One strong criticism of IQ tests that have been brought up time and time again is their over-reliance on language. Test takers who are not fluent in the language in which the test is administered are at an immediate and an extreme disadvantage. In addition, subjects who are familiar with the language in which the test is being administered are more than likely familiar with the types of questions that are on the exam. This familiarity may be an explanation as to why people who are unfamiliar with the language do poorly when compared to those who are familiar with the language.
Once researchers, test administrators and psychologists realized that these problems were prevalent, the "culture fair IQ test" was developed. According the Lillienfield text, this test focuses more on abstract reasoning items that don't depend on language. The goal of these tests is to erase the cultural influences within the system.
I agree that thee culture fair test example given in the book is very fair, using explicitly symbols that have no meaning to any specific culture. However, many culture fair IQ tests are not like that. One test that I found online (the link to it is at the bottom of my post) features many items that are questions based on numbers, knowing number order and number value. Also, it includes questions featuring the English alphabet and requires knowledge of their order and possibly knowledge of simple English worlds. This kind of test is in no way culturally fair because it requires a tremendous amount of knowledge of the English alphabet and number system.
I think that Culture-Fair IQ tests are good in theory; but are unfortunately quite difficult to achieve in reality. The only way most people know how to communicate is through a language of some sort - whether it be English, Spanish or Sign Language. In my opinion, trying to create a test that features no language is a feat that is very difficult to master.
Here's the link to the Culture-Fair test I found online! Try it for yourself!