Lecture 1: Deloitte
Staffing → different processes, but interrelated
1. Recruitment → Identifying, attracting
2. Selection → Assessing, evaluating
3. Employment → Decision making, final matching
Challenges Deloitte: Different candidates, optimal candidate journey, fun, non-biased
Validity → Measure what you want to measure
- Low validity: CV screening
- Medium validity: Personality test (Big 5), unstructured interviews
- High validity: IG/EQ, structured interview, holistic interview, case interview, game,
agility test
Tutorial 1
Article: Clarck & Watson – Constructing validity: Basic Issues in scale development
Goal scale development → create a valid measure of an underlying construct
1. Clear conceptualization of your construct
2. Initial item pool should be overinclusive and item wording needs attention
3. Structural validity: item selection and psychometric evaluation
4. The goal is unidimensionality rather than internal consistency in selecting scale items.
Structural validity: The extent to which a scale’s internal structure (the inter item
correlations) parallels the external structure of the target trait (correlations among nontest
manifestations of the trait)
Test construction strategies:
• Criterion-based methods
o A test response should be considered verbal behavior in its own right, with
non-test correlates to be discovered empirically – test developers embraced
criterion keying as a method that permitted a wide range of practical problems
to be addressed in an apparently straightforward manner
• Internal consistency methods
o Most used method for item selection in scale development is some form of
internal consistency analysis. Factor analytic methods were used most
frequently when the target construct was conceptualized as multidimensional
and subscales were desired.
o Issues
▪ Factor analytic results provide information, not answers or solutions (a
tool that can be used wisely or foolish)
▪ There is no substitute for good theory and careful thought when using
these techniques.
1
, • Item Response theory (IRT)
o Based on the assumption that test responses reflect an underlying trait (or set
of traits) and that the relation between response and trait can be described for
each test item by a monotonically increasing function called item characteristic
curve (ICC)
▪ Individuals with higher levels of the trait have higher expected
probabilities for answering an item correctly
The initial data collection → inclusion of comparison (anchor) scales + sample considerations
Psychometric evaluation
• Analysis of item distributions → before the more complex analyses.
o Identify and eliminate items that have highly skewed and unbalanced
distributions. Highly unbalanced items are undesirable
▪ When most respondents answer similarly, items convey little
information
▪ Owing to their limited variability, these items are likely to correlate
weakly with other items in the pool and therefore will fare poorly in
subsequent structural analyses
▪ Items with extremely unbalanced distributions can produce highly
unstable correlational results
o It is desirable to retain items showing a broad range of distributions.
• Unidimensionality, internal consistency and coëfficient alpha
→ shows which items you need to delete from the pool
o Internal consistency: Refers to the overall degree to which the items that
make up a scale are intercorrelated
o Homogenity/unidimensionality: the scale items assess a single underlying
factor or construct
▪ Internal consistency is a necessary but not sufficient condition for
homogeneity or unidimensionality (A scale cannot be homogeneous
unless all of its items are interrelated, but as we illustrate later, a scale
can contain many interrelated items and still not be unidimensional
Average inter item correlation (is a straightforward measure of internal consistency)
The average interitem correlation alone cannot establish the unidimensionality of a scale:
A multidimensional scale actually can have an acceptable level of internal consistency.
It is necessary to examine the range and distribution of these correlations as ell.
Guideline is that all of the interitem correlations should fall in the range of 0.15 to 0.50
The attenuation paradox →Increasing the internal consistency of a test beyond a certain
point will not enhance its construct validity and, in fact, may occur at the expense of
validity
Article: Sackett – Individual Differences and their measurement
Article focuses on 3 individual differences domains:
1. Knowledge, skill and ability
2. Personality
(incl integrity, emotional intelligence, stable motivation attributes, creativity)
3. Vocational interest
Cognitive ability was prominent in early research, but questions about limits began early.
Currently: the field agrees on hierarchical structure with general ability on top and more
specific abilities below.
Knowledge + skill → part of research in many domains, are likely to be the most proximal
determinants of differences in job performance, but only if certain conditions apply
Personality and vocational interest → waves of optimism and skepticism
2
, Lecture 2: What about recruitment?
Types of recruitment:
1. Corporate recruitment: In-house, often part of HRM, hiring manager & candidate
2. Agency recruitment: External, independent from HRM, customer & candidate
Define recruitment goals → Always start with the organization strategic goals
• Goals for attraction: what KSAO needed?
• Goals for speed: How fast do we need to attract and when to start?
• Time frame:
Open/targeted approach
• Targeted recruitment
o Find applicants with specific KSAO’s, prepare specific message + approach,
helps to find and attract passive job seekers, essential for complex and
specific jobs
• Open recruitment
o Cast a wide net with little effort, diverse set of applicants, essential when large
numbers of applicants needed, time consuming
Organization + Administration Decisions
• In-house or agency
o In-house: Large organizations
▪ Know the culture, signaling effect to applicants of people-oriented
culture, cheaper
o Agency: SME’s or complex vacancies
▪ Specialized in a niche area, large networks available, rapid access and
presentation of candidates
• Centralized or Decentralized
o Centralized: Specific recruitment team
▪ Efforts are not duplicated across organization, concentrated expertise,
consistent policy execution.
o Decentralized: individual business units
▪ Ultimately the decision makers, more responsive to specific needs,
often used in universities and research departments
• Timing: Lead time concerns, time sequence concerns
• Process flow & Record Keeping
o You have to keep track of the process + information system to deal with all the
data
• Budget and ROI
o Budget is crucial: Applicant tracking systems + use of metrics
▪ Cost per applicant or cost per hire.
Potential applicant reactions
• Reactions to job or organization
o Job and organization are ‘products’ you are selling.
o Job-level:
▪ Opportunities for growth
▪ Challenge + development
▪ Salary
▪ Interesting work characteristics
3
, o Organization-level
▪ Prestige
▪ Treating employees well (prosocial orientation)
▪ Social environment (good colleagues)
▪ Supportive culture (value fit)
▪ Emphasis on individual work and ambition
• Reactions to recruiters → Theoretical perspectives
o Critical contact theory
▪ Limited information about actual job characteristics → Difficult to make
a meaningful comparison between jobs → recruiter characteristics +
behavior form applicant perceptions
o Signaling theory
▪ Recruiter = signal about the organization
▪ Information about likelihood of job + job / organization characteristics
o Elaboration Likelihood Model (ELM)
▪ Information processing via central vs peripheral processing
▪ Lack of information: peripheral processing via recruiter
• Reactions to recruiters → Empirical evidence
o Job characteristics are more important than recruiter characteristics
Recruiter is less important when there is more information available
o Influence of recruiter on attitudes are stronger than on behavior: Important for
branding and talent pooling.
o Certain recruiter behaviors are key!
▪ Credibility (expertise + trustworthiness)
▪ Warmth (enthusiasm, empathy, personable)
▪ Show an interest in the candidate
o Recruiters should be knowledgeable → process and procedures, jobs, legal
environment etc.
• Reactions to recruitment process
o Fairness is key (organizational justice theory)
▪ Closely related to job requirements
▪ Opportunity to demonstrate ability to do the job
o Delays in the process undermine reactions
▪ Expression of interest: organizational response
▪ On-site visit: job-offer
▪ Leads to unfavorable impression of organization (not themselves)
o Reaction to website is crucial
▪ Now often the first exposure to organization
▪ Speed and follow-up online applications are key
▪ Specific job information needs to be provided
▪ Confidentiality should be protected
• Reactions to diversity issues:
o Attracted to own demographic group → woman and minorities are more
attracted to organizations that explicitly endorse diversity + have role models
for leadership position.
Communication + media
• Employer branding → Marketing process
(branding is key for competitive advantages)
o Employer brand → identifier, het logo
o Brand equity → Added value of product by its brand name (Rolex)
▪ Employer branding is crucial in tighter labor markets and for passive
candidates
4
Staffing → different processes, but interrelated
1. Recruitment → Identifying, attracting
2. Selection → Assessing, evaluating
3. Employment → Decision making, final matching
Challenges Deloitte: Different candidates, optimal candidate journey, fun, non-biased
Validity → Measure what you want to measure
- Low validity: CV screening
- Medium validity: Personality test (Big 5), unstructured interviews
- High validity: IG/EQ, structured interview, holistic interview, case interview, game,
agility test
Tutorial 1
Article: Clarck & Watson – Constructing validity: Basic Issues in scale development
Goal scale development → create a valid measure of an underlying construct
1. Clear conceptualization of your construct
2. Initial item pool should be overinclusive and item wording needs attention
3. Structural validity: item selection and psychometric evaluation
4. The goal is unidimensionality rather than internal consistency in selecting scale items.
Structural validity: The extent to which a scale’s internal structure (the inter item
correlations) parallels the external structure of the target trait (correlations among nontest
manifestations of the trait)
Test construction strategies:
• Criterion-based methods
o A test response should be considered verbal behavior in its own right, with
non-test correlates to be discovered empirically – test developers embraced
criterion keying as a method that permitted a wide range of practical problems
to be addressed in an apparently straightforward manner
• Internal consistency methods
o Most used method for item selection in scale development is some form of
internal consistency analysis. Factor analytic methods were used most
frequently when the target construct was conceptualized as multidimensional
and subscales were desired.
o Issues
▪ Factor analytic results provide information, not answers or solutions (a
tool that can be used wisely or foolish)
▪ There is no substitute for good theory and careful thought when using
these techniques.
1
, • Item Response theory (IRT)
o Based on the assumption that test responses reflect an underlying trait (or set
of traits) and that the relation between response and trait can be described for
each test item by a monotonically increasing function called item characteristic
curve (ICC)
▪ Individuals with higher levels of the trait have higher expected
probabilities for answering an item correctly
The initial data collection → inclusion of comparison (anchor) scales + sample considerations
Psychometric evaluation
• Analysis of item distributions → before the more complex analyses.
o Identify and eliminate items that have highly skewed and unbalanced
distributions. Highly unbalanced items are undesirable
▪ When most respondents answer similarly, items convey little
information
▪ Owing to their limited variability, these items are likely to correlate
weakly with other items in the pool and therefore will fare poorly in
subsequent structural analyses
▪ Items with extremely unbalanced distributions can produce highly
unstable correlational results
o It is desirable to retain items showing a broad range of distributions.
• Unidimensionality, internal consistency and coëfficient alpha
→ shows which items you need to delete from the pool
o Internal consistency: Refers to the overall degree to which the items that
make up a scale are intercorrelated
o Homogenity/unidimensionality: the scale items assess a single underlying
factor or construct
▪ Internal consistency is a necessary but not sufficient condition for
homogeneity or unidimensionality (A scale cannot be homogeneous
unless all of its items are interrelated, but as we illustrate later, a scale
can contain many interrelated items and still not be unidimensional
Average inter item correlation (is a straightforward measure of internal consistency)
The average interitem correlation alone cannot establish the unidimensionality of a scale:
A multidimensional scale actually can have an acceptable level of internal consistency.
It is necessary to examine the range and distribution of these correlations as ell.
Guideline is that all of the interitem correlations should fall in the range of 0.15 to 0.50
The attenuation paradox →Increasing the internal consistency of a test beyond a certain
point will not enhance its construct validity and, in fact, may occur at the expense of
validity
Article: Sackett – Individual Differences and their measurement
Article focuses on 3 individual differences domains:
1. Knowledge, skill and ability
2. Personality
(incl integrity, emotional intelligence, stable motivation attributes, creativity)
3. Vocational interest
Cognitive ability was prominent in early research, but questions about limits began early.
Currently: the field agrees on hierarchical structure with general ability on top and more
specific abilities below.
Knowledge + skill → part of research in many domains, are likely to be the most proximal
determinants of differences in job performance, but only if certain conditions apply
Personality and vocational interest → waves of optimism and skepticism
2
, Lecture 2: What about recruitment?
Types of recruitment:
1. Corporate recruitment: In-house, often part of HRM, hiring manager & candidate
2. Agency recruitment: External, independent from HRM, customer & candidate
Define recruitment goals → Always start with the organization strategic goals
• Goals for attraction: what KSAO needed?
• Goals for speed: How fast do we need to attract and when to start?
• Time frame:
Open/targeted approach
• Targeted recruitment
o Find applicants with specific KSAO’s, prepare specific message + approach,
helps to find and attract passive job seekers, essential for complex and
specific jobs
• Open recruitment
o Cast a wide net with little effort, diverse set of applicants, essential when large
numbers of applicants needed, time consuming
Organization + Administration Decisions
• In-house or agency
o In-house: Large organizations
▪ Know the culture, signaling effect to applicants of people-oriented
culture, cheaper
o Agency: SME’s or complex vacancies
▪ Specialized in a niche area, large networks available, rapid access and
presentation of candidates
• Centralized or Decentralized
o Centralized: Specific recruitment team
▪ Efforts are not duplicated across organization, concentrated expertise,
consistent policy execution.
o Decentralized: individual business units
▪ Ultimately the decision makers, more responsive to specific needs,
often used in universities and research departments
• Timing: Lead time concerns, time sequence concerns
• Process flow & Record Keeping
o You have to keep track of the process + information system to deal with all the
data
• Budget and ROI
o Budget is crucial: Applicant tracking systems + use of metrics
▪ Cost per applicant or cost per hire.
Potential applicant reactions
• Reactions to job or organization
o Job and organization are ‘products’ you are selling.
o Job-level:
▪ Opportunities for growth
▪ Challenge + development
▪ Salary
▪ Interesting work characteristics
3
, o Organization-level
▪ Prestige
▪ Treating employees well (prosocial orientation)
▪ Social environment (good colleagues)
▪ Supportive culture (value fit)
▪ Emphasis on individual work and ambition
• Reactions to recruiters → Theoretical perspectives
o Critical contact theory
▪ Limited information about actual job characteristics → Difficult to make
a meaningful comparison between jobs → recruiter characteristics +
behavior form applicant perceptions
o Signaling theory
▪ Recruiter = signal about the organization
▪ Information about likelihood of job + job / organization characteristics
o Elaboration Likelihood Model (ELM)
▪ Information processing via central vs peripheral processing
▪ Lack of information: peripheral processing via recruiter
• Reactions to recruiters → Empirical evidence
o Job characteristics are more important than recruiter characteristics
Recruiter is less important when there is more information available
o Influence of recruiter on attitudes are stronger than on behavior: Important for
branding and talent pooling.
o Certain recruiter behaviors are key!
▪ Credibility (expertise + trustworthiness)
▪ Warmth (enthusiasm, empathy, personable)
▪ Show an interest in the candidate
o Recruiters should be knowledgeable → process and procedures, jobs, legal
environment etc.
• Reactions to recruitment process
o Fairness is key (organizational justice theory)
▪ Closely related to job requirements
▪ Opportunity to demonstrate ability to do the job
o Delays in the process undermine reactions
▪ Expression of interest: organizational response
▪ On-site visit: job-offer
▪ Leads to unfavorable impression of organization (not themselves)
o Reaction to website is crucial
▪ Now often the first exposure to organization
▪ Speed and follow-up online applications are key
▪ Specific job information needs to be provided
▪ Confidentiality should be protected
• Reactions to diversity issues:
o Attracted to own demographic group → woman and minorities are more
attracted to organizations that explicitly endorse diversity + have role models
for leadership position.
Communication + media
• Employer branding → Marketing process
(branding is key for competitive advantages)
o Employer brand → identifier, het logo
o Brand equity → Added value of product by its brand name (Rolex)
▪ Employer branding is crucial in tighter labor markets and for passive
candidates
4
