# Ordinal response with monotonic predictor

**URL:** <https://discourse.mc-stan.org/t/ordinal-response-with-monotonic-predictor/13165>\
**Category:** brms\
**Created:** [February 19, 2020, 11:52am UTC](https://discourse.mc-stan.org/t/ordinal-response-with-monotonic-predictor/13165 "2020-02-19T11:52:32Z")\
**Posts on this page:** 1\
**Showing post:** 10

<div class="post-metadata">

**Author:** ![ssalimi](https://yyz2.discourse-cdn.com/flex030/user_avatar/discourse.mc-stan.org/ssalimi/32/7119_2.png) [@ssalimi](https://discourse.mc-stan.org/u/ssalimi)\
**Post date:** [February 19, 2020, 2:28pm UTC](https://discourse.mc-stan.org/t/ordinal-response-with-monotonic-predictor/13165/10 "2020-02-19T14:28:49Z")

</div>

> [@torkar](#):
>
> p ← get\_prior(bodn ~ mo(chf) + (1 | id), family = cumulative, data = bodn) p$prior[1] ← “normal(0,1)” # sets a N(0,1) on \beta p$prior[3] ← “normal(0,2.5)” # for each cutpoint p$prior[18] ← “cauchy(0,2.5)” # for sd p$prior[21] ← “dirichlet(2,2)” # sets weakly regularizing dirichlet for 1-3 Likert scale response

I ran the model using the above priors. The good thing is that I got simplex results. However, the Rhats are not good.  
here is the results

```no-highlight
Data: bodi (Number of observations: 2622) 
Samples: 4 chains, each with iter = 2000; warmup = 1000; thin = 1;
         total post-warmup samples = 4000

Group-Level Effects: 
~idno (Number of levels: 898) 
              Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS Tail_ESS
sd(Intercept) 4.13 0.14 3.86 4.41 1.01 457 1116

Population-Level Effects: 
                  Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS Tail_ESS
Intercept[1] -11.12 0.50 -12.11 -10.17 1.01 949 1834
Intercept[2] -8.07 0.30 -8.67 -7.51 1.01 703 1816
Intercept[3] -5.67 0.21 -6.10 -5.28 1.02 422 1057
Intercept[4] -3.56 0.18 -3.92 -3.20 1.03 262 803
Intercept[5] -1.82 0.17 -2.15 -1.50 1.04 219 514
Intercept[6] -0.06 0.16 -0.38 0.24 1.04 125 498
Intercept[7] 1.40 0.16 1.08 1.71 1.04 130 560
Intercept[8] 2.97 0.17 2.64 3.30 1.03 172 687
Intercept[9] 4.49 0.19 4.13 4.87 1.03 259 983
Intercept[10] 5.97 0.21 5.57 6.40 1.02 373 1267
Intercept[11] 7.60 0.25 7.14 8.08 1.02 554 1719
Intercept[12] 9.49 0.31 8.92 10.10 1.01 650 2034
Intercept[13] 11.21 0.41 10.43 12.04 1.01 1412 2826
Intercept[14] 13.22 0.66 12.04 14.59 1.00 2611 2885
       mochf 0.50 0.08 0.34 0.65 1.00 915 1622

Simplex Parameters: 
                      Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS Tail_ESS
      mochf1[1] 0.16 0.09 0.03 0.36 1.00 2083 2314
      mochf1[2] 0.84 0.09 0.64 0.97 1.00 2083 2314

```

Samples were drawn using sampling(NUTS). For each parameter, Bulk\_ESS  
and Tail\_ESS are effective sample size measures, and Rhat is the potential  
scale reduction factor on split chains (at convergence, Rhat = 1).

The good news is I got the simplex result but not good Rhat  
Should I change the priors’ values?  
Thanks

---

_[View the full topic](https://discourse.mc-stan.org/t/ordinal-response-with-monotonic-predictor/13165)._