Menu
Home
Log in / Register
 
Home arrow Psychology arrow The Wiley Blackwell handbook of the psychology of recruitment, selection and employee retention
Source

The nature of the measurement of the predictor

A third class of explanation focuses on the measurement of individual differences as a contributing factor to the size of the score differences. Many authors have argued that the way we measure the construct introduces construct-irrelevant variance to the assessment (e.g., Chen & Gardener, 2005). Construct-irrelevant variance refers to any extraneous variance captured by an assessment that is not due to the construct of interest, but to the assessment capturing additional constructs or to the assessment’s methodology (Messick, 1995). Unlike random error, construct-irrelevant variance systematically impacts assessment scores for individuals or groups. Consequently, it compromises the construct validity of assessments and hence our ability to draw inferences from assessment scores (Binning & Barrett, 1989; Messick, 1995). Sources of construct-irrelevant variance in intelligence assessments include indeterminable items (Freedle, 2003), cultural content (Freedle, 2003; Freedle & Kostin, 1997; Helms- Lorenz et al., 2003; Malda et al., 2010; Scherbaum et al., 2015), differential prior knowledge and exposure (Agnello, Ryan & Yusko, 2015; Fagan & Holland, 2002, 2007; Ortiz et al., 2012; Scherbaum et al., 2015) and content not relevant to the intended domain (Helms-Lorenz et al., 2003; Malda et al., 2010; Scherbaum et al., 2015). For example, some research has examined the use of content requiring previously acquired knowledge that is not related to the test domain impacts test performance. Fagan and Holland (2002, 2007, 2009) have conducted a series of studies examining whether score differences between race groups on cognitive ability tests could be attributed to differences in the ability to process information (i.e., fluid intelligence) or differences in prior exposure to the acquired knowledge that test items use. They found no differences in test performance when knowledge required by the test items was unfamiliar to both groups and there was an equal opportunity to learn it. However, when the knowledge required by the test items was such that it was believed to be common and previously acquired by all test-takers, the typical Black-White score differences emerged. These findings are likely to have implications for understanding score differences for national culture groups.

 
Source
Found a mistake? Please highlight the word and press Shift + Enter  
< Prev   CONTENTS   Next >
 
Subjects
Accounting
Business & Finance
Communication
Computer Science
Economics
Education
Engineering
Environment
Geography
Health
History
Language & Literature
Law
Management
Marketing
Mathematics
Political science
Philosophy
Psychology
Religion
Sociology
Travel