SYE6047 Chap.6 Basic Quality Improvement Tools
Basic Quality Improvement Tools
Tools answer different questions
The seven basic quality tools are useful because each gives structure to a particular evidence problem. A check sheet creates consistent observations. A histogram displays the distribution of a quantitative variable. A Pareto diagram ranks categorical contributions. A cause-and-effect diagram organises candidate causes. A scatter diagram examines association between paired variables.
A control chart preserves time order and tests process behaviour. A graph shows change or comparison without necessarily applying statistical limits.
Collection precedes display
A check sheet fixes categories, event rules, place and period before counting begins. Poor categories create poor charts no matter how polished the output.
Categories should be mutually interpretable, and an other category should be reviewed rather than allowed to absorb the most informative events. Stratification by shift, supplier or product can expose patterns hidden in a total.
Priority is not causation
A Pareto diagram sorts categories from greatest to least contribution and adds a cumulative curve.
It identifies where observed burden is concentrated under the chosen measure. The tallest bar is not automatically the root cause: it may be an outcome category, combine several mechanisms or carry low severity.
Fishbone diagrams then widen candidate thinking across relevant branches, but their arrows remain hypotheses until tested.
Association is not mechanism
A scatter diagram can show direction, form and unusual observations in paired data. It does not establish that one variable causes the other, especially when time, product mix or a third variable affects both.
The best tool answer names the question, selects the display, and states the boundary on interpretation.
What this chapter covers
- 01
Check sheets and operational categories
- 02
Histograms and distribution shape
- 03
Pareto bars and cumulative share
- 04
Cause-and-effect diagrams
- 05
Scatter diagrams
- 06
Control charts
- 07
Line and comparison graphs
- 08
Tool selection by question
Worked example · free
Read a Pareto diagram without inventing causation
- 4Calculate total and cumulative percentages.
- 2Identify the concentration correctly.
- 3State a next investigation without claiming cause.
Key terms
- Check Sheet
- A structured form for recording defined events consistently at the point of observation.
- Histogram
- A display of the frequency distribution of a quantitative variable across intervals.
- Pareto Diagram
- Ordered category bars combined with a cumulative contribution curve.
- Fishbone Diagram
- A structured display of possible causes grouped into useful families.
- Scatter Diagram
- A plot of paired measurements used to examine association, form and unusual points.
Basic Quality Improvement Tools FAQ
When should I use a histogram instead of a bar chart?
Use a histogram for ranges of one quantitative variable, where adjacent bins represent a scale. Use bars for distinct categories. Sorting histogram bins by height would destroy the distribution's meaning. Preserve bin boundaries and units so shape, spread and unusual gaps remain interpretable, while category bars may be ordered to support a separate priority question. State the observation period on either display.
Does the 80/20 rule always hold in a Pareto chart?
No. It is a heuristic about concentration, not a required mathematical pattern. Calculate the actual cumulative contributions and make the priority statement supported by the observed data.
Can a scatter diagram validate a cause?
It can support or weaken a causal hypothesis, but association alone is insufficient. Consider temporal order, confounding, measurement error and a mechanism, then seek a comparison or intervention that distinguishes alternatives.
Exam move
Build a tool-choice table with four columns: data type, question, display and forbidden conclusion. Redraw a small Pareto chart and scatter diagram by hand, then write one supported sentence and one sentence the display cannot support. This makes interpretation faster than memorising tool names alone.
Start a second table with the verbs count, distribute, prioritise, explore, associate and monitor, and place each tool under its primary decision. For a check sheet, define categories, exposure, period and missing-data treatment before tallying. For a histogram, preserve numerical order; for a Pareto diagram, sort categories and calculate cumulative share; for a line or control chart, preserve time.
On a fishbone, label every branch possible until another method validates it. On a scatter plot, identify a plausible third variable and an unusual observation before discussing the pattern. After each chart, write the next unanswered question and choose the tool or investigation able to answer it.
This sequence prevents a visual from being treated as a conclusion and prepares both tool-recognition items and longer project answers that must connect collection, diagnosis and action.
