I work in a geochemistry lab that isolates elemental carbon form rocks dating from impact periods in the earth's history that may have resulted in animal extinctions. Basically someone postulates that an extraterrestrial impact occurred at time X. Some one sends us rocks from time period X and we demineralize them, and oxidize orgnaics leaving elemental carbon. We then use an SEM to look at the sample looking for soot which has a unique aciniform shape. If we find soot at a bunch of geographic sites, we say that the impact caused wide spread fires that resulted in the animal extinctions.
As for not knowing what you are doing, it's ok to ask questions. One day talking to my PI she ask's me if I knew what a supernova was. I told her no and she told me what it was. If you have a question, you can always ask your PI or the other senior members of your lab group. When I started working with my PI she gave me a stack of past publications, told me to read them over the weekend, and come in with questions on Monday. The other option is to go to the internet. Google what you are doing and get informed. Then, approach your PI and say "when it says........does that means I do......?" If you get info from the internet, always cross check with your PI! I think most PIs would agree: they'd rather you ask questions then screw up.
My lab consists of me and my PI so if I got a question, I go straight to her. She has never had a problem with me asking questions.