Chapter 6: Staffing decisions Flashcards
Staffing
Activities concerning the inflow, throughput and outflow of personnel.
Goal: The right person at the right time and right place in the organization, in order to execute the planned activities and projects.
Quality of staffing
Influence the organizational performance.
- Formal Job analysis.
- Internal selection for important positions.
- Validated selection instruments.
What are the stakeholders in the staffing process?
- Line management (gather information from applicatns usaually in the form of one or more interviews. The line manager will be expected to supervise the new member).
- Co-workers ( They have to work with the new member).
- Applicants
- Other; Personnel department, recruitment&selection agency, society, applicant’s family.
Selection Ration (SR)
Is a ratio that shwos how many people are hired and how many people are assessed. So a low SR is good because the company then has a lot of choise and can choose the best candidate.
Good fit between person and job
Good fit between abilities/skills and job demands has a positive influence on work performance.
Misfit in staffing outcome
Can lead to negative consequences.
Criterion validity
Evaluation of selection procedures
Especially predictive validity based on test scores, draw accurate conclusion now about performance in the future.
Utility analysis
Is a technique that assesses the economic return on investment of human resources interventions such as staffing and training.
Base rate
The percentage of the current workforce that is performing sucessfully. If the base rate is very high than a new member is not capable of changing the rate a lot.
IF you combine different test methods you can get the highest validity.
What are two types of forecasting errors?
- False positives: Accepted, but performed poorly.
- False negative: Rejected, but would have performed successfully.
What are two types of correct forecasting?
- True positives: Accepted, and performed successfully.
- True negatives: Rejected, and would have performed poorly.
Shift of cut-off score from X to W leads to: Decrease of false positive (A) and Increase of false negative (B).
Shift of cut-off score from X to Z leads to: Decrease of false negatives (C) and increase of false positives (D).
What are the three factors where division of prediction errors is dependent on?
- Validity of selection instrument.
- Base rate; number of suitable candidates for the job.
- Selection ratio: number of vacancies dividied by the number of applicants (the lower the score, the better).
What is the criterion-referenced?
Determining cut-off score
The absolute desired level of performance.
- Performance of current employees.
- Supervisor and/or expert judgements of desired performance.
What is the norm-referenced?
Determing cut-off score
Relative (e.g. based on average).
- Order
- Percentiles
- Standardized scores
How to combine information from different sources?
- Clinical method: Subjective, intuitive.
- Statistical method: Based on quantitative data.
Clinical vs Statistical method?
- Availability of empirical data.
- Statistical method is better in general; the same predictors are used for all candidates, explicit formula, predictors are weighted.
Compensatory systems
Selection systems
Allow low scores in some tests of an applicant to be compensated by his or her high scores in other tests.
Non-compensatory systems
In which an individual has no opportunity to compensate at a later assessment stage for a low score in an earlier stage of the assessment process.
- Hurdle system
- Multiple hurdle system
Hurdle system
Only when an applicant achieves the lowest required score of a test, his or her other scores can be considered in a follwoing compensatory manner.
Multiple-hurdle system
With every test a minimum should be achieved for further consideration so no compensation is possible.
Multiple regression analysis
Is a method of analysis that results in an equation for combining test scores into a composite based on the correlations among the test scores and the correlations of each test score with the performance score.
Cross-validation
Is a process used with multiple regression techniques in which a regression equatin developed on a first sample is tested on a second sample to determine if it still fits well.