Page 1 of 1
MNL model specification & interpretation of the parameters
Posted: 26 Apr 2022, 07:24
by Yashin
Dear Concern,
Is it possible to identify which of the parameters are statistically significant for the thresholds of p-value?
The estimation procedure in Apollo reports the model fit indices such as AIC, BIC, etc.
Are there any other model fit indices in Apollo that can be reported, to understand the incremental model fit other than AIC,BIC?
Is there another way to include the p-value in the estimation procedure?
Re: MNL model specification & interpretation of the parameters
Posted: 26 Apr 2022, 23:28
by stephanehess
Hi
I suggest you have a detailed look at the manual, available at
http://apollochoicemodelling.com/manual.html
Stephane
Re: MNL model specification & interpretation of the parameters
Posted: 27 Apr 2022, 11:37
by Yashin
Dear Sir,
I have looked at the manual, the P-value is to be printed out in the output settings of the model. however, I tried it doing so but it usually gives an error.
Here is the code:
model_09 = apollo_estimate(apollo_beta, apollo_fixed, apollo_probabilities, apollo_inputs)
# ################################################################# #
#### MODEL OUTPUTS ####
# ################################################################# #
# ----------------------------------------------------------------- #
#---- FORMATTED OUTPUT (TO SCREEN) ----
# ----------------------------------------------------------------- #
apollo_modelOutput(model_09,printClassical = TRUE, printPVal = TRUE,
printT1 = FALSE, printDiagnostics = TRUE, printCovar = FALSE,
printCorr = FALSE, printOutliers = FALSE, printChange = FALSE)
But this gives the following error:
Error in apollo_modelOutput(model_09, printPVal = TRUE) :
unused argument (printPVal = TRUE)
Re: MNL model specification & interpretation of the parameters
Posted: 27 Apr 2022, 19:46
by stephanehess
Hi
you need to use
Code: Select all
apollo_modelOutput(model_09, modelOutput_settings=list(printPVal = TRUE))
not
Code: Select all
apollo_modelOutput(model_09, printPVal = TRUE)
Stephane