Oxide drainage in the steel industry

The principal objective of a project executed by students in a chemometrics course was to optimize a procedure to decrease the drainage time of oxides used in the steel industry. Four factors were 

The values calculated for the contrasts strongly indicate that there are only two important factors (1 and 4, i.e., solvent and time), and that the interaction between them is also important. Treating factors 2 and 3 as inert ones, we obtain a 2 full factorial design in duplicate (Table 4A.6), from which we can extract an estimate of pure error and consequently a measure of the significance of the effects. 

The pooled variance of the duplicate runs is 44.91. The standard deviation of a response is the square root of this value, 6.70. The variance of an effect is half that of the pooled variance, 22.45, and therefore its standard error is 4.74. Since the estimate of the pooled variance has 4 

Trypanosoma genus, which cause Chagas disease. The results obtained are reproduced in Table 4A.7. 

Since the signs of columns 12-15 in the table do not correspond to any real factors, we can use the values of their calculated contrasts as estimates of standard errors of the effects (assuming, of course, that all interactions are negligible). We can obtain a pooled estimate of the 

---