beadexperiment Blog

Calculating process capability is a foundational skill in the world of quality management. For many practitioners, Deming’s Red Bead Experiment is an insightful introduction to the statistical phenomena behind variation, system constraints, and performance measurement. But how do you translate the outcomes of the Red Bead Experiment into meaningful process capability calculations? In this article, we’ll guide you step-by-step through the technical analysis, using concepts familiar to Six Sigma, Lean, and continuous improvement professionals.

What Is Process Capability?

Process capability is a statistical measure that quantifies how well a process can produce output within specification limits. In manufacturing and service environments, this typically involves calculating how much natural variation exists in a process and whether the process can reliably meet customer requirements. Key indicators include Cp, Cpk, Pp, and Ppk.

When you bring the Red Bead Experiment into play, you create a simplified—but powerful—model of a process plagued by unavoidable variation. Unlike a “real” factory, your process in the Red Bead Experiment is virtually locked: workers can’t improve their technique, supervisors’ instructions don’t matter, and the only “solution” is to change the process itself (i.e., the bead mix).

Given this fixed system, let’s explore how to apply process capability analysis.

Step 1: Gathering Your Data

In a standard Red Bead Experiment, there are typically:

• Six workers
• Eight draws per worker (historically four days × two samples per day, or adjust as performed)
• Each draw involves 50 beads
• 20% red beads in the bin

Workers draw beads at random, following the process prescribed. Inspectors count the red beads (defects) for each sample, and the results are recorded.

Over the course of the experiment, you’ll collect a table of “# of red beads in each sample” for each worker and each round. This dataset is your goldmine for process capability calculations.

Step 2: Understanding the Theoretical Distribution

Since the beads are well-mixed (and assuming diligent randomization), each draw of 50 beads from a collection containing 20% red is a straightforward statistical scenario:

• Expected defective rate (proportion of reds): 0.20 (20%)
• Expected number of red beads per sample: 50 × 0.20 = 10
• Variance for binomial process: n × p × (1-p) = 50 × 0.20 × 0.80 = 8
• Standard deviation: √8 ≈ 2.83

This binomial model underpins our statistical analysis.

Step 3: Calculating Index Parameters: Cp, Cpk, Pp, Ppk

These process capability indices are designed for continuous data, but they translate to attribute data (like counts of red beads) with care.

Define the Specification Limits

In a production context, the customer (or management) may specify, “No more than 5 red beads per sample,” or perhaps a maximum allowable defect rate of 8%. For demonstration, let’s assume the upper specification limit (USL) is 5 red beads per sample, and the lower specification limit (LSL) is 0 (no negative defects).

Calculate the Process Mean (μ) and Standard Deviation (σ)

From your experimental data, calculate the mean number of red beads per draw and the standard deviation.

For example, after 48 draws across all workers:

• Mean number of red beads per sample: Suppose you recorded these numbers: 12, 9, 10, 11, etc. Add all, divide by the count.
• Standard deviation: Use the formula for sample standard deviation.

Cp (Process Capability Index)

Cp compares the width of the process spread to the width of the specification:

$$ Cp = \frac{USL - LSL}{6σ} $$

If your standard deviation is 2.8 (from above) and USL–LSL is 5–0 = 5:

$$ Cp = \frac{5}{6 × 2.8} \approx 0.30 $$

A Cp < 1 means the process cannot meet requirements—predictable from the Red Bead setup!

Cpk (Process Capability Index, Centered)

Cpk adjusts for the process mean not being centered in the tolerance:

$$ Cpk = min \left( \frac{USL - \bar{x}}{3σ}, \frac{\bar{x} - LSL}{3σ} \right) $$

If your average draw is 10 red beads:

$$ Cpk = min \left( \frac{5 - 10}{3 × 2.8}, \frac{10 - 0}{3 × 2.8} \right) = min ( -0.60, 1.19 ) = -0.60 $$

A negative Cpk signifies the process mean is well outside acceptable limits.

Pp and Ppk

Pp and Ppk use the overall standard deviation (not just within subgroups) and are calculated similarly but typically over longer periods.

• Pp: Overall process spread
• Ppk: Worst-case tail performance

Given the Red Bead process’s underlying instability and lack of improvement, Pp and Ppk will mirror Cp and Cpk.

Step 4: Interpreting the Results

With Cp and Cpk so low (even negative!), it is mathematically clear that the process is not capable. What does this mean for managers?

• No amount of retraining, warning, or worker motivation can improve Cp/Cpk—because the process itself is the constraint.
• All variation is inherent to the system.
• Adjustments (tweaking, firing “bad performers”) are numerically meaningless under stable conditions.

The Red Bead process demonstrates in hard numbers that until the process is changed (reducing the percentage of red beads, better sampling design, or another fundamental system shift), process capability will not improve.

Step 5: Using Process Capability Analysis for Improvement

Here’s how you can turn these results into practical lessons for your organization or training session:

• Quantify baseline process capability with real numbers. Build understanding of statistical variation.
• Show mathematically why worker-focused “interventions” cannot improve results.
• Discuss what would be necessary to achieve a capable process (for example, reducing the red-bead fraction from 20% to 2%).
• Use the Red Bead Experiment’s process capability analysis as a springboard to discuss process redesign, robust system changes, and the Deming philosophy of management.

Conclusion

Process capability analysis within the framework of Deming’s Red Bead Experiment is a striking way to show why focusing on the system—not the worker—is the foundation for quality improvement. By calculating Cp, Cpk, and related indices, continuous improvement teams can reinforce one of the most powerful lessons in all of quality management: You cannot inspect or motivate your way to excellence when the system itself is fundamentally limited.

Ready to see these concepts in action? Try our virtual Red Bead Experiment with your remote team and bring process capability data to life! Visit beadexperiment.com to learn more and access more tools for quality management and continuous improvement.