Question

Based on the cost of obtaining the sample information and its efficiency, is it worth obtaining it? Why? Please include ALL the relevant information for this decision and use specific numbers. Your answer must discuss EVSI, net EVSI (The EVSI minus the cost of the SI), the Efficiency of the Sample Information (abbreviated "E") and the Risk Profile.

E	U	E (With Cost)	U (With Cost)
						Decision using the Expected Opportunity Loss (EOL)	13.2 (B)
B	44	-29	39	-34
N	24	-11	19	-16		Decision using the minimax regret approach.	20 (N)
S	18	4	13	-1
Probability (State)	0.6	0.4
Probability (State/F)	0.81	0.19				Decision with the optimistic Approach	44 (B, E)
Probability(State/X)	0.4	0.6
						Decision with the Conservative Approach	4 (S,U)
Probability (F)	0.65		EVPI	28
Probability(X)	0.35		EVwPI	14.8		E(B)	14.8
						E(N)	10
						E(S)	12.4
EV(Node 2)	25.13		EV(Node 8)	14.8		EV(If more Accurate Research is conducted, Node 1)	17.9445
EV(Node 3)	12.35		EV(Node 9)	10
EV(Node 4)	10.34		EV(Node 10)	12.4
						EV(If Standard Research is conducted)	14.8
EV(Node 5)	-4.8
EV(Node 6)	-2
EV (Node 7)	4.6

Answer 1

EVSI = EV with sample information (Assuming no cost to gather it) - EV of best decision without sample information

= EV with sample information + Cost of the sample - EV of best decision without sample information

= 17.9445 + 5 - 14.8

= 8.1445

Net EVSI = EVSI - Cost of Sample information

= 8.1445-5

= 3.1445

Efficiency = Net EVSI * 100/EVPI

= 3.1445*100/28

=11.23%

As efficiency is low and also the Net EVSI is less than the cost of sampling, it is not worth obtaining it.