[NAGDU] Using positive reinforcement instead of punishment

Danielle Sykora dsykora29 at gmail.com
Tue Dec 19 15:10:10 UTC 2017


You've gotten some good suggestions already. Here are some ways I have
reduced leash corrections while working with my dog, a lot of which
are similar to Tami.

When my dog makes a work error and I am able to rework it, I simply
stop, turn around and walk back a few steps, and rework. I don't leash
correct, because the act of stopping and reworking the obstacle tells
my dog he has done something wrong. For very minor errors or things I
can not rework at the time, I use the phrase "watch it". Said in a
negative tone it is a correction that I use exclusively with tasks
(obstacle avoidance and stopping at changes in elevation); said in a
warning tone, it tells my dog that he is starting to make a mistake or
that I know something ahead is tricky and needs to be approached very
carefully. If you find you need to rework something more than a time
or two, then it most likely means that either the dog is distracted by
something in the environment or the dog doesn't fully understand what
you are asking.

For distractions, it really depends on what the dog responds best to
and what type of distraction it is. If at all possible, I try to
minimize how much my dog can be distracted in the first place. For
example, Thai struggles with sniffing and scavenging, so he wears a
Halti frequently while working. This allows me to feel his head
movements more easily, thus allowing me to intercept the distraction
sooner. If my dog does get distracted, typically I start with a verbal
reminder like "leave it" or something said in a firm tone--the word
doesn't really matter. Then I give a directional leash cue, or put
very light pressure on the collar away from the distraction (E.G.
light pressure to the right if the dog is looking at something on the
left). Some other ways to handle distractions are to stop and do some
obedience, or to turn around and then  walk past the distraction
again. I do still use leash corrections sometimes, when other methods
do not work and I know the correction will actually have an effect.

I use the clicker for teaching new behaviors and strengthening
behaviors that my dog is struggling with. I give treats every time I
click, but I only give treats randomly otherwise.  Treats are a
fantastic training tool, but something people frequently don't realize
is that treats are just one tool in the positive reinforcement
toolbox.

Danielle and Thai


On 12/19/17, sonfire11--- via NAGDU <nagdu at nfbnet.org> wrote:
> You can do many things with clicker training. Recently, I clicker trained
> Alec to take me across a very busy and distraction prone parking lot. He
> had
> to hit a specific point on the other side, so using the clicker to help
> pattern him helped. Now, he takes me to that point without fail. Another
> good use for clicker training is to build confidence during a route. Alec
> was a little hesitant during a 3 mile route we take every other day. Using
> the clicker helped him remember what I was asking him to do.
> When you rework something like missed obstacles, it is considered a work
> error. Every dog is different, but the normal process is to correct, show
> him the error, then rework. Seeing Eye will show you how to do these.
> However, you must learn how your dog responds to corrections/praise during
> work errors, then act accordingly. For instance, Alec bumps me into a pole.
> First, he gets a light leash correction. I then sit him down next to the
> pole, tap it and tell him fui in a displeasing tone. After that, we rework
> the problem by going about 20 feet back down the street and walk through
> the
> problem spot again. Usually, the dog will get it on the first time through.
> However, it can take up to 3 or so times.
> I don't have a disproval marker during clicker training. The normal leave
> it, fui, leash corrections, and control methods apply. Although, I might
> have to retrain him on a code word for "let's go now" because anyone who
> says the words "ok" and "ready" in the same phrase clues him into worktime
> and he stands up and gets annoyed when we don't leave right away.
>
> -----Original Message-----
> From: NAGDU [mailto:nagdu-bounces at nfbnet.org] On Behalf Of Yiska via NAGDU
> Sent: Monday, December 18, 2017 5:23 PM
> To: NAGDU Mailing List, the National Association of Guide Dog Users
> <nagdu at nfbnet.org>
> Cc: Yiska <ichoosechrist2 at gmail.com>
> Subject: [NAGDU] Using positive reinforcement instead of punishment
>
> Hello,
>
> I am wondering about clicker training and positive reinforcement when it
> comes to guide work. How do you get a dog to do the right task after
> messing
> up. For example the dog misses an overhead obstacle. Normally one would
> leash correct and then rework. How would you rework if the dog does not get
> it the first or fourth time? Normally you would increase corrections. Also
> for those who clicker train do you have a disapproval marker like no? I am
> trying to use much more positive training but struggled with my last dog.
> From what I have learned from youtube there are lots of basic tasks for
> clicker but not much past that.
>
> I get my dog in January from the seeing eye and am very excited.
> Yiska
> _______________________________________________
> NAGDU mailing list
> NAGDU at nfbnet.org
> http://nfbnet.org/mailman/listinfo/nagdu_nfbnet.org
> To unsubscribe, change your list options or get your account info for
> NAGDU:
> http://nfbnet.org/mailman/options/nagdu_nfbnet.org/sonfire11%40gmail.com
>
>
> _______________________________________________
> NAGDU mailing list
> NAGDU at nfbnet.org
> http://nfbnet.org/mailman/listinfo/nagdu_nfbnet.org
> To unsubscribe, change your list options or get your account info for
> NAGDU:
> http://nfbnet.org/mailman/options/nagdu_nfbnet.org/dsykora29%40gmail.com
>




More information about the NAGDU mailing list