- For a test to be fair, it must be valid and reliable.
- Validity is the characteristic of a test that measures what it is supposed to measure.
- Reliability is an indication of how precise a student´s score is.
- Sources of Variance include true score variance and Error variance.
- True variance is the actual difference across examinees in the trait or ability being measured.
- Error variance is variability due to any score differences not attributable to the Construct of interest.
- There are two types of error: systematic error and Random error.
- Systematic error is error that occurs in a predictable fashion for all, or a group, of test takers in the trait or ability being measured.
- Random error occurs by chance and is not consistent or predictable.
- Each set of test scores has its own Standard error of measurement, which tells us the standard amount of error that can be expected from a test and is derived from a formula that uses a reliability statistic.
- The issue of test fairness and test bias has become one of growing public concern since the Civil Rights Act of 1964. Fairness is not a technical concept but a broad concept that is based on philosophies of test use, social, and personal values.
- Judgment and empirical methods may be used to identify bias items. There are several statistical methods that can be used to help determine whether a test is fair or not.
- Differential Item Functioning (DIF) analysis is one such method that allows test developers to ascertain statistically whether or not individual items perform differently in important ways for relevant subgroups.
- A bias committee is an example of a judgment method for detecting bias. In this method, a group of reviewers representative of the population of examinees analyze test items and, based on their own judgment, assess the fairness them.
- Universal design, in terms of testing, refers to a test design that can serve a diverse population without having several different accommodating forms.
- Accommodations in testing are adaptations made on a test to accommodate special needs, such as providing the test in Braille or on audio. A test with accommodations is not modified in any other way from the original test.
- A test with Modifications is one in which the original form of the test has been modified to such an extent that it cannot be considered an accommodated version of the original test.
- Alternate Assessments are assessments with modifications. These tests are often designed for students who have an Individual Education Program (IEP) and who meet specific criteria qualifying them for the alternate version compared to the original version of a test.
|
