![Picture](/uploads/4/8/0/3/48037207/5343124.jpg?414)
To create a test, developers must meet these 3 main criteria for making a standardized test: Standardization, Reliability, and Validity.
- Standardization: This defines meaningful scores by comparison with the performance of a pretested group. Meaning, that this is used to accurately evaluate the test taker's performance accurately. This also works in correlation with the normal curve which utilizes the bell shape to describe the distribution.
- Reliability: This is the extent to which a test yields consistent results. To check a test's reliability researchers retest or split the test in two halves and distribute them. When they split them the test results need to have a strong correlation and be similar to each other. If the results aren't similar then the test is not reliable.
- Validity: This is the extent to which a test measures or predicts what it is supposed to. There are 2 aspects of validity: predictive and content validity.
- Predictive: The success with which a test predicts the behavior is it designed to predict.
- Content: The extent to which a test samples the behavior that is or interest.