0

I wonder how should I report the results of categorical variables with reference level in linear model?

The response variable of my model is Duration (time taken to reach 25C for an animal model). My explanatory variables are Lamp types (3 lamp types), Distance and etc.

For instance, in the image below. Distance 30 is being compared to distance 20 (20cm and 30cm respectively), so should I say "there was also a significant difference in heating efficiency between distances (t= 22.52, P=<0.0001), with distance 30 taking 8.96min more than distance 20" ?

For Lamp types, as there are three levels I did a post hoc Tukey test. Should I directly report the p values and parameter estimates from the Tukey test and ignore the one in the summary()?

Many thanks!

Summary(model) results

Post hoc tukey test

kjetil b halvorsen
  • 77,844
Sammi
  • 1

1 Answers1

0

This is really a FAQ---but so far no good reference (duplicate) answer. The short answer is that the omitted reference level has an implied coefficient of zero. It would be useful if software presented the results that way!

For a post with more details see What to do in a multinomial logistic regression when all levels of DV are of interest?

kjetil b halvorsen
  • 77,844