Employee Selection:

Reference & Testing

 References
• Reference
– The expression of opinion , either orally or through a written
checklist, regarding an applicant’s ability, previous
performance, work habits, character, or potential for future
success.
• Reference Check
– The process of confirming the accuracy of information
provided by an applicant
• Letter of Recommendation
– Letter expressing opinion regarding an applicant’s ability,
previous performance, work habits, character, or potential
for future success. Content and format are determined by
letter writer (usually)
Why Require References?
 Why Require References?
1. Check for resume fraud
2. Find new information about the applicant
3. Check for potential discipline problems
4. Predict future performance
 1. Checking for Resume Fraud
• Why Check? • Obtaining Missing
– 1/3 resumes contain information
inaccurate info – unintentional omission
– over 500,000 people – strategic omission
have bogus degrees – deceptive omission
• Verifying Information • Alternative methods
– truth – bogus application items
– error – social security reports
– embellishment – hire professional
– fabrication reference checkers
 2. Finding New Information
About the Applicant
• Types of Information • Alternative Measures
– personality – psychological tests
– interpersonal style – letters of
– background recommendation
– work habits – biodata
• Problems – resumes
– references seldom agree – interviews
– people act in different
ways in different
situations
 3. Checking for Potential
Discipline Problems
• Criminal Records-sexual harassment, poor
attendance and violence
• Prevent negligent hiring
– Previous employers
– Motor vehicle records
– Military records
– Credit reports
– Colleges and universities
– Neighbors and friends
 Criminal Records
• Obtained from local and state agencies
• Check with each jurisdiction
• Only convictions can be used (EEOC Decision No. 72-
1460)
– “Reasonable amount of time” between release and
decision to hire
– In using convictions, employer must consider
• Nature and gravity of offense
• Amount of time that has passed since the conviction and/or
completion of the sentence
• The nature of the job held or being sought
 Credit Checks
• Purpose
– Predict motivation to steal
– Determine character of applicant
• Fair Credit Reporting Act
– Order through a Consumer Reporting Agency (CRA)
– Provide written notice to applicant to you will be checking credit
– Get applicant’s written authorization to check credit
– If adverse action is to be taken
• Provide applicant with “Pre-adverse Action Disclosure” which includes
copy of credit report
• Inform applicant that they will not be hired due to credit check and
provide name of CRA and notice of applicant rights to appeal within 60
days
 4.Predicting Future Performance
• References are not good predictors of performance
– Uncorrected validity is .18
• References are not reliable (r = .29)
– High correlation between two letters written by the same person for
two people than between letters written by two people for the same
person
– They say more about the person writing the letter than the person
being written about
• References are lenient
– Fewer than 1% of applicants are rated below average!
 Problems with References- 1.
Leniency?
1. Applicants often choose their own
references
2. Applicants often have the right to
see their files
3. Former employers fear legal
ramifications
– Charged with defamation of character
4. Confidentiality concerns
 Potential Legal Ramifications
• Negligent hiring
– Potential employer
(seek)
• Invasion of privacy
• Negligent reference
– Former employer
(provide)
• Defamation
– Libel (written)
– Slander (oral)
– Self-publication
 Problems with References- 2.
Knowledge of the Applicant
• May not know applicant well
• May not have had
opportunities to observe
behavior
• Employees act differently
around supervisors
 Problems with References- 3.
Reliability
• Lack of agreement between two people
providing the reference for the same person
– Supervisor vs. Professor
• Research indicates reference reliability is only
.22 (Aamodt & Williams, 2005)
– High correlation between two letters written by the same person
for two people than between letters written by two people for
the same person
– They say more about the person writing the letter than the
person being written about
 Problems with References- 4.
Extraneous Factors Surrounding
the Reference
• Reference giver’s • The words used by the
ability to articulate reference giver
– Letters containing – she has no sexual oddities
specific examples were that I am aware of
rated higher than a – I have an intimate and
general one caring relationship with
• The extent to which the applicant
the referee remembers – Jill is a bud that has
the applicant already begun to bloom
 Some Recommendations
1. Increase Structure
– Job analyses
2. Explicitly stating relationship with the person
being recommended
3. Be honest
4. Let applicant see your reference
 Trait Method of Evaluating
Letters of Recommendation
• Peres and Garcia (1962)
• The Technique
– Read each letter
– Highlight traits in each letter
– Place each trait into one of five categories
• Mental agility (openness to experience)
• Vigor
• Urbanity (Extroversion)
• Cooperation-Consideration (Agreeableness)
• Dependability-Reliability (Conscientiousness)
– Total the number of traits per category
– Divide the number of traits per category by the total number of traits
Dear HR Director,

Ms. Rachel Green asked that I write this letter in support of her
application as an assistant manager and I am pleased to do so. I
have known Rachel for six years as she was my assistant in the
accounting department.

Rachel is one of the most popular employees in our agency as she

is a kind, outgoing, sociable individual. She has a great sense of
humor, is extroverted, and is very helpful. In completing her
work, she is independent, energetic, and motivated.

Mental Ability:0 Urbanity: 5 Vigor: 3

Cooperation: 2 Dependability: 0
Dear HR Director,

Ms. Monica Geller asked that I write this letter in support of her
application as an assistant manager and I am pleased to do so. I
have known Monica for six years as she was my assistant in the
accounting department.

Monica always had her work completed accurately and

promptly. In her six years here, she never missed a deadline.
She is very detail oriented, critical, and methodical in her
problem solving approach. Interpersonally, Monica is very
caring and helpful.

Mental Ability:0 Urbanity: 0 Vigor: 0

Cooperation: 2 Dependability: 6
 Validity of the Trait Method
Trait Category Graduate GPA Teaching Ratings
Mental agility .15* -.09
Vigor - .16* - .03
Urbanity .02 .19*
Cooperation - .07 .09
Dependability .02 -.18
 Personnel Selection Methods
• Training & Education • Skills
• Experience – Work Samples
– Applications/Resumes – Assessment Centers
– Biodata – References
– Interviews • Personality & Character
• Knowledge – Personality Tests
• Ability – Integrity Tests
– Cognitive • Medical
– Physical – Medical Exams
– Perceptual – Psychological Exams
– Drug Testing
Predicting Performance Using
Training and Education
 Ratings of Training
1. Education
2. Work-Related Training
3. Military
Does Education
Predict
Performance?
 Summary of Meta-Analyses
Meta-analysis Occupation K N ρ

Ng & Feldman (2009) Police .28-.38

Aamodt (2002) Police 38 9,007 .34

Vineberg & Joyner (1982) Military 35 .25

Hunter (1980) USES data 425 32,124 .10

Hunter & Hunter (1984) base
Schmidt & Hunter (1998)
Dunnette (1972) Entry level 15 .00
petroleum
 Education and Incremental Validity
• Schmidt & Hunter (1998) say no
– Cognitive ability (r = .51)
– Cognitive ability and education (r = .52)
 Validity of GPA
• GPA is a valid predictor of performance on the
job, training performance, starting salary,
promotions, and grad school performance
• GPA is most predictive in the first few years after
graduation (Roth et al., 1996)
• GPA will result in high levels (d=.78) of adverse
impact (Roth & Bobko, 2000)
• People with high GPAs
– Are intelligent (r = .50; Jensen, 1980)
– Are conscientious (r = .34; Bevier et al., 1998)
 Validity of GPA
Meta-Analysis Results
r ρ
Work-Related Criteria
Job performance (Roth et al., 1996) .16 .36
Training performance (Dye & Reck, 1989) .29
Promotions (Cohen, 1984) .16
Salary (Roth & Clarke, 1996)
Starting salary .13 .20
Current salary .18 .28
Graduate School Performance (Kuncel et al., 2001)
Grades .28 .30
Faculty ratings .25 .35
 Predicting Performance Using
Applicant Knowledge-Job
Knowledge Tests
• Taps job-related knowledge
• Good validity (ρ = .48)
• Face valid
• They could be scored in
multiple choice format or
essay format
• Can have adverse impact
Predicting Performance Using
Applicant Ability
 Cognitive Ability Tests
• Taps
– Learnability or coachbility of the applicant
– Better decision making by better
information processing
• High validity (ρ = .51)
• Predicts training and job performance
for all jobs (Hunter, 1986)
• The more complex the job, the better
cognitive ability tests predict
performance
 Cognitive Ability Tests
Strengths
– Highest validity of all selection
measures (ρ = .51)
– Easy to administer
– Relatively inexpensive
– Most are not time consuming
 Cognitive Ability Tests
Weaknesses
– Likely to cause adverse
impact
– Low face validity
– Not well liked by applicants
– Frequently challenged in
courts
– Difficult to set a passing
score
 Perceptual Ability Tests
• Perceptual Ability
(Fleishman & Reilly
(1992)
– Vision (near, far, night,
peripheral)
– Depth perception
Machinist,
cabinet maker, – Glare sensitivity
die setter – Hearing (sensitivity,
– auditory attention, sound
localization)
 Psychomotor Ability Tests

• Psychomotor Ability (Fleishman & Reilly (1992)

– Dexterity (finger, manual) Police officer,
– Control precision carpenter,
– Multilimb coordination sewing machine
– Response control operator
– Reaction time
– Arm-hand steadiness
– Wrist-finger speed
– Speed-of-limb movement
 Physical Ability
• Used for jobs with high physical demands
• Three Issues
– Job relatedness
– Passing scores
– When the ability must be present
• Two common ways to measure
– Simulations
– Physical agility tests
 Physical Ability
Physical Abilities (Fleishman & Reilly, 1992)
– Dynamic strength (strength requiring repetitions)
– Trunk strength (stooping or bending over)
– Explosive strength (jumping or throwing)
– Static strength
– Dynamic flexibility (speed of bending or stretching)
– Extent flexibility (Degree of bending or stretching)
– Gross body equilibrium (balance)
– Gross body coordination (coordination)
– Stamina
Police officer, fire fighter, and
life guard-job simulations,
dragging a hose, climbing a
ladder
 Physical Ability
Controversies
1. Job Relatedness: Some out of shape cops perform
safely and at high levels
– Technological advances. Pepper spray study (Sollie &
Sollie, 1993)
2. Passing Score
– Adverse Impact: Can’t use relative standards
– Passing score are set are minimum levels needed to
perform a job
3. When abilities must be present: may not require
the ability until in the academy or on the job
Predicting Performance Using
Applicant Skill
 Work Samples
• Applicants perform tasks that replicate actual
job tasks
• Advantages
– Directly related to the job
– Good criterion validity
• Verbal work samples (ρ = .48)
• Motor work samples (ρ = .43)
– Good face validity
– Less adverse impact than cognitive ability
– Provide realistic job previews
• Disadvantages
– Can be expensive to develop and maintain
 Assessment Centers
What are They?
• A selection technique that uses multiple job-related
assessment exercises and multiple assessors to
observe and record behaviors of candidates
performing job-related tasks
 Guidelines for Assessment Center Practices
Joiner (2000)
• Based on job analysis • Use multiple assessors
• Use multiple • Assessor training
assessment exercises • Assessors must
• Simulations prepare reports
• Behavioral • Overall judgment
Observations must be based on integration of
documented information
 Assessment Center Exercises
• Leaderless group discussions
• In-basket technique
• Simulations
– Situational exercises
– Work samples
• Role plays
• Case analyses and business
games
 Evaluation of Assessment Centers
Reliability
– Can have low inter-rater agreement among raters
– Test/retest reliability pretty high (.70)
Validity
– Uncorrected .28
– Corrected .38
– Good face validity
Evaluation of Assessment Centers
Weaknesses
– Very expensive
– Time consuming
– Can have low inter-rater agreement
– Behaviors can overlap into several
dimensions
– Safety of candidates for some work
samples
When are assessment centers most appropriate?
– Most useful for promotion rather than selection
– When candidates have some knowledge of the job
– When you have the money to develop and maintain
assessment centers
– When you have the time and trainers
Predicting Performance Using
Prior Experience
 Experience Ratings
• Past behavior predicts future behavior
– Experience is a valid predictor of future
performance (ρ = .27; Quinones et al.,
1995)

• Types of Experience
– Work
– Life
 Experience
• Evaluated through:
– Application blanks
– Resumes
– Interviews
– Reference checks
– Biodata instruments
 Experience
• Considerations
– How much experience?
– How well did the person perform?
– How related is it to the current job?
 Experience Predicts Best…
• Credit prior work experience only:
– In the same occupational area as that in which performance is to be
predicted
– In the performance of tasks or functions that have direct
application on the job
• Recency of experience should be used as a decision rule for awarding
credit only when justified on a case-by-case basis
• Credit for duration of work experience should be limited to a few
years.
• High prediction up to about 3 years of experience, declining to low
prediction for more than 12 years of experience.
 Biodata
A selection method that considers an
applicant’s life, school, military,
community, and work experience
 Example of Biodata Items

Member of high school student government?

Yes No
Number of jobs in past 5 years?
1 2 3-5 More than 5
Transportation to work:
Walk Bus Bike Own Car Other
 Development of Biodata Items
1. Choose a job
2. Create pool of potential biodata items
3. Choose a criterion to measure behavior
4. Prescreen items and test on employees
5. Retest items on second sample of
employees
Development of Biodata Items

1. File Approach
– Obtain information through personnel files
2. Questionnaire Approach
3. Criterion Groups
– Quantity, absenteeism, performance, turnover
 Biodata Scoring
Vertical Percentage Method Variable Long Short Differen Unit
1. Percentage is calculated for Tenure Tenure ces in % Weight
each group on each item (%) (%)
2. Percentage of particular Education
response for the low group
is subtracted from the
percentage of the same High 40 80 -40 -1
School
response in the high group
to obtain weight for that
Bachelor 59 15 +44 +1
item ’s
3. One weights are assigned
then weigh information and Masters 1 5 -4 0
sum it up
4. Composite scores are
correlated with criterion
 Biodata Standards
Gandy & Dye, 1989; Mael, 1991
Good Biodata Items Bad Biodata Items
Historical Future or Hypothetical
How old were you when you got What position do you think you will
your first paying job? be holding in 10 years?
External Internal
Did you ever get fired from a job? What is your attitude toward friends
who smoke marijuana?
Objective Subjective
How many hours did you study for Would you describe yourself as
your bar exam? shy?
First-hand Second-hand
How punctual are you about coming How would your teachers describe
to work? your punctuality?
 Good Biodata Items Bad Biodata Items
Discrete Summative
At what age did you get your driver’s How many hours do you study during an
license? average week?
Verifiable Non-verifiable
What was your grade point average in How may servings of fresh vegetables do
college? you eat everyday?
Controllable Non-controllable
How many tries did it take you to pass How many brothers and sisters do you
the CPA exam? have?
Equal Access Non-equal Access
Were you ever class president? Were you ever captain of the football team?

Job Relevant Not job relevant

How many units of cereal did you sell Are you proficient at crossword puzzles?
during the last calendar year?
Noninvasive Invasive
Were you on the tennis team in college? How many young children do you have at
home?
 Strengths of Biodata

– Good validity (r = .36, ρ= .51)

– Can predict for variety of
criterion measures
– Easy to administer
– Relatively inexpensive
– Fairly valid
– Can have good face validity
 Weaknesses of Biodata

– Low face validity

– Can invade privacy
– Items can be offensive
– Expensive to develop
– Not always practical to
develop
 Validity Issues
• Shrinkage?
• Good validity but not sure why
• Validity seems to drop when items based
rationally (job analysis) rather than
empirically
 Differences in Personality
Inventories
• Types of Personality Inventories
– Measures of normal personality
– Measures of psychopathology
• Basis for Personality Dimensions
– Theory based
– Statistically based
– Empirically based
• Scoring
– Objective
– Projective
Five-Factor Model (The Big 5)
Openness to Experience
– imaginative, curious, cultured
Conscientiousness
– organized, disciplined, careful
Extraversion
– outgoing, gregarious, fun-loving
Agreeableness
– trusting, cooperative, flexible
Neuroticism (emotional stability)
– anxious, insecure, vulnerable to stress
 Validity of Personality
Meta-Analysis

Hurtz & Donovan Barrick & Mount Tett et al. (1991)

(2000) (1991)
Dimension Observed True Observed True Observed True
Openness .03 .06 .03 .04 .18 .24
Conscientiousness .15 .24 .13 .22 .12 .16

Extroversion .06 .09 .08 .13 .10 .13

Agreeableness .07 .12 .04 .07 .22 .28

Neuroticism - .09 - .15 - .05 - .08 - .15 - .19

 Comparison of Meta-Analyses
Conscientiousness
Hurtz & Donovan Barrick & Mount Tett et al. (1991)
(2001) (1991)
Types of studies Only those Any test that could Only studies in
included in meta- developed to tap be assigned to a Big which a Big 5
analysis Big 5 5 dimension dimension was
hypothesized to be
related to
performance
k 42 123 7
n 7,342 19,721 450
Observed validity .15 .13 .12
 Evaluation of Personality
Strengths
– Relatively cheap
– Easy to administer
– Little adverse impact
– Predicts best when based on a
job analysis
Weaknesses
– Scale development
– Validity
– Faking
 Interest Inventories
• Tap an applicant’s interest in particular
types of work or careers
• Poor predictors of job performance (ρ = .10)
• Better predictors of job satisfaction
 Integrity Tests
• Estimate the probability that applicants
will steal money or merchandise
• Used mostly in retail, but gaining
acceptance for other occupations
 Types of Integrity Tests

Electronic Testing
• Polygraph testing

Paper and Pencil Testing

• Overt
• Personality based
 Polygraph Testing
• Polygraph (lie detector) is a machine that
measures the physiological responses that
accompany the verbal responses an individual
makes to a direct questions asked by
polygraph operator.
Limitations of the Polygraph
• Emotions other than
guilt can trigger
responses
• Countermeasures used
to avoid detection
• Frequency of false
positives
• Frequency of false
negatives
 Legal Guidelines for Polygraph
Testing
Employee Polygraph Protection Act of 1988
makes it illegal to:
• Directly or indirectly require an employee to take a
polygraph
• Use, accept, refer to, or inquire about the results of any
polygraph test of any applicant or employee
• Discharge, discipline, discriminate against, or deny
employment or promotion to (or threaten such actions)
against any prospective or current employee who
refuses, declines, or fails to take or submit to a
polygraph
 Legal Guidelines for Polygraph
Testing
The following are exempt from these prohibitions

– Private employers providing security services

– Employers who manufacture, distribute, or dispense

controlled substances

– Federal, state, and local government employees.

 Paper and Pencil Integrity Tests
Overt integrity tests
• Directly ask for attitudes about theft and
occurrences of theft behavior
Personality based measures
• Measure traits linked to several theft related
employee behaviors that are detrimental to the
organization
 Overt Integrity Tests
Rationale is to measure job applicants’ attitudes and cognitions
toward theft that might predispose them to steal at work,
especially when both the need and opportunity to steal are
present. More Reliable & Valid
Validity has been determined by comparing scores with:
• Polygraph test results
• Self-admissions of thefts
• Shrinkage
• Future theft
 Other Behaviors Integrity Tests Can
Predict
• Drug and alcohol abuse
• Vandalism
• Sabotage
• Assault behaviors
• Insubordination
• Absenteeism
• Excessive grievances
• Bogus workers compensation claims
• Violence
 The Validity and Reliability of Integrity
Tests
Validity
• Theft
• .41 for predicting probability of theft by employees
• Performance (Ones et al. 1993)
• Observed = .21
• True = .34
Reliability
• Reports of test-retest reliabilities between .90-.70
 Evaluation of Integrity Tests
• Advantages
– Good validity (ρ = .34)
– Inexpensive to use
– Easy to administer
– Little to no racial adverse impact
• Disadvantages
– Males have a higher fail rate than females
– Younger people have a higher fail rate than older people
– Failure has a negative psychological impact on
applicants.
 Conditional Reasoning Tests
• Designed to reduce faking
• Applicants are given a series of statements and asked to select the
reason that justifies each statement
• Aggression predicts CWB (r = .16) and job performance (r = .14)
• Aggressive individuals tend to believe
– most people have harmful intentions behind their behavior (hostile
attribution bias)
– it is important to show strength or dominance in social interactions
(potency bias)
– it is important to retaliate when wronged rather than try to maintain a
relationship (retribution bias)
– powerful people will victimize less powerful individuals (victimization
bias)
– evil people deserve to have bad things happen to them (derogation of
target bias)
– social customs restrict free will and should be ignored (social discounting
bias).
 Graphology
• Concept
– A person’s handwriting is a reflection on his or her personality and
character
• Use
– 6,000 U.S. organizations
– 75% of organizations in France
– 8% of organizations in the United Kingdom
• Evaluation
– Few studies
– Validity depends on the writing sample
• Autobiographical (r = .16, p = .22)
• Non-autobiographical (r = .09, p = .12)
 Drug Testing
Forms of Testing
– Pre-employment testing
– Random selection at predetermined times
– Random selection at random times
– Testing after an accident or disciplinary action
Responses to the Presence of Drugs
– 98% of job offers withdrawn
– Current employees who test positive
• 25% are fired after a positive test
• 66% are referred to counseling and treatment
 Two Stages of Drug Testing
• Initial screening of hair or urine
– Cheaper method ($30 for urine, $50 for hair sample)
– Enzyme Multiplied Immunoassay Technique (EMIT)
– Radioimmunoassay (RIA)
• Confirmation test
– Typically used only after a positive initial screening
– Thin layer chromatography/mass spectrometry
– More expensive
 Other Exams
1. Psychological Exams
2. Medical Exams
– After the offer has been made
– For jobs requiring public safety
 Typical Corrected Validity Coefficients for
Selection Techniques
Method Validity Method Validity
Structured Interview .57 College grades .32
Cognitive ability .51 References .29
Biodata .51 Experience .27
Job knowledge .48 Conscientiousness .24
Work samples (verbal) .48 Unstructured interviews .20
Assessment centers .38 Interest inventories .10
Integrity tests .34 Handwriting analysis .02
Situational judgment .34 Projective personality tests .00
 Summary of Types of Selection
Methods
1. Training and Education 4. Knowledge
• Degree • Job Knowledge Tests
• GPA 5. Experience
2. Ability • Application Blank
• Resumes
• Cognitive • Interviews
• Perceptual • Reference Checks
• Pychomotor • Biodata
• Physical 6. Personality and
3. Applicant Skills Character
• Personality Tests
• Work Samples
• Integrity Tests
• Assessment Techniques • Interest Inventories
• Resume
 Rejecting an Applicant
• A personally addressed signed letter
• Appreciating for applying
• Complimenting qualifications
• Some information about the individual who
got hired
• A wish of luck
• A promise to keep application on file