Measurement Issues in Selection

1
Making Selection Decisions And Final Offers
2
Wrapping Up the Selection Process Three Issues

• How do we combine information from evaluating
  applicants to make final decisions?
• What issues are important to consider in the
  employment contract?
• What issues are important to consider in the job
  offer process?

3
Making Final Selection Decisions Goals

• Make decisions that are valid by minimizing false
  positive and false negative decisions about
  applicants
• Make decisions that have the least adverse impact
  on different categories of employees
• Make decisions in a timely manner
• Make decisions that promote customer satisfaction
  (i.e., managers and applicants)

4
What do we do with information from multiple
selection methods?

• How do we combine scores on multiple predictors
  (e.g., interview ratings, work sample ratings,
  intelligence test scores)?
• How do we determine the line at which we divide
  qualified and unqualified applicants (i.e., the
  cut score)?
• How do we make our final choice?

5
Methods of Combining Information

• Compensatory models
• Multiple cutoffs (non-compensatory)
• Multiple hurdles (non-compensatory)
• Combination approaches
• Profile Matching

6
Compensatory Models

• Scores of all predictors are added up to yield
  total score
• Assumption strengths in some areas can offset
  weaknesses in others
• Weighting of predictor scores
• Subjective methods clinical prediction and
  rational weighting
• Objective methods unit weighting and multiple
  regression

See p. 2, 1 on Hiring Plant Manager handout
7
Are Objective or Subjective Methods Better?

• Methods that use more objective combinations of
  predictor information seem to be more accurate
  and efficient to use (i.e., minimize error
  bias)
• Implications for decision makers
• Use objective methods as much as possible in
  collecting and combining
• However, objective methods are resisted by both
  decision makers and applicants

8
Multiple Cutoffs Method

• Applicants are measured on each predictor
• Applicants are rejected if any of their scores
  fall below a certain minimum cutoff score
• Assumes
• Minimum amount of each predictor needed for
  successful job performance
• Non-compensatory model

See p.2, 2
Intelligence Test 12.5 Work Sample
17 Interview 14
9
Multiple Cutoff Method

• Advantages
• Narrows pool of applicant to those with at least
  minimal qualifications on each predictor
• Easy to understand and implement

• Disadvantages
• Need to measure all applicants on all
  predictors--expensive!!
• Only produces set of applicants who are minimally
  qualified--how to decide amongst those who pass
  cutoffs (cannot rank order)?

10
Multiple Hurdle Method

• Each applicant must meet minimum cutoff or hurdle
  for each predictor before being considered on
  next one
• Sequential passing of hurdles required--applicants
  are dropped at each stage
• Non-compensatory model

See p.2, 3
First hurdle Intelligence Test 12.5 Second
hurdle Work Sample 17 Third hurdle
Interview 14
11
Multiple Hurdle Method

• Advantages
• Less costly than others--dont have to measure
  applicant on each predictor
• Useful when minimal qualifications are necessary
  on some predictors
• Useful when selection procedures are costly

• Disadvantages
• Reduces the validity for each subsequent
  predictor due to range restriction in applicants
• Takes a lot of time to implement--may lose
  applicants in the process
• Cannot rank order remaining candidates

12
Combination Method

• Each applicant is measured on each predictor
• Those below minimal cutoff scores on any
  predictor are rejected (multiple cutoff method)
• Then use multiple regression to predict
  performance with applicants remaining in pool
  (multiple regression method)
• Therefore, rank ordering of applicants can be
  done based on level of predicted performance
  scores

See p. 3, 4 on handout
13
Combination Method

• Advantages
• All applicants must have minimal qualifications
• After determining minimal qualifications,
  predictors can compensate for each other
• Allows rank ordering from best to least qualified
  applicant

• Disadvantages
• Very costly as all applicants must be screened on
  all tests
• Not cost effective with large applicant pools

14
Profile Matching

• Measure current successful employees across
  several predictors
• Average their scores to get overall profile of
  scores needed for successful performance
• Measure applicants on predictors and compare
  their score profiles with successful job
  incumbents
• Select those applicants who best match the
  profiles of successful employees

See p. 3, 5 on handout
15
Profile Matching

• Advantages
• No arbitrary cutoffs are set for decision making
• Can rank order applicants based on how similar
  they are to ideal profile
• Useful when there is an ideal profile for success

• Disadvantages
• Determining the KSAOs in which most successful
  and least successful job incumbents differ
• Range restriction
• Assumes one best profile that does not change
  over time
• Does not account for validity of predictors

16
How do we make our final choice?

• Requires evaluating final scores and deciding to
  whom to extend job offer
• Cut scores have important consequences
• Three approaches
• Minimum competency
• Top-down selection
• Score Banding

17
Minimum Competency

• Cut score is based on minimum qualifications
  needed to perform the job
• For compensatory methods, the minimum competency
  total score would be calculated
• For multiple hurdles and cutoffs, minimums have
  already been used in getting pool of final
  candidates all would be considered minimally
  qualified
• Cannot be used with profile matching

18
Top-Down Selection

• Rank order applicants scores from highest to
  lowest
• Extend job offers starting from top of list and
  proceeding downward until positions available are
  filled
• Problems
• legally cannot be used with content validation
• tends to have adverse impact on blacks and
  hispanics

19
Score Banding

• Banding is establishing ranges of equivalent
  scores
• All applicants whose scores fall in the band are
  assumed to be equally qualified (differences in
  scores within band assumed to be due to
  measurement errors)
• May construct fixed or sliding bands
• Selecting within a band can be based on criteria
  outside of testing (e.g., race, gender, or some
  other discretionary criterion)

20
Score Banding

• Advantages
• Flexibility in hiring decisions
• Can utilize other criteria such as demographics,
  attendance data, special skills, etc. that are
  important but not measured in predictors
• Social and financial implications of having
  diverse workforce are considered

• Problems
• Use of protected status as criterion for
  selection within band may be illegal (Section 106
  of 1991 Civil Rights Act)
• Loss of economic utility in selection
• Weights used with subjective criteria to select
  within band are unknown
• Fairness questioned if someone with lower score
  selected over one with higher score

21
Methods of Final Choice

• Random Selection
• Luck of the draw
• Ranking
• Discretionary assessments used to rank order
  final candidates
• Grouping
• First choices, acceptable, non-acceptable

22
Practical Advice on Selection Decision Making

• Major decisions you will have to make
• 1. Should a sequential or non-sequential process
  be used?
• 2. Should the decision be based on a
  compensatory or non-compensatory model?
• Should one select applicants from top-down, set
  minimal cutoffs, or from within bands?
• What kinds of discretionary methods should you
  use (e.g., diversity goals)?

23
What variables help you make these three
decisions?

• Consider the job performance needed, selection
  ratios, number of applicants expected, resources
  and time you can devote to the decision making
  process
• Use objective methods over judgmental ones
• Consider applicant face validity and fairness
• Consider affirmative action/diversity/glass
  ceiling goals
• Periodically audit selection procedures and
  decisions

24
Employment Contracts

• Requirements for legally enforceable contract
• Offer by employer specific and clear terms and
  conditions of employment
• Acceptance by applicant offer must be accepted
  on the terms and conditions specified
• Consideration exchange of promises (inducements
  and contributions specified)

25
Employment Contracts

• Parties employee, independent contractor, role
  of agents
• Forms written are favored over oral employment
  at will concerns consistency between oral
  written
• Disclaimers limits on employee rights
  (employment at will, reference checking, false
  statement warning)
• Contingencies that must be fulfilled (e.g.,
  passing medical exams, work authorization, etc.)
• Reneging and unfulfilled promises (breach of
  contract, promissory estoppel, fraud claims)

26
Job Offers

• Attempt to induce offer receiver to enter into
  employment relationship
• Consider monetary and non-monetary inducements,
  labor market conditions, qualifications of offer
  receiver, organizational vacancies, timing
  issues, legal issues

27
Content of Job Offers

• Starting dates
• Hours of work
• Acceptance terms
• Duration of Contract
• Compensation rates
• Benefits
• Perks
• Hiring bonuses or skills premiums

• Relocation Package
• Severance Package
• Confidentiality clauses
• Noncompete agreement
• Methods of resolution of employment disputes
• Golden handcuff agreements

28
Job Offer Process

• What are competitors doing in this area and what
  are likely reactions of offer receivers?
• Is offer negotiable? If so, what aspects of it
  are negotiable? What are the upper and lower
  bounds of negotiable aspects?
• How should offer be presented? Dry, sterile
  approach vs. sales pitch approach
• Handling rejected applicants quickly and
  humanely