Here's a breakdown:
* Predictor Measure: This is the measure used to predict future performance. It could be a test score, a job interview evaluation, or any other factor.
* Criterion Measure: This is the measure used to evaluate actual performance. It could be job performance ratings, sales figures, or any other relevant outcome.
* Contamination: This occurs when the criterion measure is influenced by the predictor measure. This means that the criterion is not a true reflection of performance but is artificially boosted by the predictor.
Example:
Imagine a company uses a personality test as a predictor for job success. The test asks questions related to conscientiousness, which is often associated with good performance. However, the company's performance evaluation system also heavily emphasizes punctuality and task completion, which are directly influenced by conscientiousness.
In this scenario, the performance evaluations are contaminated by the personality test. Employees who score highly on conscientiousness in the test are likely to receive higher performance ratings, even if their actual job performance is average. This creates a false positive correlation between the test score and job performance.
Consequences of Criterion Contamination:
* Overestimation of predictor's validity: The predictor appears more effective than it actually is.
* Poor selection decisions: The company might select candidates based on the contaminated criterion, leading to hiring individuals who may not be truly suitable for the job.
* Misleading insights: The contamination can obscure the true relationship between the predictor and the actual performance.
How to Avoid Criterion Contamination:
* Develop independent criterion measures: Use different measures to evaluate performance that are not influenced by the predictor.
* Control for potential biases: Be aware of any factors that might influence the criterion measure and take steps to control for them.
* Use multiple predictors: Rely on multiple predictors to assess performance, rather than relying solely on a single predictor.
* Conduct validation studies: Regularly validate the predictors and criterion measures to ensure they are not contaminated.
By understanding and avoiding criterion contamination, you can ensure that your performance evaluations are accurate and that your selection decisions are based on valid and reliable information.