Chapter 2

Validity and reliability 

can be conflicting needs for assessment techniques and procedures. The most reliable assessments will be pencil and paper tests in which each item measures only a single aspect of a skill and which give each tester a numerical mark. But the most valid assessments will be on those that collect a lot of information about performance on several aspects of a skill. When validity increased, reliability decreased.

validity: the highest content of the validity of a test, it's likely to be an accurate measure of what is supposed to measure, a test in which the main areas identified in the specification are presented or not represented at all. such a test is likely to have a detrimental effect. areas that have not been tested are likely to become areas ignored in teaching and learning. Too often the content of the test is determined by the ease of testing rather than what is important to test. the best protection against this is to write the test specifications and to ensure that the content of the test is a true reflection of this, content validation should be performed while developing a test, you should not expect to complete the test used.

Reliability depends on the type of test and the time it could take to take, ie, the most important desicion for reliability is: if we are to reject someone on your opportunity to study abroad because of their score on a language test, then we must be very sure that your score would not have been very different if it was been tested for a day or two before or after. to explain the coefficient is contuniacion reliability can get to use and estimate possible differences of this kind.

The first requirement is to have two sets of scores to compare. The most obvious way to get these is to get a group of subjects and take the same test twice. This is known as the test-retest method. taking the test is too soon after the first, then subjects tend to remember the elements and responses, so that the probability of responses and ran them falsely elevated reliability. If too much time between tests and the other, then learning has taken place, and the ratio will be lower than it should be. When the time is very long it is unlikely that subjects take the same test twice, and this is also likely to have a negative effect on the coefficient. These effects are reduced somewhat by the use of different forms of the same test, however, the shape of the tests is not the same in many occasions.

Reliability is important because it verifies the form of how it should be built the test scores compared to administration in the form in which it can be administered on two separate occasions for the same group of students and the results will be similar.