In Generalizability Theory, If You Have a Variance Component Estimate

In Generalizability Theory, if you have a variance component estimate that is extremely small, what does that tell you?

A) That component is not an important source of variance.
B) It would be difficult to generalize across the different levels of that component.
C) You need to collect more data on that component.
D) That component must be fixed.