Correlation does not equal causation. I know this is a picky distinction in some cases, but not in this one, because so many factors come into play. We have no way of knowing if this "slight correlation" is because of some other reason, or even if the correlation exists at all if it's so slight.
Also, just logic-wise, it's more likely that more guns=more deaths by guns than the opposite. Simply because it makes it much easier and faster to kill someone. "Good guy with a gun preventing gun deaths" does happen, but it's counter-intuitive and sadly, rare. More data is needed.
And the NRA bans all studies about gun violence/gun control measures. ALL. You can see why it's difficult to reach a solution on this matter.