Partial Budgets & Decision Analysis

Question 1

What are types of costs?

Answer 1

1. fixed costs (paid regardless of which choice is taken; w/ a short time horizon these can usually be ignored)
2. variable costs (vary depending on which action is taken)
3. opportunity costs (costs incurred by not making some other choice; usually assume that we will borrow the needed funds)

Question 2

What are partial budgets and benefit cost analysis for?

Answer 2

• attempt to forecast financial events that follow a management decision
• considers a limited aspect of the total enterprise (only variable costs)
• areas that are irrelevant to the decision are ignored
• used to compare 2 or more different options
• current program often used as a baseline

Question 3

what are partial budgets?

Answer 3

• simply calculate the total change in revenues & total change in expenses
• net revenue change - net expense change = net profit
• only consider those costs that will change depending on which option you choose
• spreadsheets allow you to play “what if” games

Question 4

What is decision analysis?

Answer 4

systematic method that evaluates:
- future events
- consequences of actions
- likelihood of upcoming occurrences
- value of various endpoints
- Note: money is not a major component of this analysis
- most of the focus is on probabilities & logic, money just helps us keep score!

Question 5

What are the advantages of decision analysis?

Answer 5

• explicit (forces us to formulate a problem by it’s component parts)
• quantitative (forces us to make precise statements about probabilities, costs, & revenues)
• prescriptive (helps us choose a course of action)
• allows us to play “what if” games (sensitivity analysis - change 1 factor & see if decision changes)

Question 6

What are the disadvantages of decision analysis?

Answer 6

• simplistic view of a complex biological system
• decision analysis is an approximation of reality
• forecasting is extremely difficult
• can never create a complete model of all aspects of a problem
• risk is not considered
• best used for commonplace decisions that are made over & over

Question 7

What are the 4 steps of decision analysis?

Answer 7

1. identify the problem & choices
2. structure the decision prob
3. characterize the info needed
4. choose a preferred course of action

Question 8

What is a decision tree?

Answer 8

• graphic arrangement of the flow of events
• “tree” branches from L to R & is read from L to R
• branches occur @ each alternate action or chance event
• probability of each outcome can be considered
• starts w/ a decision “node”
• usually represented by a square

Question 9

What is a decision node (choice node)?

Answer 9

• indicates a clinical event over which the decision maker has control
• exhaustive: must include all possible choices to be considered
• exclusive: only 1 choice is possible @ node

Question 10

each branch of the decision node can lead to 1 of 3 options:

Answer 10

1. another decision node
2. a chance node
3. a terminal branch or outcome

Question 11

what is a chance node?

Answer 11

• essence of a decision tree
• uncertain pts in a decision tree
• branches of a chance node represent all possible outcomes that can result @ that point
• probabilities are assigned to each branch
• probabilities of all branches from that node must add up to 1
• can lead to a decision node, a chance node, or a terminal branch

Question 12

What are paths in a decision tree?

Answer 12

• represents a stream of events over time
• bounded by a decision node & by a final outcome event
• in btwn can be a number of decision nodes & chance nodes
• each path in a tree is unique & represents alternative choices & events that can occur

Question 13

what are terminal branches?

Answer 13

• far right of a decision tree
• each has an outcome specific to that chain of events of that branch
• outcome value is assigned to the terminal branch

Question 14

What are outcome values?

Answer 14

• various endpoints must be differentiated to be of use
• usually done by assigning a cash value
• in companion animals, human medicine, etc., the end points may be utilities rather than $ values

Question 15

What are expected monetary values?

Answer 15

• if you had only a 20% chance of receiving a $500 bill in a game, how much would you be willing to pay to play that game?
• expected monetary value of that action is 0.2($500) = $100
• utility theory can be used to generate non cash outcomes & EMVs

Question 16

How do you fold back the decision tree?

Answer 16

• work backwards from R to L
• “back up to a decision node”
• calculate an EMV for each pathway
• costs along each pathway must be taken into account
• most advantageous value is assigned to that decision node & other choice is “pruned” out
• expected value = value of the branch x probability it will occur
• expected value @ the node = sum of expected values of all branches @ that node
• to max profit: best decision is the branch w/ the highest expected value