WORKSHOP: Conclusions

CONCLUSION:  In lieu of fresh sample, which we have never had the opportunity to test, scent training should continue, focusing on search duration, building test stamina and the ability to easily transfer between different scents.

CONCLUSION:  In training, stop the exercise after the first false positive.  When you return to training, begin where the dog was successfully completing the task.

CONCLUSION:  Conduct training on the scent wheel, but the scent option can be placed in a variety of objects: plastic test tubes, small metal containers, film cannisters, etcl, as well as glass shakers.  For variety, conduct training on other types of stationary testing devices that give the option of selecting the "hot" sample from dummies.

CONCLUSION: During training using other scents, each container should hold either cotton rounds or gauze squares, including the container holding the target scent.

WORKSHOP: Discussion - Final Cold Test?

Is there empirical value in conducting a final cold test?

• only when BCM?
• on every test, even TC?
• with the same live sample?
• with a fresh sample?
• Same sample/dummies in clean shakers?  cans? different device?
• Single set?
• Single session?

Is there value in changing the testing equipment within a single test?

• Change shakers? cans? 
• daily?  between sessions?

There are two things at play during all of our resting

• Sample viability
• Test validity

Testing equipment should contain 2 sets each of both cans and shakers.

Syn re-do clothing test; clean equipment, same sample.

WORKSHOP: Problem Solving Discussion

TRAINER PROBLEM:  Overcoming the "trainer ego" - the urge to try it one more time.

"Once is a fluke, twice is a pattern."

In SAP training, we will assume that one False Positive is an emerging pattern, and train towards accuracy.

1

3

Fresh vs Frozen
We don't know what it is that the dog is indicating.
We can no longer test old sample.  We must locate sources of fresh sample for testing.
In the interim, we will test for testing stamina and duration of search.

4

Types of Scent

5

Medications
Need specific information as to the medications taken daily
as well as medications given at the time of the seizure.
To insure that dog is not indicating to a medication.

6

7 

Nothing There indicator
We don't need a Nothing There indicator.
We will be training with scents that we know are there.
I want my dog to have duration of search and keep going until he identifies the scent.
One False Positive and we end the session w/no reward.

8

Scent Wheel Nothing There

10

Keep Working/Build faith in the dog that something is always there.
To eliminate guessing.

11

Future Training

For training purposes, we will use a variety of identifiable scents that we know are there.

And, finally, all testing containers will hold either gauze or cotton pads, including the container with the scent (tea or coffee or spice, etc)

13

Donor Concerns
We are looking for day old scent for testing purposes.
Shipping scent between countries is no longer acceptable.
Ideally, a donor could provide saliva sample, skin swipe sample, clothing sample.
Additionally we need non-event samples of all three for Test D.
Sue will pursue the pre-ictal donor for sample.
We must develop professional contacts.

WORKSHOP: Discussion

The SAP began with Lily the 25th of March, 2012.  This week the three participants met to evaluate testing procedures and discuss methods of improvement.

We held a two day workshop attended by the two primary SAP training teams, Jan and Nad, Sue and Syn.  The workshop was held at Lynn's home - important to note because the testing that we did was in Lynn's living room (distraction) with all attendees present (distraction).

We tested a "new to us" sample - unadulterated clothing provided by Donor 3.  The clothing had not been refrigerated or frozen and was 47 days old at first testing.  Videos of each dog working with this sample can be found on their individual pages (Dog 2 and Dog 3).  All of the testing done by Dog 2 was done in Lynn's living room.  Dog 3 completed some testing at Lynn's and some at home, in his normal testing environment.

Two important considerations emerged from the workshop: 1) what are we actually testing? (what is it that the dogs are indicating?) and 2) ways of improving our testing protocol to better answer that question.  One new element was added to the testing protocol: the ritual.

Ritual: a single set of trainer's choice may be introduced once per session.  If used, the ritual should be documented in the test record.

DISCUSSION TOPICS:

What is it that the dogs are indicating?

A discussion of ways to improve the testing protocol.

How does "scent fatigue" impact upon the testing?

Measuring sample strength vs dog interest.

Training continuity.

Testing continuity

Peripheral consideration: Although Lily did not participate in the testing, she was in the environment and exposed to live seizure scent.  I will notate the dates on her medical calendar.


48 minute discussion divided into 5 parts (unedited; no video)

• TRAINING protocol:  (reference)
• Before the testing of live SS begins, the dog must first be taught how to play the "find the scent" game.  And, be introduced to the Scent Wheel. 
• The criteria of switching from training mode to testing mode should be 
• eager enthusiasm for the game; 
• being able to find 
• alert to 3 different scents on the scent wheel
• clear and consistent alerts.
• During this training process it is assumed that distractions (as defined above) will be used so that the dog understands all the parameters of the game.  
• A and B are NOT training; they introduce the scent to the dog
• A and B will be done whenever a new sample is to be tested
• What is a RITUAL?
• ritual = 10 reps of trainer's choice (using seizure scent)
• When is a ritual used?
• maximum: 1 ritual per session as needed
• should be documented in record when it occurs

• If success rate of any set drops below 90%
• you may introduce a single ritual before resuming test
• if success rate then drops below 90% again
• conclude the test (TC)
• retrain at trainer's discretion
• begin test @ INTRO A/B
• Test may be repeated only ONCE before it is concluded that the Baseline Criteria has not been met
• Discussion of Baseline Criteria:  
• This discussion was held w/Steve White and Dr Dana Sue Hardin
• 90%+ per set or 90%+ session average
• because they each felt the BC should be met EACH session, the current protocol was established
• Here we are discussing if the discrepancy occurs in the final set, is that SAMPLE ERROR or DOG ERROR?
• Are we testing the dog or are we testing the sample?
• Testing stamina is taught first in the initial TEACHING of the "find the scent" game
• Through experience and practice
• We must do the testing in the dog's optimum condition testing environment
• Free from distractions
• Where the environment as well as the presence of the scent wheel tell the dog that the game is on
• NMT one session at any given time

• Dog error or Sample error?
• significance of WHEN in the sequence of reps the FPs (false positives) occur
• repeating FP teaches the dog to do it wrong 
• in any set, if dog misses TWO, stop the test  (TC: Test Concluded)
• Trainer has the option 
• to re-teach (using the testing sample)
• to begin again at Intro A/B
• to do a testing ritual and begin test again
• During the repetition of the test, if the dog misses TWO at any point, the test is ended  (TC) because the Baseline Criteria has not been met
• PROTOCOL for Ending a Test
  When do we conclude a test?
  
  • When the test meets the Baseline Criteria, of course.
  • Because we don't begin having the dog testing live sample until we are confident that she has a thorough understanding of all tests performed on the test wheel, we will consider the scent to not be viable when the dog has completed 3 subsequent sets with an AB of 80% success rate or lower.
  Whenever a test is concluded (TC) without meeting the Baseline Criteria, using the current Sample we will re-introduce the test in a teaching model.  If, after completion of the entire scope of re-teaching the test, the SS appears to have a viable, detectable scent, the test may then be repeated.
  
  
  Repeated tests will be indicated within parenthesis.  i.e.  SS Test I:C (2)
  
  
  Any test may be repeated once per Sample before it is considered to be not viable for training purposes.
• Clarification of terms:
• training
• testing
• ritual
• repetition
• 10 reps = 1 set
• set
• 3 sets = 1 session
• session
• 2 sessions = 1 test  (BCM: Baseline Criteria Met)
• tests per sample
• Intro A/B
• Tests C/D

• Scheduling testing:
• sets and/or sessions must be consecutive (nothing in between)
• AGE of SAMPLE
• The longer between sets and/or sessions, the more the sample changes
• age of sample
• storage of sample
• Each session should be completed within 24 hours
• If I have 24 hours to complete a session, then Trainer Judgement can dictate WHEN the sets are tested
• CONDITION of SAMPLE
• All sample tested at room temp
• MOST (but not all) sample refrigerated between tests

• Original BC remains the same
• We have added the option of a single training RITUAL
• When we are testing, the dog's nose does the talking.
• We should not correct the dog for a FP (false positive)
• When the dog gives a FP, she will be reset and then sent to the wheel again
• "It's my job to show up with a dog who is ready to do the job."

WORKSHOP: Success....and then Conundrum

Having met Baseline Criteria on the test of 49 day old clothing sample, we now presented Syn with a new sample of the same clothing, placed in a fresh shaker on the wheel with the same dummies used in Test D.

Syn was unable to indicate the new sample.

So we then began experimenting.......

1

COMMENT:

2

COMMENT:

3

COMMENT:

4

COMMENT:

5

COMMENT:

6

COMMENT:

7

COMMENTS:

8

COMMENTS:

9

COMMENTS:

10

COMMENTS:

11