# What's the difference between specifying prior for general intercept vs using family =?

**URL:** <https://discourse.mc-stan.org/t/whats-the-difference-between-specifying-prior-for-general-intercept-vs-using-family/25042>\
**Category:** brms\
**Tags:** prior-choice\
**Created:** [October 29, 2021, 3:03pm UTC](https://discourse.mc-stan.org/t/whats-the-difference-between-specifying-prior-for-general-intercept-vs-using-family/25042 "2021-10-29T15:03:08Z")\
**Posts on this page:** 2\
**Page:** 1

<div class="post-metadata">

**Author:** ![mavavilj](https://yyz2.discourse-cdn.com/flex030/user_avatar/discourse.mc-stan.org/mavavilj/32/12812_2.png) [@mavavilj](https://discourse.mc-stan.org/u/mavavilj)\
**Post date:** [October 29, 2021, 3:03pm UTC](https://discourse.mc-stan.org/t/whats-the-difference-between-specifying-prior-for-general-intercept-vs-using-family/25042/1 "2021-10-29T15:03:08Z")

</div>

What’s the difference between specifying prior for general intercept vs using `family =`?

In the documentation they read almost like specifying the same thing (the prior of the response)?

---

<div class="post-metadata">

**Author:** ![martinmodrak](https://yyz2.discourse-cdn.com/flex030/user_avatar/discourse.mc-stan.org/martinmodrak/32/133_2.png) [@martinmodrak](https://discourse.mc-stan.org/u/martinmodrak)\
**Post date:** [November 10, 2021, 8:41pm UTC](https://discourse.mc-stan.org/t/whats-the-difference-between-specifying-prior-for-general-intercept-vs-using-family/25042/2 "2021-11-10T20:41:08Z")

</div>

Those are two very different things - family describes how the whole linear predictor (including intercept) gives raise to observed data (transformations, whether those are count or real valued, the noise model).

You might benefit from reviewing some introduction to generalized linear models, R. McElreath course is IMHO particularly fun and useful: [GitHub - rmcelreath/statrethinking\_winter2019: Statistical Rethinking course at MPI-EVA from Dec 2018 through Feb 2019](https://github.com/rmcelreath/statrethinking_winter2019)

Best of luck with your modelling!