topic Lost DFs Error in Discussions

Lost DFs Error

Abby — Fri, 09 Jun 2023 00:55:47 GMT

Hi, Good day all,

I have an issue running a fit model for my experiment. I have gone through similar problems posted on the platform but I can't seem to pull through the situation.

I have an experiment with 3 main factors [Stations (n = 2), treatment (n = 14) and time (n = 4)] and 10 replicates per sample. When I run a fit model, I get biased, Lost DFs error (kindly see attached).

I am not sure how to go ahead with this error, since it is vital for me to see the effect test and interactions.

I have also attached a sample of my data to see where I may be getting something wrong.

Re: Lost DFs Error

statman — Sun, 16 Oct 2022 16:03:48 GMT

Welcome to the community. The lost DF's is usually because you have a model with more degrees of freedom than are available in the data set. You need to enter the appropriate model for this data and data type (categorical factors).

Are you sure the 10 "measures" for each treatment are replicates? Or are they repeats?

Re: Lost DFs Error

Abby — Mon, 17 Oct 2022 07:37:38 GMT

The 10 measures are not repeats but replicates

Re: Lost DFs Error

Mark_Bailey — Tue, 18 Oct 2022 15:14:49 GMT

I think it is because you are missing a combination. You do not have Time=1 and Treatment=4 data, so the interaction effect is a problem for estimation with least squares regression.