5-7 Flashcards
(11 cards)
What is Expert Judgement? Explain “Experts” and “Intuition”.
Expert Judgement is using knowledgeable people to estimate project costs, especially when data is limited or the project is new/complex.
What Makes Someone an “Expert”?
➤ Real Experience: They’ve worked on similar projects before
➤ Good Track Record: Their past estimates were usually accurate
➤ Deep Knowledge: They understand key cost drivers
➤ Pattern Recognition: They spot similarities between projects
An expert isn’t just a title—it’s proven skill in making good estimates.
What is “Intuition”?
➤ A gut feeling based on experience
➤ Knowing when an estimate feels too low or high
➤ Quickly spotting red flags others miss
It’s not magic—just years of experience processed rapidly by the brain.
Simple Explanation:
Experts use experience and gut feeling to judge costs when numbers alone don’t tell the whole story.
Can We Trust Expert Judgement?
Yes, When:
➤ Expert has real experience with similar projects
➤ You use multiple experts, not just one
➤ You combine judgement with other methods
➤ Project is similar to what they’ve done before
No, When:
➤ Expert is overconfident or biased
➤ Under pressure to give a “good” answer
➤ Project is totally different from their experience
➤ They have a poor accuracy record
Common Problems:
➤ Optimism bias: Thinking all will go smoothly
➤ Anchoring: Stuck on first numbers heard
➤ Recent memory: Focus on last project only
Bottom Line:
Expert judgement is helpful but imperfect—use with other methods and multiple opinions.
Simple Explanation:
Experts help a lot but can make mistakes; combining views and data improves reliability.
Flashcard 3: How Do We Identify Real Experts?
Strategy 1: Check Their Track Record
➤ Accuracy of past estimates
➤ Projects delivered on budget
➤ Number of similar projects done
Strategy 2: Test Their Knowledge
➤ Understand cost drivers
➤ Explain their reasoning clearly
➤ Know industry and technology
Strategy 3: Ask Other Experts
➤ Recommendations from peers
➤ Reputation in the industry
Strategy 4: Look for Honesty About Uncertainty
➤ Say “I don’t know” when uncertain
➤ Explain assumptions
➤ Give ranges, not just single numbers
Red Flags:
➤ Self-proclaimed experts with no proof
➤ Always 100% confident
➤ Can’t explain their thinking
➤ Old or narrow experience
Simple Explanation:
Real experts prove their skills through past success, knowledge, honesty, and peer recognition.
How Do Other Cost Estimation Methods Depend on Expert Judgement?
Name the 4 methods
and what they need to check
Analogous Cost Estimation:
➤ Experts pick comparable past projects
➤ Adjust for differences
➤ Validate final estimate
Example: Expert decides if a North Sea wind farm is similar to a Norwegian one.
Unit Cost Approach:
➤ Experts select appropriate cost drivers ($/MW, $/km)
➤ Judge applicability of historical data
➤ Decide on averaging or regression methods
Parametric Cost Estimation (Regression):
➤ Experts select variables to include
➤ Prevent data mining by theory-based choices
➤ Validate model results make sense
Monte Carlo Simulation:
➤ Experts provide optimistic, likely, pessimistic estimates
➤ Choose probability distributions
➤ Identify uncertain variables
Why Experts Are Needed:
➤ Numbers lack full context
➤ Every method makes assumptions needing expert input
➤ Experts catch errors in models
➤ New projects need expert adaptation
Key Point:
Expert judgement isn’t separate from other methods—it’s what makes them reliable and accurate.
What are the steps involved in Three-Point Estimation? Give an example.
Phase 1: Generate three estimates using expert judgement
➤ Most optimistic (a): Best-case scenario cost
➤ Most likely (m): Regular expert estimate
➤ Most pessimistic (b): Worst-case scenario cost
Phase 2: Choose statistical distribution
➤ Triangular: Simple average → (a + m + b) / 3
➤ PERT: Weighted average → (a + 4m + b) / 6 (more weight on most likely)
Phase 3: Calculate expected value
➤ Triangular: (a + m + b) / 3
➤ PERT: (a + 4m + b) / 6
Example: Driving time from home to work
➤ a = 8 min (optimistic)
➤ m = 15 min (most likely)
➤ b = 90 min (pessimistic)
PERT expected time = (8 + 4×15 + 90) / 6 = 26 minutes
Simple Explanation:
You estimate best, worst, and most likely cases, then combine them with weights to get a realistic expected cost/time.
What is the difference between “Expected” cost and “Most likely” cost?
Most likely cost (m):
➤ The mode of the distribution
➤ The highest peak of the probability curve
➤ The single most common outcome
Expected cost:
➤ The mean (average) of the distribution
➤ Calculated using triangular or PERT formulas
➤ Represents the overall average outcome
Key Insight:
➤ If distribution is symmetric (m midpoint between a and b), mode = mean
➤ If asymmetric, mode and mean differ
➤ PERT usually gives expected cost closer to mode than triangular
Simple Explanation:
“Most likely” is the single most probable value; “expected” is the weighted average considering all possibilities.
How does Monte Carlo Simulation differ from Three-Point Estimation?
Monte Carlo Simulation:
➤ Handles multiple uncertain variables
➤ Can use any statistical distribution
➤ Runs thousands of random trials to show full outcome range
➤ Requires software/computer simulation
Three-Point Estimation:
➤ Limited to one uncertain variable
➤ Uses only triangular or PERT distributions
➤ No actual simulation—calculations are direct formulas
➤ A quick shortcut for simple uncertainty
Key Point:
➤ With a single variable, three-point estimation matches Monte Carlo results without needing simulation
Simple Explanation:
Monte Carlo is a powerful, complex simulation; three-point is a quick math shortcut for simple uncertainty.
Why use Three-Point Estimation instead of just relying on expert guesses or averages?
Main Purpose:
➤ Adjust for cognitive biases, especially over-optimism
Problem with pure expert judgment:
➤ Experts often too optimistic (planning fallacy)
➤ Fast, automatic thinking (System I) leads to ignoring bad outcomes
How Three-Point Estimation helps:
➤ Forces consideration of worst-case scenarios
➤ Counters human tendency to underestimate risks (Taleb’s insight)
➤ Not just averaging random guesses but systematically adjusting estimates
➤ Results in more realistic, less biased cost estimates
Simple Explanation:
It fights optimism bias by making experts think about worst cases, so estimates aren’t unrealistically low.
Why must children of a WBS(Work Breakdown Structure) node be mutually exclusive and collectively exhaustive? What happens if violated?
WBS (Work Breakdown Structure) organizes all project work into a hierarchy.
Children of each node must be:
Mutually exclusive: no overlap Collectively exhaustive: nothing left out
This ensures complete coverage with no duplication.
Violating this leads to confusion, missed tasks, or wasted effort.
If violated:
Overlapping tasks cause double work or budgeting errors Missing tasks lead to delays, cost overruns, or incomplete deliverables
What are the four common mistakes made when creating a Work Breakdown Structure (WBS)?
Lack of Proper Coding Scheme:
➤ Hierarchy unclear without codes or indentation → relationships between levels unclear.
Violation of the 100% Rule:
➤ Children not mutually exclusive or collectively exhaustive → overlaps or gaps in work.
Insufficient Chunking:
➤ Not broken down to manageable work packages → hard to estimate/manage tasks.
Activities Not Deliverable-Oriented:
➤ Vague tasks (e.g., “carpet – new office”) instead of clear deliverables (e.g., “clean carpet in new office”).
Simple Explanation:
WBS errors come from unclear hierarchy, overlaps/missing work, big vague tasks, and unclear deliverables.
Why is it important to model correlated stochastic variables in Monte Carlo simulation for project cost?
Improved Answer:
In a project, many tasks share common risks and resources—for example, multiple tasks might rely on the same materials, labor market, or weather conditions. Because of this, their costs tend to move together: if material prices go up, all tasks needing those materials likely become more expensive.
If we ignore these connections and treat each cost as if it changes independently, the simulation will underestimate the overall uncertainty. It might show some costs going up while others go down, which rarely happens in reality. This leads to unrealistic and overly optimistic cost estimates.
By modeling correlations, the simulation better reflects how costs actually behave together, capturing the true risk of the project’s total cost. This helps project managers prepare more accurate budgets and contingency plans, reducing the chance of surprises.
