A finished BUS-660 Topic 2 decision tree analysis example, with decision and chance nodes distinguished, every branch given a probability and the tree rolled back. Searches like "bus 660 topic 2 assignment example", "bus660 topic 2 sample" and "bus-660 topic 2 example" land here.
What a finished BUS-660 Topic 2 decision tree analysis looks like
The finished example draws a problem that genuinely needs a tree, meaning one where a later choice depends on how an earlier uncertainty resolved. Decision nodes and chance nodes are marked differently and never confused, since one is chosen and the other is not. Every set of branches leaving a chance node sums to one, which the example checks explicitly. Payoffs sit at the ends. The rollback then proceeds from right to left, taking expected values at chance nodes and maxima at decision nodes, with the arithmetic shown at each. The recommendation is stated as the first action rather than as the whole path, because everything after the first branch depends on what happens.
How a BUS-660 Topic 2 example is structured
The example constructs a tree and solves it backwards. It opens with the decision sequence and explains why order matters here, which is the justification for using a tree rather than a table. A second section draws the structure, marking decision and chance nodes distinctly and describing the branches in text so the tree is readable without the diagram. A third attaches probabilities and verifies that each chance node sums to one. A fourth places payoffs at the terminal points with their derivation. A fifth performs the rollback from right to left, showing the value computed at every node. A closing section states the immediate recommendation and describes what the firm would do at the next decision under each way the uncertainty could resolve.
A problem that needs a tree
A later choice depending on how an earlier uncertainty resolved is what a payoff table cannot represent.
Node types marked distinctly
One node is chosen and the other is not, and confusing them produces a rollback that computes the wrong thing.
Branch probabilities verified
Every set leaving a chance node is checked to sum to one, which catches a whole class of setup errors.
Rollback shown node by node
Expected values at chance nodes and maxima at decision nodes, with the value written at each point.
The recommendation is the first move
Everything after the opening branch depends on what happens, so the advice is what to do now.
Where marks go in BUS-660 Topic 2
Taking a maximum at a chance node is the structural error that invalidates a whole tree, and it happens whenever the two node types are not clearly distinguished. A second failure is branch probabilities that do not sum to one, which the example catches by checking and which otherwise propagates silently through every value. Papers lose marks for describing a tree that only a diagram could explain, since a reader without the image should still be able to follow the structure from the text. Recommending an entire path rather than a first action misrepresents how sequential decisions work. Trees drawn for problems with no sequence in them use a heavy tool where a table would have been correct and clearer.
Get a BUS-660 Topic 2 example written to your instructions
Send the BUS-660 Topic 2 problems and the rubric from your classroom, with the sequential decision your section supplied. We write a custom example to those criteria, with node types marked distinctly, branch probabilities verified, the rollback shown at every node and the recommendation stated as a first move, in 24 to 48 hours. The first is free.
BUS-660 Topic 2 questions, answered
When should I use a tree instead of a payoff table?
When the decision has a sequence in it. A table handles one choice made against uncertainty; a tree handles a choice, then an uncertainty resolving, then another choice that depends on the result. If nothing in your problem happens after something else, a table is simpler and clearer, and using a tree anyway adds structure without adding information.
Why does the rollback go backwards?
Because you can only value a decision once you know what each branch leads to. Working from the terminal payoffs leftward means that by the time you reach a decision node, every option in front of it already has a value attached. Working forwards would require valuing a choice before knowing its consequences, which is exactly the information the tree exists to organize.
What do I do differently at each node type?
At a chance node you take the expected value, weighting each branch by its probability, because you do not control which occurs. At a decision node you take the best value available, because you do control it. Getting these the wrong way round is the commonest error in the topic and it silently produces a completely wrong recommendation.