Given certain evidences, I ‘ought’ to believe certain things. I am intellectually responsible for drawing certain conclusions, given certain pieces of evidence. . . If I ought to believe something, then I must have the ability to choose to believe it or not believe it. If one is to be rational, one must be free to choose her beliefs in order to be reasonable. . . But such deliberations make sense only if I assume that what I am going to do or believe is ‘up to me’ - that I am free to choose and, thus, I am responsible for irrationality if I choose inappropriately. 
However, it is a necessary presupposition of rationality and rational pursuits (such as philosophy) that rationality is possible. Therefore, determinism, which rules out libertarian freedom, is necessarily false (not just contingently false, i.e. not simply possibly true but actually untrue, but not even possibly true). If determinism is necessarily false, any world-view that requires determinism to be true must also be necessarily false. Naturalism and physicalism both imply determinism. Therefore both naturalism and physicalism are necessarily false: ‘It is self-refuting to argue that one ought to choose physicalism. . . on the basis of the fact that one should see that the evidence is good for physicalism. . .’  "
A common fallacy used by metaphysical naturalists is argue that "reasons" are causes too; thus you're not free of unavoidable causes. The fallacy consists in to identify "reasons" (that applies to conceptual matters) with physical causes (that only applies to physical or empirical phenomena). Reasons aren't physical causes, they're the foundations of logical conclusions and decisions. For example, in the physical world, the cause X of Y phenomenon is previous to it (i.e., no physical causes is posterior than the effects of it).
However, conceptual reasons can be posterior than conclusions or actions (in fact, most people have preconceived ideas, and then seek for "reasons" to support them... a phenomenon known as "rationalizing"). Thus, physical causes are previous (or, in some cases, simultaneous) to effects. But conceptual reasons can be actually previous, simultaneous or posterior to any conceptual conclusion (only in a strict logical sense, can be asserted that reasons must be previous to a conclusion, i.e conclusions and actions should be based on previous rational reasons).
Scientific evidence support the conclusion that people who doesn't believe in free will are more prone to cheat (because the feel no personal responsability to their actions... after all, these actions were unavoidable!). According to this recent study: "it is well established that changing people’s sense of responsibility can change their behavior. But what would happen if people came to believe that their behavior was the inevitable product of a causal chain beyond their control -- a predetermined fate beyond the reach of free will?
Surprisingly, the link between fatalistic beliefs and unethical behavior has never been examined scientifically -- until now. In two recent experiments, psychologists Kathleen Vohs of the University of Minnesota and Jonathan Schooler of the University of British Columbia decided to explore this knotty philosophical issue in the lab, and they figured out an innovative way to do it.
Vohs and Schooler set out to see if otherwise honest people would cheat and lie if their beliefs in free will were manipulated.
The psychologists gave college students a mathematics exam. The math problems appeared on a computer screen, and the subjects were told that a computer glitch would cause the answers to appear on the screen as well. To prevent the answers from showing up, the students had to hit the space bar as soon as the problems appeared.
In fact, the scientists were observing to see if the participants surreptitiously used the answers instead of solving the problems honestly on their own. Prior to the math test, Vohs and Schooler used a well-established method to prime the subjects' beliefs regarding free will: some of the students were taught that science disproves the notion of free will and that the illusion of free will was a mere artifact of the brain's biochemistry whereas others got no such indoctrination.
The results were clear: those with weaker convictions about their power to control their own destiny were more apt to cheat when given the opportunity as compared to those whose beliefs about controlling their own lives were left untouched.
Vohs and Schooler then went a step further to see if they could get people to cheat with unmistakable intention and effort. In a second study, the experimenters set up a different deception: they had the subjects take a very difficult cognitive test. Then, the subjects solved a series of problems without supervision and scored themselves. They also "rewarded" themselves $1 for each correct answer; in order to collect, they had to walk across the room and help themselves to money in a manila envelope.
The psychologists had previously primed the participants to have their beliefs in free will bolstered or reduced by having them read statements supporting a deterministic stance of human behavior. And the results were just as robust. As reported in the January issue of Psychological Science, this study shows that those with a stronger belief in their own free will were less apt to steal money than were those with a weakened belief.
Although the results of this study point to a significant value in believing that free will exists, it clearly raises some significant societal questions about personal beliefs and personal behavior."
It's wise to keep pseudosskeptics (professionals or amateurs) away from you. Beware of them!