I am running a regression model in R and I want to interpret the variance explained percentage, as well as, R-squared value.
result1<- lme(values~ gas*price, random = ~1|ID, data=dt) r2beta(model=result1,partial=TRUE,method='sgv') MuMIn::r.squaredGLMM(result1)
Effect Rsq upper.CL lower.CL Model 0.15 0.312 0.032 gas 0.08 0.323 0.003 gas:price 0.063 0.123 0.001 price 0.001 0.234 0.000
R2m R2c 0.0596562 0.7687985
I am not quite sure if I understand the output correctly. To my understanding, this whole model can explain 15% of the variance. Furthermore, gas can only explain 8% of variance. On the other hand, when we look at the conditional r2c, which is the percentage of both fixed and random effect, we see high percentage, 0.76. May I ask if someone can explain how to interpret these two different output? I would also like to ask why do Model and R2c result also differ while both are using whole model?