In general, the term expected value of sample information $($EVSI$)$ means the difference between the expected value of an optimal strategy

a$.$ without any sample information and the "best" expected value based on sample information.

b$.$ based on uncertain information and the "best" expected value with perfect information.

c$.$ with perfect information and the "best" expected value with uncertain information.

d$.$ based on sample information and the "best" expected value without any sample information.