FB is right of course, except that it's possible for the string measure to come out pretty close even if the seat and head tubes are twisted in relation to each other. His two methods together tell you if the rear triangle or the fork have problems. The reason I suggested sighting along the wheels is that it is a simple one-step process that allows anyone to tell if there is a problem. If your wheels run in line with each other then you can be fairly sure everything is OK. If not there are either frame or wheel dish problems.