Week 10 Slides
Assigned Readings Kacmar et al. (2013) Organizational Politics Kuncel & Sackett (2013) Assessment Centers
Utility Selection ratio: number of people hired / number of people available to hire Baserate: ratio of people that would be successful compared to the total available. Either compared to Those that are hired Those that are available
Cut-scores Relative Cut Scores (Norm-Referenced) Distributive comps Known-Groups Method Fixed Cut Scores (Absolute Cut scores) Angoff Method Multiple Hurdles vs. Compensatory More expensive last
Cutoff Scores Predictive Yield Discriminant Analysis Book Mark Method Takes into account selection ratio and base rate Discriminant Analysis Book Mark Method
Overall Assessment Cutoffs Clinical Prediction Subjective Unit Weighting Total score Rational Weighting Multiple predictors by a predetermined weight Multiple regression Weights are set by a statistical procedure
Expectancy data Hit: correct classification (Hit Rate) Those that do well on this test do well on the job Miss: incorrect classification (Miss Rate) False Positive: Those that do well on this test do poorly on the job False Negative: Those that do poorly on this test do well on the job
Content & Face Validity Do the items match up with the definition? Do they adequately assessing the testing universe Is the content validity ratio high enough? If there is an issue with content validity, what are the implications for the organization in regards to using it?
Factorial Validity What is our expected factor structure? Is the construct homogenous? Exploratory Factor Analysis Interpret Scree Plot Is the reliability estimate high enough? Is it a problem if the factor structure is not supported?
Content Validity Ratio Judges rate item on scale of importance Essential, important, not-essential CVRi = (ne – (N/2))/ (N/2) CVR = value of the item n = number of experts saying the item is essential N = total number of experts
Construct Validity Convergent vs. Discriminant Two Ways to assess Multi-trait Multi-Method matrix Correlation matrix with theoretically related, unrelated, and criterion-related variables
Criterion Validity Two ways Correlation Table Multiple Regression Results R2: How much variance accounted for by predictors Beta values: How much unique variance accounted for Is multicollinearity and issue and is the measure redundant with the current battery of tests?
Additional Issues Adverse Impact Face Validity and Faking Determine if it is an issue and what can be done, if anything, about it. Face Validity and Faking Is the method and cost of administration feasible for the current organization?