how good is a test?
What do you mean by "good?"
What kind of test?
Basically a test measures what the student knows about the questions on that particular test.
For instance, my seventh-grade students always did very well on their final exam -- almost everyone got at least an 80%. My goal was for them to be able to identify about 150 countries, mountain chains, and bodies of water on a world map.
However -- they knew well in advance which geographical features would be on the test. The test was also in a matching format. Was it a good test? In my eyes, yes. But if they hadn't known what was going to be on the test -- and if they had to recall the names of all of these places -- I know they wouldn't have done very well at all.
Many thing go into a "good" test.
Criteria for test "goodness" involve test objectivity, validity, reliability and standardization.
Here are some of my class notes:
Good tests require certain qualities.
I. Objectivity indicates consistency among scores, minimal scorer bias (r approaches +1 and is >. 90). Contrast with subjective measures.
II. Reliability indicates the consistency of scores when give to same individuals (r approaches +1 and is >.90). It requires good level of objectivity.
III. Validity means that a test "measures what it claims to measure." Although valid test needs to be reliable, reliability does not ensure validity. Valid if it correlates with criterion measure (ACT scores vs. grades example).
IV. Standardization (Q student, p. 383) involves two aspects.
A. Tests run with same procedures and conditions each time given.
B. The above allows the use of norms, comparison standards used to judge a specific score, where one person scores in relation to the standardization group. Most standardized tests use middle class, WASP norms.
I hope this helps a little more.
To determine how good a test is, we often look at its reliability and validity. Reliability refers to the consistency and stability of the test, meaning that it consistently measures what it is intended to measure. Validity, on the other hand, refers to how accurately the test measures what it is supposed to measure.
To evaluate the reliability of a test, one common method is test-retest reliability. This involves administering the same test to a group of individuals at two different time points and comparing the results. If the scores are similar across the two administrations, the test is considered reliable.
To assess the validity of a test, there are several types of validity that can be examined. Content validity involves determining if the test assesses the specific content or skills it is intended to measure. Criterion-related validity involves comparing the test scores to an established criterion or outcome to determine if the test is accurately predicting or correlating with that criterion. Construct validity is concerned with how well the test measures the underlying theoretical construct it is designed to measure.
Additionally, it's important to consider the norms or standards against which the test results are compared. Norms provide a reference point for interpreting individual scores by comparing them to a representative sample of individuals who have taken the test before.
In summary, the goodness of a test can be evaluated by analyzing its reliability, validity, and the appropriateness of the norms used for interpretation.