Chapter 6: Selection Flashcards
What is Selection?
The process of choosing individuals with relevant qualifications to fill existing or projected job openings. —> To maximize “hits” and avoid “misses.”
Hits
Accurate predictions
Misses
Inaccurate predictions
2 types of misses
- False positives
- False negative
Who typically make the final decisions about hiring people in their unit?
Line managers
Steps in the selection process
- Completion of application
- Initial interview
- Employment testing
- Preliminary selection in HR department
- Supervisor/team interview
- Background investigation
- Hiring decision
Quality of selection tests
- Relaibility
- Validity
- Interrater Reliability
Reliability
The degree to which interviews, tests, and other selection procedures yield comparable data over time
- Scale of 0.0 - 1.0
- CONSISTENCY of measure
- Considered bad when < 0.7
Validity
The degree to which a test or selection procedure measures a person’s attributes and predicts future behavior and how well it measures it. The selection process should be able to predict how well a person performs on the job.
- Accuracy of measure
- Considered bad if < 0.3, good if 0.3 - 0.5, excellent if > 0.5
- Criterion-related, content, construct
Testing RELIABILITY of the test
3 tests of reliability
1. Test-retest
- Stability over time
2. Internal-consistency
- Degree test measures single construct
- Split half
- Cronbach’s alpha
3. Inter-Rater
- Correspondence between multiple raters
Testing VALIDITY of the test
- Rational Approaches
- Content
- Face
- Construct
Convergent
Divergent - Empirical Approach
Criterion-related validity
- Predictive
- Concurrent
Content validity
- Evaluated by SMEs
- Extent predictor assesses job-relevant behavior
Face Validity
- Do applicants see link between test and job requirements?
- Assessed buy job applicants
Construct Validity
- Convergent and divergent validity is collected
- Evaluated by experts
Empirical Approach: Criterion-Related Validity
Determines a relationship between selection test scores and job performance
- Predictive
- Concurrent
Predictive
Predictive
- Test job applicants before they are hired.
- Later on the job, measure the performance of those hired and compare it to their initial test score.
Concurrent
- Test existing employees while simultaneously measuring their job performance
Validity Coefficients
The correlation (r) between a predictor and a criterion.
- r > 0.5, great prediction
- r between 0.3 and 0.5, good prediction
- r < 0.3, not great prediction
- r2 indicates variance accounted for by test.
Incremental Validity
Degree to which validity of total selection method increases with each new test
Ways of using multiple predictors
- Independent predictors of performance (criterion)
- Correlated predictors of criterion (performance)
Value of the test
Utility
- Degree to which information from selection method enchances bottom line effectiveness
- Does the test increase the accuracy of the selection process
Interviews
Structured:
- Standardized questions, probes, conditions, & response codes
Unstructured:
- Unstandardized open-ended questions, no planned response codes
Situational Interviews
- Hypothetical problems posed during the interview
- Key assumption:
Intentions predict actual behavior - Key to successful interview:
questions should be in the form of dilemmas
Behavioural Inteviews
- Question ask about applicant’s experience
- Key assumption:
best predictor of future behaviour is past behaviour - Key to successful interview:
Do not accept general answers that describe typical behaviour
Legality of the selection method
- Hiring based on up-to-date job analysis
- Use valid predictors for performance
- Recruit for qualified and diverse candidates
- Apply human rights equity and legislation
Making hiring decision
- Clinical strategy
- Statistical strategy
- Multiple hurdle approach
- Banding
Clinical strategy
Subjective evaluation of all information and making an overall judgment
Statistical strategy
Use a formula which includes all valid predictors and hiring candidates with the highest score
Multiple hurdle approach
Multiple predictors are used, and candidates only move on to the next phase if they pass the previous one. Hire those candidates with the highest score on the last hurdle.
Banding
Multiple predictors are used, but applicants who fall within a range of scores are considered equal.
Why do interviews stay here to be implemented?
- It is practical
- Serves other purposes
- Greath faith and judgment
Types of structured interviews
- Situational interviews
- Behavioural description interviews
Methods for administering interviews
- Panel and sequential interviews
- Video interviews
- Computer-Administered interviews
Panel interview
An interview in which a board of interviewers questions an observes a single candidate. Interviewers pool their observations and rating scores after the interview → reach a consensus about the suitability of the candidate. The advantages of this method are:
- Reduction in bias (there are multiple inputs)
- Discrimination is minimized
- Applications are more likely to accept decisions made by panels
Sequential Interview
A format in which a candidate is interviewed by multiple people, one right after another. Different interviewers who have a vested interest in the candidate’s success.
Video Interviews
Interviews conducted via videoconferencing or over the web
Computer-Administered (Automated) Interviews
Interviews in which the questions are administered to applicants via computers; the interview can be conducted at a firm’s facilities using kiosks, online, or via phone.
Can’t ask candidates follow-up probing questions
Perceived less favorable by applicants
Used mainly as a complement rather than a substitute for live interviews.
Interviewer best practices
- Understand the job
- Establish an interview plan
- Establish and maintain rapport and listen actively
- Provide information as freely and honestly as possible
- Use questions effectively
- Separate facts from inferences
- Recognize biases and stereotypes
- Avoid the “halo error”
- Control the course of the interview
- Standardize the questions asked
Model’s of statistical approach
- Compensatory
- Multiple cutoff
- Multiple hurdle
- Selection ratio
Compensatory Model
A selection decision model in which a high score in one area can compensate for a low score in another.
Multiple Cutoff Model
A selection decision model that requires an applicant to achieve a minimum level of proficiency on all selection dimensions.
Multiple Hurdle Model
A sequential strategy in which only the applicants passing the cutoff score at an initial evaluation stage go on to the next stage. The process continues through several stages (hurdles). Useful when testing procedures are expensive and when there is more time to make the hiring decision.
Selection Ratio
The number of applicants to be selected compared with the total number. If the selection ratio is low, only the most promising applicants will be hired. When the ratio is high, very little selectivity will be possible because even applicants with mediocre abilities must be hired if the firm’s vacancies must be filled.
Cross-Validation
Verifying the results obtained from a validation study by administering a test or test battery to a specific sample. Administering a test or battery of tests to a specific sample of people to verify the results obtained from a more extensive validation study.
- Organizations can assume that cognitive ability tests will predict performance in their workplace, but they must collect their own data to confirm that this relationship does in fact exist in their specific work environment
Criterion-Related Validity
The extent to which a selection tool predicts, or significantly correlates with, important elements of work behaviour.
Concurrent Validity
The extent to which test scores (or other predictor information) correlate with criterion data obtained at about the same time from current employees.
Types of Tests
- Job knowledge tests
- Work Sample Tests
- Cognitive ability tests
- Polygraph tests
- Personality and interest inventories
- Honesty and Integrity Tests
- Physical ability tests
- Medical Examinations
- Drug testing
- Assessment centres
