Dear all,

I have a data set where I want to predict the outcome `Effort`

depending on a number of predictors (8 in total).

My assumption is that I should use `family=gaussian`

so,

```
fit <- brm(bf(Effort ~ Input + Output + Enquiry + File + Interface + Added +
Changed + Deleted),
family=gaussian,
prior=c(prior(normal(0,1), class=b)),
cores=4, chains=4, data=foo)
```

ran fine, but when I did a posterior predictive check it looked ridiculous (`Effort`

and all predictors are `>=0`

in the data set). Here is the data (`N=555`

).

So, I thought that I was mistaken and picked `Poisson`

instead, but many divergences and BFMIs with a simple model such as `Effort ~ 1 + Input`

, didnāt make me any wiser (and I even checked the results from using `zi`

).

Can someone please explain to me what is going on. Iām missing something obvious but Iām sure some of you know what that might be.

Not that it matters but:

```
> packageVersion("brms")
[1] ā2.7.0ā
```