SPC Data Collection & Interpretation 2 — Questions and Answers
Question 1: A quality engineer wants to track the number of defects per unit on an assembly line. Which type of data is this?
- Attribute data (Correct answer)
- Continuous data
- Variable data
- Interval data
Correct answer: Attribute data
Count of defects per unit is attribute (discrete) data because it represents countable occurrences rather than measured values.
Question 2: When collecting data for SPC, a rational subgroup is best described as:
- A sample collected under the same conditions to minimize within-group variation (Correct answer)
- The largest sample size feasible for the process
- A random selection from the entire production run
- A subgroup that spans multiple shifts to capture all variation
Correct answer: A sample collected under the same conditions to minimize within-group variation
A rational subgroup is formed so that variation within the subgroup is minimized and only represents common cause variation, making between-subgroup variation easier to detect.
Question 3: Which sampling strategy is MOST appropriate when a process has known periodic cycles or patterns?
- Systematic sampling aligned to cycles (Correct answer)
- Simple random sampling
- Convenience sampling
- Judgment sampling
Correct answer: Systematic sampling aligned to cycles
Systematic sampling aligned with known cycles ensures the pattern is captured rather than aliased or missed during data collection.
Question 4: On a check sheet used for defect collection, a tally mark represents:
- One occurrence of a specific event or defect type (Correct answer)
- The total defect rate for the shift
- The control limit for that category
- A coded inspector ID
Correct answer: One occurrence of a specific event or defect type
Each tally mark on a check sheet records one occurrence, enabling easy frequency counts by category.
Question 5: A dataset has values: 10, 12, 11, 100, 13. The value 100 is most likely a(n):
- Outlier that should be investigated before analysis (Correct answer)
- Valid data point that must always be included
- Median of the dataset
- Common cause variation
Correct answer: Outlier that should be investigated before analysis
A value far removed from the others is an outlier and should be investigated for measurement error or a special cause before using the data.
Question 6: The primary purpose of stratifying data during collection is to:
- Separate data into meaningful subgroups to identify hidden sources of variation (Correct answer)
- Reduce the total sample size needed
- Eliminate the need for control charts
- Ensure all data follows a normal distribution
Correct answer: Separate data into meaningful subgroups to identify hidden sources of variation
Stratification splits data by factors like shift, machine, or operator to reveal variation sources that would be obscured in pooled data.
Question 7: Which measure of central tendency is LEAST affected by extreme values in a dataset?
- Median (Correct answer)
- Mean
- Range
- Standard deviation
Correct answer: Median
The median is a positional measure that is resistant to outliers, unlike the mean which is pulled toward extreme values.
SPC Data Collection & Interpretation 2
A quality engineer wants to track the number of defects per unit on an assembly line.
Which type of data is this?