In principle, the answer here is as precise as it justifies the effort involved in the design and application of the process.
In fact, it is the same as measuring other “measurement objects”, e.g. the length.
For example, if I have defined a piece of length as the shortest abstant between the two points A and B, I can measure how large the distance between the points is.As a “measuring instrument” I may have nothing but a matchbox available in the first attempt (I had in my pocket, cost 0.20€ and I had no ways). This fits in length (by repeated stringing together) perhaps almost 10 times in the distance between the two points. If I estimate the length of the matchbox I know to be about 5 cm, the distance between A and B would now be “measured” about 50 cm (5 cm • 10 = 50 cm). Certainly relatively inaccurate, but also very cost-effective. In the second attempt, I use money to buy a “measuring instrument” (e.g. a “Seca W60092 mechanical rolling measuring belt 206” for approx. 42.18 €, for which I had ways and had to go to a shop). Now I measure with roll measuring tape the route of A and B. Best fair I measure three times myself and ask a friend of mine also to measure twice. So we have made a total of five measurements and can with relative accuracy consider the mean as a “measurement result” (49.8 cm + 49.7 cm + 49.6 cm + 49.8 cm + 49.6 cm) / 5 =): 49.7 cm. Apparently more accurate, but more expensive and expensive. In the third attempt, we spare no expense and effort: I call a friend from Fraunhofer ILT and ask him to provide me with a UKT laser as a “measuring instrument”. With this brand new million expensive device, you can now take really very accurate measurements (which is good, e.g. in ophthalmology in the Femto-LASIK process). By means of the laser I measure the distance of A and B very precisely and get the following value: 49,68737482672 cm.
With three different, different elaborate “measuring instruments” the straight distance between points A and B was determined:
Measurement: approx. 50 cm (little effort)
Measurement: 49.7 cm (average effort)
Measurement: 49,68737482672 cm (huge effort)
This can be described and compared in the same way with the “measurement instrument” IQ test.There are few elaborate, medium and extremely elaborately constructed processes. What of it is as good (in the best case) ALWAYS determines the question (and not the best marketing or availability or …).
At best, I advise to increase quality by repeating the measurement (here errors in the arithmetic mean are removed) and variety of procedures (here “means” the possible errors of the procedures among themselves are exhausted).So the IQ value will be relatively accurate whenever several rather elaborately constructed IQ tests are used and are performed several times by different test conductors. The evaluation should be carried out by independent persons and the data should be calculated by averaging. Thus, the actual reading will come very close to the actual value…
However, quite usually a relative size, since the “good” measured value is always given with a confidence interval (confidence interval) – this takes into account error probability.Thus, 100% accuracy is neither possible nor necessary.