How to Check a Regression Fit
podcast episode with speaker Fred Schenkelberg
When fitting a line or curve to data, it’s a model. When modeling, it is worth remembering the George Box quote, “Essentially all models are wrong, but some are useful.” Yet, how do we separate a useful model from one that isn’t useful? One step in finding a helpful regression model is to consider the residuals. BTW: Weibull analysis is another term for regression analysis.
Residuals in regression analysis are the differences between the data and the model-predicted values. When the regression ‘fits’ the data, the residuals should represent the naturally occurring experimental (measurement) errors. These should be well-behaved differences that tend to fit a normal distribution centered on zero.
Let’s explore what residuals are, where they come from, and how to evaluate them to detect whether the fitted line (model) is adequate. Not checking or using a poor model is a recipe for major errors when making decisions. Checking residuals is one step to validating a model, yet it’s quick and easy to accomplish.
This Accendo Reliability webinar was originally broadcast on 10 May 2022.
- Social:
- Link:
- Embed:
To view the recorded video, visit the webinar page.
Related Content
Regression Metrics episode
How do I Control a Process That Trends Naturally Due to Tool Wear? article
Are You a Keyboard Jockey? article
Reliability Engineering Statistics course
Starting with Mathematical Foundations with Fred Schenkelberg episode
Making Use of Reliability Statistics
Let’s find the motivation to use reliability statistics and find the resources to learn the statistical tools necessary to succeed.
R Software and Reliability
Let’s explore R software's many capabilities concerning reliability statistics from field data analysis, to statistical process control.
Reliability Distributions and Their Use
Let’s explore an array of distributions and the problems they can help solve in our day-to-day relaibility engineering work.
Practical Application of DOE
Perry discusses the basics of DOE (design of experiments) and fundamentals so you can get started with they useful product development tool.
Fundamentals of Sample Size Determination
Let's discuss the 6 basic considerations to estimate the necessary sample size to support decision making.
Fundamentals of Measurement System Analysis
When we make a measurement, we inform a decision. It’s important to have data that is true to the actual value.
Creating Effective Reliability Graphics
One of the first things I learned about data analysis was to create a plot, another, and another. Let the data show you what needs attention.
PDFs, CDFs, and other ‘Fs’ — What the hell are they?
If you want a really easy introduction or review of these functions that help inform a decision – then check out this webinar.
Discrete Distributions
Sometimes we have to work out how many of them we need (if they make up a fleet) or how many spare parts we need to keep them running.
Why We Use Statistics
Let’s explore the ways we use, or should use, statistics as engineers. From gathering data to presenting, from analyzing to comparing.
How to Check a Regression Fit
Let’s explore what residuals are, where they come from, and how to evaluate them to detect if the fitted line (model) is adequate or not.
Basic Mathematical Symbols and Stuff
This webinar is a light (re)introduction into common mathematical symbols used in many engineering scenarios … including reliability.
Confidence in Reliability
Reliability is a measure of your product or system. Confidence is a measure of you. But we often forget this.
Practical Measurement Systems Analysis for Design
How to calculate Gage discrimination - the more useful result for a design situation, and even how to use it for destructive tests.
What is the Weibull Distribution?
For those who conduct reliability data analysis … or turning a jumble of dots (data points) into meaningful information
Where does the Bell Curve come from?
It is not just a ‘pretty shape’ that seems to work, It comes from a really cool physical phenomena that we find everywhere.
Fundamentals of Hypothesis Testing
Let’s examine a handful of parametric and non-parametric comparison tools, including various hypothesis tests.
Understanding (how bad) the Exponential Distribution (is)
You need to have a good idea of the probability distribution of the TTF of your product when it comes to reliability engineering.
Leave a Reply