Kenett and Steinberg (1987) described a two-level factorial experiment con-

ducted by students to study the time required to boil 1 qt of water. Factors were A= ame level (low or high), B=pan size (small or large), C=pan cover

(none or glass cover), and D=salt added to water (no or yes).

(a) If the standard deviation in boiling time (tested at the same conditions)

was found to be ^=0.236 minutes, use the shortcut formula to determine how many experiments you will need to perform in order to have power of 0.95 for detecting eects of size
=0.50 minutes. Would this answer change if you decided to only perform an experiment with 3 of the 4 factors?

(b) Create a list of experiments in random order for performing these ex-
periments.

(c) Actually perform the experiments by boiling the water and collect the data.

(d) Analyze the data to determine which eects and interactions are signif-
icant.

(e) Interpret and explain any signicant eects you nd.