In an experiment to study the yields of a chemical process under four factors A, temperature; B, pressure; C, catalyst concentration; and D, stirring rate, each factor was set at two levels only. Only one replicate of the experiment could be run. The yields are

(1) 20 a 14 b 18 ab 12 c 18 ac 14 bc 22 abc 14 d 16 ad 10 bd 16 abd 10 cd 8 acd 9 bcd 13 abcd 10 Obtain the effects of the treatment combinations. Assuming that all the second- and third-order interactions are not significant, use their combined sums of squares as an estimate of the residual sum of squares and analyze the data. Obtain a normal probability plot of the estimates. Can you confirm that the second- and third-order effects are not significant?