liveBook · Manning

Project variance

As explained in Statistics Every Programmer Needs

Project variance is the sum of the duration variances of the activities on a project's critical path under an independence assumption.

Project variance measures uncertainty in the planned completion time by summing the variances of activities on the critical path, assuming statistically independent activity times. For an activity with optimistic estimate a and pessimistic estimate b, calculate variance as ((b − a) / 6)². A larger value indicates less predictable duration. section 13.4

Use project variance to convert schedule uncertainty into a completion-time probability:

  • Sum only the variances of critical-path activities. section 13.4
  • Take the square root of project variance to obtain the project standard deviation. section 13.4
  • Compute z = (due date − mean project duration) / standard deviation, then evaluate the normal cumulative probability. section 13.4

How Statistics Every Programmer Needs covers it

How deep it goes
Derives activity and project variance, standard deviation, z-scores, and normal probabilities, but assumes independence and normal completion times.
Examples
Uses an automated reporting tool project, activity estimates, a 29-week mean, and a 30-week due date.
What it assumes
Assumes activity estimates, a critical path, statistical independence, and a normal completion-time model.
What you can do after
Apply it via scipy.stats to calculate the probability of completing a project by a specified due date.
How it connects
Connects critical path, standard deviation, z-score, and normal probability distribution.

Related concepts

This explainer was generated by Manning's knowledge graph from the text of Statistics Every Programmer Needs. The section links point to the parts of the book it draws on.