Job Analysis - Chapter 9

1. Job description

2. Job classification

3. Job Evaluation

4. Job design/redesign

5. Personnel requirements/specifications

6. Performance appraisal

7. Worker training

8. Worker mobility

9. Efficiency/safety

10. Workforce planning

11. Legal/quasi-legal requirements

The placement of each job into a category on the basis of similarity so that each job goes into the category that it most resembles.

1. Job-oriented system - focus on outcome of a job (e.g., a product or a service)

2. Worker-oriented systems - focus on the observable human requirements or demands made by the job

3. Attribute requirements - focus on the KSAOs required by the job

4. Overall job systems - focus on job title and type of work performed

5. Compensation - clustered into bands or levels based on salary

The movement of people between jobs within an organization. typically a large number of jobs are involved in this type of study with incumbents and supervisors as sources of data.

Descriptors depend on the job in question (e.g., analysis of production and assembly job in a time-and-motion study). Methods can focus on machines, tools, equipment and work aids required by the job.

1. Engineering Approach - design of equipment and processes of task accomplishment

2. Personnel Psychology Approach - selection of people with traits suited for the work

3. Industrial/social Approach - development of programs to motivate employees to behave safely

Making sure an organization will have people with the needed skills both now and in the future. This requires taking stock of where an organization is and where it expects to be in terms of jobs/requirements and people/skills.

Can vary by the purpose or human resource application to plan for legal challenges. Tasks and work activities are useful descriptors for legal requirements because people usually think of jobs in terms of tasks.

Operational Status

Off-the-shelf availability

Occupational versatility/suitability

Standardization

Respondent/user acceptability

Amount/availability of job analyst training

Sample size

Reliability

Cost

Quality of outcome

Time to completion

1. Time and budget (deadline constraints)

2. Project staffing (someone to lead project and someone to execute; consultant or internal agent)

3. Acceptability (respondent/user acceptability, quality of outcome, time to completion)

The amount of error or disagreement contained in the responses. Most commonly estimated using alpha (Chronbach's alpha). For example, alpha is used to test when all items or judges are expected to give similar results or ratings.

The quality of the responses in relation to what we want to know or infer

The extent to which the job analysis adds incrementally to the effectiveness or efficiency of individual or system-level interventions derived from it

The average distance of the ratings from the mean.

How much study results are likely to vary given variability and sample size. Calculated by the standard deviation of the variable divided by the square root of the sample size.

Refers to a function of those judgments that are identical and those that are different.

1. Percentage agreement - the number of ratings for which the judges agree divided by the total number of judgments

2. r(wg) - a comparison of the observed standard deviation with the standard deviation that would be observed if incumbents responded randomly to scale (i.e., a within-group correlation)

Refers to two sources of variance

1. Variance due to random errors

2. Variance due to systematic differences among the items of interest

1. Simplest true index of reliability is the correlation coefficient, r

2. Intraclass correlations (generalizability theory) - looks at the correlation between measures taken under different conditions to see how one set generalizes to the other

1. Human Judgement

2. Specific item content

3. Changes over time

Correlation is used when summarizing relationships between two variables with a single number.

Regression is used when making numerical predictions of the value of one variable when given values of another variable

1. Factor Analysis - pulls variable correlations from matrix and show similarity of items across jobs

2. Multidimensionality Scaling - Similar to factor analysis

3. Cluster Analysis - Groups grabs rather than items

1. Conformity pressures

2. Extremity Shifts

3. Motivation loss

4. Impression management

5. Social desireability

6. Demand effects

1. Information overload

2. Heuristics

3. Categorization

4. Carelessness

5. Extraneous information

6. Inadequate information

7. Order and contrast effects

8. Halo

9. Leniency and severity

10. Method effects