- In testing we face a conundrum. Standardizing all that we can enhances the consistency of scores (reliability) but may make the test more difficult for some people in ways not directly relevant to the knowledge and skill being measured (thus threatening test Validity). An obvious example would be requiring a blind person to take a printed test, rather than a Braille or oral version. Where does the balance lie between these two considerations?
- Some people believe multiple-choice tests are biased. Others believe any tests given as a separate and time-limited activity are biased. Is there a way to perform large-scale assessment of students that you think would not be open to charges of bias? How would you design such a program? Are there any particular challenges with implementing such a program?
- As you read this lesson, think about the next major test you are going to develop for one of your classes. Are any of the techniques described hear ones you might want to use for a classroom test?
|
