Science applied to paranormal research
Science in paranormal investigation
Most evidence for the paranormal is anecdotal. This means that someone has related it as an experience. That's not to imply that they are lying or misremembering, simply that their experience is subjective. Anecdotal evidence has its place but there must always be other kinds as well before a subject can become scientific. Unfortunately, many subjective experiences are a result of the way the brain works (see also here). Even when experiences are objective, they are frequently misreported.
Thus, there is a need to verify anecdotes with instrumental measurements. Though instruments have proliferated in field investigations (like vigils) in recent years, they have not always been applied scientifically. It is often claimed, for instance, that ghostly activity is accompanied by a jump in the local magnetic field. However, where is the published evidence for this? Were other possible sources of field changes eliminated? And how was 'ghostly activity' verified (for instance, was a ghost actually seen or did a medium claim a ghost was present, perhaps)? Also, what kind of EMF meter was used? What was its range, accuracy and sensitivity? Was it properly calibrated? Were the readings recorded to a computer or just seen by an observer (which brings in people again!). How were baselines decided? Without answers to such crucial questions, the evidence remains anecdotal.
Reproducing and repeating
A crucial cornerstone of science is that it does not rely on who does it. It should be possible for any one, reading a scientific paper, to follow the same method used and get the same results. Obviously, this isn't always strictly practical but, where it is, it is crucial.
This has been a big problem in academic parapsychology. Often results DO appear to depend on who does the experiment. Worse, an experiment may work one day and not the next. This is one of the reasons why many mainstream scientists are sceptical of the subject.
However, this is no reason to give up on science (after all, parapsychologists haven't!). It just means we have to be more imaginative in the ways we use instrumentation. Professional scientists are always using their imagination to think of new ways to extract information. As a result, they can tell you the state of a bunch of atoms on the far side of the universe. So, it shouldn't be so difficult to say what's going on in someone's haunted house.
A central part of science are theories. These are explanations of how particular things work, to the best of current scientific knowledge. Theories are provisional - they can always be modified or even overthrown by new evidence.
Paranormal researchers should make themselves aware of relevant scientific theories that apply to their investigations (eg. how magnetism works). Such theories have been repeatedly tested and cannot simply be rejected unless there is compelling new scientific evidence otherwise.
A scientific theory must be falsifiable. That means there must be a test that would demonstrate that it is not true. Without such a test the theory would 'explain' everything and so, effectively, explain nothing.
A hypothesis is, essentially, a working, untested theory. For instance, the idea that ghosts produce magnetic fields would be a hypothesis. To become an accepted theory it would need to be compellingly demonstrated.
To do serious paranormal research, it is a good idea to take such relatively simple hypotheses and attempt to test them. It is much easier to concentrate on as few variables as possible. Set out to prove that ghosts 'exist' and you are on a hiding to nothing! There are too many variables, starting with - what exactly IS a ghost?
Are vigils scientific?
You can apply the scientific method to any research, so there is no reason why vigils shouldn't be scientific. A good start would be to ask advice from a professional scientist and, if you're really lucky, persuade one to accompany you. However, this option won't be open to most people.
For a vigil to be scientific, you really need to plan exactly what you are going to do, in detail. It is no good just turning up with lots of people and equipment and expecting to do scientific research. You need a fixed plan that everyone must adhere to at all times.
It is not sufficient to say that 'in my opinion, it is likely something is paranormal'. You have to be able to demonstrate to someone who wasn't there, that you have effectively eliminated all other possibilities.
You should aim to provide continuous, time-synchronised readings that show that a possible natural cause did NOT happen (to eliminate it). Try to imagine having to prove to someone who wasn't there that what you say happened, really DID happen. Unless you can demonstrate it, anyone reading your report will simply say 'you didn't eliminate this possibility'. If any possible natural explanation is not eliminated then your results are interesting but neither conclusive or persuasive.
You should ensure you understand your instruments (in particular their accuracy, range and sensitivity) and establish effective baselines for them.
In science, most experiments rely on eliminating all possible variables except two - the one you know about and can measure, like magnetism, and the unknown one, like the appearance of an apparition. The idea is to measure the relationship, if there is one, between the two variables. You may need to measure other variables simultaneously in order to show that they did NOT change and so were NOT involved.
So, if you want to show that EMF meters can detect ghosts, you'll need to have someone see a ghost while an EMF meter is measuring continuously. In addition, you'll need to eliminate other possible causes of a jump in magnetic fields. And, to be sure, you'll need to do it repeatedly (with different people and at different locations) to show that it's not just a coincidence! Luckily, there are simpler things you can research than this. Just make sure you never include more than one 'paranormal' unknown at a time. You cannot explain the paranormal in terms of the paranormal.
Is it haunted?
When doing research comparing 'haunted' and 'non-haunted' locations, researchers are often not critical enough of the evidence. It is easy to read an EMF meter but what level of anecdotal evidence do you accept that somewhere is haunted? That 'variable' is difficult to measure so set standards beforehand (eg. repeated similar reports by independent witnesses over several years) and stick to it.
One of the cornerstones of the scientific method is peer review. The idea is that other researchers in the same field review your work and pass comments on it. It is 'peer' review, rather than 'expert' review, because at the edges of knowledge where science operates, there are no experts with all the answers.
In our field, peer review could mean showing your vigil (or other research) results to a fellow investigator you respect, for comment. Showing the report to a scientist in a relevant field would be even more valuable.
The problem with peer review, from a researcher's perspective, is that the reviewer may criticize your work! Obviously, this can be personally annoying but it is designed to make you think. If you feel the criticisms are unjustified then you can say so. It could be that there is simply information missing from your report. Often, though, the criticisms, which are meant to be positive, will be valid. In that case you may need either to (a) rewrite your report or (b) do more research to fill in the gaps. You should not take such criticisms personally. They are intended to help you improve the credibility of your research not to undermine it.
This may sound like a grim process but it is necessary. The idea of peer review is that two (or more) minds are better than one. No one can think of everything that might affect some research but more minds can pick up all the important points.
Peer review is the way scientific publication work. Someone submits a paper and it is sent out for peer review. The comments are sent back to the researcher who may choose to amend or withdraw their paper. Usually, after suitable amendments, the paper is accepted by the scientific journal and becomes part of scientific knowledge.
Build on existing...
Science builds on existing knowledge. Obviously, sometimes researchers make discoveries that proves previous theories wrong. However, to do this you will need impeccable methods, results and the support of peers. Any new discovery also needs to be repeated.
In general, though, researchers quote existing scientific papers in support of their work. So, you should know who's done what in your field before you attempt to extend it. There are various parapsychological journals (as well as journals from fields related to your specific research) you may want to consult before attempting any serious scientific work in this field. Your research should build on earlier results. There's no point redoing what others have already done. You can quote their results in your reports to show why you don't need to redo certain things.
It is important to develop a talent for critical thinking. It is not something that comes naturally to most people.
If you're doing some research, don't just consider the way you expect things to work. You need to critically analyse your hypotheses, the way you work, how you are getting your results and what you do with them. There are some hints on developing this faculty here.