I’ll begin with the definition. Reliability as used by test makers has little or no to do with our know-how of the dictionary definition of reliability. So what DO they suggest?
When a check is deemed to be dependable, it genuinely way that the test will deliver the equal consequences when given to the equal character at extraordinary times. Testing of reliability is executed in considered one of three approaches:
Administer the identical take a look at at one of a kind times.
Administer half the test each at two exceptional times.
Administer two ‘paperwork’ of the test at exceptional instances.
A dependable check have to give the equal final results each time. (Test consequences do, of route, tend to ‘improve’ with practice, however that can be considered pretty without problems.) A first-rate deal of attention is given to the layout of checks that can be proven, via one of the above techniques, to be reliable. Why is that this vital? This type of take a look at reliability bears little resemblance to the kind of factors one generally buddies with being ‘dependable’. It says nothing approximately the take a look at being ‘useful’ for any motive at all. So one wonders why it is something this is measured and examined so carefully.
Simply positioned, test designers will tell us that if a take a look at has any chance by any means of being beneficial, it should first be dependable. Their good judgment is inescapable. There is virtually no way we will anticipate to rely upon take a look at scores to suggest ANYTHING meaningful if the take a look at gives appreciably special outcomes for the same person on special attempts.
But what about the query of a test being the ‘reliable’ form of ‘dependable’? This is the point of testing to start with, right? Unfortunately, dependability is a moving goal this is very hard to HIT, so test designers are compelled to accept what they name “Validity”. There are without a doubt 3 extraordinary measures of validity:
Criterion validity – the take a look at is correct while measured towards some unique scale, along with a list of preferred mathematical capabilities.
Content validity – the test consists of all the fabric it is designed to cover.
Predictive validity – the check results can be test iq used to make accurate predictions of destiny overall performance.
“Ah!” you say! Now we are getting someplace! Yes, in terms of testing, validity great approximates what WE consider as reliability. Some assessments, consisting of those taken for admission to secondary and put up-secondary schools, were shown to be ‘legitimate’ whilst the consequences are used to expect pupil fulfillment, but this nonetheless does now not mean that the take a look at is a good degree of INTELLIGENCE.
It may also appear bizarre, but the validity of an intelligence take a look at is often measured through how carefully its effects replicate those acquired from ANOTHER intelligence take a look at. What, exactly, either test measures is difficult to outline, and it’s miles an plain reality that no IQ take a look at yet devised can expect destiny fulfillment with any accuracy. It is cheap to say that a battery of numerous IQ checks can deliver a picture of innate talents in an individual, however no score, not even a hard and fast of rankings, can virtually quantify that large, nebulous aspect we call intelligence.
In view of all of this, my recommendation might be to move beforehand and take intelligence checks (or have your children take them). There isn’t any damage in this sort of checking out so long as you recognise that you have to now not take the rankings VERY critically.