aka Common Myths about Science That Bug The Hell Out Of Me:

A scientist develops a great new invention in his lab on his own.
It used to be possible, but the days of great geniuses and polymaths like Newton and Descartes are past. Nowadays, it needs a team to develop anything. Most scientists work in teams, and so do most engineers. Someone might still have this amazing idea and make a breakthrough, but a complete prototype on his own? Not possible, as this involves several different areas of competence (see the point just below, too). Also, having people with similar competences to check what you do is usually a Good Idea, if only to make sure you’re not making any mistakes.
Just for the record, a simple prototype for a demonstration, built from scratch, should require at least a dozen people to handle the various aspects of the job. And that’s a bare minimum–like your student association building a robot for a competition. A company would have far, far more people designing the thing.

On a related subject: the scientist who knows everything about every field. He/she was trained in aeronautics, but is also a dab hand at biology, and chemistry too, when needed (I’m looking at you, Sam Carter).
Again, the days of polymaths is past. It’s possible to have vague knowledge of a lot of subjects, but to be able to make deep and sophisticated calculations in various divergent fields… You can’t be proficient in more than 2-3 connected areas (the BF had a wider education than most, and is still only proficient in physics and somewhat knowledgeable in computer science. He sucks in biology or mechanics. I’m good at computer science, reasonable in applied maths electronics, and suck at everything else).

The scientist(s) who has this great and amazing idea, and builds a prototype in a few days or a few weeks. Frequent bonus: the prototype survives field use and turns out to be perfectly operational.
Here’s the deal: developing anything is a long and drawn-out process, and field conditions are not a joke (sand that gets everywhere, weird temperatures… Your average materials are often going to take it badly). Building a prototype, even as part of a team, is more likely to need a year than a few weeks. And I’ll eat my hat if that hastily-conceived prototype is actually up to field conditions unless God takes a personal hand in the matter…

The aforementioned prototype is taken for an experiment, and no one keeps any backup anywhere. When it’s destroyed, people complain that they won’t be able to rebuild it.
Er, yes, OK. Sometimes it has happened. But this is BAD planning. Most companies/army research centres have backups and document every step of the prototype production. Not being able to rebuild it at all smacks of incompetence.

The scientist is setting up an experiment in field conditions, make modifications to the setup, and takes ages to relaunch the experiment (this usually happens when the bad guys are firing on the scientist’s position)
If you have, say, an electrical circuit and you’ve just rewired it with a few components, you don’t actually need to spend ages typing on the computer to make it work. It should be the equivalent of flicking a switch, and if you need more than a few seconds, I’ll start wondering about your actual skillset…

On a related subject, the scientist sets up an experiment in field conditions, and appears to have no idea what they’re going to do when the experiment fails.
Experiments have a protocol. They are actually prepared. You just don’t show up with your new shiny equipment and start fiddling with it in the thick of the action. Otherwise, the likelihood is that it won’t work, or worse, that you’ll fry something. And if you’re a good scientist/engineer, you’ve considered the fact that it might not work and have thought of one if not several backup solutions. Fiddling is all well and good, but the sad fact is that it’s seldom effective.

The scientist solves what looks like an amazingly complicated problem to the profane, but is actually quite a basic problem.
This is obviously a case of writerly misdirection and/or lack of research, but it’s really annoying when you happen to know a little about the subject matter (and nothing kills my trust in an SF show faster than this). The most recent example of this was Sam Carter bragging about writing an algorithm that searched through a database for words composed of a particular set of 18 symbols. This is a trivial problem (all the more so if the database is sorted, which they’ll have done if they have any brains).
A bonus case of this is when people start misusing scientific concepts. If I hear “logarithmic decrease” again, I’ll scream (a logarithm is a function that actually increases, and it’s also the one that has the slowest possible growth, so I’m not quite sure what a logarithmic decrease is supposed to be except obfuscation).

See, this is why I can’t watch most shows that are all about the cool science. I need characters to distract me 🙂

What about you? Any common misconceptions that drive you up the wall when you watch movies/series?