| Technique | Type | Best Used When | Factors Supporting Choice | Example Scenario |
|---|---|---|---|---|
| Ratio-Based | Metric-Based | Historical data is available from similar past projects | ✅ Data availability❌ Expert availability not mandatory✅ Fast❌ Limited error margin✅ Works in stable contexts | Avg. 15 test cases per requirement used for estimating new CRM system |
| Extrapolation | Metric-Based | Trends/patterns from large datasets can be used for predictions | ✅ Historical trend data✅ Modeling skills❌ Experts not required❌ Time-consuming✅ Error margins if modeled well | Based on defect detection rate trend across 3 previous releases |
| Three-Point Est. | Metric-Based | Need to account for uncertainty in effort estimates | ✅ Uncertainty estimation✅ Modeling knowledge❌ Heavy expert dependency✅ Good for volatile contexts | Effort estimation = (Optimistic + 4×Most Likely + Pessimistic) ÷ 6 |
| Delphi Method | Expert-Based | Expert team available; no reliable historical data | ✅ Expert availability❌ Historical data not needed✅ Collaborative❌ Can be slow✅ Flexible error range | Experts anonymously estimate test effort, then iterate toward consensus |
| Planning Poker | Expert-Based | Agile teams needing collaborative and quick estimation for backlog items | ✅ Expert availability❌ Requires minimal modeling✅ Time-efficient✅ Fits Agile contexts❌ Historical data not mandatory | Scrum team uses poker cards to estimate testing points for a sprint backlog |
| Wideband Delphi | Expert-Based | Complex projects in sequential (e.g., Waterfall) models needing structured expert collaboration | ✅ Expert availability❌ Data not essential❌ Time-consuming✅ Good in traditional models✅ Suitable for high complexity | Multiple experts discuss in rounds to estimate testing effort for a government system |
| Factor | What It Affects | Example |
|---|---|---|
| Data Availability | Needed for metric-based techniques like ratio or extrapolation | Past defect density data to forecast new release testing effort |
| Expert Availability | Needed for expert-based techniques like Delphi or Poker | Experienced testers available for estimation via collaborative sessions |
| Modeling Knowledge | Required for three-point and extrapolation methods | Ability to use statistical models for uncertainty or trends |
| Estimation Error | Consider if you want a range or average for uncertainty | Three-point estimation provides mean ± deviation |
| Time Constraints | Some techniques are quick (Planning Poker), some are slower | Use lightweight methods in tight schedules (e.g., Agile sprint planning) |
| Context (SDLC Model) | Influences technique type (Agile vs. Waterfall, etc.) | Planning Poker suits Agile; Wideband Delphi fits Waterfall |
🔹 Short Scenario-Based (K4)
Q1. A team working in Agile wants a lightweight, collaborative way to estimate the testing effort for each user story. What technique should be used?
A) Extrapolation
B) Three-point estimation
C) Planning Poker
D) Ratio-based estimation
✅ Answer: C) Planning Poker
Q2. You’re managing a project with detailed historical data and repeatable modules. You want a fast and data-driven way to estimate test cases. Which approach suits best?
A) Delphi
B) Wideband Delphi
C) Extrapolation
D) Planning Poker
✅ Answer: C) Extrapolation
🔹 Long Scenario-Based (K4)
Q3. You are the Test Manager for a project that has no reliable historical data, but domain experts with similar project experience are available. The project follows a traditional Waterfall lifecycle. Which estimation technique should you choose?
A) Ratio-based estimation
B) Planning Poker
C) Wideband Delphi
D) Three-point estimation
✅ Answer: C) Wideband Delphi
Explanation: Traditional SDLC + expert availability → Wideband Delphi is ideal.
Q4. Your team is using Scrum with bi-weekly sprints. There is a cross-functional team, and estimation needs to be fast and involve everyone. What technique is most appropriate?
A) Ratio-based estimation
B) Delphi Method
C) Extrapolation
D) Planning Poker
✅ Answer: D) Planning Poker
Explanation: Agile + collaboration + short cycles = Planning Poker.
Q1. Your test team is new, historical data is unavailable, and estimations must be based on discussions with stakeholders and senior testers. Which technique best suits this context?
A) Ratio-based estimation
B) Planning Poker
C) Delphi Method
D) Three-point estimation
✅ Answer: C) Delphi Method
Q2. In a Waterfall project with high-level requirements, experts are available, and management wants a structured, collaborative way to estimate. Which method should be used?
A) Wideband Delphi
B) Planning Poker
C) Extrapolation
D) Ratio-based estimation
✅ Answer: A) Wideband Delphi
Q3. You are working on a new software product with no historical data. However, the team has previously delivered similar projects. Which method could be most appropriate?
A) Extrapolation
B) Planning Poker
C) Delphi Method
D) Ratio-based estimation
✅ Answer: C) Delphi Method
Q4. Which technique requires mathematical knowledge and helps calculate standard deviation to manage uncertainty in test effort estimation?
A) Delphi
B) Three-point estimation
C) Wideband Delphi
D) Ratio-based estimation
✅ Answer: B) Three-point estimation
Q5. You have complete historical data from five earlier releases and want a repeatable, data-driven model to forecast this cycle’s effort. What’s the best approach?
A) Wideband Delphi
B) Extrapolation
C) Planning Poker
D) Delphi
✅ Answer: B) Extrapolation
Q6. In an Agile team sprint planning meeting, developers and testers jointly estimate testing effort using cards with values 1, 3, 5, 8, etc. Which technique is being used?
A) Planning Poker
B) Ratio-based
C) Three-point
D) Delphi
✅ Answer: A) Planning Poker
Q7. When is ratio-based estimation NOT suitable?
A) When historical data is available
B) When similar past projects are available
C) When working on a brand-new product with no prior metrics
D) When estimating based on user stories
✅ Answer: C) When working on a brand-new product with no prior metrics
Q8. You are asked to explain estimation error and provide a range instead of a single value. What technique supports this?
A) Delphi
B) Extrapolation
C) Three-point estimation
D) Ratio-based estimation
✅ Answer: C) Three-point estimation
Q9. In a planning session, each team member gives an estimate privately, then estimates are discussed anonymously to reach a group consensus. This is an example of:
A) Ratio-based estimation
B) Wideband Delphi
C) Three-point estimation
D) Planning Poker
✅ Answer: B) Wideband Delphi
Q10. You are managing a distributed Agile team with limited expert availability and no past data. The team wants a fast and lightweight estimation method. Which approach is best?
A) Delphi
B) Planning Poker
C) Three-point estimation
D) Extrapolation
✅ Answer: B) Planning Poker
No comments:
Post a Comment