Combining Test Data MANA 4328 Dr. Jeanne Michalski

Slides:



Advertisements
Similar presentations
STAFFING. KEY ASSUMPTIONS ä People differ ä Jobs differ ä Goal? ä ä Requires ä.
Advertisements

Managing Human Resources, 12e, by Bohlander/Snell/Sherman © 2001 South-Western/Thomson Learning 5-1 Managing Human Resources Managing Human Resources Bohlander.
Combining Test Data MANA 4328 Dr. Jeanne Michalski
Developing a Hiring System OK, Enough Assessing: Who Do We Hire??!!
Chapter 10 Decision Making © 2013 by Nelson Education.
Simple Regression Equation Multiple Regression y = a + bx Test Score Slope y-intercept Predicted Score  y = a + b x + b x + b x ….. Predicted Score 
Marketing 334 Consumer Behavior
PowerPoint Slides developed by Ms. Elizabeth Freeman
Chapter 7 Using Multivariate Statistics P173 Multiple Regression Multiple Correlation – What’s the difference between regression and correlation? Validity.
Strategic Staffing Chapter 9 – Assessing External Candidates
III Choosing the Right Method Chapter 10 Assessing Via Tests p235 Paper & Pencil Work Sample Situational Judgment (SJT) Computer Adaptive chapter 10 Assessing.
Staffing Chapters
REVIEW I Reliability Index of Reliability Theoretical correlation between observed & true scores Standard Error of Measurement Reliability measure Degree.
DECISION-MAKING AND UTILITY METHOD SELECTION OBTAINING ACCEPTANCE.
Developing a Hiring System Reliability of Measurement.
OS 352 2/28/08 I. Exam I results next class. II. Selection A. Employment-at-will. B. Two types of discrimination. C. Defined and methods. D. Validation.
Multivariate Data Analysis Chapter 4 – Multiple Regression.
Effect of Selection Ratio on Predictor Utility Reject Accept Predictor Score Criterion Performance r =.40 Selection Cutoff Score sr =.50sr =.10sr =.95.
Chapter Five Selection © 2007 Pearson Education Canada 5-1 Dessler, Cole, Goodman, and Sutherland In-Class Edition Management of Human Resources Second.
1 Report Tile UNITED STATES OFFICE OF PERSONNEL MANAGEMENT Principles of Assessment.
1 Chapter 7 Staffing Decisions Copyright © The McGraw-Hill Companies, Inc.
Part 5 Staffing Activities: Employment
Today Concepts underlying inferential statistics
Tools for Successful Selection Developing a Hiring SYSTEM.
DEFINING JOB PERFORMANCE AND ITS RELATIONSHIP TO ASSESSMENTS.
Week 14 Chapter 16 – Partial Correlation and Multiple Regression and Correlation.
Chapter 12 Inferential Statistics Gay, Mills, and Airasian
Dessler, Cole, Goodman and Sutherland Fundamentals of Human Resources Management in Canada Chapter Five Selection © 2004 Pearson Education Canada Inc.,
Standardization and Test Development Nisrin Alqatarneh MSc. Occupational therapy.
Hires 5 Offers 10 Interviews 40 Invites 60 Applicants 240 Adapted from R.H. Hawk, The Recruitment Function (New York: American Management Association,
Hires 5 Offers 10 Interviews 40 Invites 60 Applicants 240 Adapted from R.H. Hawk, The Recruitment Function (New York: American Management Association,
Selection Decisions MANA 5341 Dr. George Benson
CHAPTER 4 Employee Selection
Managing Human Resources, 12e, by Bohlander/Snell/Sherman © 2001 South-Western/Thomson Learning 5-1 Managing Human Resources Managing Human Resources Bohlander.
Selection 1- Measurement 2- External. Organization Strategy HR and Staffing Strategy Staffing Policies and Programs Staffing System and Retention Management.
Chapter Seven Measurement and Decision-Making Issues in Selection.
Part 5 Staffing Activities: Employment
Dessler, Cole and Sutherland Human Resources Management in Canada Canadian Ninth Edition Chapter Six Selection © 2005 Pearson Education Canada Inc., Toronto,
Hires 5 Offers 10 Interviews 40 Invites 60 Applicants 240 Adapted from R.H. Hawk, The Recruitment Function (New York: American Management Association,
Selection. Types of Assessment Biographical Information Interviews Work Samples Letters of Recommendation Psychological Tests.
Educational Research Chapter 13 Inferential Statistics Gay, Mills, and Airasian 10 th Edition.
Chapter 9 Selection Tests McGraw-Hill/Irwin Copyright © 2009 by The McGraw-Hill Companies, Inc., All Rights Reserved.
IMPORTANCE OF STATISTICS MR.CHITHRAVEL.V ASST.PROFESSOR ACN.
Measurement MANA 4328 Dr. Jeanne Michalski
Topic #5: Selection Theory
Strategy for Human Resource Management Lecture 15
©2011 Cengage Learning. All Rights Reserved. May not be scanned, copied or duplicated, or posted to a publicly accessible website, in whole or in part.
Educational Research Inferential Statistics Chapter th Chapter 12- 8th Gay and Airasian.
Hiring Decisions MANA 4328 Dennis C. Veit
© 2013 by Nelson Education1 Decision Making. Chapter Learning Outcomes  After reading this chapter you should:  Appreciate the complexity of decision.
© 2013 by Nelson Education1 Selection I: Applicant Screening.
Chapter 6 Staffing Decisions.
Human Resource Selection, 8e
Yield Pyramid Hires 5 Offers 10 Interviews 40 Invites 60
CHAPTER 4 Employee Selection
Week 14 Chapter 16 – Partial Correlation and Multiple Regression and Correlation.
Human Resource Staffing and Performance Management
MANA 4328 Dr. Jeanne Michalski
MANA 4328 Dennis C. Veit Measurement MANA 4328 Dennis C. Veit 1.
Week 10 Slides.
Chapter 7 Using Multivariate Statistics
Evaluating Recruiting Methods
MANA 4328 Dr. George Benson Combining Test Data MANA 4328 Dr. George Benson 1.
MANA 4328 Dennis C. Veit Measurement MANA 4328 Dennis C. Veit 1.
Evaluating Recruiting Methods
Chapter Six Selection 6 Human Resources Management in Canada
Developing a Hiring System
Personnel decisions Study Unit 4.
DM’ing with Multiple Predictors
Chapter 7: Selection.
Presentation transcript:

Combining Test Data MANA 4328 Dr. Jeanne Michalski

Selection Decisions  First, how to deal with multiple predictors?  Second, how to make a final decision?

Developing a Hiring System  OK, Enough Assessing:  Who Do We Hire??!!

Summary of Performance-Based Hiring  Understand job requirements and performance expectations  List competencies, KSAO’s that predict performance  Match attributes with selection tools  Choose/develop each tool effectively  Make performance-based decisions

Interpreting Test Scores  Norm-referenced scores  Test scores are compared to applicants or comparison group.  Raw scores should be converted to Z scores or percentiles  Use “rank ordering”  Criterion-referenced scores  Test scores indicate a degree of competency  NOT compared to other applicants  Typically scored as “qualified” vs. “not qualified”  Use “cut-off scores”

Setting Cutoff Scores  Based on the percentage of applicants you need to hire (yield ratio). “Thorndike’s predicted yield”  You need 5 warehouse clerks and expect 50 to apply. 5 / 50 =.10 (10%) means 90% of applicants rejected  Cutoff Score set at 90th percentile  Z score 1 = 84 th percentile  Based on a minimum proficiency score  Based on validation study linked to job analysis  Incorporates SEM (validity and reliability)

Selection Outcomes PREDICTION PERFORMANCE No PassPass Regression Line Cut Score 90% Percentile

Selection Outcomes PREDICTION High Performer Low Performer True Positive True Negative Type 2 Error False Positive Type 1 Error False Negative PERFORMANCE No HireHire

Selection Outcomes PREDICTION High Performer Low Performer PERFORMANCE UnqualifiedQualified Prediction Line Cut Score

Dealing With Multiple Predictors “Mechanical” techniques superior to judgment 1. Combine predictors  Compensatory or “test assessment approach” 2. Judge each independently  Multiple Hurdles / Multiple Cutoff 3. Profile Matching 4. Hybrid selection systems

Compensatory Methods Unit weighting P1 + P2 + P3 + P4 = Score Rational weighting (.10) P1 + (.30) P2 + (.40) P3 + (.20) P4 = Score Ranking RankP1 + RankP2 +RankP3 + RankP4 = Score Profile Matching D 2 = Σ (P(ideal) – P(applicant)) 2

Multiple Regression Approach  Predicted Job perf = a + b 1 x 1 + b 2 x 2 + b 3 x 3  x = predictors; b = optimal weight  Issues:  Compensatory: assumes high scores on one predictor compensate for low scores on another  Assumes linear relationship between predictor scores and job performance (i.e., “more is better”)

Multiple Cutoff Approach  Sets minimum scores on each predictor  Issues  Assumes non-linear relationship between predictors and job performance  Assumes predictors are non-compensatory  How do you set the cutoff scores?

Multiple Cutoff Approach  Sets minimum scores on each predictor  Issues  Assumes non-linear relationship between predictors and job performance  Assumes predictors are non-compensatory  How do you set the cutoff scores?  If applicant fails first cutoff, why continue?

Test 1Test 2 Interview Background Finalist Decision Reject Multiple Hurdle Model Fail Pass

Multiple Hurdle Model  Multiple Cutoff, but with sequential use of predictors  If applicant passes first hurdle, moves on to the next  May reduce costs, but also increases time

Profile Matching Approach  Emphasizes “ideal” level of KSA  e.g., too little attention to detail may produce sloppy work; too much may represent compulsiveness  Issues  Non-compensatory  Small errors in profile can add up to big mistake in overall score  Little evidence that it works better

Making Finalist Decisions  Top-Down Strategy  Maximizes efficiency, but may need to look at adverse impact issues  Banding Strategy  Creates “bands” of scores that are statistically equivalent (based on reliability)  Then hire from within bands either randomly or based on other factors (inc. diversity)

Banding  Grouping like test scores together  Function of test reliability  Standard Error of Measure  Band of + or – 2 SEM  95% Confidence interval  If the top score on a test is 95 and SEM is 2, then scores between 95 and 91 should be banded together.

Applicant Total Scores

Information Overload!!  Leads to:  Reverting to gut instincts  Mental Gymnastics

Combined Selection Model Selection Stage Selection TestDecision Model Applicants  Candidates Application BlankMinimum Qualification Hurdle Candidates  Finalists Four Ability Tests Work Sample Rational Weighting Hurdle Finalists  Offers Structured InterviewUnit Weighting Rank Order Offers  Hires Drug Screen Final Interview Hurdle

Alternative Approach  Rate each attribute on each tool  Desirable  Acceptable  Unacceptable  Develop a composite rating for each attribute  Combining scores from multiple assessors  Combining scores across different tools  A “judgmental synthesis” of data  Use composite ratings to make final decisions

List of Critical Attributes

Performance Attributes Matrix

Who Do You Hire??

Improving Ratings 1. Use rating system  Unacceptable  Did not demonstrate levels of attribute that would predict acceptable performance  Acceptable  Demonstrated levels that would predict acceptable performance  Desirable  Demonstrated levels that would predict exceptional performance

Categorical Decision Approach 1. Eliminate applicants with unacceptable qualifications 2. Then hire candidates with as many desirable ratings as possible 3. Finally, hire as needed from applicants with “acceptable” ratings  Optional: “weight” attributes by importance

Sample Decision Table

More Positions than Applicants

More Applicants than Positions

Selection  Top Down Selection (Rank) vs. Cutoff scores  Is the predictor linearly related to performance?  How reliable are the tests? 1. Top-down method – Rank order 2. Minimum cutoffs – Passing Scores

Final Decision  Random Selection  Ranking  Grouping  Role of Discretion or “Gut Feeling”

Summary of Performance-Based Hiring  Understand job requirements and performance expectations  List competencies, KSAO’s that predict performance  Match attributes with selection tools  Choose/develop each tool effectively  Make performance-based decisions