STATS_GAGE_RNR

Measurement System Analysis (Gage R&R) extension for IBM SPSS Statistics.
Implements AIAG MSA 4th Edition methods for evaluating gauge repeatability and reproducibility across Crossed, Nested, Type 1, and Expanded study designs.

---

Features

• Study types — Crossed, Nested, Type 1 Gage Study, Expanded (multi-factor)
• Estimation methods — ANOVA, Xbar-R, REML (via lme4)
• Output tables — Variance Components, Study Variation (6σ/5.15σ), NDC, Gage R&R Metrics, Control Chart Parameters, Operator Summary, Part Summary, Confidence Intervals (Wald / Parametric Bootstrap), Operator Bias & Consistency, SNR & Discrimination, Misclassification Risk, Measurement Uncertainty Budget (ISO GUM)
• Charts — Components of Variation, R Chart, Xbar Chart, By-Part, By-Operator, Interaction Plot, Run Chart, Operator Bias Chart, Residuals, Discrimination Chart, Heatmap
• Interactive HTML report — Plotly-based report opens in browser on every run
• Generate Study Worksheet — randomised or systematic run order for Crossed / Nested designs
• Tolerance analysis — % Study Var, % Tolerance, or both; Cg / Cgk for Type 1

---

Requirements

• IBM SPSS Statistics 28 or later
• R 4.4.1 or later
• R packages:
  • lme4 (required — REML estimation and Expanded GRR)
  • jsonlite (required — HTML report data serialisation)
  • ggplot2, plotly, htmlwidgets, htmltools (optional — enhanced HTML report)

---

Installation

Via Extension Hub (recommended)

1. Open IBM SPSS Statistics
2. Navigate to Extensions → Extension Hub
3. Search for STATS GAGE RNR and click Install
4. Restart SPSS — the procedure appears under Analyze → Quality Control → Measurement System Analysis (Gage R&R)

Manual installation

1. Download STATS_GAGE_RNR.spe
2. Navigate to Extensions → Extension Bundles → Install Local Extension Bundle
3. Select the downloaded .spe file and click Open
4. Restart SPSS

---

Usage

Dialog

Analyze → Quality Control → Measurement System Analysis (Gage R&R)

Assign your measurement, part/sample, and operator variables, select a study type and analysis method, then click Run.

Syntax

STATS GAGE RNR
  VARM=Measurement VARP=Part VARO=Operator
  STUDYTYPE=CROSSED  METHOD=ANOVA
  /OPTIONS SIGMA=SIXSIGMA ESTIMATION=REML INTERACTION=YES REMOVEP=YES INTALPHA=0.25
           LSL=0 USL=10 TOLMETHOD=TOTALTOL
  /OUTPUT  ANOVA=YES VARCOMP=YES METRICS=YES NDC=YES CCPARAMS=YES
           OPSTATS=YES PARTSTATS=YES CI=YES OPBIAS=YES SNR=YES
  /CHARTS  COMPONENTS=YES RCHART=YES XCHART=YES BYPART=YES
           BYOPERATOR=YES INTERACTION=YES RUNCHART=YES
  /STUDYINFO GAGENAME='Torque gauge' BY='A. Saraswathy' SPEC='ISO 9001'.

Key subcommands

Subcommand	Purpose
STUDYTYPE	CROSSED | NESTED | TYPE1
METHOD	ANOVA | XBARR
/OPTIONS ESTIMATION	REML | ANOVAVAR
/OPTIONS SIGMA	SIXSIGMA | FIVE15SIGMA
/OPTIONS INTALPHA	Interaction pooling α threshold (default 0.25)
/GENERATE	Generate study worksheet: PARTS= OPERATORS= REPLICATES= RUNORDER=RANDOM|SYSTEMATIC

Full syntax reference and option descriptions are available in the extension help file (STATS_GAGE_RNR.htm) — click ? in any dialog.

---

Validation

A complete QA test suite (STATS_GAGE_RNR_QA_Test_Rules.docx) covering 36 test cases across 12 sections is included in the repository.
Five reference CSV datasets are provided for reproducing all expected outputs:

File	Design	Expected %GRR
GRR_DS1_crossed_good.csv	10 × 3 × 2 Crossed	≈ 2.6% (Acceptable)
GRR_DS2_crossed_poor.csv	10 × 3 × 2 Crossed	≈ 33% (Unacceptable)
GRR_DS3_nested.csv	30 parts nested in 3 operators	σ_EV ≈ 0.12
GRR_DS4_type1.csv	1 part × 1 operator × 25 reps	Cg ≈ 0.85, Cgk ≈ 0.67
GRR_DS5_expanded.csv	10 × 2 ops × 2 days × 2 reps	Day effect ≈ +0.05

---

Statistical references

• AIAG Measurement Systems Analysis Reference Manual, 4th Edition (2010)
• Burdick, Borror & Montgomery, Design and Analysis of Gauge R&R Studies, ASA-SIAM (2005)
• Wheeler, EMP III: Evaluating the Measurement Process, SPC Press (2006)
• ISO/IEC Guide 98-3 (GUM) — Measurement uncertainty

---

License

GPL ≥ 2.0

---

Contributors

• Aruna Saraswathy