How Reliable Are Intelligence Tests
I am going to start with the definition. Reliability as utilized by test makers has very little to do with our understanding of the dictionary definition of reliability. So what DO they mean?
When a test is deemed to be reliable, it simply implies that the test will give the identical outcomes when given to the identical person at completely different times. Testing of reliability is completed in one of 3 ways:
Administer the identical test at two totally different times.
Administer half the test every at two totally different times.
Administer two 'varieties' of the test at two completely different times.
A reliable test must give the identical end result every time. (Test outcomes do, after all, are inclined to 'enhance' with apply, however that can be taken into consideration fairly easily.) A great deal of attention is given to the design of tests that may be shown, by one of the above methods, to be reliable. Why is this necessary? This sort of test reliability bears little resemblance to the type of things one usually associates with being 'relyable'. It says nothing about the test being 'helpful' for any objective at all. So one wonders why it is something that's measured and tested so carefully.
Merely put, test designers will tell us that if a test has any probability in any way of being useful, it should first be reliable. Their logic is inescapable. There's merely no way we will anticipate to depend on test scores to indicate ANYTHING significant if the test offers significantly totally different results for the same individual on different attempts.
But what concerning the query of a test being the 'dependable' type of 'reliable'? This is the point of testing to start with, right? Sadly, dependability is a moving goal that could be very hard to HIT, so test designers are forced to settle for what they call "Validity". There are literally three totally different measures of validity:
Criterion validity - the test is accurate when measured in opposition to some particular scale, corresponding to a list of desired mathematical skills.
Content validity - the test contains the entire materials it's designed to cover.
Predictive validity - the test outcomes can be used to make accurate predictions of future performance.
"Ah!" you say! Now we're getting someplace! Yes, when it comes to testing, validity best approximates what WE think of as reliability. Some tests, such as these taken for admission to secondary and submit-secondary schools, have been shown to be 'legitimate' when the outcomes are used to predict student success, however this nonetheless does not imply that the test is an effective measure of INTELLIGENCE.
It might seem strange, but the validity of an intelligence test is often measured by how intently its outcomes mirror those obtained from ANOTHER intelligence test. What, precisely, either test measures is troublesome to define, and it is an undeniable incontrovertible fact that no practice iq test test but devised can predict future success with any accuracy. It's reasonable to say that a battery of a number of IQ tests may give an image of innate abilities in a person, but no rating, not even a set of scores, can actually quantify that broad, nebulous thing we call intelligence.
In view of all of this, my advice would be to go ahead and take intelligence tests (or have your children take them). There isn't a hurt in this sort of testing as long as you realize that you should not take the scores VERY seriously.