# Regression Model Analysis

Six Sigma – iSixSigma Forums Old Forums General Regression Model Analysis

Viewing 6 posts - 1 through 6 (of 6 total)
• Author
Posts
• #42166

melvin
Participant

Can someone help me with the following:1-What is T Test, what does is measure and when is it considered good or bad? (Significant)2-What is F Test, what does is measure and when is it considered good or bad?3-What is R Square, what does is measure and when is it considered good or bad.4-What is Adjusted R Square, what does is measure and when is it considered good or bad.5-What is P-Value, what does is measure and when is it considered good or bad.6-What is standared Deviation, what does is measure and when is it considered good or bad.

0
#132988

clb1
Participant

1. t-test – measures difference between two means
2. F-test – measures the difference between two standard deviations
3. R2 – is a crude summary of the variance observed in the data that is explained by terms in a model.
4. Adjusted R2 – same as 3 only modified to take into account model parameters.
5. P value is a measure of significance
6. Standard deviation is a measure of the scatter of the data about a mean.
As for good or bad – it depends on what you are doing and the kind of data you have – there are a bunch of rules of thumb but each one has a list of “yeah-but’s” a mile long.  I’d recommend you check you class notes to determine what your teacher considers to be “good or bad”.

0
#132989

Peppe
Participant

Bob,  significant mean that a test have been done and it produced a result ‘p’ < of 'x'.
It is called significant for p<0,05 ; Very significat for p< 0,01 ; highly significant for p<0,001
Rgs, Peppe

0
#133028

melvin
Participant

Thanks guys for the help

0
#133029

melvin
Participant

Thanks guys for the help

0
#133184

McMurray
Participant

The F-test measures the difference between 2 variances.  R^2 is the coefficient of determination and tells you how much of the response is explained by the variables in your regression model.  R2-adj accounts for the DOF, not your model parameters.
And there are no degrees of significance, a statistical parameter is either significant or not based on the level of significance chosen for the test.

0
Viewing 6 posts - 1 through 6 (of 6 total)

The forum ‘General’ is closed to new topics and replies.