OC Curve Calculator (Operating Characteristic)

About the OC Curve Calculator (Operating Characteristic)

The operating characteristic (OC) curve is a graph that shows the probability of accepting a lot as a function of the actual lot quality (defect rate). It is the most powerful tool for evaluating how well a sampling plan discriminates between good and bad lots.

For any given sampling plan (n, Ac), the OC curve plots the acceptance probability at each possible defect rate from 0% to 100%. A steep OC curve indicates a plan that sharply distinguishes between acceptable and unacceptable quality. A flat curve means the plan offers little discrimination, accepting good and bad lots at similar rates.

This calculator computes acceptance probabilities at multiple defect rates for a given single sampling plan, allowing you to evaluate producer's risk (α), consumer's risk (β), and overall plan effectiveness.

By calculating this metric accurately, production managers gain actionable insights that drive continuous improvement efforts and strengthen overall operational performance across the shop floor.

When This Page Helps

The OC curve reveals the true performance of your sampling plan. Without it, you cannot know how likely you are to accept lots of various quality levels, or whether your plan adequately protects against poor quality.

How to Use the Inputs

1. Enter the sample size (n) of your sampling plan.
2. Enter the accept number (Ac).
3. Review the table of acceptance probabilities at different defect rates.
4. Identify the probability at your AQL (should be ~95% for good producer's risk).
5. Identify the probability at your LTPD (should be ~10% or less for adequate consumer's risk).
6. If risks are unacceptable, increase sample size to steepen the OC curve.

Example Calculation

Tips & Best Practices

• Compare OC curves of different plans to select the one that best balances inspection cost and quality risk.
• An ideal OC curve would be a step function — accepting all lots at AQL and rejecting all at LTPD.
• Increasing sample size steepens the OC curve, improving discrimination.
• Use the Poisson approximation when n ≥ 16 and p ≤ 0.1 for faster calculations.
• Plot OC curves for normal, tightened, and reduced plans to understand switching scheme behavior.
• Share OC curves with stakeholders to set realistic expectations about sampling plan performance.

Reading the OC Curve

The x-axis shows the lot defect rate (proportion defective, p). The y-axis shows the probability of acceptance, P(accept). At p = 0, P(accept) = 1 (perfect lots are always accepted). As p increases, P(accept) decreases, eventually approaching 0.

Comparing Sampling Plans

Overlay OC curves of candidate plans to compare them visually. The plan with the steepest drop in the critical quality region offers the best discrimination. Balance this against the cost of larger samples.

OC Curve and Switching Rules

In Z1.4, switching from normal to tightened inspection is equivalent to moving to a steeper OC curve with more protection. Understanding the OC curves at each level helps quality managers set appropriate switching criteria.

Frequently Asked Questions

• What is producer's risk on the OC curve?

  Producer's risk (α) is 1 minus the acceptance probability at the AQL. It is the probability of rejecting a lot whose quality is at the acceptable level. Typically set at 5% (P(accept) = 95% at AQL).

• What is consumer's risk on the OC curve?

  Consumer's risk (β) is the acceptance probability at the LTPD (lot tolerance percent defective). It is the probability of accepting a lot whose quality is at the rejectable level. Typically set at 10%.

• How do I make the OC curve steeper?

  Increase the sample size while adjusting the accept number proportionally. Larger samples provide more information about lot quality, narrowing the transition zone between acceptance and rejection.

• Does the OC curve change with lot size?

  For large lots (N >> n), the OC curve depends primarily on sample size and accept number, not lot size. For small lots where n is a significant fraction of N, the hypergeometric distribution applies.

• Can I use OC curves for double sampling plans?

  Yes. The OC curve for a double sampling plan accounts for the two-stage decision process. It typically steep slightly faster than a single plan with the same average sample number.

• What is the ideal shape of an OC curve?

  Ideally, the curve would be a vertical line at the critical defect rate — accepting everything below and rejecting everything above. In practice, increasing n approaches this ideal.