# Problem in using ordered logstic contribution

**URL:** <https://discourse.mc-stan.org/t/problem-in-using-ordered-logstic-contribution/32082>\
**Category:** Modeling\
**Tags:** rstan\
**Created:** [July 14, 2023, 7:25am UTC](https://discourse.mc-stan.org/t/problem-in-using-ordered-logstic-contribution/32082 "2023-07-14T07:25:19Z")\
**Posts on this page:** 2\
**Page:** 1

<div class="post-metadata">

**Author:** ![yz1308786079](https://avatars.discourse-cdn.com/v4/letter/y/e56c9b/32.png) [@yz1308786079](https://discourse.mc-stan.org/u/yz1308786079)\
**Post date:** [July 14, 2023, 7:25am UTC](https://discourse.mc-stan.org/t/problem-in-using-ordered-logstic-contribution/32082/1 "2023-07-14T07:25:19Z")

</div>

[edit to remove tt format]

Dear All, my dependent variable is altruistic punishment amounts for offenders, for example, the participant could choose from 0 to 50, with an interval of 5. Thus, the participant has 11 choices in this scenario. Also, I would like to apply the utility desicion theory in my model, which means calculating and comparing utility for different choice and making thier desicion in this end.  
For the purpose of this experiment, I am going to use an ordered logistic distribution as the likelihood distribution. However, it is diffcult for me to understand the eta and c(cutpoints) in the rstan function reference. If I want to use it, how do I translate the utility of the different options into the parameter values needed for the ordered logstic distribution?  
Thank you!

---

<div class="post-metadata">

**Author:** ![Bob\_Carpenter](https://yyz2.discourse-cdn.com/flex030/user_avatar/discourse.mc-stan.org/bob_carpenter/32/9230_2.png) [@Bob\_Carpenter](https://discourse.mc-stan.org/u/Bob_Carpenter)\
**Post date:** [July 27, 2023, 9:45pm UTC](https://discourse.mc-stan.org/t/problem-in-using-ordered-logstic-contribution/32082/2 "2023-07-27T21:45:55Z")

</div>

> [@yz1308786079](#):
>
> how do I translate the utility of the different options into the parameter values needed for the ordered logstic distribution?

Usually you would fit the parameter values given data. The ordered logistic just chooses the outcome between 1 and 11—it doesn’t give you any other kind of value. If you want a value based on the choice, you’d need something like an 11-element vector of values.

Stan’s parameterization of the ordinal logit is pretty much the standard one, so maybe the Wikipedia page will help: [Ordered logit - Wikipedia](https://en.wikipedia.org/wiki/Ordered_logit)