Introduction to Control Charts

Overview

• Control charts, also known as Shewhart charts, honor Dr. Walter Shewhart.
• Transitioning focus from Run charts to Control charts.

Run Chart Review

• Axes: Time on horizontal (X-axis), metric on vertical (Y-axis).
  • Examples: patients, months, days, days of the week.
• Median: Used as the center line in Run charts, denoted as X with a tilde.
• Plot: Data over time with median indicating the 50th percentile.

Control Chart Basics

• Axes: Time on horizontal, measure of interest on vertical.
• Mean as Center Line: Replaces median with the mean, denoted as X-bar (X with a line over it).
• Control Limits:
  • Upper Control Limit (UCL) and Lower Control Limit (LCL).
  • Define variation in the process.
  • Tighter variation = closer limits; wider variation = farther apart limits.
• Sigma Limits:
  • Also known as SL in software, marked with a symbol sigma (σ) often with a hat.
  • Not standard deviations but estimates of dispersion.

Understanding Control Limits

• Standard Deviation vs. Sigma Limits:
  • Standard Deviation: Single statistic for average dispersion.
  • Control Limits: Process boundaries changing over time.
  • Control limits are not standard deviations.

Data Requirements

• Run Chart: Can be created with ~10 data points.
• Control Chart: Requires at least 15 data points, preferably 20.
  • More data needed due to sensitivity of mean to point-to-point variation.
  • Longer data collection time required (e.g., monthly data).

Practical Application

• Improvement Initiatives:
  • Start with Run charts when data is limited.
  • Transition to Control charts as more data is collected.

Next Steps

• Introduction to analyzing and interpreting Control charts.
• Further details on statistical process control can be found in specialized literature.

Conclusion

• Understanding the differences and requirements for Run and Control charts is essential for effective data analysis in improvement projects.