Quality Data Leads to Quality Conclusions - Seguridad Mania.com - España y América Latina

 

Making decisions based on data is at the heart of Lean Six Sigma and statistical analysis. But before we can learn about our process from the data, we need to understand how much our measurement system impacts our data. If the measurement system creates a lot of variation, then it will mask the impact of our X’s on the Y.

 

In this 90-minute online session, we will take a deep dive into measurement system analysis, or MSA. We will explore standard tools to assess our measurement system, specifically looking at measurement accuracy and precision, and the quality of the data we measure.

 

Agenda topics:

 

The intent behind MSA. Everything we do with our data, from assessing capability to drawing practical conclusions is based on the premise that our data is “good.” Data validation is critical. We need to focus on what we are measuring and the ways in which this could be distorted.

 

Assessing variable measurement system. We cannot just complete a measurement study with any statistical software and call it done. We need to strive to assess accuracy (how well we hit the target) and precision (how consistently). We will briefly look at scenarios where it is appropriate to conduct variable Gage R&R and what we can learn from it. Key topics will include: repeatability, reproducibility, resolution and bias. Unusual situations will also be considered.

 

Assessing attribute measurement systems. Attribute data brings with it its own difficulties and great need for validation, often due to its vague and subjective nature. In practice, we find many problems in addition to accuracy and precision issues. We will discuss how the application of attribute MSA allows us both to assess and improve our processes. In addition, we will go beyond MSA and explore how attribute assessment can also be used to improve the classification of multiple category appraisals.

 

Utilizing data validation questions. Essential to assessing your measurement system is knowing how accurate and precise your data is. There are some instances where a traditional measurement study is not possible. We will explore this exception and see how data validation questions can help us to assess our systems.