1
$\begingroup$

I need to interpret this summary. I have a response variable with 5 levels (response) and an explanatory variable (taxon) with 2 levels (raptor and others). Can someone help me?

Call:

multinom(formula = response ~ taxon, data = data, trace = FALSE)

Coefficients:
          (Intercept) taxonraptor
CA         -1.7045477  -0.2410289
Colission  -1.7045071 -12.1472903
E          -0.4519443  -2.1870233
NR         -3.0912802  -8.2838816
OA        -14.3087225  11.6707716
$\endgroup$

1 Answer 1

0
$\begingroup$

Difficult to read your results. You need to provide more info about your dependent variable. If you've estimated a multinomial logit model, then you should expect to see 5 parameters:

  • 4 constants which capture the effects of being one level relative to the baseline/reference

  • 1 effect of "taxon" on the level probability

Now it could also be the case that by default "multinom" estimate level-specific effects for "taxon" - In this case you would have four 4 constants + 4 taxon-related effects. But again, your results are difficult to read.

$\endgroup$
1
  • $\begingroup$ As a side note, this model does not seem very interesting. A simple 2-way contingency table would do the job. $\endgroup$
    – Nicolas K
    Feb 7, 2019 at 13:39

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.