There is no way to create an intelligence test that is not biased. Intelligence is at least partly a cultural phenomenon. A language-based test will be biased toward the culture that created it. There are also different types of intelligence. A person with good spatial sense may do well on the graphic tests, but an otherwise very intelligent person without spatial sense would not.
Absolutely. These tests box people in to "one" correct answer. They do not provide an opportunity to qualify one's answers. They are also biased in that SES factors into them. One might show a picture of a thermostat which is associated with central heating. What if the child has always lived in a home that uses wood heat? Vocabulary is a big part of these tests, and if the child lives in a home where reading and literature is not available, he or she won't have the vocabulary. This is not relative to intelligence; it is relative to knowledge.
I think that Standardized IQ testing is biased and Pohnpei makes a very important point about Howard Gardner and multiple intelligences. This is a much more modern way to look at knowledge and how the brain works than traditional IQ tests. If we add the work of Susan Kovalik and brain-based learning to this idea, we also accept that a brain can only process information if it can connect it to something already in the brain. Therefore you take a student who has great musical intelligence and teach them math through music and suddenly their mathematical IQ score goes up.
Also, again to echo the above post. All standardized tests are socioeconomically, racially and ethnically biased. They assume a certain set of knowledge. You need this basic cultural knowledge to understand what the question is asking. If you are missing this knowledge than you can't answer the question, even if you have the intelligence and skill to, because you can't understand the question. Some of the simplest forms it takes are in the names, places or products mentioned in questions in everything from English to Math. If these are unfamiliar to you, the question becomes daunting. IQ tests assume a mid-class white childhood experience. Whether this is intentional or not, I don't know, but it does exist.
My answer to this depends on what sort of bias you are asking about.
I am convinced that IQ tests are biased against people who think creatively. An IQ test is meant only to measure a person's ability to come up with a single, preconceived answer. A person who can think of multiple ways to answer a question may do worse on an IQ test even if, one can argue, that person is more intelligent.
I also think that it is possible that IQ tests are biased against people with different intelligences. Gardner has argued that people have many different kinds of intelligence. IQ tests, on the other hand, measure only a very specific kind of intelligence (lingustic and logical-mathematical intelligence).
Finally, I do believe (but cannot prove) that IQ tests are biased against poor people. I do believe that poorer people (and racial minorities are overrepresented in this group) are typically less educated. I do not, however, believe that this means they are less intelligent. The fact that IQ tests tend to say that they are less intelligent makes me believe the tests are biased.