Validation of Existing Time Standards
Check time standards against comparable data, accounting for method, rating, allowances and variability.
Updated:
Validate a standard by comparing equivalent quantities
A standard time describes defined work under a specified method and conditions. It is not an activity percentage, order elapsed time, or attendance record. Observed waiting does not itself prove that an operation's basic time is incorrect; the loss may lie outside the standard's work content.
Time study does not inherently select ideal cycles, and sampling does not eliminate observer effects. Both need defensible design and documentation.
Review the specification before measuring
Check the work unit, method, product, quality, equipment, skills, performance rating, allowances, and date. Identify setup and contingencies included per unit versus those planned separately.
If the method changed, distinguish verifying the old standard under its conditions from establishing a new one. Do not mix the objectives. Examine materials, breakdowns, and mix before attributing a discrepancy to individual performance.
Design complementary evidence
Use element times when task duration and variation are needed. Use sampling to estimate activity distribution over a representative period. Converting that distribution into time per unit also requires compatible labor time and output, justified rating, and defined allowances.
ILO work-study guidance treats method and measurement as related techniques. Occupancy is not a substitute for rating and does not determine an allowance.
Teaching comparison
Assume 2,400 labor-minutes, 1,200 good units covering the same scope, estimated productive share 0.70, rating factor 1.05, and a 12% allowance on normal time:
- Estimated productive time per unit: 2,400×0.70/1,200 = 1.40 min.
- Normal time: 1.40×1.05 = 1.47 min.
- Proposed standard time: 1.47×1.12 = 1.6464 min/unit.
These are example assumptions, not customer results. Against an existing 1.60-minute standard, the difference would be about 2.9%. Being within a supposed universal 5% threshold does not establish validity. Examine uncertainty and the decision's agreed tolerance.
If allowances are defined as a fraction of final standard time, divide by 1−a rather than multiply by 1+a. Do not apply both conventions or double count delay already included.
Precision and discrepancies
An interval for p reflects sampling uncertainty under its model, not all uncertainty in output, rating, or allowances. NIST explains binomial intervals. Correlated data need suitable analysis; reaching nominal n is insufficient.
Separate changes in method, operating conditions, and measurement error. Investigate before changing planning or incentives. Document which conclusions are supported and which remain hypotheses.
Decide and retain traceability
Keep the previous standard, proposal, sources, period, limitations and approving responsibilities. A review can confirm, revise or defer a standard; it does not make all nonproductive time an avoidable cost.
Observed percentages do not themselves redefine recorded working hours or pay rights. Statistical observations do not establish automatic legal compliance.