Thats kinda odd, the first thing I always do if I doubt my readings is test the resistance of the probes Im using. Shouldnt read more then an ohm when touched together, otherwise maybe a bad connection.
Were you taking these readings with the circuit powered up? That'll produce some wierd readings, weak batteries can do the same. Jerran has a good point also, check the leads. I personally don't think analog meters are any more reliable than digital, (that'll start something) :twisted:
If there is a transistor or diode in the circuit you are measuring. it would explain the strange readings. The apparent resistance of a semiconductor junction varies with the current in a non linear manner.
You don't mention if the resistor is in circuit or not, you really need it out of circuit to test it reliably - you may well get different readings on different ranges if it's in circuit, as it could be affected by other components.
As a hobbyist I find that I have to warm up, so to speak, when making measurements. I'll take a minute to measure some things that are more or less "known" to me - the voltage of a new battery, the resistance of some resistors, etc. With the scope I'll check to see if I can make sense of the calibration voltage then proceed with making measurements. That tests my skills as well as the integrity of probes and leads - possibly the instrument itself.
Yes, it can read through other components as well, depending on the exact way the meter works it may read through semiconductors, and also with highish values like you are talking about leakage currents through electrolytics can alter readings as well. The only sure way is to remove one end of the resistor before testing - if you don't do that, try testing with the meter both ways round - and see if it reads differently, if it does you obviously have a problem with it reading through something else.
I have the same prob in my digital multimeter , already changed fuse and battery but the problem remains . The resistors aren´t connected , the probes are OK ! Any clue ?
The things that I might do would relate to verifying that the DVM is functioning properly. I'd check the voltage of some good batteries to see if the readings are reasonable. Check some other resistors - some low and some higher. I've had several of the lower cost DVMs fail because of the range/function switch. There was little that I could do about it on those models.