500 450 475 Friday 350 600 475 Mean 425 525 Day of Week