In a population with 5 items (X1, X2, X3, X4, and X5), a researcher has developed a function F() (see below) to draw a simple random sample (every item has the equal probability of being chosen) which includes one item.

output = F(X1,X2,X3,X4,X5)

The output is the randomly chosen item, and the inputs are X1 to X5.

If the researcher now wants to draw another random sample in which the probability of choosing X1 to X3 are the same, but the probability of choosing X4 and X5 is twice as much as choosing X1 to X3. How should the researcher use F() to carry out this?