Chapter Seven Measurement and Decision-Making Issues in Selection
Copyright © Houghton Mifflin Company. All rights reserved. 7–27–2 Chapter Outline Statistical Methods in Selection Reliability Validity Decision Making in Selection Utility of a Selection System
Copyright © Houghton Mifflin Company. All rights reserved. 7–37–3 The Selection Process Measurement –Ensuring that selection tests are reliable and valid Decision making –Combining information about a person to make hiring decision Evaluation –Making sure hiring decisions increase firm efficiency and profitability
Copyright © Houghton Mifflin Company. All rights reserved. 7–47–4 Statistical Methods in Selection Correlation Analysis –Degree of linear relationship between variables –Commonly measured by the Pearson product moment correlation coefficient Regression Analysis –Best fitting line equation
Copyright © Houghton Mifflin Company. All rights reserved. 7–57–5 Reliability Consistency of measurement Systematic Error versus Random Error Methods of Measuring Reliability –Test-Retest –Interrater Reliability –Internal Consistency
Copyright © Houghton Mifflin Company. All rights reserved. 7–67–6 Validity Content Validity –Are items a representative sample –Based on job analysis –Determined by judgment of job experts –Lawshe method
Copyright © Houghton Mifflin Company. All rights reserved. 7–77–7 Validity (cont’d) Criterion-Related Validity – does a selection test adequately predict an outcome –Concurrent Validation Current employees Predictor and criterion measured at same time Problem = representativeness of sample Problem = restriction of range
Copyright © Houghton Mifflin Company. All rights reserved. 7–87–8 Validity (cont’d) Predictive Validation –Job applicants –Predictor measured (time 1) –Applicants hired without predictor influence –Criterion measured (time 2) –Problem = sample size needed –Problem = time between T1 and T2
Copyright © Houghton Mifflin Company. All rights reserved. 7–97–9 Special Concerns in Validity Test Fairness –Mean differences in test scores –Equitable treatment –Differential prediction –Differential validity Validity Generalization
Copyright © Houghton Mifflin Company. All rights reserved. 7–10 Decision Making in Selection Additive Models –Higher score = better score –Compensatory model Multiple Cutoff –Minimum score for each test –Non-compensatory model Multiple Hurdle –Sequential testing –Minimum score for each test –Non-compensatory model
Copyright © Houghton Mifflin Company. All rights reserved. 7–11 Decision Making (cont’d) Profile Matching –Ideal pattern of scores Adjusting Test Scores of Minority Group Members –In most cases illegal –Fixed bands –Sliding bands
Copyright © Houghton Mifflin Company. All rights reserved. 7–12 Utility of a Selection System Efficiency of Selection Types of selection decisions –Quadrant A: True-Positive Decisions –Quadrant B: False-Negative Decisions –Quadrant C: True-Negative Decisions –Quadrant D: False-Positive Decisions
Copyright © Houghton Mifflin Company. All rights reserved. 7–13 Utility (cont’d) Selection Efficiency Is Affected By: –Validity –Selection Ratio –Base Rate of Success Taylor-Russell Tables –Standard Deviation of Performance in Dollars –Costs Associated with Selection
Copyright © Houghton Mifflin Company. All rights reserved. 7–14 Review Statistical Methods in Selection Reliability Validity Decision Making in Selection Utility of a Selection System