Scoring Biodata: Is It Rational to Be Quasi-Rational?

Applying theory to the black box: A model for empirically scoring biodata

“…The benefits of empirical keying, or at least some quasi‐rational version of empirical‐keying, are not in dispute (Cucina et al, , ; Karas & West, ; Mitchell & Klimoski, ). Given large enough samples and consistency in test content, some degree of empirical keying will improve the functionality of biodata scales.…”

“…Although biodata can be developed rationally to measure specific constructs (e.g., Mumford et al, ; Oswald et al, ; Reiter‐Palmon & Connelly, ; Stricker & Rock, ), applying some degree of empirical keying has been shown to result in improved prediction of desired criteria (Cucina, Caputo, Thibodeaux, & MacLane, ; Cucina, Caputo, Thibodeaux, MacLane, & Bayless, ; Karas & West, ; Mitchell & Klimoski, ; Reiter‐Palmon & Connelly, ). However, such empirical keying leaves users with predictive test content but at the expense of a thorough understanding of what the scale is measuring and in what contexts the developed scale scores' predictiveness will generalize.…”

“…To this end, biodata instruments have been broadly examined via meta‐analysis in terms of their criterion‐related validity across a range of outcomes without much regard to what those biodata scales were assessing (e.g., Bliesener, ; Hunter & Hunter, ). On the other hand, studies have taken a more construct‐centric approach as well (e.g., Cucina et al, ; Karas & West, ; Mumford et al, ; Oswald et al, ; Reiter‐Palmon & Connelly, ; Sisco & Reilly, ; Stricker & Rock, ), highlighting the value of this method, not only because it can measure many types of constructs but also because it is less fakable than other self‐report measures (Ramsay, Schmitt, Oswald, Kim, & Gillespie, ; Sisco & Reilly, ).…”

“…Biodata scoring methods can broadly be differentiated into the three major categories: (a) rational keying, (b) empirical keying, and (c) hybrid blends of the two (Cucina et al, , ). Rational keying, although allowing for a clearer understanding of construct measurement and potentially increased legal defensibility (Sharf, ), has been shown to generally produce weaker criterion‐related validity than empirical keying of biodata inventories (Cucina et al, , ; Karas & West, ; Mitchell & Klimoski, ; Reiter‐Palmon & Connelly, ). For empirical keying, although items may be written with intended constructs in mind, item responses are scored as a function of how they covary with important criteria such as job performance.…”

Considering Age Diversity in Recruitment and Selection: An Expanded Work Lifespan View of Age Management

“…In practice, biodata questions can be quite broad (e.g., past experiences in general, personality, attitudes, interests); however, more experiencebased biodata measures may be particularly helpful in this given context (e.g., focusing on past work experiences, education, and training). Moreover, biodata measures can be scored using both rational and empirical means or with hybrid approaches (e.g., unit-weighting) that combine desirable qualities of each method (see Cucina et al 2013 ).…”