For tonight's class you read a lot about statistics and how to determine a student's true score. However, what does that mean? How do we know if a student's performance on a test is accurate?
To help shed light on this, there is an old saying, 'A test can be reliable and not valid; but a test cannot be valid and not reliable'. What does this mean?
To help shed light on this, there is an old saying, 'A test can be reliable and not valid; but a test cannot be valid and not reliable'. What does this mean?
A test can be reliable by producing the same results over a given period of time. However, the information in the test cannot be valid which means that it does to vailidify the information that it was meant to test. A test that is not valid cannot be reliable due to the fact that the invalid information will produce incorrect, unreliable scores.
ReplyDeleteA test can yield consistent results but not measure what it is intended to measure. If a test does a good job of measuring what is intended, it will be repeatable.
ReplyDeleteReliabilty means that it consistantly produces the same results meaning it could produce invalid results every time. However validity means that the test measure what it says it is supposed to measure each time.
ReplyDeleteA true score is never really known. We have to figure in a students obtained score and calculate an error of measurement.
ReplyDeleteTests can be reliable and not valid, because students have bad days. Perhaps the student took the test on a bad day and scored 15-20 points lower than he/she would normally. In this instance the test would be reliable, because it is proven to accurately test other students. The test would not be valid, because the student had a bad day and didn't truly show his/her full potential.
Tests cannot be valid and not reliable, because every test needs to be reliable. Reliability is the foundation of testing. If the test is not reliable it cannot a valid way of testing.
When looking at a test, it can be reliable because the test might have the student giving the same score over a certain period of time, but when thinking about validity does the test really score what it is meant to score? Then when the test tests what it is meant for, but the scores of the students change over time, then you have an issue with the test. So we should always have a test that will give the same scores and it should test what we want it to test.
ReplyDeleteAnd when we think on a student's performance on a test is accurate, we should look at the students past scores. For example, if we have a test on multiplication facts every week and the student does well and keeps consistent and then it starts to go down, maybe we had an error in the grading or we should look at what the student has going on.
The true score is taking a students base score and then taking into account a standard error of measurement. A test always has room for error and this needs to be taken into account when calculating a students grade. This means that a students true score will be within a range of scores.
ReplyDelete'A test can be reliable and not valid; but a test cannot be valid and not reliable.' This means that the reliability, which is the dependability of a test across a time or different items. However, valididity is the quality of a test, the test measures what is says its going to measure. So if a test is reliable it doesn't matter if it is valid because it is consistant but if a test has validity and does what it says its going to do it must be reliable to give validity to the test at all.
A student's true score can not be fully determined. An obtained score is the score a student earns on a specific test on a certian day, but because all tests have some standard error or measure,the obtained score may not reflect the knowledge of the student about a certain topic or are of study. Tests can be reliable, in that, the outcome will be the same or close to the same, for certain material. However, a vaild test will test the material intended.
ReplyDelete