1. Criterion Related Validity Two Types
1.1. Concurrent Validity (measures that can be given at the same type with another correlated test.
1.2. Predictive Validity relates to foreshadowing of how students may encounter the examine. Predicts who may do well on test.
2. Types of Validity
2.1. Content Validitiy
2.2. Examine test questions
2.3. Do the questions connect with the content studied?
2.4. Content, is the content at the correct level for the students?
2.5. " A test with good content validity evidence matches or fits the instructional objectives" Kubiszyn, T. & Borich, G. (2013
3. Construct Validity Evidence
3.1. If a test has been developed to measure a new subject and does not have any support of being valid. A test must must be constructed.
3.2. The subject material has to relate to the theory or the purpose of the desired end result
4. Interpreting Validity
4.1. Recognize the extent to measure assessments
4.2. Content Validity should be based upon subject matter taught should also parallel assessment
4.3. Concurrent and Predictive Validity most likely will many different variables that may affect the outcome "Adequacy of validity evidence depends on both strength of the validity and purpose of the test" Kubiszyn, T. & Borich, G. (2013
5. Type of Reliability
5.1. Reliability is consistency
5.2. Three basic methods to ensure reliable scores
5.2.1. Test-Retest
5.2.1.1. The test is given twice. Educators can correlate the scores
5.2.2. Alternative Form
5.2.2.1. Two forms of test can be used and if both forms of testing are given, the correlation between the the two sets of assessment scores is determined. If the scores have a large gap, the scores are unreliable.
5.2.3. Internal Consistency
5.2.3.1. In general. the assessment is designed to measure a sole concept. If it is one concept then the whole test should be internally consistent.
5.2.3.1.1. Split Halves involves splitting the test into two equal halves and determines the correlation between them.(Kubiszyn, T. & Borich, G. (2013)
5.2.3.1.2. Odd-Even Reliability is necessary to balance the difficulty of the test questions. Odd number questions will have difficult questions as well as Even numbered test questions. Historically, easy questions were the first have of the test and more difficult questions were at the end of the test.
5.2.3.1.3. Kuder-Richardson Method
5.3. Interpreting Reliability
5.3.1. Consider how long the test is.
5.3.2. Test reliability is the aspect of test quality concerned with whether or not a test produces consistent results
5.3.3. Group variability affects test score reliability
5.3.4. How difficult the test items are also affects score reliabiity