Experimental protocol for water dowsing
1 This protocol is designed to test the proposition that a particular individual ("the Dowser") is able to divine the presence of a quantity of water ("the Target") when it is concealed from sight, exclusively by means of dowsing.1.1 The process as a whole ("the Experiment") will consist of a number of identical iterations of the same dowsing procedure ("the Trial").2 The Experiment will be administered by four individuals, who will operate in two pairs ("the Observers").
1.2 Each Trial will be completed by the Dowser, and also by an independent individual agreeable to all parties ("the Control"). The Control will operate under exactly the same conditions as the Dowser, but will attempt to locate the Target by means of guesswork alone.
1.3 Each Trial will produce a clear result (success or failure) for both the Dowser and the Control.
1.4 The Target shall consist of four pints of ordinary tap water, sealed in a bottle.
2.1 One Observer in each pair will be nominated by the Dowser, the other will be an independent observer who may be skeptical of the Dowser's claimed ability.3 The Experiment will be conducted in a room ("the Venue") with no internal windows or similar apertures, and no external ground floor windows or similar apertures.
2.2 One pair of Observers will remain with the Dowser and the Control throughout the Experiment, and will not be present while the Trial is being set up.
2.3 One pair of Observers will set up the Trial, but will not be present while either the Dowser or the Control is performing the Trial.
2.4 All Observers will conduct themselves in accordance with both the letter and the spirit of this protocol; any failure to do so will invalidate the results of the Experiment, and may result in the Experiment being terminated immediately.
3.1 The Venue will not contain any microphone, security camera or other surveillance device.4 The Dowser will be permitted to invite a reasonable number of supporters to attend the Trial.
3.2 Video and/or audio recording equipment may be present in the Venue subject to the agreement of all parties. Any such equipment shall not be connected in any way to any system external to the Venue. It is suggested (subject to the agreement of all parties) that such equipment should be left running uninterrupted and untouched by any party throughout the Experiment.
3.3 It is the explicit intention of this protocol to prohibit any means whereby any person, not present in the Venue while the Trial is being set up, could possibly ascertain the whereabouts of the Target.
4.1 The supporters may accompany the Dowser at all times throughout the Experiment.5 The Venue will contain a number of possible sites where the Target could be concealed ("the Locations").
4.2 All supporters are required to enter and exit the Venue along with the Dowser, in accordance with the requests of the Observers as required by this protocol.
4.3 Any failure by the Dowser or any supporter to follow the requests of the Observers or otherwise behave outside the letter or the spirit of this protocol will invalidate the results of the Experiment, and may result in the Experiment being terminated immediately.
5.1 The number of Locations present will be agreed in advance in accordance with 12.1 below.6 The Dowser will be afforded full access to the Venue before the Experiment begins, to ensure total familiarity with the surroundings and all arrangements stipulated by this protocol.
5.2 Each Location will consist of the following items:
5.2.1 A bucket, with marks on the floor to indicate its position for the duration of the Experiment.
5.2.2 An equal amount of sand inside each bucket, of a quantity sufficient to prevent the bucket from moving easily.
5.2.3 A plain square black cloth placed over the bucket, with its edges parallel to the walls.
5.2.4 A card bearing a single letter attached to the floor next to the bucket for the duration of the Experiment, to identify each Location uniquely.
5.3 All items used will be procured in advance, and should as far as possible be identical to each other.
5.4 The pair of Observers responsible for setting up the Trial shall equip themselves with an opaque container containing a set of identical tokens ("the Randomiser"), as follows:
5.4.1 One token to represent each Location, bearing the letter corresponding to that Location.
5.4.2 Some blank tokens, which bear no letter nor any other identifying mark.
5.4.2.1. The number of blank tokens will be agreed in advance in accordance with 12.3 below.
6.1 At this time, the Dowser should set up and conduct a series of test Trials, using the actual Target and Locations. Such test Trials shall not form part of the result of the Experiment, but will allow the Dowser to verify that no aspect of the Venue interferes with dowsing ability.7 The Experiment will consist of a number of identical iterations of the Trial.
6.2 At the conclusion of the test Trials, the Dowser will sign a declaration to the effect that:
6.2.1 Every aspect of this protocol has been understood and agreed by the Dowser.
6.2.2 All the stipulations of this protocol have been effected correctly.
6.2.3 The Experiment represents a fair process for testing the Dowser's claimed ability.
6.2.4 No environmental or other circumstances are present which may interfere with that ability.
6.2.5 No intentionally fraudulent technique of any kind will be used during the Experiment.
7.1 The number of iterations of the Trial will be agreed in advance in accordance with 12.2 below.8 The Trial will be set up in accordance with the following procedure by the pair of Observers responsible for setting up the Trial, who shall:
7.2 Each Trial will follow the same series of stages:
7.2.1 The Observers responsible for setting up the Trial will set up in accordance with 8 below.
7.2.2 One party (as determined in 7.3 below) will complete the Trial in accordance with 9 below.
7.2.3 The other party will then complete the Trial separately, in accordance with 9 below.
7.2.4 The Observers responsible for setting up the Trial will check the Venue in accordance with 10 below.
7.3 Prior to the commencement of the Experiment, the Observers responsible for administering the Trial will toss a coin to establish which party (the Dowser or the Control) will first complete the Trial in the initial instance. In the second instance of the Trial, the other party will complete the Trial first, and thereafter the two parties shall alternate.
8.1 Enter the Venue alone, noting the exact time when they do so and so informing the Observers responsible for administering the Trial using a simple prearranged signal.9 The Trial will be conducted in accordance with the following procedure by the Observers responsible for conducting the Trial, who shall:
8.2 Remove all buckets from the Locations, then draw one token blindly from the Randomiser.
8.3 If the token drawn corresponds to a Location, push the Target into the sand inside a bucket and position it on the marks at the corresponding Location; or if the token drawn is blank, remove the Target from the Venue altogether.
8.4 Make a written record, countersigned by both observers, of the Location number where the Target has been placed, if any ("the Answer").
8.5 Replace the token in the Randomiser.
8.6 Place buckets containing only sand on the marks at all Locations other than that which may contain the Target.
8.7 Ensure that the arrangement of cloths and buckets at all Locations is identical, and is in accordance with the protocol.
8.8 Vacate the Venue, having competed the setup procedure within a prearranged length of time.
8.8.1 The time allowed for completing the setup procedure will be agreed in advance in accordance with 12.4 below.
8.8.2 In the event of any unforeseen difficulty preventing completion of the setup process within the agreed time, the Observers responsible for setting up the Trial shall notify the Observers responsible for administering the Trial that the setup procedure is not yet complete, using a simple prearranged signal which does not allow for any conversation with any other party.
8.9 Not re-enter the Venue until it has been vacated by everybody present during the Trial.
8.10 Never introduce into or remove from the Venue any object of any kind (other than the Target), nor move any object which is present in the Venue (other than those which constitute the Locations).
8.11 Never make any attempt to communicate in any way with any party involved with the Experiment, other than pre-arranged signals to other Observers as allowed by this protocol.
9.1 Ensure that no person other than the Observers responsible for setting up the Trial enter the Venue while the procedure for setting up the Trial is being followed.10 After the Trial has been concluded, the Observers responsible for setting up the Trial shall:
9.2 After the agreed length of time for setting up the Trial has elapsed, and the Observers responsible for setting up the Trial have left the Venue, invite the first party due to complete the Trial (in accordance with 7.3 above) to enter the Venue and inform them that the Trial may commence.
9.3 Ensure that the Dowser has adequate opportunity to investigate all Locations using any dowsing technique, including the use of dowsing rods, crystals or any other similar objects.
9.4 Ensure that the Dowser, the Control and any supporters do not touch any part of any Location during the Trial.
9.5 Ensure that the Dowser, the Control and any supporters do not employ any form of apparatus or technique (electronic, magnetic or otherwise) capable of detecting the Target by any means other than dowsing, either during the Trial or at any other time during the Experiment.
9.6 Ensure that both the Dowser and the Control record in writing the number of the Location which is believed to contain the Target, if any ("the Prediction").
9.7 Ensure that the Prediction is countersigned both by the party who made it and by both Observers.
9.8 Ask the party who made the Prediction to leave the Venue, then invite the other party (in accordance with 7.3 above) to enter the Venue and repeat steps 9.3 to 9.7 above.
9.9 Ensure that the Venue is vacated at the conclusion of the Trial.
9.10 Inform the Observers responsible for setting up the Trial that it has been completed, using a simple prearranged signal which does not allow for any conversation between the Observers responsible for setting up the Trial and any other party.
10.1 Re-enter the Venue alone.11 The Experiment shall consist of repeating the Trial over a number of iterations, as agreed in advance in accordance with 12.2 below. The outcome of the Experiment shall then be resolved as follows:
10.2 Check that all Locations are still arranged in precisely the same manner as they were when set up.
10.3 Check that the Target is still present in the bucket corresponding to the recorded Location.
11.1 The Dowser, the Control, all Observers and any supporters will assemble in the Venue.12 This protocol does not stipulate several specific details. These will be negotiated and agreed in advance between the Dowser and all Observers. They include:
11.2 An opportunity will be given for any person present to highlight any concerns they may have that the conduct of the Trial was not in accordance with both the letter and the spirit of this protocol.
11.3 The Dowser will sign a declaration to the effect that:
11.3.1 All the stipulations of this protocol were effected correctly during the Experiment.
11.3.2 The Experiment represented a reasonable test of the Dowser's claimed ability.
11.3.3 No environmental or other circumstances were present which interfered with dowsing ability.
11.3.4 No intentionally fraudulent technique of any kind was used during the Experiment.
11.4 The countersigned written records of first the Control's Prediction, then the Dowser's Prediction, and finally the Answer for each Trial shall be revealed by the Observers, in the order in which the Trials were conducted.
11.5 If a Prediction coincides exactly with the Answer of a given Trial, then that Trial shall be considered a success for the party who made that Prediction; otherwise it shall be considered a failure.
11.6 All countersigned declarations, Predictions and Answers, along with any video and/or audio recordings of the Experiment, shall be made freely available for all involved parties to inspect, copy and/or make public as they see fit.
12.1 The number of Locations to be present in the Venue (this number shall be between four and ten).Bruce Walton, UK-Skeptics © 2007
12.2 How many iterations of the Trial shall constitute the Experiment (this number shall be at least twelve, and should vary in inverse proportion to the number of Locations present).
12.3 The number of blank tokens present in the Randomiser (this number shall be at least two, shall be smaller than the number of Locations, and should vary in direct proportion to the number of Locations present).
12.4 The length of time required for the Observers to set up the Venue for each Trial (this duration will be agreed contingent on relevant practical considerations concerning the Venue).
.
Re: Dowsing test: protocol for a water dowsing test.
Hello everyone,
In the spirit of peer reviewing, we would very much welcome opinions and constructive criticism concerning this protocol.
You can post you thoughts here, or if you prefer feel free to PM me (the author) to discuss it.
I had a couple of initial thoughts
1. Is a "Control" subject really necessary? It may sound good in the interests of thoroughness, but it would lengthen the test and may not contribute anything to the statistical analysis (although I am not a statistician, so I could be wrong).
2. If the location is not allowed any internal or external windows, doesn't this exclude pretty much anywhere? I guess you could use a lecture theatre or meeting room with no windows, but it seems like quite a restriction.
I agree with Araneus that a control subject is not necessary. It is extremely easy to work out the probablities involved for this sort of test, which is essentially all a control would establish, except that the control would need a very large number of trials. Having a control in this manner would open all sorts of excuses for the dowser - "I only did as well as chance but the control did even worse, so I must have something". In addition, the control adds nothing at all. If a dowser claims they can get 80%, the presence of a control subject does not affect this. The dowser's results will be exactly the same either way.
I also question the "no windows" rule. I can see the reasoning behind it, but I am not convinced it is actually workable in practice. A Randi-style test with just a cordoned off area in a park should be sufficient. If anyone actually manages to pass then stricter rules might be needed, but I haven't heard of any tests where this has actually been a problem.
I would also suggest that the protocol could be made more general by replacing the bottle of water with just "the substance". Water is not the only thing dowsers claim to find, and this protocol would work just as well with a lump of gold or a box containing whatever they claim to be able to find.
In section 6, I think you need to specify the details of the open trial more clearly. Rather than just saying that the dowser should conduct some test trials, make it explicit that these will be exactly the same as the real trials but with the knowledge of where the target is.
Thanks for the thoughts, AraneusOriginally Posted by Araneus
1. Good point - when I first wrote this there was no control, so it was more akin to a JREF-style confrontational test. However (at Dr B's suggestion) we did decide to add a control. The presence of a control makes it feel more of a valid experiment, and less like a debunk. The reasoning is that assessing the outcome against chance alone may not allow for other envoronmental factors which might lead people to select a certain location disproportiately often - slope on the floor, temperature in different parts of the room, or whatever.
As you say, it might lengthen the test considerably - although I should think we'd want someone as the control who would not hang around over making their guess too much!
There could be a case for having an alternate version without the control. Opinions welcome!
2. It actually says "no external ground floor windows or similar apertures". The idea was to preclude the possibility of passers-by disrupting proceedings, or even accomplices giving clues. We could leave this sort of thing more vague, of course - but one of the great things about strong protocols is that you set stringent standards to start with, and then adjust for the particular circumstances of an actual venue (for instance, if there was a ground floor window but it opened into an internal courtyard which was not open to the public).
Now that you mention it, I suppose than an ingenious fraud with an accomplice with binoculars in an overlooking building could possibly manage to cheat by observing through first floor windows. Observers would have to be vigilant for things like that ...
Thanks for the thoughts, Cuddles. Useful stuff.
See my reasoning above (we just cross-posted) , but yes I see your point. This is definitely something worthy of discussion.
IIRC, Randi did have a problem with controls when in an open space. I think that with our likely resources, an indoor venue is much easier to regulate. See also my comments above - I'd have thought we want to start from a high standard when the practicalities of an individual venue are being negotiated.
You must be psychic- that version already exists!
OK, good idea. I think this should be included.
Now, I know I’m not known for my critical thinking on this forum – but I AM making an effort!
Where would you get your “dowsers”? I noticed in the French studies (also indoors) they looked to me like a bunch of university students. I would think (dare I?) that there are many people who THINK they can “dowse”. If you used commercial dowsers, they would charge you, making it an expensive “test”.
Also, why does it have to be indoors with “contrived” targets? Surely if you took your dowsers to a natural place – or even city streets where known (mapped) targets are and asked them to find whatever, this would make it a proper test? You could still have your controls using guesswork. And your observers.
The difference would be that those “setting up” the test would (by a random method) choose the place (the research team would have already collected the plans/maps for several different places in various cities, or towns or villages/whatever, in different counties – or even countries!!! ).
Your dowsers, controls, observers would be bundled into a vehicle and taken to a “real” place. The chance of a cheater setting up co-conspirators to “signal” them, or pre-studying the location if it were randomly selected seems slim.
I know this will have loads of “flaws” – I am not a scientist. I know scientists like to set up experiments in laboratory setting where they can control ALL the conditions – but perhaps, in this case, that’s the problem?
Get rid of the supporters. If you REALLY CAN dowse, you don’t need supporters!
Also, if your dowsers are using CRYSTALS! I would rule them out. That sounds new agey to me.
Dunno really, just some pre-gardening day thoughts…skeps are allowed to shoot me down in flames on this - I really don't know what I'm talking about!
M
The purpose of the protocol is to provide a method for testing dowsers who choose to apply to take the test, it is not a proposal to "go out and find" dowsers to test.
Also, why does it have to be indoors with “contrived” targets? Surely if you took your dowsers to a natural place – or even city streets where known (mapped) targets are and asked them to find whatever, this would make it a proper test? You could still have your controls using guesswork. And your observers.
Because it wouldn't be controlled -- for example, a "natural" target is likely to have a large number of other clues that a dowser could use to identify water. Performing the test in laboratory conditions removes this difficulty.
Get rid of the supporters. If you REALLY CAN dowse, you don’t need supporters!These parts are concessions to the dowsers, who would otherwise claim that they can't dowse when surrounded by "negative energy" or some such. By allowing them to perform under their preferred conditions, you remove the post hoc objection that the conditions weren't right (although they will no doubt make this sort of claim anyway).Also, if your dowsers are using CRYSTALS! I would rule them out. That sounds new agey to me.
Hi Michelle, and thanks for your thoughts.
Don't worry, we'd never dream of wasting UKS' resources on paying dowsers!
It has happened from time to time that dowsers have approached us, expressing an interest in establishing sound evidence for their claimed abilities. the purpose of this protocol is to provide us with a ready-made robust way of responding to such approaches.
The central claim made by most dowsers is that they can locate water using their dowsing rods alone. If this claim is true, it should work just as well indoors as outdoors, and just as well with tap water in a bottle as water collecting underground.
It is possible to run tests outdoors of course - but the difficulty is that all possible sources of external clues to the location of water must be eliminated, otherwise the test does not directly address the claim. This means that just heading off into a field or down the street is a non-starter - there are loads of mundane environmental clues which would betray possible sources of water by means other than dowsing.
A clue-free environment is most readily available indoors, and of course there are also practical issues (like the weather, and the accuracy of 'known water' information) which can be most easily addressed in such a setting.
I am not a scientist either. However, I do understand the scientific method sufficiently to understand that controlling all the conditions is absolutely necessary to meaningful experiments. You need to be clear exactly what proposition you are testing, and eliminate all extraneous influences - otherwise the results will not address the intended phenomenon.
How could controlled conditions be a "problem" exactly?
The reason the supporters are there is because the dowser for whom this protocol was originally constructed had got dowsing to 'work' in front of a crowd of people, and I was trying to reproduce the circumstances of their claim as closely as possible in a controlled way. The experiment would probably run mores smoothly without them, I agree. However, having some mates along might make the whole thing a bit less threatening for the dowser, and thereby increase the chances that they would go through with it.
What do others think about this?
What have you got against new agers? One brand of woo is as good or as bad as any other, as far as I can see. They all have exactly the same chance of being right ...
ETA: Araneus and I cross-posted here - but we make much the same points (more verbose, in my case!)
Control condition.
With a dowsing-type test the scoring can be done with binomial statistics as each trial can only result in a right/wrong condition. This means that the expected scores by chance can be calculated exactly. So a statistical baseline can be worked out in advance. This means a control condition is not necessary.
On the other hand, relying completely on a purely statistical baseline would mean we'd have to be certain that the test was 100% watertight. As this protocol was developed to be reusable (it can be adapted to different locations, targets, etc.) we'll never be sure that each time it's used the conditions are being controlled for 100%.
So, a control condition would help to account for, and help us identify, any weaknesses in the testing that show up. A control condition is not absolutely necessary but it is something we should use if we can.
As an example:
The Aussie skeptics did a 'dowsing challenge' with Randi in 1980 (I think). The video is on youtube somewhere.
Now, the results were of 3 different tests and they were pooled, and showed nothing statistically significant. But, in the water dowsing test alone, the dowsers actually scored well above chance (z-score ~2.5 IIRC).
Now, it could be that the dowsers actually did do really well or the tests could have been flawed. There was no control group. How do we know if the dowsers scored well or if the test was flawed if there's no control group? Well, we can't say. That's why a control would be useful if we can include it.
.
Mmm – yes, I see. I think my problem is that I understood dowsing as an ancient “skill” – not new age really. The idea has been hi-jacked by them for “predicting” all manner of things – not only water! If a dowser approached you to test his claims – I’d be VERY happy with your test because it will fail. God help us all if it should be proved to “work” – imagine scores more psychics “dowsing” for Madelaine McCann (which they probably actually ARE.)
My error is in thinking that you wanted to “test” whether the ancient skill may have something to it. I would think it has little to do with rods or twigs, but some ability to pick up clues in the environment – if there’s no “environment”, I doubt that it works. But as a test for some errant “dowser” who happens along here, I think it’s a great protocol!
I’ll go away now and weed some more….
M
You make it sound like the test is "crocked" in some way so that failure is certain. That is not my intention. If dowsing ability is real, the test will produce positive results - and it should therefore be a fair way of examining the question experimentally.
Exactly, you've got it
I have no interest in testing the proposition that some people are better that others at guessing the location of water based on environmental clues, and therefore the protocol does not address that question. I am interested in people making guesses of this sort, but then claiming they are based on a paranormal power. By separating the claimed power from the environment, you can test one isolated from the other and thereby expose dowsing for what science tells us it is - guesswork, expressed through the ideomotor effect, then misrepresented as a paranormal ability.
No No! I didn't mean that at all! I apologise. But my psychic powers predict failure is certain!
M
It's OK Michelle, I know you didn't mean to imply that! I was concerned that any lurking dowsers might get the wrong end of the stick ...
Jocky
One of the frequent criticisms against Randi's tests is that all the testers are picked by him and are blindly loyal to him. This is not true of course, but it does raise an important point. It is far better to have friends/supporters of the dowser present, and preferably involved with the testing, than to have a closed tests with only skeptics present. Ideally, all official roles should be done in pairs, one skeptic and one believer. In reality this is probably not at all practical, but the more supporters are around, the less the skeptics can be accused of cheating.
Posting Permissions
Reply With Quote
Bookmarks