# Slight specification changes resulting in warnings and unexpected results

**URL:** https://discourse.mc-stan.org/t/slight-specification-changes-resulting-in-warnings-and-unexpected-results/14095
**Category:** Modeling
**Created:** [April 5, 2020, 10:04am UTC](https://discourse.mc-stan.org/t/slight-specification-changes-resulting-in-warnings-and-unexpected-results/14095 "2020-04-05T10:04:00Z")
**Posts on this page:** 1
**Showing post:** 11

<div class="post-metadata">

### Author: ![Mike](https://yyz2.discourse-cdn.com/flex030/user_avatar/discourse.mc-stan.org/mike/32/6159_2.png) [@Mike](https://discourse.mc-stan.org/u/Mike)
#### Post date: [April 6, 2020, 7:03pm UTC](https://discourse.mc-stan.org/t/slight-specification-changes-resulting-in-warnings-and-unexpected-results/14095/11 "2020-04-06T19:03:58Z")

</div>

Thanks everyone this is all very helpful, great community here. As an initial attempt to fix it I tried constraining gamma to be positive and estimating more priors but this didn’t seem to change much.  
I think I found a potential solution here: [Non convergence issue on polytomous IRT model](https://discourse.mc-stan.org/t/non-convergence-issue-on-polytomous-irt-model/12576/7)

That said, I’m getting in to territory that’s over my head and need to turn my attention to other things at the moment. So I’ll take a few days to read up, test a few solutions, and post back here with what I find/follow up questions.

---

_[View the full topic](https://discourse.mc-stan.org/t/slight-specification-changes-resulting-in-warnings-and-unexpected-results/14095)._