cancel
Showing results for 
Show  only  | Search instead for 
Did you mean: 
Submit your abstract to the call for content for Discovery Summit Americas by April 23. Selected abstracts will be presented at Discovery Summit, Oct. 21- 24.
Discovery is online this week, April 16 and 18. Join us for these exciting interactive sessions.
Choose Language Hide Translation Bar
peace
Level I

lack of fit test issues

Hello, I urgently need an answer before the deadline of my task. any quick answer will be highly appreciated.  the lack of fit test p-value and F2019-11-04 (5).png ratio are missing. what could this means. I have attached a screenshot of the results.

1 ACCEPTED SOLUTION

Accepted Solutions

Re: lack of fit test issues

The Lack of Fit test is based on an analysis of variance. The test statistic is the F ratio. It compares the mean square for the pure error to the mean square for the lack of fit by ratio. Your mean square for pure error is 0. There is no variation between the replicate values.

View solution in original post

3 REPLIES 3

Re: lack of fit test issues

The Lack of Fit test is based on an analysis of variance. The test statistic is the F ratio. It compares the mean square for the pure error to the mean square for the lack of fit by ratio. Your mean square for pure error is 0. There is no variation between the replicate values.

peace
Level I

Re: lack of fit test issues

Thanks Mark. In other words, all I can interpret from the 'lack fit results' is that there is no variation between the replicants? What justification can I use to support the claim that there is no lack of fit then?

Re: lack of fit test issues

The lack of fit test, when possible, provides some evidence to decide if the estimated errors (residuals) in the current model are entirely random effects or a mixture of random and fixed effects. If it is the latter case, then the problem is model bias. All models should be validated with new empirical observations. Use the selected model to predict the outcome for factor levels / combinations that were not included in the data set that trained the model. Do the predictions confirm?