Suppose that different amounts of fertilizer had been used in
different fields that were involved in trials of different types of
wheat seeds.
(a) Explain how the use of different amounts of fertilizer
confounds the experimental analysis in the absence of
randomization.
(b) Considering the regression models with
explanatory variables, if the amount of fertilizer
were known, how could an analyst adjust for this
confounding effect?
(a) Choose the correct answer below.
A.
Yield is uncorrelated with fertilizer use.
B.
Fertilizer use is not uniform for each type of seed.
C.
Seed variety is highly correlated with fertilizer use.
D.
Differences in yield could be due to fertilizer rather than seed
variety.
(b) Choose the correct answer below.
A.
Add a quantitative explanatory variable for the amount of
fertilizer used.
B.
Add a quantitative explanatory variable for whether fertilizer
was used.
C.
Add a categorical explanatory variable for the amount of
fertilizer used.
D.
Add a categorical explanatory variable for whether fertilizer
was used.