Wired contact failure

I've got a wired contact that is randomly failing. I get numerous trouble signals for a few hours about once or twice a month. It's also caused an alarm while we were asleep--that was fun.

In PC Access the value fluctuates between 160-200 when it is giving a trouble signal (no resistors installed). I assume I need to replace the sensor, but thought I would check to see if anyone else has had this issue before I tear into the window casing.

Thanks in advance.

Could be a cable issue or a crushed reed. Both are symptomatic of that. I'd pull the contact and twist the pairs and let it soak to see which is the culprit

Here had a similar issue with one exit door.  Values fluctuated and I was getting exit beeps when not opening or closing the door.
 
I had originally installed the sensor under the threshold of the bottom of the door 10 years prior.
 
I could see under the door frame in the basement and had noticed that the wires were very tight. 
 
I took the bottom of the threshold apart.  The frame of the door had shifted just enough to pull on the wire taking it apart some.  It had been soldered and I was using shrink tubing on it.  Rewired and replaced the contact sensor any how and all was good after that.

The house is only about 3 months old. The contact is on a window that has probably been opened twice. I'm really hoping it's not a short in the wall somewhere from a nail or screw--if that were the case though, I don't think I'd be getting values in the 160-200 range.

DEL, I assume you are suggesting that I take the reed switch out of the equation by shorting the wires at the contact and see if that resolves my issue?

It is easy to short out the leads on the reed switch if they have terminals.  If not you might have to cut in to the wires.
 
Use a laptop connected to PCA watching the values as you check right there at the switch.
 
Could be the distance from the magnet to the switch.  The switches are inexpensive and easy to replace.

I've seen humidity and insects cause similar issues, typically when wiring is exposed, loose, or has an issue.  Could be as simple as dew is dripping into a bad connection or ants on the hunt for food... 
 
A nail or screw in the wall can cause a similar issue as well; as the metal may be providing a connection.  Often during the day, a house will move, expand and contract from heat, cold, humidity, sun beating down on a wall or roof, even the wind can move a house just enough to have a screw or nail goof a connection (or unrelated to your post cause a random fire in the case of an electrical circuit). You may be surprised how walking around beating on a wall or two might show you where a screw/nail issue is.
 
You could always try a few tests before digging into the window frame.
 
Disconnect the wire run at the panel and put an ohm meter on the connection. With an assistant, open/close the window quite a few times and look for consistent readings. Now try using a better magnet to simulate opening/closing window. Using a spray bottle, mist the area where the sensor is. Check again and then again 2 hours later.  Lastly, walk around tapping walls (with some force) along the wire run. Lastly try shorting the window switch wires and check that the ohm meter is consistent and 0 ohms or near that value.
 
If you have any oddity with any of the above tests, and you will likely have found your issue and if not, the cause could likely be your wiring or the panel connection.

I would disagree slightly with the methodology of removing the loop from the panel and strictly going with a meter first and foremost unless we're talking analog meters. Plenty of times the issue occurs too fast for even high resolution DMM's to catch.
 
The first issue would be to eliminate the contact and see if the issue follows that portion. It's the easiest and least destructive, and we're only talking $5 or less. Drywall screw and diagonal cutters are great for pulling contacts with zero damage to the surrounding material.
 
The contact should be hermetically sealed, unless the reed is cracked or cheap unsealed (old school) contacts were used. Under drilled hole and swelling wood, sure, that could cause an issue. As could a poor splice (twist and pray method, with tape) or cable stapled too tight, but the easiest to prove/disprove for an end user would be to remove the contact, twist the pair and let the cable "soak" for a period of time. Insects should be of little consequence to a cable, properly installed and terminated/spliced.
 
Favorite issue I ran into was a batch of contacts with bad alnico magnets  steadily failing after install.

The house is only about 3 months old. The contact is on a window that has probably been opened twice. I'm really hoping it's not a short in the wall somewhere from a nail or screw--if that were the case though, I don't think I'd be getting values in the 160-200 range.
 
The above said and if new then the wiring / GC warranty should cover the repair of this stuff.  I would personally go back to the GC and tell them to fix it.
 
Or
 
If you are unfamiliar with this stuff then I would subcontract a company to look see at the issues.
 

John999 - you described tearing into the window casing. Is the switch subsurface or surface mount?
 
pete_c seems to always have the best point - turn it back over to the general contractor/installer.

Finally got this fixed...bad solder joint in one of the window contacts. PITA to repair but fixed. Thanks to all for the suggestions.

BTW, I opted against going to the GC because they installed a Honeywell system that I ripped out and replaced with the OPII. I assumed they would want nothing to do with the wiring after that even though it had nothing to do with the panel swap.

Good news John!