Table of Contents
Attribute Chart Guide
Learn how to track defect rates and attribute data in 5 simple steps. Create p, np, c, and u control charts to monitor your process.
Start Attribute ChartWhat Is the Attribute Chart Tool?
Attribute control charts are used to monitor processes where you count defects or classify items as pass/fail rather than measuring a continuous dimension. Unlike variable charts (XmR, X-bar R) which track measurements, attribute charts track proportions, counts, or rates of defectives and defects over time.
Chart Types Explained
- p Chart (Proportion Defective)
- Tracks the fraction of defective items in each sample. Use when sample sizes vary and you classify items as good or defective. Example: percentage of solder joints that fail inspection per lot.
- np Chart (Number Defective)
- Tracks the count of defective items per sample. Use when sample size is constant. Example: number of defective labels out of 200 labels per shift.
- c Chart (Defect Count)
- Tracks total defects per inspection unit or lot. Use when the opportunity area is constant. Example: number of paint blemishes per car body.
- u Chart (Defects per Unit)
- Tracks the defect rate per unit when the opportunity count varies. Example: scratches per square meter of glass when panel sizes differ.
When Should I Use This Tool?
- When your data is pass/fail, good/bad, or defect counts rather than continuous measurements
- To monitor defect rates over time and detect shifts in process performance
- When validating that a corrective action reduced the defect rate
- For visual inspection results, go/no-go gage checks, or acceptance sampling data
Before You Start
Make sure you have the following ready:
- Count data: number of defectives or defects per sample (at least 2 samples, 20+ recommended)
- Sample sizes (n) for p, np, and u charts -- how many items or opportunities were inspected per sample
- Know which chart type fits your data: p/np for defective items, c/u for defect counts
1 Setup - Choose Your Chart Type
Give your study a name, choose the chart type (p, np, c, or u), and optionally add notes for context. Each chart type card shows a brief explanation to help you decide.
- Title
- A descriptive name for your study, e.g. "Solder defect rate - Line 3" or "Label inspection - Shift A".
- Chart Type
- Select p (proportion), np (number defective), c (defect count), or u (defects per unit). If unsure, p chart is the most common starting point.
- Notes (optional)
- Additional context such as process conditions, operator, or reason for the study.
Click to view screenshot
2 Enter Your Data
Enter your attribute data by typing values directly into the table or pasting from a spreadsheet. The columns change based on the chart type you selected.
Columns by Chart Type
- p / np Chart
- Enter defectives (d) and sample size (n) for each sample. Defectives must be less than or equal to n.
- c Chart
- Enter defect count (c) for each sample. No sample size needed -- the opportunity area is assumed constant.
- u Chart
- Enter defect count (c) and the number of units/opportunities (n) for each sample.
Click to view screenshot
3 View Your Control Chart
The tool calculates control limits and plots your data. Points outside the control limits are highlighted as out-of-control signals.
- Center Line (CL)
- The average proportion, count, or rate across all samples. This is the process baseline.
- Upper Control Limit (UCL)
- The upper boundary of expected variation. Points above UCL suggest an assignable cause.
- Lower Control Limit (LCL)
- The lower boundary. For attribute charts, LCL is often zero or very close to zero.
Click to view screenshot
4 AI Coach Analysis
The QC Coach reviews your chart results and provides an AI-powered interpretation. It highlights out-of-control points, trends, and suggests areas to investigate. You can ask follow-up questions for deeper insight.
Click to view screenshot
5 Report and Export
Review your complete report with chart image, summary statistics, and data table. Download a professional PDF report to share with your team, include in CAPA documentation, or attach to audit records.
Click to view screenshot
Understanding Your Results
An attribute control chart tells you whether your defect rate is stable (in statistical control) or changing over time. A stable process has all points within the control limits with no obvious patterns. This does not mean the defect rate is acceptable -- only that it is predictable. To improve, you need to change the process itself.
What Do Out-of-Control Points Mean?
A point above UCL means the defect rate was unusually high for that sample -- investigate what was different (new material, operator change, equipment issue). A point below LCL (rare for attribute charts) means an unusually good result -- find out why so you can replicate it.
Tips and Best Practices
- Collect at least 20-25 samples before drawing conclusions. Fewer samples make the control limits unreliable.
- Use a p chart when sample sizes vary. Use np only when every sample has the exact same size.
- For defect counts (scratches, dents, errors), use c when the inspection area is constant and u when it varies.
- Investigate every out-of-control point. Even if the root cause seems obvious, document it for future reference.
- Recalculate control limits after implementing a process change. Old limits may no longer reflect the current process.