Add Spike and slab prior for Claret-Bruno model

[gowerc]

I believe the idea here is that the prior helps guard against issues of non-identifiability for the c parameter in cases with few data points. I believe the prior essentially means in these cases the c parameter just becomes a fixed effect (need to double check literature / references though).

[gowerc]

@mercifr1 - Are there any references for this in particular?