There is a popular paranormal theory that ghosts produce magnetic fields when they appear.
There are several objections to this theory. Undoubtedly the most important is that there is no obvious evidence to back it up. Until it has been persuasively demonstrated that a magnetic field appears whenever a ghost is seen, the theory is little more than speculation. 
Even if a carefully controlled study does demonstrate a link between ghost sightings and magnetic field disturbances, it would still be necessary to decide whether it is cause or effect. It has, for instance, been shown in the laboratory that certain magnetic fields can induce hallucinations in some people. So, it is possible that the appearance of magnetic fields is actually stimulating hallucinations of ghosts rather than that ghosts are producing magnetic fields.
One argument sometimes quoted against the idea that ghosts produce magnetic fields is that those proposing the theory don't explain exactly how ghosts can produce such fields. This is really an argument that the theory is incomplete rather than that it is incorrect. If it can be shown that ghost do produce magnetic fields then it is up to science to explain it. It is science's job to explain observations, not dismiss them out of hand.
A scientific theory must explain whatever reliable observations are available and make testable predictions about things that have not yet been recorded. This is crucial because there could be many competing theories explaining an observation. The only way we have to decide which is true is to do tests based on predictions from the competing theories.
So, if a theory predicted that the magnetic fields accompanying the appearance of a ghost would always be of a specific frequency and duration, this could be looked for to see if it was correct. To do this, the theory would have to explain how ghosts produced magnetic fields.
What people commonly call 'paranormal theories' are not the same as scientific theories. They don't usually explain things in terms of existing knowledge nor do they generally make testable predictions.
They are usually too vague to be of any practical use.
Are some reports of hauntings caused by magnetic hallucinations? Ghostly hallucinations have been induced in certain people in the laboratory (by Michael Persinger) by very low frequency, weak, complex magnetic fields. If such fields occured in haunted locations, could they account for some reports of ghosts?
Jason Braithwaite, who devised the MADS project, has given the name EIF (experience inducing fields) to the weak, complex magnetic fields required to produce hallucination.
In early research at Muncaster Castle in Cumbria, suitable conditions were discovered for the production of EIFs at a haunted location. If anyone moved slightly in the 'haunted bed' (where several independent witnesses have reported hearing a child crying) they could induce EIFs in their brain. This is because the bed was found to cause a strong distortion to the local magnetic field. This distortion leads to a high magnetic gradient across the bed so that anyone moving, even slightly, will be subjected to significantly varying fields.
An interesting question is - why should different people experience the same hallucination at the same place on separate occasions? Hallucinations typically have varying content between different people. Two possible answers are (1) that they react in the same way to identical magnetic fields or that (2) there is some additional stimulus providing the subject matter for similar hallucinations. For instance, a sound or sight characteristic of the location may be incorporated into a hallucination. This is similar to the way that you may incorporate the sound of an alarm clock into your dream.