This is true; IQ tests are questionable enough when used with different groups of humans. You’ll never be able to make one which can fairly test different species. (An interesting aside – researchers once thought that a certain kind of fish wasn’t very “smart”, until they realised that they weren’t placing the stimulus where the fish could see it. They simply assumed it was in the fishs’ field of vision.)
However, this doesn’t stop comparative psychologists from trying. I already mentioned two of the areas they look at to get some estimation of intelligence – Theory of Mind and language. Another poster mentioned the ability to recognise one’s self in a mirror and tactical deception. There have been many ingenious studies on this, and a lot of them indicate that chimps are capable of intentional deception (which is a key piece of evidence for ToM – after all, you can’t intentionally deceive someone without being aware of their erroneous thoughts). But there are plenty of skeptics who point out ways in which the chimp could be performing this way without trying to intentionally deceive the other chimp/researcher. When it comes to proving another species has ToM, ultimately we may never be able to, because we will never actually know what goes on in their minds.
Another way of looking at intelligence in non-human primates is through tool use. Most chimp societies in the wild make tools for various purposes to suit their environments. These include things like sticks to fish out termites or ants, leaf sponges to soak up water in hard-to-get places, and the use of particular types of rocks to break open seeds and nuts. They also examine useful behaviours such as washing sandy food. Researchers are particularly interested in tool-use not just because of the proficiency it requires, but also to look at how these techniques might be socially transmitted, i.e., how do chimps learn these behaviours from one another. Mimicry? Imitation? Purposeful teaching (e.g. to offspring)? They’ve also looked at this in chimps in captivity, most notably with Washoe, a chimp who (mostly successfully) taught her adopted chimp baby how to sign.
Lastly (I could go on forever about this), a more quantitative way of looking at “intelligence” in animals is to infer it through relative brain size. Comparative psychologists do this through a method called allometric scaling. Basically, for each particular group of animals one wishes to compare (e.g. primates, mammals), brain size is plotted against body size in such a way that you can determine which animals in the group stand out with relatively large brains for their size, and which lag behind the rest of the group. In the primate example, we see that humans have brains three times larger than what we would expect of an ape of human bodily size. And yes, this method sure does have its limitations. But it is at least some sort of quantitative measure of brain size, which is closely related to intelligence (in inter-species terms at least).
Given all the various information that has accumulated about the intelligence-related abilities of chimps, researches generally summarise that the most intelligent chimps are roughly similar to 3-year-old humans.
Btw, all of this and more is explained in the book I referenced above (Byrne’s “The Thinking Ape”). Very interesting stuff. Almost makes me want to go on and do a PhD in comparative psychology.