High-range I.Q. tests are often taken in a non-native language. This may result in a disadvantage, and I have tried to estimate the size thereof for several situations. These estimates assume that one knows the non-native language well, but has learnt it after the "critical period" for learning languages, so after the onset of puberty (the "critical period" lies before the onset of puberty).
The test is vocabulary-based and may or may not be knowledge-based, and prohibits the use of reference aids. Typical examples are the Concept Mastery Tests (by Lewis Terman), the W-87 (used by the ISPE), and the Word Classification Test (Schmies).
In this situation, one will score 30 to 35 I.Q. points below one's true level in verbal ability, assuming of course one does not commit fraud by using reference aids (in which case one will score far above one's true level in verbal ability even in a non-native language). This disadvantage exists even if one is highly intelligent and apparently fluent in the non-native language in question, so that others say, "Oh, but your mastery of this (to you) non-native language is so good, you must surely do well on this test!", those others therewith demonstrating how counter-intuitive and deceptive the nature of a non-native vocabulary disadvantage is.
The disadvantage lies for a major part in common everyday household words that are well known to natives of even average intelligence. Contrary to what some might expect, the disadvantage does not lie mainly in esoteric, high-brow or scientific words, as those often originate from Latin or Greek and transcend language barriers, becoming more or less "culture free". The points one loses on such (esoteric, high-brow, scientific) words one would mostly have lost in a test in one's native language as well. This explains why, by cheating with dictionaries, one does not merely offset one's disadvantage, but scores far above one's true level even in a foreign language on this type of test.
The test is vocabulary-based and may or may not be knowledge-based, and allows the use of reference aids. This is typical for verbal analogies and other verbal item types in high-range tests.
In this situation, one will score 0 to 5 I.Q. points below one's true level in verbal ability. In other words, this type of test is virtually "culture free", and if one scores low on it one will score low on a similar test in one's native language as well.
The relatively small disadvantage one may still have lies partly in idiomatic and pronunciation matters, and partly in information that is specific for one country, not well known in the rest of the world and not important enough to show up in high-quality reference works. For instance, trivial facts about the country's national sports, or about movies, books or famous persons specific for that country.
The test is anything else than vocabulary-based and knowledge-based and allows the use of reference aids.
In this situation, one will score at one's true level in the relevant ability (on the understanding that a test without vocabulary and knowledge items will give a poorer indication of general intelligence than a test that does contain such items, as the verbal aspect is near indispensable in I.Q. testing). This - that one will score at one's true level on the given understanding - is so because the language aspect is now only in the test's instructions, and understanding those is a matter of reading comprehension. Contrary to vocabulary, reading comprehension transcends language barriers; it is not impaired in a non-native language learnt after the "critical period". One has the same reading comprehension in the foreign language as in one's native language, provided reference aids are allowed.