Scientific versus intuitive decision-making, the value and limitations of data in decision-making, opportunity cost, and how to construct and interpret a decision tree including expected values.

Calculate the expected value (EV) at each chance node, then subtract the cost of the option.

New product EV $= (0.6 \times 500{,}000) + (0.4 \times 150{,}000) = 300{,}000 + 60{,}000 = \pounds360{,}000$. Net of cost: $360{,}000 - 200{,}000 = \pounds160{,}000$.

Extension EV $= (0.7 \times 250{,}000) + (0.3 \times 100{,}000) = 175{,}000 + 30{,}000 = \pounds205{,}000$. Net of cost: $205{,}000 - 80{,}000 = \pounds125{,}000$.

Recommendation: on expected value the new product is higher ($\pounds160{,}000$ versus $\pounds125{,}000$), so the numbers favour launching. But the new product needs far more cash up front and carries more risk, and the probabilities are estimates, so a cautious firm might prefer the cheaper extension. Markers reward correct EV at each node, correct netting of cost, the right numerical choice, and a comment that recognises risk and the limits of the data.

A decision tree quantifies options using probabilities and expected values, but its limitations matter.

The probabilities are estimates, often subjective, so a small change in them can flip the recommended choice; the monetary outcomes are forecasts that may prove wrong; the technique ignores qualitative factors (the firm's objectives, ethics, staff and reputation) and the time value of money; and it can give false confidence in a single number. It also assumes the manager has identified all the options and outcomes. So a tree is a useful aid that structures a decision, not a decision-maker in itself. Markers reward developed limitations (unreliable probabilities, ignored qualitative factors, false precision) explained rather than listed.