In some cases it helps to first try to understand exactly why they might have some misconception. If you can do that, though not always possible, you can more easily find a way to guide them.
An example: I once had a student that evaluated math expressions strictly left to right without considering operator precedence. It was very hard to convince him that he was getting the wrong answers and why. So 2 + 3 * 4 was 20 in his computations.
After some explorations I discovered that there were two things that got him to this situation and reinforced his thinking. The first was bad teaching at some point in his past. He was convinced by some instructor that this was correct. Then, he happened to buy and use a calculator that didn't take account of precedence so his earlier misconception was reinforced daily.
It took a while to convince him that, while he made correct inferences based on his available evidence, it was the evidence that was the cause of his issue, not a personal failing. But until I understood that, and could communicate it, along with the proper rules, he resisted my efforts as somehow undermining him. He hated me for a while, I think, until we worked it out.