ai_ratio

Helper function to convert mean subgroup differences to AI ratios (Newman et
al., 2007). Called by calc_out().

internal

Estimates Pareto-optimal solution for personnel selection
with 3 objectives using Normal Boundary Intersection (NBI) algorithm
introduced by Das and Dennis (1998) <doi:10.1137/S1052623496307510>.
Takes predictor intercorrelations and predictor-objective relations as
input and generates a series of solutions containing predictor weights
as output. Accepts between 3 and 10 selection predictors. Maximum 2
objectives could be adverse impact objectives. Partially modeled after
De Corte (2006) TROFSS Fortran program
<https://users.ugent.be/~wdecorte/trofss.pdf> and updated from
'ParetoR' package described in Song et al. (2017)
<doi:10.1037/apl0000240>. For details, see Study 3 of Zhang et al. (2023).

Chelsea Song

rMOST

Estimates Pareto-Optimal Solution for Hiring with 3 Objectives

Yesuel Kim

ai_ratio function

<dl><dt>d</dt>
<dd>Mean subgroup difference of predictor(s)</dd>
<dt>sr</dt>
<dd>Selection ratio in the full applicant pool</dd>
<dt>p</dt>
<dd>Proportion of minority group in the full applicant pool</dd></dl>

Arguments

ai_ratio — ai_ratio

<dl>

<dt>d</dt>
<dd>Mean subgroup difference of predictor(s)</dd>


<dt>sr</dt>
<dd>Selection ratio in the full applicant pool</dd>


<dt>p</dt>
<dd>Proportion of minority group in the full applicant pool</dd>

</dl>

ai_ratio: ai_ratio

Description

Usage

Arguments