You answer this in literally the next sentence:
I have a weird way of checking, but it works well for me. The first thing I do is to check through for about 3 minutes, just to see if I've answered the question correctly. No checking of calculations. Just look for if everything's in the right form (e.g. in terms of y in question 7), and
if your answer is something sensible. It doesn't take long at all, so it's worth doing this check even in practice papers.
I then do more in-depth checks afterwards, maybe trying alternative methods lightly in pencil (and rubbing them out afterwards) to see if I get the same answer. Checking the details of your working is hard, since I often find myself thinking the same things while checking through, hence alternative methods being helpful.
It's also good to check your value back in the original question. e.g. i
n Q6, use a graphical calculator to see if your tangent line is actually a tangent. Or if you find a point on a curve, put it into the equation of the curve to see if it does work with the equation. I actually do this before checking my method, since if there is a problem it's clearly highlighted by something not being quite right. It identifies the questions you need to prioritise checking, which is useful if you're short of time.