Haunted house experiment

cavedave

Say you were invited to stay the night in a "haunted house"
Let us say there are two classes of phenomena that might be observed
Normal phenomena: clanging pipes, noises from wood contraction, animals etc
How can you prove these are normal without a control? You could not have another similar house to act as control because you cannot prove that it is not haunted.

Abnormal phenomena: ghost in white sheet, blood coming from ceilings, spinning heads etc
Now say you are recording the events to avoid observer bias. How can you discount fraud? Most experiments seem to rely on the fact that the mice do not pretend to be sick
or that the atoms will not alter their path just to play a trick etc.

So phenomena that might be explained by normal events or by fraud need to be ruled out. Is this possible and so is a haunted house stay over a useful experiment? Even if you could not reasonably guarantee objectivity, if you got such an invite would you go?

robindch

Well, a lot of thought needs to go into doing this in a proper way, so as to discount observer bias, selection bias and other common problems which plague research. Here are a few suggestions, assuming that you have one house which is allegedly haunted, and a few that are not:

Randomized + controlled: Select a large group of houses and choose, randomly (ie, with a dice) from amongst them, a smaller set to provide the control. Make sure that the "haunted" house doens't stand out in any way. Assign each house an anonymous number. (see about control here).

Double-blind: get people who are unfamiliar with ghost-hunting and give them written instructions on what things to look for (banging pipes, blobs on photos or whatever), and some easy way of recording the instances properly. Give the numbered list of houses to your friends and ask them to assign each house number to further friends of theirs who are unfamiliar with ghost-hunting and unfamiliar with you. This way, there can be no expectation set up in the minds of each group as to what to expect and you should get an accurate picture of what's going on.

Decide beforehand on a publishing protocol and a success criterion and publish both -- ie, tell the world what you are doing, how you will report it and then follow through. Publishing bias (publishing only what's decided is successful) is common in some areas of research, homeopathy to take one example. Be prepared to accept critisicm of the method and, if it's reasonable, be prepared to update your experimental protocol.

Additional control/fraud control -- you will also need to test whether your researchers are reporting correctly. So, go to a few of the houses and carry out some of the things which you asked them to report. See whether what they reported matches what you did.

Repetition -- this has to be repeated many times and consistent results gained from different groups of houses and different groups of researchers.

ooops -- ten past seven. gotta run. There's plenty more to write on this topic, but the above should give you a few pointers to doing this well enough that you can have accurate, unbiassed results. It's not easy. At least, when it's done properly.

Darragh29

I'd definitely go on something like this if I ever had the opportunity. I think the "control" part of the project you need to focus on would be the people who are sent into the house as opposed to the houses themselves. If they are proper sceptics, as I would consider myself to be, things like banging pipes and other things/noises that I would call characteristics of a house, would not be interpreted as paranormal activity. However, if I saw something with my own eyes that had no other explanation other than a paranormal one, I would have to revisit my sceptism, for example if I was an transparent apparition of someone walking through a wall or poltergeist activity...

6th

Darragh29 wrote:

things like banging pipes and other things/noises that I would call characteristics of a house, would not be interpreted as paranormal activity.

Darragh29 wrote:

or poltergeist activity...

So if banging etc is discounted what would you consider poltergeist activity?

6th

robindch wrote:

Additional control/fraud control -- you will also need to test whether your researchers are reporting correctly. So, go to a few of the houses and carry out some of the things which you asked them to report. See whether what they reported matches what you did.

And if they dont match are any findings discount because they dont fit in with what you find? (not "you" robin but whoever is carrying out the test).

As someone who goes to "haunted" houses etc a fair bit I personally dont think there is any effective way to "test" them. I wish there was but its all so subjective.

cavedave

friends of theirs who are unfamiliar with ghost-hunting and unfamiliar with you. This way, there can be no expectation set up in the minds of each group as to what to expect and you should get an accurate picture of what's going on

I think everyone has some familiarity with ghost hunting. I have seen films and heards stories as my sole exposure to it but still have a rough idea of the sorts of things you would be looking for.
Clever idea on the control btw. I cannot figure out whether paranormal believers and admitted skeptics should be left out of the experiment, iff the control works correctly it should not matter.

And if they dont match are any findings discount because they dont fit in with what you find? (not "you" robin but whoever is carrying out the test).

I would say yes. If the testers failed to notice you banging pipes at 2:30am they are unlikely to witness and "paranormal" events that occur.

As someone who goes to "haunted" houses etc a fair bit I personally dont think there is any effective way to "test" them. I wish there was but its all so subjective.

I have unsuccesfully waited for badgers to come out of a set. I still believe in badgers, you might not be able to prove a house is not haunted but you could say "we observed nothing in the test" or "we observed these unexplained events"

6th

I think you might think i'm taking a believers view on this, I'm actually saying you cant prove a house is haunted or not haunted.

As for badgers, huh? Of course badgers exist theres plenty of places you can go see them, theres lots of footage and you can even google them

pH

6th wrote:

I think you might think i'm taking a believers view on this, I'm actually saying you cant prove a house is haunted or not haunted.

Of course you can prove a house is not haunted.

What you're saying is equivalent to 'You can't prove a house doesn't have a family of elephants living it it', to which I reply 'rubbish'.

All you need is a machine or a reliable mechanism that can detect ghosts/hauntings, bring it to the house and turn it on. If it doesn't detect any ghosts then you've proven that the house isn't haunted.

6th

pH wrote:

All you need is a machine or a reliable mechanism that can detect ghosts/hauntings

Wow, where can I get one of these?

Thats like saying of course you can turn yourself into a fish ... all you need is a reliable mechanism that can turn people into fish.

enda1

6th wrote:

Wow, where can I get one of these?

Thats like saying of course you can turn yourself into a fish ... all you need is a reliable mechanism that can turn people into fish.

Now you're getting it!

6th

Its you that dont get it, I'm not one of these people who is trying to go around convincing people ghosts exist. Like I said above I dont think there is a way to "prove" that a house is haunted going by our current means.

cavedave

Check out my new phone line
Irish Skeptics Live

pH

enda1 wrote:

Now you're getting it!

cavedave wrote:

Check out my new phone line
Irish Skeptics Live

6th wrote:

Its you that dont get it, I'm not one of these people who is trying to go around convincing people ghosts exist. Like I said above I dont think there is a way to "prove" that a house is haunted going by our current means.

But what you are really saying is "we have no way of detecting a ghost". When you have the means to detect them, only then does the question of whether a house is 'haunted' become answerable.

robindch

> And if they dont match are any findings discount because they dont fit in
> with what you find? (not "you" robin but whoever is carrying out the test).

Well, you will have defined and debugged your research protocol and you will have published the success and failure criteria, in strict and literal detail, beforehand. It should be a straightforward exercise to see if what actually happened during the test meets the pre-defined criteria for "paranormal activity".

BTW, while the above is just a few basic ideas to think about, it's worth bearing in mind how the professionals in medical research do this. There was a case six months or so ago in a multi-million dollar trial of some new drug in the USA, where the trial was rigorously designed -- one group synthesizes the drug, one administers the drug, one monitors the results and a fourth group analyzes the results. Trial management can communicate only with each group, not with the test patients. No group can communicate with any other group. The problem arose because one afternoon two lads met and got talking in an airport in germany about their work; turned out they were both medical researchers and after some backing and forthing they realized they were members of different groups in the same clinical trial, and that this communication would invalidate the trial. Both reported the meeting and the last I heard was that the trial may have to be rerun, on account of a chance meeting, at a cost of millions of dollars. This is how a properly conducted trial is run and this is the level of control which has to be exerted in order to remove trial bias. If you are not prepared to undergo this level of rigor, then there's really very little point in undertaking any research as your results will probably be unreliable.

> I wish there was but its all so subjective.

...and if, on the other hand, you believe this to be so and you are unable to produce any reliable physical indicators for paranormal activity, then you should consider whether what you are looking for can be considered "real" in any genuine sense of the word. Subjectivity implies a belief-centered existence. Perhaps it only seems to exist because you believe it does?

Finally, psychologists have reasonably good and reasonably consistent explanations for many of the events that seem to happen in cases of "paranormal activity" -- you should research these first, to ensure that you remove the possibility of any of these occurring during your own research, thereby invalidating it also.

.

cavedave

If you are not prepared to undergo this level of rigor, then there's really very little point in undertaking any research as your results will probably be unreliable.

The more I think about this the more problems pop up. Particularly considering that you are unlikley to prove much either way.

The thing is the relatively easy experiments are on people. Dowsing ,esp and such should be fairly easy to test. But no one considers these fun, either they believe in these and do not want to test them or they do not and find the whole idea weird. Strangly all the scientists I have asked have thought the idea of staying in a haunted house appealing but very few view testing astrology or esp the same way.

robindch

> Dowsing ,esp and such should be fairly easy to test.

Dowsing and ESP are the most regularly tested of all the paranormal stuff and unfortunately, while people think that they have some gift, it turns out on closer investigation, 100% of the time, that it's nothing that can't be explained by wishful thinking or bad analysis on the part of the dowser.

> but very few view testing astrology or esp the same way.

Astrology has been tested plenty of times and nothing at all has been found and most researchers don't believe that anything more can be gained by any further testing. I'm not aware of paranormal research being any different from astrology -- it's been tested thoroughly too and nothing at all has been found, at least, when the research has been done in a completely rigorous fashion. Wishful thinking on the part of the researchers, though, is common, which is why great lengths have to be gone to in order to eliminate it as a source of false positive results.

Personally, I think it'd be cool if any or all of all of these things were real -- the implications for the nature of the world and the rules which govern it, would be remarkable. But unfortuantely, it seems beyond any resonable level of doubt, that there's nothing more to them than well-meaning, but misplaced, trust.

6th

For the OP, interesting bit off a read.

http://news.bbc.co.uk/2/hi/science/nature/3046179.stm

Darragh29

6th wrote:

So if banging etc is discounted what would you consider poltergeist activity?

Something that I can see with my two eyes (as distinct from something I can hear with my two ears), and find absolutely no physical explanation for, for example something levitating, something heavy moving with no source of effort causing the object to move, or an apparition. I'm a skeptic, but I would love to observe something that I simply could not explain, like a ghost apparition, or some poltergiest activity that forced me to re-evaluate my understanding of the physical world... Although I consider myself a skeptic, I would love nothing more than to see something with my own two eyes and have to say to myself, "how the fu*k did that just happen, that's not possible!". And to still have to say that to myself a month after I saw it...