Lecture 5 Flashcards
In the context of three-point estimation, what is “most likely” and “expected cost? And how are these concepts different?
- Most likely: refers to the mode of the statistical distribution. In other words, it is the highest point of the profitability density functino (PFD), i.e. te peak of the statistical distribution.
- Expected: is the result of an assumed distribution (Triangular or Beta). Can be calculated using a formula with the three estimates: most likely, optimistic and pessimistic. Two commonly used formulas are Triangular and Beta distribution. You can also use techniques like Monte Carlo simulation. The expected cost considers not only the most likely scenario, but also the potential variability and uncertainty associated with the project.
In summary, the “most likely” cost represents the single point estimation that is considered most likely, the “expected” cost takes the profitability distribution of outcomes, providing a more comprehensive view of potential costs.
Why is three-point estimation called three-point estimation?
The defining characteristics of the two distributions (Triangular and Beta) is that they can be defined in terms of three paramters. The characteristics of being defined by three paramters is why we call this technique for three point estimation.
Requires three input parameters: most likely, optimistic and pessimistic.
Three-point estimation consolidates different estimates made by an inidividual, which are usually based on said person’s intuition. As such, this has led some to question what is the purpose of using three-point estimation even is. Explain the benefits of using three-point estimation.
Benefits of using three-point estimation:
* Can be used as a tool for minimizing the probability of some of the cognitive biases for occuring with the use of expert judgment. Expert judgment is prone to cognitive biases, i.e. systematic error of judgment caused by a reliance on system 1 (intution) type of cognition.
* Helps you understand the risks involved. If you only think about the best-case scenario, you might be caught off guard if things don’t go as planned.
If we use the same input parameters, will the output from three-point estimation based on tringular and PERT distribution be different? Explain how and why.
Whether or not PERT and Triangular three-point estimation yield the same output, depends on the relationship between input values.
* If c- a = b - c (i.e. if the distribution is symmetric) then the output will be equal.
* If c - a != b - c (i.e. if the distribution is asymmetric) then the output will not be equal.
Specifically, the output based on Triangular will be greater than the output based on PERT. In other words, the output will be closer to the most likely input when using PERT distribution.
Why does three-point estimation restrict itself to only using the Triangular and PERT distribution?
Three-point estimation typically restricts itself to using the triangular and PERT distributions because they offer a balance between simplicity and accuracy. These distributions are widely understood, straightforward to apply, and have well-defined mathematical properties. While there are other probability distributions available, such as normal or beta distributions, these may introduce unnecessary complexity without significantly improving the estimation process. Therefore, triangular and PERT distributions are favored for their practicality and effectiveness in project management contexts.
(Liker ikke dette svaret. Har bare hentet fra ChatGPT da jeg ikke helt forstår spørsmålet)
