You need to read up on the difference between 'incident' light metering and 'reflected' light metering. Incident light metering, measures the intensity of light falling ON your subject; reflected light metering measures the intensity of light reflected OFF your subject.
If you use a camera's in-built meter, that will almost always be a 'reflected light' meter reading; and most SLR's & DSLR's meters are actually metering 'TTL' or Through The Lens. (Some older SLR's, and many range-finders had more simple external meters above the lens, but with a fixed angle of view, independent of that of the lens, they might be measuring light from a much wider or smaller portion of the scene than the lens, hence not be so accurate.. not that that's particularly relevant here). If you take an 'incident light reading', measure how bright the light falling on your subject is, and make your exposure settings based on that... theory is, that you'll get an exposure that's more like what you see; dark things will come out dark, bright things come out bright, where-as, if you took a reflected light reading, bright stuff will chuck more light at the meter, dark stuff less, so if your scene has more light or dark stuff in it than 'average' the meter is likely to suggest an exposure value that will make for over or under exposure.
OK; f16-Sunny... old rule of thumb for 'metering by eye'.. and it works! No meter needed! But guide says, on good sunny day, with clear sky, your exposure value will be a shutter speed 1/ISO value @ f16... if its a bit cloudy, open up a stop, if the sky's actually a bit dark, open up two stops; if its starting to get dark, ad its a bit cloudy, open p three stops... you get the idea!
So, we have an almost f16-Sunny day,crack out the old hand-held meter, and take an incident meter reading of the light falling from the sky, and yup, meter suggests EV10, or ISO100, F16& 1/100th.. great.... now, point it at the scene, take the inver-cone off and take a 'reflected light' reading, now the needle suddenly falls two stops to EV8... suggesting f11 & 1/50th at ISO100, because the subject is a dark stand of trees on a mountain side, and they reflect less light than 'average'. Point the meter down a little or turn around, and ow the meter needle jumps two stops to EV12, begging f22 & 1/200th @ ISO100, because the subject is ow a white sow covered bit of hill, reflecting more light than 'average'... or it jumps to EV14, because you are pointing it at the sun, and your subject is strongly back-lit! So you try it again, with the inver-cone back on, to get a 'better' icident reading, remembering you are supposed to put your meter i the same light as the subject.. so you walk to the woods and try it there, and meter reading comes up to EV9.. walk back over to where. you pointed it at the hill-side that gave a reflected reading of EV 12, and it comes up to EV11..
Many meter readings, all perfectly 'accurate'.. but you have a range of maybe four or five stops of readings.. you can only use one... which one do you 'trust'? Which one is 'right'?
Answer is none of them... you trust your eyes! You evaluate the scene ad how bright or dark the ambient light is, and how bright or dark your subject, and you make your OWN call as to what is most likely to give a 'good' exposure... exposure meters are very good at measuring light levels, but they are bludy thick when it comes to knowing what they are looking at! That's where you have to do your bit!
But you probably need a little bit of help & you probably don't have the luxury of a hand-held meter to make independent incident or reflected light meter readings; you only have the cameras in-built TTL reflected light meter; so you use a Grey-Card to make a 'pretend' incident meter reading with the cameras TTL meter.... you put the grey-card in the same light as your subject, and take a reflected light meter reading off it with the camera. The Grey-Card is a known 'average' brightness, neither light or dark, so if your meter is being 'fooled' by an over or under average brightness scene to under or over exposure, grey-card meter reading should help you decide if you need to add any compensation.