# Should I report the results with random effects or not (brms, re\_formula = NULL vs NA)?

**URL:** <https://discourse.mc-stan.org/t/should-i-report-the-results-with-random-effects-or-not-brms-re-formula-null-vs-na/16280>\
**Category:** Modeling\
**Tags:** interpret-results\
**Created:** [June 29, 2020, 3:02pm UTC](https://discourse.mc-stan.org/t/should-i-report-the-results-with-random-effects-or-not-brms-re-formula-null-vs-na/16280 "2020-06-29T15:02:10Z")\
**Posts on this page:** 1\
**Showing post:** 4

<div class="post-metadata">

**Author:** ![Max\_Mantei](https://yyz2.discourse-cdn.com/flex030/user_avatar/discourse.mc-stan.org/max_mantei/32/7912_2.png) [@Max\_Mantei](https://discourse.mc-stan.org/u/Max_Mantei)\
**Post date:** [July 8, 2020, 5:35pm UTC](https://discourse.mc-stan.org/t/should-i-report-the-results-with-random-effects-or-not-brms-re-formula-null-vs-na/16280/4 "2020-07-08T17:35:19Z")

</div>

> [@di4oi4](#):
>
> And these are the predictions for the predictor I am interested in. Study question is: is this predictor associated with outcome variable (received treatment hours) in the whole country. We are not interested in county/regional level effects.

Then I’d say use `re_formula = NA` (the default for `marginal_effects`, right?), i.e. Panel A, because these will not include the uncertainty from the county/region levels. Panel B would answer the question of what marginal effects could look like in a new, not previously seen county/region.

In cases like this I think it helps to spell out the mathematical model to grasp the implications of those choices. (Also, you should probably think about the `conditions` of the marginal effect and see if those make sense. Here you need domain expertise, so I won’t be able to help with this.)

Hope this helps (at least my first paragraph, haha)!  
Cheers,  
Max

PS: You need to fix your plot axis’ labels ;)

---

_[View the full topic](https://discourse.mc-stan.org/t/should-i-report-the-results-with-random-effects-or-not-brms-re-formula-null-vs-na/16280)._