3.2 Measurement Flashcards
What is a predictor
Definition: Variables used to forecast employee performance.
Example: Previous work experience in years, scores on a job-specific skills assessment, and results from a personality test. These predictors are selected because they are believed to have a relationship with how well an employee will perform in their job.
What is a criterion?
Definition: The outcome variable that the model aims to predict.
Example: Employee performance rating, which could be a composite score based on factors like productivity, quality of work, teamwork, and adherence to deadlines. This is the variable the employer is interested in predicting to make informed hiring decisions.
What is a construct?
Definition: An abstract concept or theoretical idea that is relevant in the employment context.
Example: Intelligence/Job competency, which is a construct that encompasses various skills and attributes necessary to perform a job effectively. This construct is not directly observable but is inferred from various indicators like performance in tasks, communication skills, and problem-solving abilities.
What is a measure?
Definition: Tools or methods used to assess a construct or predictor.
Example: For measuring the construct of job competency, an employer might use a combination of measures like structured interviews, job simulations, and reference checks. To assess the predictor of job-specific skills, a standardized skills assessment test might be used
What are the 5 criteria for evaluating effective staffing tools?
VULGR
-Validity
-Utility
-Legality
Generalizability
Reliability
(T/F) Reliability is necessary for validity but not sufficient
T
What is reliability?
The extent to which a measure is free from random error.
Consistent response
Upper-limit of correlation coefficient (standardized measure of association; r) is the product of each measures reliability.
What is validity?
The extent to which performance on the selection device/test is associated with performance on the job. (tests what we want)
What are two ways validity is reduced?
- Deficient – selection device does not measure all the important aspects.
- Contaminated – selection device measures irrelevant aspects.
What are the 3 ways of measuring validity according to Uniform Guidelines on Employee Selection Procedures?
Criterion-related, Content-related, Construct-related
What is criterion-related validity?
Most important. reflects relationship b/t score on selection device and score on criterion.
o Correlation between two sets of scores assessed (correlation/validity coefficient)
o Range = -1.00 to +1.00
o Effect size = strength of relationship:
r = .5 is strong, .3 moderate, .1 is weak.
d = .8 is strong, .5 is moderate, .2 is weak.
o Direction = whether positive or negative
What is content-related validity?
involves using expert opinions that the items, questions, or tasks used in a selection tool are representative of the kinds of situations, problems, tasks that occur on job.
o When developing measures: Look at job analysis, compare current/proposed measures to KSAOs, develop measures relevant to job components.
o Ex. Covering all aspects of the final exam equally.
What is predictive validation?
uses the test scores of all applicants and looks for a relationship between the scores and future performance of applicants hired
What is concurrent validation?
consists of administering a test to people who currently hold a job, and then comparing their scores to existing measures of job performance
What are examples for why correlation does not equal causation?
Reverse causation (Violent kids play violent video games)
3rd Variable (Poor parenting)
Reciprocal (A and B cause each other)
Coincidence (Drownings/Nick Cage films)
