It should be a simple enough thing to create a controlled experiment. In one room, we have the subject, attached to various and sundry medical monitoring devices. In another, we have someone randomly watching either pornography or pictures of dogs playing poker. After a statistically significant number of runs, the medical monitoring ought to detect any harm done to one person by another person's viewing of pornography (or, for that matter, by another person's viewing of dogs playing poker).
At least you were able to answer the question. Others were smart enough to avoid it because when stated explicitly the problem becomes obvious.
Decisions are made by the human will, which by its very nature is free. Under a materialist rubric*, true freedom cannot exist. Everything must be determined. So if there exists a causal relationship between pornography and violence, it must be a one-to-one correspondence which would be obvious under your laboratory conditions.
But people have free will and according to the grace that they have received can resist temptation to varying degrees. Therefore we cannot expect to see a one-to-one correspondence between exposure to pornography and violence.
_________________________________________________
*We can know with certainty that we do not live in a materialist world (and that the existence of free will is a logical possibility). In a theoretical materialist world, both the assertion that "truth exists" and the assertion that "truth does not exist" are equally "valid" since both are equally the result of "matter in motion." There would exist no objective basis for determining the validity of either assertion.
In this actual world, the world we live in, we know many truths with certainty: "truth exists," "the whole is greater than its parts," "1 + 1 = 2," "the good is to be done and evil avoided." Since we know with certainty that truth exists, we know that we logically cannot be living in a strictly material world where truth could not be known with certainty.