Standard deviation measures how much an estimate’s optimistic and pessimistic values spread out around the PERT weighted-average mean, quantifying the uncertainty behind a time or cost estimate. A small standard deviation signals a tight, predictable estimate; a large standard deviation signals significant uncertainty requiring more contingency planning.
What Is the Standard Deviation Formula?
The standard deviation formula used in project estimating is SD = (P − O) / 6, where P is the pessimistic estimate and O is the optimistic estimate. This is a simplified approximation drawn from the same Beta distribution logic behind the PERT formula, not the full statistical standard deviation formula taught in a general statistics course.
How Do You Calculate Standard Deviation?
A project manager calculates standard deviation by subtracting the optimistic estimate from the pessimistic estimate and dividing the result by six.
Returning to the server migration task: Optimistic (O) = 8 hours, Most Likely (M) = 14 hours, Pessimistic (P) = 30 hours.
SD = (30 − 8) / 6 SD = 22 / 6 SD = 3.67 hours
This estimate carries a standard deviation of 3.67 hours around its PERT mean of 15.67 hours, calculated earlier using the PERT formula.
How Does Standard Deviation Relate to the PERT Mean?
Standard deviation and the PERT mean work together: the PERT mean provides the single expected estimate, while standard deviation quantifies how much that estimate could realistically vary. Neither figure alone gives a complete picture — the mean without a measure of spread hides how much confidence to place in it.
How Do You Build a Confidence Interval Using PERT and Standard Deviation?
A confidence interval is built by adding and subtracting multiples of the standard deviation from the PERT mean, producing a probability-based range instead of a single number.
Using the server migration estimate (PERT mean = 15.67 hours, SD = 3.67 hours):
- 68.3% confidence interval (±1 SD): 15.67 − 3.67 to 15.67 + 3.67 = 12.00 to 19.34 hours
- 95.5% confidence interval (±2 SD): 15.67 − 7.34 to 15.67 + 7.34 = 8.33 to 23.01 hours
- 99.7% confidence interval (±3 SD): 15.67 − 11.01 to 15.67 + 11.01 = 4.66 to 26.68 hours
A project manager can now state, with 95.5% confidence, that the migration task will take between 8.33 and 23.01 hours — a far more useful statement to a stakeholder than a single 15.67-hour estimate presented without any range.
What Do the Standard Deviation Confidence Bands Mean?
The standard deviation confidence bands follow the empirical rule: approximately 68.3% of outcomes fall within 1 standard deviation of the mean, 95.5% fall within 2 standard deviations, and 99.7% fall within 3 standard deviations.
- ±1 SD — approximately 68.3% probability, roughly 34.1% on either side of the mean.
- ±2 SD — approximately 95.5% probability, roughly 47.7% on either side of the mean.
- ±3 SD — approximately 99.7% probability, roughly 49.85% on either side of the mean.
When Is Standard Deviation Most Useful in Project Management?
Standard deviation is most useful for repeatable, similar-type work with enough historical data to establish a meaningful spread, and least useful for unique, one-time deliverables with no comparable data. A manufacturing line producing identical units repeatedly generates a reliable standard deviation; a one-time creative campaign with no comparable prior work does not.
How Does Standard Deviation Appear on the PMP Exam?
The PMP exam tests standard deviation through direct calculation questions and through interpretation questions requiring confidence-interval probabilities, often pairing it with a PERT mean calculation as a distractor. A common exam trap presents both O, M, and P values and asks specifically for the standard deviation, tempting candidates to calculate and submit the PERT mean instead.
Sample Question 1: An activity has an optimistic estimate of 15 days, a most likely estimate of 22 days, and a pessimistic estimate of 41 days. What is the standard deviation? A) 4.33 B) 26 C) 13 D) 7.67
Sample Question 2: What is the approximate probability that an activity finishes within 2 standard deviations of its PERT mean? A) 68.3% B) 95.5% C) 99.7% D) 50%
Sample Question 3: A project’s PERT mean duration is 60 days, with a standard deviation of 5 days. What is the 68.3% confidence interval for completion? A) 50 to 70 days B) 55 to 65 days C) 45 to 75 days D) 58 to 62 days
Sample Question 4: An activity has an optimistic estimate of 12 days, a most likely estimate of 18 days, and a pessimistic estimate of 30 days. What is the standard deviation of this estimate? A) 19 B) 3 C) 6 D) 9
Answers: 1) A — SD = (41 − 15) / 6 = 26 / 6 = 4.33. 2) B — approximately 95.5% of outcomes fall within 2 standard deviations of the mean. 3) B — 68.3% confidence interval = 60 ± 5 = 55 to 65 days. 4) B — SD = (30 − 12) / 6 = 18 / 6 = 3; the PERT mean for this data (19) is a common distractor, not the standard deviation.
Frequently Asked Questions
Is Standard Deviation Part of the Official PMBOK Guide Lexicon?
Standard deviation does not currently appear in PMI’s online lexicon, though the concept has appeared in past editions of the PMBOK Guide and remains a standard topic in PMP exam preparation. Its absence from the current lexicon does not diminish its relevance to estimate uncertainty and risk analysis.
Can Standard Deviation Be Calculated Without PERT?
Standard deviation as used in project estimating requires the same optimistic and pessimistic inputs PERT uses, so the two calculations are built from the same underlying data even though they answer different questions. PERT produces the expected estimate; standard deviation quantifies the uncertainty around it.
What Is the Difference Between Standard Deviation and Variance?
Standard deviation and variance both measure spread around a mean, but variance is standard deviation squared, expressed in squared units rather than the original unit of measurement. Project estimating relies on standard deviation specifically because it stays in the same units — hours, days, or dollars — as the original estimate.
Why Doesn’t Standard Deviation Work Well for One-of-a-Kind Projects?
Standard deviation relies on enough comparable data points to establish a meaningful spread, and a unique, one-time deliverable with no comparable prior work provides no such data. Applying the formula to a single estimate with no historical pattern behind it produces a number without real statistical meaning.
