Specify RR3 Data Model

Specify a data model where the outcome variable has been randomly rounded to base 3.

Usage

set_datamod_outcome_rr3(mod)

Arguments

mod

An object of class "bage_mod", created with mod_pois(), mod_binom(), or mod_norm().

Value

A modified version of mod.

Details

set_datamod_outcome_rr3() can only be used with Poisson and binomial models (created with mod_pois() and mod_binom().)

Random rounding to base 3 (RR3) is a confidentialization technique that is sometimes applied by statistical agencies. RR3 is applied to integer data. The procedure for rounding value \(n\) is as follows:

• If \(n\) is divisible by 3, leave it unchanged

• If dividing \(n\) by 3 leaves a remainder of 1, then round down (subtract 1) with probability 2/3, and round up (add 2) with probability 1/3.

• If dividing \(n\) by 3 leaves a remainder of 1, then round down (subtract 2) with probability 1/3, and round up (add 1) with probability 2/3.

If set_datamod_outcome_rr3() is applied to a fitted model, set_datamod_outcome_rr3() unfits the model, deleting existing estimates.

See also

• datamods Overview of data models implemented in bage

• mod_pois(), mod_binom(), mod_norm() Specify a model for rates, probabilities, or means

Examples

## 'injuries' variable in 'nzl_injuries' dataset
## has been randomly rounded to base 3
mod <- mod_pois(injuries ~ age:sex + ethnicity + year,
                data = nzl_injuries,
                exposure = popn) |>
  set_datamod_outcome_rr3() |>
  fit()
#> Building log-posterior function...
#> Finding maximum...
#> Drawing values for hyper-parameters...