RQ 4 — Scope: Work Breakdown Structure and Estimation
Covers L4a (the work breakdown structure) and L4b (estimation), part by part. Summaries at SN L4a and SN L4b; the detail is in the Scope area.
Answer each in a short paragraph, claim first.
1. Lecture Questions — L4a
1.1 What a WBS is
1. Compare the two halves of the 100% rule. Who pays when each is broken, and why does “if in doubt, include it” follow?
2. A colleague calls a network of tasks the WBS. What is missing, and why must the WBS come first?
3. Why does a Scrum scope tree stop at the story, and what in Scrum is rolling wave?
1.2 Orientation
1. Why does a sponsor read a deliverable-oriented WBS more easily than a phase-oriented one? Use the ambulance-dispatch system.
2. How does “nouns at the top, verbs at the leaves” reconcile Wysocki’s case for activities with PMI’s warning?
3. Why can the grouping at the top of a WBS decide how many managers a project needs?
1.3 Product plus enabling work
1. Why does a WBS contain elements that are not parts of the product?
2. Why does NASA want the WBS to follow the integration strategy, and what goes wrong if it does not?
1.4 Building a WBS
1. Why are reused vendor code and a risk mitigation both work that belongs in the WBS?
2. When do you stop decomposing, and why should depth differ between branches?
1.5 Rolling wave
1. Why is a planning package more than a placeholder, and why is the replanning itself a work package?
1.6 Keeping the WBS alive
1. A project has two sponsors and three releases. Why put one dimension in the tree and tag the rest?
2. Lecture Questions — L4b
2.1 Why estimates go wrong
1. Surveys put the typical overrun near a third; Standish says 89%. What should a manager do with each figure?
2. Why keep estimate, target and commitment as three separate words?
3. The budget is fixed at 600 hours before anyone estimates. What does that imply, and how do you check it?
2.2 Sizing the product
1. Why estimate size before effort, and why count functions rather than lines of code?
2. When is counting preferable to analogy or judgment, and why is an estimate from a count still a range?
2.3 Expert judgment
1. What separates the six families of estimation methods, and why does that make them easy to combine?
2. Your company keeps a project database nobody can search. Which estimating strategy does that hurt, and why?
3. Why do bias fixes change what estimators see instead of asking them to try harder?
2.4 Estimating as a group
1. Why can matching cards in a Planning Poker round still hide disagreement?
2. Which Poker weakness does each rule of Eduardo’s Variation target?
3. Why is Planning Poker a poor fit for an architecture or infrastructure item, and what would you do instead?
4. What is the advantage of story points, and what is the trade-off?
2.5 Parametric models
1. Why must a published model be calibrated before you trust it?
2. Why does effort grow faster than size, and which of COCOMO’s scale factors can a manager change?
2.6 Combining methods
1. No history, incomplete requirements: which methods would you combine, and why?
2. Why read the spread of independent estimates rather than average them?
3. Reading Questions
Cite the reading you used, with a page number where you have one.
3.1 Globerson, “Impact of Various Work-Breakdown Structures on Project Conceptualization” (1994)
1. Globerson calls the WBS “the taxonomy of a project”. What does the metaphor claim that a task list cannot?
2. Why can two WBSs with identical work packages demand different organisations and management styles?
3. Why design the WBS early but keep it under configuration management, and what two kinds of chaos does Globerson warn against?
4. Globerson treats every orientation as valid; PMI insists on deliverables. Reconcile them, and say how the choice affects monitoring and reporting.
3.2 Jørgensen & Boehm, “Formal Models or Expert Judgment?” (2009)
1. Why, in Jørgensen’s account, are formal models little used despite decades of research, and how does Boehm reply?
2. The paper’s Table 1 shows the two processes sharing most activities. Why does that undercut the claim that models are objective?
3. Where does Boehm say a model earns its place even when experts are available, and does Jørgensen accept it?
3.3 Longstreet, “Fundamentals of Function Point Analysis” (2004)
1. Why are function points independent of technology, and what problem did that solve?
2. How can a function-point count improve communication with users and expose scope creep?
3.4 Grenning, “Planning Poker” (2002)
1. Grenning targeted wasted time and lopsided participation, not bias. How does the game also reduce anchoring?
2. Why does Grenning accept less precision at release level?
3.5 Jones, “Software Estimating Rules of Thumb” (1996)
1. Why do manual estimates run optimistic, and why does an estimating tool not cure it?
2. What do Jones’s exponents above and below one tell a manager, and why can no universal ratio turn size into effort?
Acknowledgments
These questions build on the Fall 2025 study-notes selection and on lectures by Eduardo Miranda and David Root; the WBS examples, the 600-hour exercise and Eduardo’s Variation are Miranda’s.
Disclaimer: AI is used for text summarization, polishing and explaining. Authors have verified all facts and claims. In case of an error, feel free to file an issue.