Which statements describe methods for estimating task duration and when each is appropriate?

• A. Analogies are the only method for all tasks.
• B. Expert judgment (calibrated estimate) vs parameter based estimates; reference class (similar tasks) vs bottom up.
• C. Estimates should never use data from similar tasks.
• D. Estimates are unnecessary for repetitive tasks.

Answer: (B)

Estimating how long a task will take depends on choosing methods that fit what you know and what data you have from past work. The best approach combines both human judgment and quantitative thinking. Calibrated expert judgment means you still rely on experienced people, but you adjust their estimates using actual results from similar projects to correct bias and improve accuracy. Parameter-based estimates turn measurable factors like size, complexity, and team productivity into a calculation of duration, giving a more objective, model-driven figure. When you have data from similar tasks, using a reference class lets you anchor your estimate in real-world performance of those tasks. If you can break the work into identifiable components, a bottom-up estimate adds up the duration of each piece for a detailed, precise plan. In practice, you’d choose the method that fits the available data: use calibrated expert judgment or a reference class when data is limited or you want anchors from similar work, and use bottom-up or parameter-based estimates when you can measure factors and detail the work. The idea that analogies are the only method or that data from similar tasks should never be used isn’t accurate, and treating repetitive tasks as if they don’t need estimates ignores how historical results improve planning over time.