Whenever someone is defending faith, or is arguing that faith and reason are compatible, they should be asked which of three common meanings of the term they are thinking of. If the exact meaning of the word isn't made clear, it is almost a given that their claims will deteriorate into a mess of equivocation.
When challenged to provide evidence for the existence of God, most theists reply that their belief is based on faith. This makes it clear that, in this context, “faith” means belief without evidence. This meaning of the word also applies to the claim that faith is needed when the evidence isn't conclusive. Or in other words, when the believer says that reason can only take one so far, and one must make the decision to believe.
The religious often also use this meaning of the word in criticisms of atheists, as when they claim that it takes greater faith to be an atheist, or that atheists believe in materialism on faith. This is when faith is described negatively as “blind faith.” However, when they themselves are accused of believing without evidence, they almost always begin claiming that faith means something like trust instead. They make analogies between having faith in God and someone having faith in their doctor's abilities, or something similar.
Now, one problem with this as an explanation of what is going on when people have religious faith is that trust, when rational, is based on evidence. It is a judgment one makes based on incomplete but hopefully adequate information. But a more serious problem with treating religious faith as if it means trust is that it doesn't make sense when faith is what makes one believe in God. For that faith is itself supposedly faith in God — and obviously one cannot trust God unless one already believes that God exists.
The word is also used in a third sense, to mean a particular doctrine or set of religious beliefs. Now, it should be a simple enough matter to keep this meaning separate from the others, but unfortunately the religious sometimes fail to do even that. For instance, in the introduction to his The Case for Faith, Lee Strobel replies to the claim that faith is believing without evidence by setting out to discover “once and for all whether the Christian faith can stand up to scrutiny.” He doesn't seem to realize that, in defending the Christian doctrine this way, he's not defending believing on faith, but is instead defending believing a particular faith on the basis of evidence. The fact that the word “faith” is used to mean both what one bases one's religious beliefs on, and what the specific contents of those beliefs are, causes Strobel and many others like him to conflate the two ideas. But perhaps this is intentional: For what they want, after all, is to claim both that there is adequate evidence and that one must have faith to believe in God — even though that's a contradiction.