The seminar was set up as a series of
mini-lectures, and between each one, we were paired up with another
dog/handler team and several auditors. The working dogs would make a
plan and then train, while the observers would help with data
collection and then have a small group discussion on how things went,
what to change on the next round, etc.
As the title implies, Maisy and I had
one of the working spots. Although she was amazing, I was not.
Sigh. I was so frustrated by the tasks set to us. Truthfully, it had
very little to do with Maisy, the tasks, or the seminar. Remember:
the seminar happened back in early October, which was about two
months after I had separated from my (now ex-) husband. There were a
lot of adjustments to be made during that time, and honestly, I
probably wasn’t in the right mind-space to be training a dog.
Each day had a separate training task.
The first day had to do with targeting; to be honest, I forget
exactly what we were supposed to be doing now, but I remember trying
to get her back feet on a piece of carpet. Utter failure. (Short
legged dogs are hard!) The second day was a bit better; Kathy showed
us a commercial of a dog who took a chicken strip from his owner,
held it until given some sauce, then dipped it in the sauce and ate
it. We were to work on this in whatever form we wanted. This went
better; I worked on having Maisy hold a dog biscuit without eating
it.
Here are some of the things that really
hit home with me. I’m not sure that these were new things,
exactly, but they were things that I understood in a new way after
the seminar.
Train in Short Sets
I’ve heard this about a billion times
before, but somehow I never do it. You get on a roll and don’t want
to stop, or just lose track of time. We got around that in the
seminar by deciding how long we were going to train (usually between
30 and 90 seconds), and then having someone time us. And even then,
some of us had a tendency to do “just one more rep.”
Training in short sets is important,
though. Part of this is because you can only concentrate for so long;
good dog training requires you to both SEE what the dog is doing and
then to MARK it with good timing. This is hard work, and something
people often underestimate. Taking regular and frequent breaks allows
you to rest between sets, and thus keep your eyes (and thumb!) fresh.
Training in short sets also allows you
to think in between. This is incredibly important when it comes to…
Criteria and When to Change It
First, keep your criteria the same
throughout each set. Since your sets are going to be short, it should
be easier to resist the temptation to change it midstream. The
problem with trying to change it on the fly is that it tends to throw
off your timing. In addition to seeing and marking, you’re now
analyzing performance and deciding what to click. I don’t know
about you, but my brain just cannot do that much at once!
Raising criteria can be done based on
the rate of reinforcement (Kathy recommends doing so when you get
into the double digits per minute), the percentage correct (most
people do so at around 80%), or by the density of reinforcement. This
last one comes in to play with duration behaviors where you can’t
get reinforcement rates in the double digits. Instead of giving one
treat per click, you give a larger amount so that the dog would end
up with roughly the same amount per minute.
Once you’re ready to raise criteria,
do so by looking at the responses you are currently accepting. You’ll
have “technically meets expectations, but nothing special” on one
end of the continuum, “average responses” in the middle, and
“outstanding!” at the other end. Raise criteria by clicking only
average-to-outstanding behaviors, and drop the “technically meets
standards.”
If something unexpected happens- either
super good or super bad- STOP. End the set and think because…
Planning is Important
Perhaps the most important thing I
learned was how specific you should be in order to train well. Many
of us had the tendency to state “I am going to work on duration
now,” but that really doesn’t say a whole lot. Instead, Kathy
challenged us to be very specific in what we mean by duration. Much
better to say “I’m going to click when she holds the dog biscuit
centered and lengthwise in her mouth without mouthing it for a
minimum of two seconds.”
In other words, know what you’re
going to click. There should be no guesswork about whether or not
something meets criteria.
So, even though I felt sad and
frustrated the entire weekend, I learned enough to make it
worthwhile. I am fortunate to have a dog who will keep working for me
even when I’m not feeling up to par. So, it was a great seminar,
and I’m so glad I got to have a working spot with Kathy!
1 comment:
The section on criteria seem relevant to an issue I'm having with Cohen's handstand. We've been working on it for ages, and we keep hitting plateaus. I feel like I'm either not raising criteria steadily enough, or I'm changing my criteria as I try to troubleshoot our problems, which results in no real progress. It can be frustrating! But it's a phenomenal behaviour for forcing me to work on my shaping prowess. Will try to modify my approach by keeping sessions shorter and criteria more consistent.
Post a Comment