R Programming Language Certification Quality Control & Assurance 5 — Questions and Answers
Question 1: In R, which `qcc` chart type uses the median moving range to estimate process variability, making it robust to non-normality?
- xbar.one chart with MMR (Correct answer)
- CUSUM chart
- EWMA chart
- T² (Hotelling) chart
Correct answer: xbar.one chart with MMR
The Xbar.one (individuals) chart can use the median moving range instead of the average moving range as a more robust estimator of σ for non-normal data.
Question 2: What R function generates a probability plot (normal probability plot) used to assess whether process data follows a specified distribution?
- qqnorm() or qqplot() (Correct answer)
- hist() with density=TRUE
- ecdf()
- plot(density())
Correct answer: qqnorm() or qqplot()
`qqnorm()` creates a normal Q-Q plot comparing data quantiles to theoretical normal quantiles; `qqplot()` extends this to any reference distribution.
Question 3: In the `SixSigma` R package, `ss.ca.cp()` computes the Cp index. What additional data argument beyond the sample vector is required?
- LSL and USL specification limits (Correct answer)
- Target value only
- Mean and standard deviation
- Subgroup size
Correct answer: LSL and USL specification limits
Cp requires both the lower specification limit (LSL) and upper specification limit (USL) to compute the ratio of the specification width to the process spread.
Question 4: When applying the `demerit()` function in R for weighted defect scoring, what differentiates a 'critical' defect weight from a 'minor' defect weight?
- Critical defects receive much higher weights (e.g., 100) vs minor (e.g., 1) (Correct answer)
- Critical defects cause chart shutdown while minor defects trigger warnings only
- Critical defects are tracked on c-charts while minor defects use p-charts
- Critical and minor use the same weight but different control limits
Correct answer: Critical defects receive much higher weights (e.g., 100) vs minor (e.g., 1)
Demerit systems assign severity weights (e.g., 100 for critical, 50 for major, 10 for minor, 1 for incidental) so that weighted defect counts reflect true quality impact.
Question 5: In R, what is the correct way to perform a multi-variate control chart (T² chart) for monitoring two correlated quality characteristics simultaneously?
- mqcc(data, type='T2') (Correct answer)
- qcc(data, type='multivariate')
- T2chart(data, vars=c('x','y'))
- hotellingT2(data)
Correct answer: mqcc(data, type='T2')
`mqcc()` from the `qcc` package handles multivariate quality control, with `type='T2'` specifying the Hotelling T² control chart for correlated variables.
Question 6: In R, when a `qcc` control chart flags a violation, which list element of the returned object contains the indices of out-of-control points?
- $violations
- $out.of.control (Correct answer)
- $signals
- $flagged
Correct answer: $out.of.control
The `qcc` object stores indices of out-of-control observations in the `$out.of.control` element, which can be extracted and examined programmatically.
Question 7: Which R approach correctly implements a two-sided tolerance interval that contains at least 95% of the population with 99% confidence?
- tolerance::normtol.int(x, alpha=0.01, P=0.95, side=2) (Correct answer)
- t.test(x, conf.level=0.99)
- qnorm(c(0.025, 0.975), mean(x), sd(x))
- confint(lm(x~1), level=0.99)
Correct answer: tolerance::normtol.int(x, alpha=0.01, P=0.95, side=2)
`normtol.int()` from the `tolerance` package computes exact normal tolerance intervals; `alpha` sets the confidence level and `P` sets the minimum proportion covered.
R Programming Language Certification Quality Control & Assurance 5
In R, which `qcc` chart type uses the median moving range to estimate process variability, making it robust to non-normality?