Ahh, sorry, I was typing on the phone and perhaps that got a bit too abbreviated.
First I look at the scene, try to judge the light and make some mental calculations of what exposure I will need.
Then I take out the light meter and see what settings it proposes I use.
If both methods arrive at the same settings I use those.
If they disagree I go through the following checklist.
- Did I use the same ISO in both methods? (Double check against the camera.)
- Did I leave the diffuser cap on the light meter or did I forget to put it in place?
- Are there any bright spots in the scene? Did they trick the light meter? Did I take them into consideration when I did my calculations?
- Am I shooting a backlit or unevenly lit scene, which part of it do I want to expose for?
- The meter did not account for reciprocity failure.
By then I can probably see where I or the meter went wrong and make a decision on how to expose the scene.