Module 6 Attribute Systems Assessment

Lesson M6-05

Text: Section 5

Duration: 30 minutes

Attribute Agreement Analysis Examples

Two examples of Attribute Agreement Analysis are provided.  We focus on interpreting the results.  The first example illustrates the creation of a worksheet template for the study.  We review the graphical output and extensive statistical output that quantifies the within appraiser agreement, between appraiser agreement, and agreement with reference (true) values.

Downloads

MSA Attribute Agreement Example Template Inspection of Taper.MTW

MSA Attribute Agreement Example Print Integrity.MTW

Module 6 Attribute Systems Assessment

Lesson M6-04

Text: Section 5

Duration: 23 minutes

Attribute Agreement Analysis Method

Lesson 4 introduces the Attribute Agreement Analysis method which may be used to assess binary, ordinal, or nominal responses.  We define some descriptive statistics that assess performance: % Agreement, Kappa, Kendall’s Coefficient of Concordance.  These statistics may be used to assess within appraiser agreement, between appraiser agreement, and agreement with known values.  The hypothesis tests associated with Kappa statistics are reviewed. The key steps and guidelines for conducting an Attribute Agreement Analysis study are presented.  Finally, acceptance criteria are discussed.

Module 6 Attribute Systems Assessment

Lesson M6-03

Text: Section 5

Duration: 19 minutes

Attribute Gage Study – Analytic Method with Minitab & Exercise

The next lesson completes the study of the Analytic Method.  Here, we use Minitab to analyze the data and interpret the results.  We focus on the interpretation of the output.  Specifically, the repeatability estimates as well as the test for bias, which focuses on the probability of accepting / rejecting parts that measure right at the specification limit.  We discuss how the bias affects the classification errors that may be made (probability of accepting non-conforming parts and rejecting conforming parts).  You are given an opportunity to apply the Analytic Method in an exercise.

Downloads

MSA Attribute Example-2.MTW

Exercise 1-4.10.MTW

Module 6 Attribute Systems Assessment

Lesson M6-02

Text: Section 5, pages 4 – 12

Duration: 19 minutes

Attribute Gage Study – Analytic Method

In Lesson 2, we introduce the Analytic Method which may be used to assess attribute measurement systems when variable measurements are also possible (for the study).  We discuss guidelines for using this method as well as the (rather specific) requirements for selecting specimens to include in the study.  We then review the details of constructing the probability plot that is used to estimate repeatability.

Downloads

MSA Attribute Example-2.MTW

Module 6 Attribute Systems Assessment

Lesson M6-01

Text: Section 5, pages 1 – 3

Duration: 18 minutes

Overview & Simple Assessment Method

We introduce this module with a review of attribute data and the different types (e.g. binary, nominal, etc.).  We then introduce a simple assessment method for any attribute system.  While this approach does not provide any statistical results, it permits an initial determination of the adequacy of an attribute measurement system.

Downloads

MSA Attribute Example-1.MTW

Module 6 Attribute Systems Assessment

Lesson M6-00

Text: Section 5

Duration: 1 minute

Attribute Systems Assessment Module Introduction

Module 6 covers several approaches for assessing attribute measurement systems.  The methods assess both accuracy and precision of attribute systems.  We cover a simple approach, the Analytic Approach (which requires variable measurements as well), and Attribute Agreement Analysis.  For each approach, we provide examples and opportunities to practice via exercises.

Module 5 Linearity & Bias Assessment

Lesson M5-03

Text: Section 4, pages 9 – 11

Duration: 15 minutes

Linearity with Respect to Precision & Exercises

In the last lesson, we revisit linearity with respect to precision.  Again, we illustrate the concepts with simple scatter plots.  We also introduce the use of hypothesis tests that may be used to objectively test two or more variances for equality.  An exercise gives you an opportunity to assess linearity with respect to precision in Minitab and a final exercise provides a summary of this module.

Downloads

Exercise 1-4.8.MTW

Module 5 Linearity & Bias Assessment

Lesson M5-02

Text: Section 4, pages 5 – 8

Duration: 24 minutes

Linearity & Bias Studies in Minitab & Exercise

In the next lesson, we perform linearity and bias studies in Minitab.  We learn how to interpret the statistical tests of significance for linearity and the tests of bias at each reference value.  We focus on interpreting the results.  An exercise gives you a chance to perform this analysis in Minitab.

Downloads

Exercise 1-4.7.MTW

Module 5 Linearity & Bias Assessment

Lesson M5-01

Text: Section 4, pages 1 – 4

Duration: 12 minutes

Linearity & Bias Assessment Overview

In Lesson 1, we review the definition of linearity with respect to accuracy.  We discuss the use of a 1-sample t test to assess bias and whether it’s statistically significant.  We look at common reasons for bias and present an example of using a simple scatter plot to visualize whether bias exists.  Finally, some reasons for nonlinear systems are presented.

Downloads

MSA Linearity Example-1.MTW

Module 5 Linearity & Bias Assessment

Lesson M5-00

Text: Section 4

Duration: 1 minute

Linearity & Bias Assessment Module Introduction

Module 5 covers linearity assessment both respect to accuracy (bias) and precision.  We illustrate the concepts using simple graphical techniques (scatter plots) and also cover the relevant Minitab options for assessing linearity.

Module 4 SPC for Measurement Systems

Lesson M4-02

Text: Section 3, pages 4 – 19

Duration: 18 minutes

Control Chart Examples

In Lesson 2, we present several examples of common control charts along with their interpretation.  Calculations are provided for reference.  Some potential causes for measurement system stability are presented and we finish with some short review questions.

Downloads

MSA SPC Example-1.MTW

MSA SPC Example-2.MTW

MSA SPC Example-3.MTW

MSA SPC Example-4.MTW

Link to the SPC-Process Capability Course

Module 4 SPC for Measurement Systems

Lesson M4-01

Text: Section 2, pages 1 – 3

Duration: 16 minutes

Measurement Systems Stability

In Lesson 1, we discuss the concept of measurement systems stability and the benefits of ongoing stability assessment is described.  The use of Statistical Process Control charts to monitor measurement systems is presented.  The types of charts commonly used to monitor location (accuracy) and variation (precision) are introduced.  Important considerations for any SPC implementation are briefly mentioned but the details are left to a course in SPC.

Module 4 SPC for Measurement Systems

Lesson M4-00

Text: Section 3

Duration: 3 minutes

SPC for Measurement Systems Module Introduction

Module 4 introduces the idea of using Statistical Process Control charts to monitor Measurement Systems performance over time.  While the details of control chart methodology and implementation are left for another course, we focus here on the concepts along with some examples of control charts that can monitor both measurement system’s accuracy and precision.

Module 3 Gage R&R Studies

Lesson M3-09

Text: Section 2

Duration: 19 minutes

Gage R&R with Other Factors (Expanded Gage R&R)

The last lesson looks at the use of expanded Gage R&R studies to assess other factors in the study (besides part and operator).  These studies may be useful if other factors (such as within part variation) may be inflating the measurement error.  We provide an example where measurement location has a big influence on the results.  Treating the location as another reproducibility component and ultimately categorizing this variation as part variation allows us to isolate the true error coming from the measurement system.

Downloads

Expanded Gage RR.MTW

Module 3 Gage R&R Studies

Lesson M3-08

Text: Section 2, pages 35 – 39

Duration: 14 minutes

Analysis – ANOVA with Significant Interactions

In Lesson 8, we explore testing for part*operator interactions.  When they are statistically significant, they are part of reproducibility and if not, the variance is added to the error term.  Significant part*operator interactions may explain why operator agreement may depend on the specific part(s) being measured.  The previous exercise is repeated allowing the interaction term to be significant here.

Downloads

Exercise 1-4.6.MTW