And if you have a weapon, what are the statistics of trying to use it against an assailant?
http://www.bradycampaign.org/
The bulk of those statistics are amazingly well manipulated. You just have to know how to ask the right question and define the terms in ways that favors your own preferred outcome. For example, studies showing that guns are rarely used to defend the home as opposed to killing someone else in the home typically define "defend" as killing an intruder. Common defenses can involve displaying a gun, firing a warning shot, wounding an intruder, or, my personal fav, that amazingly worrisome sound of chambering a round in a pump action shotgun - any burglar who doesn't instantly flee when that noise comes from the master suite likely has a death wish.
Anyone want to bet what the statistics are from (pick a country, ANY country) on the number of deaths (murder, accident, suicide) of members of a household from kitchen knives vs. the number of intruders killed with kitchen knives? Why do I suspect that these stats would show kitchen knives are much more dangerous to family members and much less dangerous to burglars than guns in the US are?
Statistically, dogs are much more likely to bite a family member than to bite a burglar (or even a postman). Still, a largish dog is VERY high up on the list of burglary deterrents.
Feel free to check the burglary rates for Switzerland - where virtually every home has a gun (and I don't think they all keep large dogs). They've also got significantly lower gun accident rates - simple evidence that proper training in how to safely handle a firearm will reduce unnecessary deaths far faster than trying to round up all the weapons from a populace containing quite a few people who won't cooperate due to either criminal tendencies or for Constitutional reasons.