Quote:
|
Originally Posted by Theseus
I don't know what Qiaozhi would like to see tested either, so I certainly wouldn't hazard a guess.
However, if you don't mind me sticking my nose in.....
I wouldn't mind seeing a DB test of the Examiner attempting to find some Gold (a nugget or a coin or a ring). Now if you have full knowledge of how a standard DB test should be conducted, then there is no sense in me describing that to you.
All I would like to see is a Pre-test of the device done on a valid target that is in plain sight of the operator, and in the same setup and area as the real test would be conducted.
Then do the DB test with a target(s) completely unknown to the operator.
Then do a Post-test, again using a valid target in plain sight and conducted as the Pre-test was done.
Does that seem reasonable? The attribute we are testing is to see if the Examiner will point towards a Gold item, and do it repeatedly so that a statistical calculation might be performed based on the number of Trials and the number of Hits, and the results would be considerably better than what one might expect from Chance Guessing.
Just my thoughts....  |
Hi Theseus,
Thank you for the input.
I wouldn't hazard a guess at what protocol of double blind testing Qiaozhi wants to see done either.
I need to know the details of how he wants the test set up before I perform it.
For the test that you would like to see, I have a 14k gold man's ring that can be used for the test. Of course, I will want to see some pre-tests as well. I will allow the operator to make dry runs while he makes whatever adjustments to the Examiner controls he needs to find the best performance. Then, when the operator has satisfied himself that he has found the best adjustment, I will film a pre-test done with the gold ring in plain sight of the operator. After the pretest is done, I can film repeated runs of double blind testing with the location of the gold ring unknown to the operator. Then I can allow the operator to repeat the pre-run test where the location of the target is known.
There are only two missing element to the test you want to see. If you can provide the missing information, then I can perform the test.
1. Can you make a statement that describes what attribute of the Examiner this test is designed to test?
2. I know a lot about double blind testing. and I know that there are thousands of protocols that are used for double blind testing of this kind. In order to provide the degree of precision you want to see, I need to know the details of how you want the test set up. For example how many test runs, how the target is hidden from view of the operator, what distances, what kind of test field setting, etc. More details will make it more likely that you will see the kind of test you want to have done.
"Does it sound reasonable? "
For my purposes, it could be very reasonable depending on what attribute is being tested. Actually my own personal testing does not require double blind testing. I never performed a double blind test on any metal detector I bought, but I did test them out and read other people's reports about them. I was able to determine what detectors were suitable for me to purchase after trying out several competing models without a single scientific test. And I have always been happy with my purchases. I also have been happy with my choice not to buy detectors that did not seem suitable after trying them out (no scientific testing of my reject detectors either).
I have no need to perform a double blind test of the Examiner for myself. I can make my determination of it's treasure locating value by testing it the same way I test any other metal detector. The only reason I am performing other tests is because I am interested to see if certain claims are true, or to act as a proxy tester for other people who want to see some test that I don't have a need to see.
If I hear back from WesP, then we would only need to know the details of how you want the test protocol and the attribute you are looking to test before we can perform the test you have in mind.
Best wishes,
J_P