I've had a great reply to an older post about confidence in doing experimental work. Rather than leave it inconspicuously at the bottom of an aging post, I think it's worthwhile a reply in a post of its own. Hope you don't mind the publicity, John. Here's the comment, to the original post.
As a classic unconfident physics student, I would like to suggest that maybe you walk your students through HOW to check their work. My first year physics teacher would do the exact same thing to me (and yes, it can be infuriating.) One day, my teacher finally said, "well, lets take a look at your data." Going through each section of data, he asked if any of this data seemed abnormal. Then he told me to recheck ALL of my equations to make sure they were done correctly. I did this, and he told me to do it again. When I finished checking for a third time, he finally said, "If you've checked it three times and the math is right and the data doesn't seem skewed it's probably right. Now do this for all your work from now on." For some reason, just being given that process of how I should work (you know, being taught) gave me much more confidence in my own work AND I make fewer mistakes since my process is to recheck my work repeatedly. This may not work for everyone, as I imagine there are a lot of people just not willing to put in the work but some of your students might just be lost and in need of a little guidance.
As an expert in something, be it physics, sailing, gardening or whatever, it's very easy to pick up on clues that others aren't going to spot - and forget that you are doing this. So asking the question of the teacher "How do you know that?" is a good way to go. There are some simple examples in experimental work I've talked about before. If a student has an answer that is around one thousand times too big or small, I'll send them away to check their units very carefully - have they used milliamps rather than amps, etc. It's experience that tells me that, and I do it in my own work too (yes, sometimes I'll make a unit mistake - in dealing with electrophysiology we usually quote membrane potentials in millivolts - and it's easy to let that 'milli' unit slip.)
Another example I remember is when I was looking at an experimental report written by a Masters student. He had a lovely straight line, suggesting he had done something correctly, but the gradient was very wrong. My intuition told me to look at what factor it was wrong by: it was sixteen times too small. And there's a clue. Sixteen times is two to the power four. Factors of two often go astray through confusion between radius and diameter of an object. Sure enough, in the equation he was trying to prove, there was a radius raised to the power four. Bingo - he's measured the diameter, but written it down as the radius. I then went down to the lab and measured this piece of equipment myself, to verify that indeed, he'd measured the diameter, not the radius.
Arguably, a Master's student should be spotting a power of two hint, but I wouldn't expect a typical undergraduate to be able to do it.
It's hard to teach this kind of thing - experience is so important in thinking like an expert. If you're a student, make sure you make the most of the experience you get. One way, as the comment points out, is to discuss your performance with the teacher - constructively, and see how they view it, and why.