Creating a test with reliability and validity is a difficult, yet important, task. When a test de...

Get Started. It's Free
or sign up with your email address
Rocket clouds
Creating a test with reliability and validity is a difficult, yet important, task. When a test determines many directions one's life could go, one would hope time and effort is put into creating a valid and reliable assessment. A test tells a student how well he or she mastered the content. This method of measuring mastery is only useful when reliability and validity measures are taken. Tests compare students to other students (for teacher purposes). Giving a test that is an accurate determination of such comparisons is important. We rely on tests for so many things in education. Taking time to ensure reliability and validity is time well spent. by Mind Map: Creating a test with reliability and validity is a difficult, yet important, task. When a test determines many directions one's life could go, one would hope time and effort is put into creating a valid and reliable assessment. A test tells a student how well he or she mastered the content. This method of measuring mastery is only useful when reliability and validity measures are taken. Tests compare students to other students (for teacher purposes). Giving a test that is an accurate determination of such comparisons is important. We rely on tests for so many things in education. Taking time to ensure reliability and validity is time well spent.

1. Validity

1.1. Content Validity

1.1.1. Questions correspond to content

1.1.2. Appropriate reading level

1.2. Criterion Related Validity

1.2.1. Concurrent

1.2.1.1. Gives a numeric value. Compares a new test to a trusted test to ensure validity of scores.

1.2.2. Predictive

1.2.2.1. Give test, then measure after period of time if it was accurate.

1.2.2.1.1. Ex. SAT says if student will do well in college. Then, gather GPA's from those students to fin correlation, if any.

1.3. Construct Validity

1.3.1. Based on what is being measured, do the test results correspond to the expectation?

2. Reliability

2.1. Test-Retest

2.1.1. Give same test twice to the same group of students. Find correlation between two tests.

2.1.1.1. Time could be a factor-take into account how long or short between test.

2.2. Alternative Form

2.2.1. Create two versions of the test that measure the same thing.

2.3. Internal Consistency

2.3.1. A single test is split into two parts. Scores from each part are correlated.

2.3.1.1. The test would need to focus on a small number of concepts, preferable one.

2.3.2. Kruder-Richardson Method

2.3.2.1. New node