Evaluate Selection Solution Validity & Decision Quality
Screening new hires, placing new employees, promoting top performers, downsizing divisions, acquiring competitors, and team leader emergence are different forms of selection decisions. Although these selection decisions are unique in many ways each involves talent assessment. For example, applicant screening and selection is often informed by multimethod assessment. These are critical assessment-based decisions as the cost of a bad hire can be 30% of salary.
The tests, measures, scales, and business simulations comprising predictive selection models are only useful if they inform HR decision making better than alternative selection protocols. Defensible applicant selection decisions require evaluative evidence of the reliability, validity, and fairness of the composite selection solution as well as the component assessment tools. The theoretic-based rationales and accumulated empirical evidence must support the specific interpretations, inferences, and actions taken by hiring managers from assessment information. Evidence is critical as even experts can fail at subjectively combining test data during hiring.
The traditional emphasis when evidencing selection solutions is on minimizing selection errors by maximizing measure reliability, accuracy, and predictor performance criteria relationships. Reliability, validity, and predictive efficiency are unarguably essential technical components of selection solutions but there are deeper considerations that are even more critical such as incremental validity, relative importance, and the net present value of alternative options. Applying decision theory to selection involves estimating the outcomes, benefits, and costs of current and alternative options in light of capital budgeting considerations such as taxation and discounting so alternative HR program investments are competed on a level playing field.
Generating quantitative evidence to evaluate selection solutions also requires research design. Alternative selection models such as multiple cutoff, multiple regression, and multiple hurdle sequential models mandate different designs and analyses to collect situationally specific data. Design and sampling are also leveraged to discover and refine the decision policies of internal stakeholders who may use different decision rules when synthesizing applicant data for hiring. Understanding the decision policies used by HR stakeholders to make hiring decisions relative to the optimal policy is required to strike the ideal balance of human and mechanical touch.
Talent Threshold’s advisors can help you evidence and improve new hire selection decisions. Our team can illuminate your local tradeoffs, define actionable options, and generate evidence that can be applied for calibrating, tuning, and delivering sophisticated solutions for selection.
To contact Talent Threshold’s organizational scientists call 407.739.0717 or email firstname.lastname@example.org.