At long last, a review from an authority in the detector-K9 community (and i really do acknowledge the fact that he is an authority!). The following just got posted by Kevin Sheldahl.

http://www.leerburg.com/scentlogix.htm

I will like to know what the forum thinks about Kevin's report and i will be happy to entertain any questions.

 Quote: Kevin Sheldahl

I am looking forward to the chance to put his products to the test in double blind tests with a number of dogs. If he does not object I will report the results.

I'll wait for the double blind test.

DFrost

 Quote: David C.Frost

I'll wait for the double blind test.

DFrost

David, funny you went to the beginning of these posts to dig up Kevin's initial comments about what he intended to do with ScentLogix. I think that Kevin has done a stellar job testing these products and writing a comprehensive report for the benefit of the whole community. Double blind tests have already been performed by very-good K9 trainers such as like Dan Reiter (the creator of the K9-BSD), Austin Moon (Deputy Sheriff in Virginia), Eric Crawford (Kansas), and Yann Lesourd (French Police, France). Now, knowing this, is there anything that you have to say? If not, I should assume that there will be some questions that you might want to ask. Alas, K9 trained on the real stuff are hitting on ScentLogix like crazy! Would you not want to know why?

You stated it was funny I had to go to the beginning for my comments relative the double blind study. Yes sir I did. In fact, it was the third post. Your post asking for comment was number one. My post with my comments was number 2. Kevins post was number three stating a double blind study was being conducted. Nobody has posted any results from double blind studies in this thread. If I've missed them, please direct to the study. Nearly this entire thread has been in the anticipation of Kevin's double blind study. The "study" posted by Deputy Moon was anything but double blind. When I think of double blind, at least the way it was conducted when I worked for SWRI, San Antonio is a bit different. How about first establishing the dogs proficiency on the actual explosives. Explosives used should be new to the dogs that are being tested and not previously handled by anyone involved in their training. Then under the same conditions the other material is tested. Of course the people involved in the product wouldn't even be at the test site. They shouldn't be the ones that hide the material, or the ones that score the results. Obviously that is a very basic description of a double blind study. You've read the study conducted by Kevin and my hats off to him for attempting the study. Be honest though Doc, if a student of yours was to write up a single blind study as the one presented, what grade would you have given it. Contrary to what you may think, I'm all in favor of something that will improve the quality of the explosives detector dogs that are working. I'm all in favor of anything that will make the job safer and easier. I'm also not a rookie in this business. I certainly don't know it all, but I have learned a few things. I'm not even trying to be argumentative. I am trying to make some sense out of this and frankly, I hope it works. Since I can't conduct the tests myself, even though you asked me if I was willing and I answered the affirmitive; I would be considerably more interested in the results of a true double blind study. Why not even have Auburn conduct the tests for you. They are fairly well renowned in the field of canine olfaction and have considerable testing to that end.

DFrost

Edited by David C.Frost (08/26/2009 06:31 AM)
Edit reason: additional remarks

 Quote: Austin Moon

I was at the NAPWDA workshop the master trainers put them in a regular problem along with explosives and narcotics. They ran the problem blind.

Is this the double blind study you were attributing to Deputy Moon?

In addition, I've not seen the results, nor was I aware of a double blind study conducted by Mr. Reiter or the French police.

DFrost

Let me answer a couple questions.

First, why use a single blind evaluation of the aides?? It is an industry standard. It is a familiar and often used training and evaluation technique, one that is widely accepted as a test of efficacy for a team.

Double blind tests are great, but it assumes that the evaluator is either involved in the system, helping, or in some other way is assisting the team that is if you assume the results would be different between the single blind test and the double blind.

David, do you have studies to demonstrate a significant difference between a single blind test of detector teams vs. a double blind test of detector teams?? If you do this brings into question virtually all qualification examinations.

Since you mention that the dogs should be pre tested, that is already done, every time they train, variation on the explosives is a ongoing part of training. The dogs were all pretested in the format used as well through regular certification examination conducted in much the same fashion as the evaluation was done. If anything familiarization with a specific kit would weigh against the tested aides not for them.

I certainly did not intend to produce a broad set of statistics for ScentLogix. I did want to start looking at these products for myself. I am leaning towards doing some more things with these products including more comparisons. It took me a while to decide what approach would answer the most questions about the aides with the teams at hand. Placing the aides into a situation where the handlers didn't need special coaching, and into an environment where the dogs had been proofed, trained, evaluated occasionally seemed the best approach. Since of course I was looking at operational efficacy here.

Frankly, I am not a compensated spokesperson for ScentLogix, I have no stake in their success except if the product produces an increase in the effectiveness of the teams I work with. That is yet to be seen and it will take time to see that regardless of the number of lab studies performed and their outcome.

If Mr. Frost isn't convinced...well frankly I just don't care, it was never my intent to convince anyone. It was my intent to look more closely at these aides, I do hope I answered some things and created more questions. Not simply about the aides we are discussing but about detection in general.

Kevin, I didn't request a double blind test. You stated you were going to conduct one. I'm out of the discussion.

DFrost

David,

I read your posts. I especially liked your point "A pseudo by any other names smells the same"

Spoken like a true old school trainer.

Frankly you probably should try this stuff before you bad mouth it. You just might find out that you were drop dead wrong. Wouldnt that be a surprise?

When trainers approach a new training concept with a closed mind or preconcieved ideas they show a certain level of competience.

In this specific case I was skeptical. I was also wrong just as you are wrong.

So to all new bomb dog trainers who find themseleves looking for training information I would recommend this product. I would also pass a caution that you always need to keep an open mind. Just about the time you think you know evevrything a new training concepts wanders in and challenges you and your old school training ideas.


Ed Frawley

Ed,

From my perspective, I've said everything in this discussion that I'm going to say. I would like to point out however, I've not "badmouthed" this product. I may be, in spite of my age, a bit more advanced than you know. At any rate, as I said before, I'm out of the discussion. Had I known that the only comments the Dr wanted was positive ones, I wouldn't have entered the discussion at all.

DFrost

David,
You are correct that I had planned to do a double blind study. But, I wanted to dovetail to the methodology the teams had been exposed to in testing, training and deploying.

My comment about not convincing you is a direct result of the insult you threw out about my article in a response to Adebimpe. It was not appreciated.
I was not commissioned to do a study on this subject, I am not student of Adebimpe's, and if you actually looked at the question I addressed in the article I think it did give a starting place in a broader evaluation of the product.
I for one feel that particularly in EDD training we as a community have not discovered all the problems in doing the training of these dogs.
Maybe, and if you read my article, it also says this, that maybe these products, and research into them can help us answer a large number of questions.
But, you choose to leave the discussion because you make the broad assumption, that it is not possible for a single blind test to yield any pertinent information?

Of course the classic reason to do double blind testing was the discovery of the placebo effect on patients. It necessitates a double blind examination.

I will also give you the possibility of a "Clever Hans" effect. A concern that is addressed in the use of scent ID dogs working a line up of scents by countries using this as corroborating evidence.

But, since the teams were not trained to work in a fashion conducive to this I chose to work with them in the fashion they train, test, and deploy. Is it perfect,no and I point that out in the article. Is it a good place to look at these products and see if they deserve a broader look at their application? I think the answer is yes.

If you are not convinced, jump in here and provide some actual data, some thing to contribute to this process that will take a long time in the private sector to work out.

I certainly don't want you to take your ball and bat and go home! Although I can do without the insults.