Stratification in Quality Control:
A Detailed Guide for Quality Professionals
Introduction
In quality management, data is everywhere — inspection records, defect counts, downtime logs, customer complaints, audit findings, and process parameters. Yet, despite collecting large volumes of data, many organizations still struggle to identify the real causes of quality problems. One major reason is that data is often analyzed in an aggregated or mixed form, hiding meaningful patterns. This is where Stratification becomes an essential quality tool.
Stratification is one of the Seven Basic Quality Tools, widely used in manufacturing and service industries to separate data into meaningful categories. By breaking down data according to sources such as machines, shifts, materials, operators, or locations, stratification helps reveal hidden trends, root causes, and variation drivers that are otherwise invisible.
In this blog, we will explore stratification in detail — its definition, purpose, principles, types, step-by-step application, real-world examples (especially from manufacturing), advantages, limitations, common mistakes, and its role in modern quality systems such as ISO 9001 and IATF 16949.
What Is Stratification?
Stratification is a quality technique used to divide collected data into subgroups (strata) based on specific characteristics so that patterns and causes of variation can be clearly identified.
In simple words:
Stratification means separating mixed data into meaningful categories to understand where problems are coming from.
For example:
Separating defects by machine
Separating rejection data by shift
Separating customer complaints by region
Separating inspection results by supplier
Without stratification, all these data points would appear as one combined total, masking the real problem areas.
Why Stratification Is Important?
Stratification plays a critical role in problem-solving and decision-making. Its importance can be summarized as follows:
1. Reveals Hidden Patterns
Aggregated data may show an average performance that looks acceptable. Stratified data can expose that one machine, shift, or supplier is responsible for most issues.
2. Supports Root Cause Analysis
Stratification is often the first step before using tools like Pareto charts, control charts, or fishbone diagrams. It helps narrow down where to focus analysis.
3. Improves Decision Accuracy
Corrective actions based on stratified data are more precise and effective because they target the actual source of variation.
4. Saves Time and Cost
Instead of applying broad and expensive solutions, organizations can focus improvements on specific problem areas.
5. Strengthens Audit Readiness
Standards like ISO 9001 and IATF 16949 emphasize data-based decision-making. Stratification demonstrates systematic analysis and control.
Basic Principle of Stratification
The core principle of stratification is:
Variation in results is often caused by variation in sources.
When different sources (machines, people, materials, methods) are mixed together, the causes of variation are hidden. By separating data by source, the relationship between cause and effect becomes visible.
Common Bases for Stratification
Data can be stratified using many criteria, depending on the process and problem. Some commonly used bases include:
|
|---|
Selecting the right stratification basis is crucial for meaningful analysis.
Step-by-Step Procedure to Apply Stratification
Step 1: Define the Problem Clearly
Start with a clear problem statement. Example:
High rejection rate in machining process
Frequent customer complaints
Variation in process output
Step 2: Identify Possible Sources of Variation
Brainstorm potential causes based on:
Machine
Man
Material
Method
Measurement
Environment
This step is often supported by a fishbone diagram.
Step 3: Decide Stratification Categories
Choose categories relevant to the problem. For example:
Rejections by machine
Defects by shift
Complaints by region
Step 4: Collect Data in Stratified Format
Design check sheets or data collection formats that capture stratification information. Example:
Defect type
Machine number
Operator name
Shift
Step 5: Analyze Stratified Data
Once data is separated, analyze each stratum individually. Use tools like:
Pareto charts
Histograms
Control charts
Step 6: Identify Focus Areas
Determine which category contributes most to the problem and prioritize corrective actions.
Real-Time Manufacturing Example
Problem Statement
A machining process shows an overall rejection rate of 4%, exceeding the target of 2%.
Without Stratification
When data is viewed as a total, no clear cause is identified. Management considers machine replacement — an expensive decision.
With Stratification
Rejection data is stratified by machine:
|
Insight
Machine C is the major contributor to rejection.
Further Stratification
Data for Machine C is stratified by shift:
|
Root Cause
Shift 2 shows abnormal rejection, leading to investigation of:
Operator skill
Tool change practices
Setup parameters
Corrective action is focused on Shift 2 of Machine C, resulting in rejection reduction to below 2%.
Stratification in Service Industries
Stratification is not limited to manufacturing. In service sectors, it is equally powerful.
Examples include:
Customer complaints stratified by service location
Call drop rates stratified by time of day
Billing errors stratified by system or operator
Delivery delays stratified by route or vendor
By separating service data into logical categories, organizations can improve customer satisfaction and operational efficiency.
Relationship with Other Quality Tools
Stratification works best when combined with other quality tools:
Check Sheet: Designed to collect stratified data
Pareto Chart: Identifies major contributors within strata
Histogram: Shows distribution within each category
Control Chart: Monitors stability of individual strata
Fishbone Diagram: Helps identify stratification criteria
Stratification often acts as the foundation for effective use of these tools.
Advantages of Stratification
Simple and easy to understand
Low cost and quick to implement
Helps identify true problem sources
Improves effectiveness of corrective actions
Supports data-driven decision-making
Applicable across industries
Limitations of Stratification
Depends on correct selection of categories
Poor data collection can lead to wrong conclusions
Over-stratification may complicate analysis
Does not solve problems by itself — analysis is required
Common Mistakes to Avoid
Stratifying data after collection without proper planning
Choosing irrelevant stratification criteria
Ignoring small but critical strata
Mixing stratified data again during analysis
Not linking findings to corrective actions
Stratification in ISO 9001 and IATF 16949
Both ISO 9001 and IATF 16949 emphasize:
Monitoring and measurement
Data analysis
Risk-based thinking
Stratification supports compliance by:
Demonstrating systematic data analysis
Identifying process risks
Supporting corrective and preventive actions
Improving process control and capability
Auditors often expect evidence of stratified data analysis in problem-solving and continuous improvement activities.
Conclusion
Stratification is a powerful yet often underestimated quality tool. By separating mixed data into meaningful categories, it transforms raw numbers into actionable insights. Whether in manufacturing or service industries, stratification helps uncover root causes, improve process performance, and strengthen quality management systems.
For quality professionals, mastering stratification is essential. It not only enhances problem-solving effectiveness but also builds a strong foundation for advanced statistical and analytical tools. When applied correctly, stratification turns data into knowledge — and knowledge into improvement.
Note: Stratification is sometimes misspelled as "starification," but the correct quality tool term is Stratification.
