I would also posit that the average person has limited knowledge of the scientific method and how it may work for or against them. Experimental design is COMPLICATED! Particularly when attempting to prove something at p = <0.001! I mean, dang! Statistics is certainly not my forte, but the probability that a medication has an effect need only be <0.05... and there's far more than a million at stake there!
Your letter, Oak, has brought me back into the discussion. Experimental design
is complicated. Let's look at famous dowsing experiments.
Randi's trucks containing water shows that no supernatural being or force helps a dowser to decide which truck water is in. Randi's bottles beneath buckets shows that no supernatural being helps a dowser decide which bottle contains water. Similarly for his pipework experiments. They are all superb experiments for disproving the supernatural. They've succeeded. Their flaw small is that someone can come along and say that their supernatural helper is better than others and nothing in the previous tests can disprove their claim.
They have not disproved dowsing in the field because the experiments were not designed to examine a field trial. Designing a field trial is admittedly difficult and is unlikely to be error free.
If, in a test there was a requirement for a dowser to find water in the field, it would be extremely difficult to eliminate the possibility of the dowser reading the landscape. So my experience of finding service trenches is the basis for a less error prone experiment that does not assume that dowsing is driven by the supernatural.
It uses trench excavations that have been refilled with a drainage pipe, providing a void, sand and gravel ballast and topped up with the original soil and turf until level. Spoil from digging is to be removed along the walkway, not out from the side, so that there is no clue where the trench was dug. As plywood can be delivered before the digging starts, the digging will be conducted from the plywood, thus preserving the trench edge which will be hidden beneath the plywood anyway. The spoil will be dumped temporarily on tarpaulin and the excess removed – along the walkway.
I might as well continue to answer some queries about the dimensions of the proposed trenches (still to be agreed by my assessor, Prof French, so could be amended). I am aware that I have found service trenches around my 40 year-old house; they would have been 0.75 metres deep as pointed out. A 15’ main sewer was even easier to detect. Remember, I am not doing research at this stage. I am in a contest and one error will see me removed from the Challenge for at least a year. So my choice of a 1 metre deep trench seems prudent.
If someone is going to dig a trench one metre deep, they will need it to be one metre wide and nearly two metres long so that they can wield a spade and shovel. The 2 metre width is useful. If I wandered off course on a 4 foot wide board I wouldn't want to miss a centrally placed hole because it was only one foot wide.
If the digging is done by volunteers I needn't worry too much about deeper holes being more expensive to dig - what is another bottle of beer amongst friends.
The real cost lies in disguising where the holes have been dug.
A small excavator is likely to leave a track to and from the hole requiring even more expensive covers which is why the experiment is designed around hand dug trenches.
I do not know the limits of detection for any of these parameters that I'm having to decide in advance, so in March (hopefully) I'm likely to go for something I know and am happy to be tested on rather than something at the limits. I am conscious of the fact that I have only one chance.
My latest ideas on disguising where the trenches have been dug and keeping costs down is that the trenches will be hidden by plywood 8’ by 4’ and formed into a walkway 4’ wide. Each plywood sheet would be marked so as to divide it in half. Each walkway would comprise 6 boards marked 0,1; 2,3; 4,5; 6,7; 8,9; and 10,0. The 0 places would be there so that I would be at the right height before reaching position 1 or after leaving position 10. Three walkways with ten position options would be similar to the testing of Swedish woman who was tested by Banachek using three sets of ten playing cards.
Oak continued “
I'll reiterate what many others here have stressed. An experiment fraught with potential design complication/expense should be subject to running one or more pilot tests to further refine said design. Coming up with protocol stipulations that seem to come out of thin air is not sound experimental design.”
I hope I’ve shown they did not come out of thin air. The protocol will be subject to pilot studies in March, weather permitting. If Prof French and Pixel 42 attend I’m sure that together we can refine the test protocol before agreeing the proper preliminary test.
