I hate arithmetic, so I just put the numbers on the lens like this: The first number, 1 is the size of the picture behind the lens - on the sensor, so I say it is 1 centimeter (the sensor is 1.6X2.4 cm so a circle about a centimeter in diameter makes a good subject size.)
Then the object in front of the lens is the second number, in this case 3.9 centimeters, vs 7.7 cm. So with the 1:3.9, I'd take a picture of something about 4 cm but if the thing were 8 cm, a 1:7.7 ratio would be fine.
For me, the trick is remembering what order the numbers are in, so I say it with the camera in front of me pointed away, and say "one" (pointing to the camera), "to three point nine" (pointing to the subject.) I dunno - the visceral part helps me keep it straight.