Statistics 462
Missing Data in the Randomized Block Design
In JMP it is possible to exclude any number of data points
from a data table. For example, to exclude Row 7 in the hardness data of
Table 5-1 proceed as follows.
JMP will mark the row to indicate that it has been excluded.
Let's analyze the data with this "missing data":
Notice that the missing data made the design not orthogonal (not every treatment combination occurs the same number of times). So the Type I
ANOVA table is not the same as the Type III (Effects) ANOVA table. However, notice that since only 1/20th of the data is missing, and the missing point is fairly typical of the rest of the data, both ANOVA tables are very close to the ANOVA tables one would get with all the data.