Either way, you do tech support long enough you learn that the more certain someone is that this or that didn't happen the more likely it is they're lying through their teeth.
Lying through their teeth? I don't think that is a fair statement at all to pre-judge without seeing the evidence. Being "certain" an action did NOT cause some event may reflect ignorance or naivety, but not deceit.
It is not lying if you are certain you are being truthful. Though it could certainly be closed mindedness and conceit if you refuse to accept the facts when presented in front of you, or deceitful if you are unwilling to support your position with collaborating evidence and just expect others to believe you because you say so.
Spilled milk on the floor and an empty glass in the kid's hand is NOT proof the kid did it, or that he is lying when he says he didn't do it.
As I noted above, I appreciate the loyalty. And no doubt the folks at Avira know their stuff and Avira do make reliable and effective products. But once again, no single-source solution is perfect, especially if users (always the weakest link in security) are involved. The best security in the world can quickly be negated if the user lets the badguy in.
I do agree, however, and stated before that RAM testers should be allowed to run for several passes (several compete test cycles) and there should be zero errors. That said, no software based RAM tester is 100% conclusive. To
conclusively test your RAM, you need to use sophisticated and
very expensive test equipment, like this
Eureka Express DDR3 DIMM Memory Module Tester. So unless you have a real DIMM memory module tester, swapping out RAM to see what happens is often the best test.