Selection, part 1

1
Selection, part 1

• OS352 HRM
• Fisher
• Feb 21, 2005

2
Agenda

• Finish material on recruiting
• Impact of legal environment on selection process
• Basic characteristics of selection tools
• Reliability
• Validity

3
Legal Requirements

• Must follow requirements of Civil Rights Act,
  ADA, etc.
• Selection procedures cannot address
• Race, gender, disabilities, age
• BFOQ exceptions
• Pre-employment medical tests
• Drug testing

4
Review Disparate treatment vs. disparate impact

• What is disparate treatment?
• What is disparate impact?
• 4/5ths rule (from EEOC guidelines)
• How do these concepts affect the selection
  process?

5
Four-fifths rule
Hired
Group
of Applicants
Selection Rate
134
47.5
Whites
282
10
27.8
36
Blacks
44
43.1
Hispanics
102
170
44.2
Men
385
Women
18
51.4
35
Is there evidence of adverse impact? Against
which group(s)?
Data source Nkomo, S., Fottler, M., AcAfee,
R.B. (2000) Applications in Human Resource
Management.
6
Griggs v. Duke Power(1971 U.S. Supreme Court)

• Duke Power required a high school diploma and
  passing score on two general aptitude tests
• Black applicants were disproportionately excluded
• Duke Power was unable to demonstrate the these
  requirements were job related and consistent with
  business necessity
• Not necessary to perform most jobs in the plant
• Covered by Civil Rights Act, 1964

7
Selection criteria

• We want to select people who will succeed
• Performance
• Turnover
• Potential
• How do we determine the right criteria? What
  should we be measuring?
• Use the job analysis information to determine
  critical KSAOs
• Knowledge/skills that are easy to learn should
  generally not be used as selection criteria

8
Consequences of Selection Decisions

• Which is worse false positive or false
  negative?
• If false positive is worse, selection criteria
  should be more stringent

False Positive
True Positive
Selection Decision
-
False Negative
True Negative

-
On-the-job Performance
9
What is a selection test?

• Any device used to determine an applicants
  appropriateness for the job
• Written tests (ability, personality, knowledge)
• Interviews
• Application blanks
• Assessment centers
• All selection tests are subject to Uniform
  Guidelines

10
Reliability

• Error in a measure
• Test retest (time)
• Parallel forms (items)
• Coefficient alpha (items) easiest to assess
• Inter-rater (raters)
• Reliability coefficients range from 0-1
• Acceptable is .7 or higher

11
Validity

• Are we measuring the right thing?
• Not a characteristic of a test, but how the test
  is used
• Content validity
• Based on results of thorough job analysis
• Relies on judgment
• Criterion-related validity
• Demonstrate that the test predicts performance
• Uses statistical techniques
• Face validity test taker perceptions of
  relevance

12
For next class

• Topic More on selection
• Benefits and drawbacks of different selection
  tools
• Writing interview questions
• Current issues in selection
• Will also start preparing for Interviewing
  Workshop on Feb 28
• Exercise 2 due at the beginning of class
• Must be word processed.
• Be sure to explain your recommendations clearly,
  using the data you have analyzed.