Selection &

Performance
Evaluation
Module layout
• This course is all about application
• Relationship between OB and HR
• Defining and measuring job performance
• Job analysis
• Review selection tools and procedures
• Best practices in performance management
What is OBHR?
What is OB/HR?
• Organizational Behavior (OB): Scientific
approach to the management of people within
an organization
• Human Resources (HR): formal systems
devised for the management of people within
an organization
• OB is to HR as physics is engineering
 Content areas
Organizational Behavior Human Resources
• Individual differences • Staffing and Training
• Attitudes & Perceptions • Compensation
• Motivation • Discrimination
• Group dynamics • Administration
• Communication • Performance MGMT
• Leadership
 Approach
• Scientific methods
model
• Simplified views of “All models
reality are wrong,
• Propose causal but some
relations
are useful.”
• Falsifiable
• Data-driven
George Box

1-6
Conclusion
• We will focus on the rule while recognizing
there’s always an exception
• Civil law not criminal law
• Greats recognize when to zig when the rule
says zag
 What is job
performance?
Job performance
• An umbrella term that refers to the collective
inputs a worker contributes to the organization
 Task Performance
Tasks, duties, and responsibilities associated
with the job

10
 Organizational Citizenship Behaviors
Voluntary employee activities that may or may
not be rewarded but that contribute to the
organization by improving the overall quality of
the setting in which work takes place.

Interperson Organizati
al onal

11
Counterproductive Work
Behaviors
Voluntary behaviors that harm the
organization or its members

Interperson
al

Organizati
onal

12
One view of management
• Manager’s job
• Improve quality of task performance
• Increase the frequency of OCB
• Decrease the frequency of CWB
• A need to understand the individual, team, and
organizational factors that contribute to them.
 Measuring
Performance Part 1
Performance management cycle

6-15
Measuring performance
• Referred to as the “Criterion Problem”
• Selection systems only operate well when we know
what we are trying to predict (Bock, Ch. 5)
• Compensation systems only operate well when we
know what we are trying to improve (Bock, Ch. 8)
 Contamination, Validity,
Deficiency

performance Actual
measure performance
Darley's Law
• “The more any quantitative measure is used to
assess performance, the more subject it will be
to corruption.”
• Good or bad, however you measure past
performance will change future performance
• Please don’t abandon performance
measurement, just factor Darley’s Law into it.
Job Analysis
Why do it?
• Describe
• Classify
• Develop
• Evaluate
• (Re)design
What is needed?
• Action Plan
• Support from above
• Cooperation from below
Steps in a job
analysis
Steps in the Process
1. Know your organization
2. Decide and communicate use
3. Determine data source & type
4. Set schedule
5. Collect data
6. Synthesize
Collecting data
• Incumbent to start
• Data exist on continuum
• Two types of data
Data Type 1: TDRs
• Elements
• Tasks
• Duties
• Generalized work activities
• *Essential job functions
Data Type 2: KSAOs
• Knowledge
• Declarative and procedural
• Skills
• Abilities
• Other requirements
Data synthesis
• Verbs
• Observable
• Edit, edit, edit
• Too vague/specific
• Frequency and importance of activity
• Alignment between KSAs and TDRs
• Combinations or subdivision possibilities
Job Analysis Example
Observe and interview
• Attend a briefing
• Monitor roadways and community
• Respond to emergency/criminal report
• Issues summons
• Arrest alleged criminals
• Complete paperwork
Responding to a bar fight
• Hear a call from dispatcher
• Respond to dispatcher
• Operate a vehicle at high speed
• Communicate with suspects
• Determine if and type of crime committed
• Physically restrain and arrest suspect
• Coordinate detention
• Complete paperwork
Attraction-Selection-Attrition
• Poison the applicant pool
• Hire bad candidates
• Train cadets improperly
• Evaluate officers on bad criteria
• Lose good officers
• Homogenize your workforce
• Establish and reinforce bad norms
• Make organizational change more difficult
Selection Tools That Stink
No.
NO!
 Ridiculous.
"Estimate how
many windows are
in New York."
“You have selected
to be on Iron Chef.
How do you prepare
your team for the
competition?
What? Of course not
Please stop…
Enjoy your lawsuit.
Selection Tools That Don't
Stink
20+ Years and counting…
 #1: Work Sample Tests
• Strongest predictor
• Experience/hard skill
positions
• Captures ability and
task performance
• Costly, but high ROI
 #2: Structured Interviews
• Excellent predictor
• Perceived as fair
• Reduces bias
• Easier comparisons
• Stay on script, but
allow follow ups
Structured Interviewing
• Training needed
• No brainteasers
• Interview your interviewers
• Mix of behavioral and situational questions
• Use a rubric (e.g., BARS)
• Keep track of outcomes
• Not just superiors in the room
 #3: Job Knowledge Tests
• Two types
• Specific knowledge
• Level of experience
• Face valid
• Better when there is a
right answer
 #4: GMA tests
• Best individual
difference predictor
• Higher complexity jobs
• Low face validity
• Numbers are somewhat
juiced
• Differential validity
 #5: (Some) Personality Tests & Traits
• Strongest traits
• Conscientiousness
• Agreeableness
• Emotional Stability
• Integrity
• Extraversion for some jobs
• Quality measures needed
• NEO-PI-R, 16 PF, IPIP
• Can fake a bit
• Context dependent
• Best aren't that predictive
 #6 Assessment center
• High face validity
• Predictive of soft skills
• Tap a number of
relevant skills and traits
• Expensive
• Quality varies
• Old wine in new bottles
• Group dynamics can
skew
 #7: EI
• Objective tests
available
• Malleable
• Less faking
• Best tests are g-loaded
• Less empirical evidence
• Unclear variance across
jobs
Selection Odds and
Ends
Where does your funnel leak?
Plan ahead for hiring
• Identify a diverse search committee
• Review, as a committee, the job description
• Identify the KSAs
• Develop well-defined rubrics
• Discuss what is and what is not appropriate
Don't ask these questions
• Are you married? Single? Divorced?
• Do you have kids? Are you planning on having
kids? How do you handle childcare?
• How long do you think you’ll work before
retirement?
• Are you in good health?
• Will you be able to manage this work given
[insert protected characteristic…]?
Definitely don't ask these!
• How do you define your race?
• Are you pregnant?
• When were you born?
• Do you attend church/mosque/synagogue?
• Are you a member of the LBGTQ+ community?
• Is there any reason you would need time off
for your disability?
Test your process
• Track everything
• Reduce/eliminate bias
• Follow up with applicants
• Follow up with new hires
• Follow up with incumbents
• As job changes, recruiting changes
Actively Recruit
• Referral programs
• Often highest yield
• Solicit your best
• Encourage poaching!
• Go to where the applicants are
• Create a talent pool
• Assess likely needs (TDRs)
• Target those with that meet needs (KSAs)
• Use it
Performance
Management
Compensating and Rewarding
Employees
• Cannot reward until we know…
• what we want
• the extent person is capable of giving us
what we want
• how well the person is giving us what we
want
Performance management cycle
What are the essentials of performance
management?

• Performance management involves

two purposes:
• It serves an evaluation purpose when it
lets people know where their actual
performance stands relative to objectives
and standards.
• It serves a developmental purpose when it
provides insights into individual strengths
and weaknesses.
Performance appraisal
• Formal procedure for measuring and
documenting a person’s work performance.
• How can we overcome subjectivity and
arbitrariness of supervisor ratings
• How can we mitigate cognitive biases
• Halo error
• Leniency bias
• Central tendency bias
• Recency bias
#1 Don’t rate them in a
bubble
Comparative Methods of Performance
Appraisal
Ranking Paired Comparisons Forced Distribution
Raters rank order Raters compare Raters place a
individuals from each person with specific proportion
best to worst on every other person. of employees into
overall each performance
performance. standard
 #2 The supervisor doesn’t know all
• 360° Evaluation
• Uses a combination of evaluations from a person’s
bosses, peers, and subordinates, as well as internal
and external customers and self-ratings.
#3 Supervisors have bad memories
• Critical incident diaries
• Written records that give examples of a person’s
work behavior that leads to either unusual
performance success or failure.
#4 Reduce subjectivity
• Rating scales
• Graphic – Lists a variety of performance
dimensions that an individual is expected to
exhibit.
• Behavioral – Adds more sophistication by linking
ratings to specific and observable job-related
behaviors.
Graphic rating scale.
A sample BARS for a customer
service representative
Alternative Approach
Counterpoint to traditional thinking…
• Buckingham and Goodall put out what is fast
becoming a highly influential HBR article
regarding performance appraisal
• Picking up on a lot of frustration with HR
specialists both within industry and academics
and demanding change
• All is based on an intervention at Deloitte
(a big integration/network solution
company)
 Frustration about time, cost, and
accuracy
• 2M hours a year
• Single summary
number for each
worker too coarse
• Ratings revealed
more about the
people who give
them
• IRR around .50
The new way!
3 Goals

• Tie it to pay
• reward high performers with raises and bonuses
• The best judge is the team leader
• Gone are 360 evals & skill assessments. In are feelings
and intentions toward an employee
• firm essentially surveys leaders about what action they
would take with each team member
• clarify what’s expected, what great work looks like,
and how each person can excel
Key differences
• Check-ins are weekly, instead of annually
• No forced distribution
• 4 questions only
Key differences
Bottom line about performance
evaluation
• You have to do it
• Stick to SMART goals and watch out for Goals
Gone Wild
• Watch out for the illusion of precision
• Experiment
• Ensure fairness
• Separate the evaluation talk from the reward
talk