[R-sig-ME] Changing reference level in clmm() ordinal regression

[Thierry Onkelinx]

Dear Jessie,

All models have an identical fit. They only differ in the
parametrisation. "a3" estimates the difference between "a3" and the
reference. Hence, changing the reference results in a different
interpretation of the parameter and thus a different estimate.

Best regards,

ir. Thierry Onkelinx
Statisticus/ Statiscian

Vlaamse Overheid / Government of Flanders
INSTITUUT VOOR NATUUR- EN BOSONDERZOEK / RESEARCH INSTITUTE FOR NATURE
AND FOREST
Team Biometrie & Kwaliteitszorg / Team Biometrics & Quality Assurance
thierry.onkelinx op inbo.be
Kliniekstraat 25, B-1070 Brussel
www.inbo.be

///////////////////////////////////////////////////////////////////////////////////////////
To call in the statistician after the experiment is done may be no
more than asking him to perform a post-mortem examination: he may be
able to say what the experiment died of. ~ Sir Ronald Aylmer Fisher
The plural of anecdote is not data. ~ Roger Brinner
The combination of some data and an aching desire for an answer does
not ensure that a reasonable answer can be extracted from a given body
of data. ~ John Tukey
///////////////////////////////////////////////////////////////////////////////////////////


Van 14 tot en met 19 december 2017 verhuizen we uit onze vestiging in
Brussel naar het Herman Teirlinckgebouw op de site Thurn & Taxis.
Vanaf dan ben je welkom op het nieuwe adres: Havenlaan 88 bus 73, 1000 Brussel.

///////////////////////////////////////////////////////////////////////////////////////////



2017-09-22 11:41 GMT+02:00 Jessie Barker <jessiebarker op gmail.com>:
> I have a couple of questions that I asked on StackExchange and someone
> suggested that I ask this mailing list. The post on StackExchange is here (
> https://stats.stackexchange.com/questions/304092/changing-reference-level-in-clmm-ordinal-regression-in-r),
> but I am summarizing it below:
>
> I’m analyzing data from a questionnaire where participants had to rank
> three answers to each question (e.g. 1 = most likely, 3 = least likely).
> They had to give a different rank to each answer, so each question has one
> answer ranked 1, one ranked 2, and one ranked 3. The set of three answers
> is the same for each questions, and there were seven questions.
>
> I want to know whether participants give different ranks to different
> answers, and whether that is affected by question. Here's my model:
>
> model1 <- clmm(rank ~ answer + answer:question + (1+answer|participant),
> data = mydata)
>
> (I don’t have question as a fixed effect, because for each question
> participants had to give a 1, 2 and 3 rank, so question alone does not
> affect rank.)
>
> My first question is whether I’ve set up the model correctly, as I don’t
> have any experience with ordinal regression. When I look at coef(model1),
> all participants seem to have the same intercepts and coefficients for the
> different answers, which is not what I thought should happen (I thought I
> was setting up a model with random intercepts and random slopes).
>
> My second question is about changing the reference level of question and
> answer. When I look at summary(model1), it uses answer a1 and question q1
> as the reference levels, so I ran the model again using different answers
> and questions as reference levels.
>
> When I run the model again using a different answer as the reference level,
> the coefficients for the fixed effects are the same, but the random effects
> and threshold coefficients are quite different.
> When I run the model using a different question as the reference level, the
> coefficients for the fixed effects are quite different, but the random
> effects are exactly the same, and the threshold coefficients relatively
> similar.
>
> Could someone please help me understand what’s going on here?
>
> Thanks in advance,
> Jessie Barker
>
>         [[alternative HTML version deleted]]
>
> _______________________________________________
> R-sig-mixed-models op r-project.org mailing list
> https://stat.ethz.ch/mailman/listinfo/r-sig-mixed-models