SEL1 PRINC_INSTR copy - Selecting Employees Selecting...

Info iconThis preview shows page 1. Sign up to view the full content.

View Full Document Right Arrow Icon
This is the end of the preview. Sign up to access the rest of the document.

Unformatted text preview: Selecting Employees Selecting Chapter 6 Chapter 6 Validity Definitions and Validity Applications Applications CONSTRUCT VALIDITY: quality of any measure such as job performance, job satisfaction, personality, etc. Applies whenever something is measured. Validity Definitions and Validity Applications Applications CRITERION­RELATED VALIDITY: empirical forecast of future job performance by a “test” measured at the time of hire. The correlation between the test and performance on the job Validity Definitions and Validity Applications Applications CONTENT VALIDITY: Sample of actual job performance prior to hiring. Validity Definitions and Validity Applications Applications VALIDITY GENERALIZATION:The degree to which a selection device developed in one situation may be generalized to another situation HOW IMPORTANT IS HOW EMPLOYEE SELECTION TO THE FIRM? THE Imagine Four Cases Valid Selection – CASE #1 NO Firm has Valid Selection Firms can predict job productivity with the same accuracy. Popula tion sample Valid Selection – CASE #2 FEW Firms (20%) have Valid Selection Most firms (80%) do not have Valid Selection Popula tion 8 0% Not Valid sample 20% Valid Valid Selection – CASE #3 MOST Firms (80%) have Valid Selection Few firms (20%) do not have Valid Selection Population 20% Not Valid sample 80% Valid Valid Selection – CASE #4 ALL firms have Valid Selection All firms equally good at hiring the best applicants Population sam ple 100% Valid CONCLUSIONS 1. Few firms w/ valid selection: Big + Those w/o : Small loss 2. Most firms w/ valid selection: Small + Few firms w/o: BIG Loss 3. What do you do when everybody has valid selection? Criterion-Related Validity Criterion­Related Validity is an empirical relationship or correlation between applicants’ test scores and measures of performance on the job. (C­R validity most frequent in selection) 0 ≤ rjp test ≤ 1.0 Criterion-Related Validity Criterion-Related Correlate ­ r ­Test Score with Job Correlate ­ r ­Test Score with Job Performance Score Test Score – measured at time of hire Job Performance – measured on the job (typically 1st year) Criterion-Related Validity JP - r jp test= 0 No Validity -Cannot forecast or predict JP from test score -E.g. unstructured interview Test Low Criterion Related Validity Low Predicted Job Performance rJP.Test Score = .30 Highest Most Likely Lowest Score ÒÓ i Test Score High Criterion-Related Validity High Predicted Job Performance rJP.Test Score = .60 Highest Most Likely Lowest Score ÒÓ i Test Score Criterion-Related Validity rjp test % JP Correctly Forecast 0 – e.g. unstructured interview .2 – e.g. conscientiousness .4 – e.g. structured interview .6 ­ e.g. general cognitive ability 0% 4% 16% 36% Content Validity A selection device is content valid to the degree that it adequately samples the domain of the job “Domain” is determined by job analysis (TDRs) Judgment made by subject matter experts Machinist Pilots – Read blueprints; make precise measurements, etc. – Aircraft Simulator – Role Play/ Performance Managers Content Validity – Advantages of Content Validity: It’s fast Small number of job incumbents Small number of Judges – Disadvantages Applicant must be able to perform job at time of hire Occupation (job) specific No estimate of quantitative strength of relationship (Criterion­related validity provides quantitative estimate) Validity Generalization Validity Validity Generalization is the degree to which a Validity Generalization selection device developed (validated) in one situation can generalize to other situations – “Situations” = ­ Professional V. G. study Required Quantitative and statistical results usually published HIGH Professional STANDARDS Good news: much work has been done Validity Generalization Validity – Once V.G. is established, no more individual validity studies are necessary – User must show that their job is the same or very similar to those in the validity generalization study. – Advantage = Fast and lower cost. By far, the greatest advantage is fast implementation, or Not Using V. G. has High Opportunity Cost Effectiveness of Known Effectiveness Selection Devices (V.G.) Selection Size Correlation r = 0 % JP Correctly Predicted 0% JP predicted 4% JP predicted r = .2 r = .4 r = .6 16% JP predicted 36% JP predicted HOW WELL DO THEY WORK? Thanks to Very Extensive Research & Validity Generalization Studies, the Validity of many Selection Devices is Known EFFECTIVENESS OF EFFECTIVENESS SELECTION DEVICES SELECTION NO VALIDITY (e.g.s) – Unstructured Interview – Handwriting analysis SOME VALIDITY (e.g.s) – – – Job Knowledge (.40) Cognitive Abilities (.60) Personality (.20) Unstructured Interviews General Information – In this type of interview, an employer might ask you questions about your interests, your educational background, your previous work experience and any other subject that they feel is important in their organization or to the job. Unstructured Interview Unstructured Limitations Limitations – Subjective evaluations are made. – Decisions tend to be made within the first minute of the interview with the remainder of the interview used to validate or justify the original decision. – Interviewers form unrealistic stereotypes concerning the characteristics required for success on the job. – Research has shown disproportionate rejection rates between minority and non­minority members using interviews. – Negative information seems to be given more weight. – Validity evidence (after hundreds of studies) show that unstructured interviews cannot predict job performance. Unstructured Interviews C­R Validity Generalizability Comments None Useful as an exchange of information only None Structured Interviews Advantages Structured Interviews Disadvantages Structured Interviews C­R Validity Generalizability Comments .4 (if highly structured: both Q & A) Interview contents are occupation specific Must do job analysis for Q & A. Use a scoring key for high validity (i.e. high structure). Use Multiple Interviewers trained in consensus decision making ...
View Full Document

This note was uploaded on 01/31/2011 for the course OBHR 428 taught by Professor Campion during the Fall '09 term at Purdue University-West Lafayette.

Ask a homework question - tutors are online