ss.rr: Gage R & R (Measurement System Assessment)

In SixSigma: Six Sigma Tools for Quality Control and Improvement

Gage R & R (Measurement System Assessment)

Description

Performs Gage R&R analysis for the assessment of the measurement system of a process. Related to the Measure phase of the DMAIC strategy of Six Sigma.

Usage

ss.rr(
  var,
  part,
  appr,
  lsl = NA,
  usl = NA,
  sigma = 6,
  tolerance = usl - lsl,
  data,
  main = "Six Sigma Gage R&R Study",
  sub = "",
  alphaLim = 0.05,
  errorTerm = "interaction",
  digits = 4,
  method = "crossed",
  print_plot = TRUE,
  signifstars = FALSE
)

Arguments

var	Measured variable
part	Factor for parts
appr	Factor for appraisers (operators, machines, ...)
lsl	Numeric value of lower specification limit used with USL to calculate Study Variation as %Tolerance
usl	Numeric value of upper specification limit used with LSL to calculate Study Variation as %Tolerance
sigma	Numeric value for number of std deviations to use in calculating Study Variation
tolerance	Numeric value for the tolerance
data	Data frame containing the variables
main	Main title for the graphic output
sub	Subtitle for the graphic output (recommended the name of the project)
alphaLim	Limit to take into account interaction
errorTerm	Which term of the model should be used as error term (for the model with interation)
digits	Number of decimal digits for output
method	Character to specify the type of analysis to perform, "crossed" (default) or "nested"
print_plot	if TRUE (default) the plots are printed. Change to FALSE to avoid printing plots.
signifstars	if FALSE (default) the significance stars are ommitted. Change to TRUE to allow printing stars.

Details

Performs an R&R study for the measured variable, taking into account part and appraiser factors. It outputs the sources of Variability, and six graphs: bar chart with the sources of Variability, plots by appraiser, part and interaction and x-bar and R control charts.

Value

Analysis of Variance Table/s. Variance composition and %Study Var. Graphics.

anovaTable	The ANOVA table of the model
anovaRed	The ANOVA table of the reduced model (without interaction, only if interaction not significant)
varComp	A matrix with the contribution of each component to the total variation
studyVar	A matrix with the contribution to the study variation
ncat	Number of distinct categories

Note

The F test for the main effects in the ANOVA table is usually made taken the operator/appraisal interaction as the error term (repeated measures model), thereby computing F as $MS_factor/MS_interaction$, e.g. in appendix A of AIAG MSA manual, in Montgomery (2009) and by statistical software such as Minitab. However, in the example provided in page 127 of the AIAG MSA Manual, the F test is performed as $MS_factor/MS_equipment$, i.e., repeatability. Thus, since version 0.9-3 of the SixSigma package, a new argument errorTerm controls which term should be used as error Term, one of "interaction", "repeatability".

Argument alphaLim is used as upper limit to use the full model, i.e., with interaction. Above this value for the interaction effect, the ANOVA table without the interaction effect is also obtained, and the variance components are computed pooling the interaction term with the repeatibility.

Tolerance can be calculaten from usl and lsl values or specified by hand.

The type of analysis to perform can be specified with the parameter method, "crossed" or "nested". Be sure to select the correct one and to have the data prepare for such type of analysis. If you don't know wich one is for you check it before. It is really important to perform the correct one. Otherwise results have no sense.

Author(s)

EL Cano with contributions by Kevin C Limburg

References

Automotive Industry Action Group. (2010). Measurement Systems Analysis (Fourth Edition). AIAG.

Cano, Emilio L., Moguerza, Javier M. and Redchuk, Andres. 2012. Six Sigma with R. Statistical Engineering for Process Improvement, Use R!, vol. 36. Springer, New York. https://link.springer.com/book/10.1007/978-1-4614-3652-2.

Montgomery, D. C. (2009). Introduction to Statistical Quality Control (Sixth Edition ed.). New York: Wiley & Sons, Inc.

See Also

ss.data.rr

Examples

ss.rr(time1, prototype, operator, data = ss.data.rr, 
	sub = "Six Sigma Paper Helicopter Project", 
	alphaLim = 0.05,
	errorTerm = "interaction",
	lsl = 0.7,
	usl = 1.8,
	method = "crossed")

SixSigma documentation built on Aug. 22, 2023, 9:11 a.m.