mullenix - cat
telling us her history. we all have a version of this story, i do not care. sounds like she was apprenticed with a choke chain guy. she rejected that. and more crap about herself that you could have read in the bio, if you read it. god this is annoying.
now she’s giving case history. still annoying. i do not care how you got into it.
defining cat. which is only useful for fear driven reactivity.
uh oh, she still uses chokes/prongs. no wonder she’s ok with cat. she couldn’t get a license with that kind of bs. very unclear what she does and doesn't use actually.
going on and on about r-. this is annoying too. if you don’t know what it is, then you maybe shouldn’t be a trainer...
defining reactivity. holy shit. if you don’t know what this is, you really, really should not be a dog trainer. or, you know, at a conference for them...
shows a vid of some lady asking the dog if they want a bath and he is reactive. she asks like 6 times, though it’s pretty clear at the first rep the dog is less than thrilled
she’s happily telling the truth - environmental factors matter. yes, yes they do. this. is. boring.
consequences of anxiety range from unhappy pooch to death, and all in between.
consequences of this whole deal for people - from unhappy people to lawsuits to quality of life suckage.
consequences to society - these are interesting - noise, bsl, other $ problems.
what else to do
- cc
- d/cc
- dri
what cat does is change emotion via behavior. cc does the reverse.
cat vs d/cc -
(as ian says, r- is the most efficient quadrant. and he isn’t wrong, but that doesn't mean it's a good idea.)
claims cat is no more aversive than d/cc as both expose dog to aversive. d/cc dog must always be kept under threshold, so no learning happens, dog must react to learn cat requires the stimulus to be aversive, while d/cc strives to keep dog calm 100%. (never does, though)
claim cat is more effective - contingency is more functionally relevant - escape/avoid is big reinforcer through d/cc underlying premise is different, escape is not motivationalin that situation
cat is faster! no, they’re the same!
how to study this ethically? of course there is no way... each dog is different
cat doesn’t generalize, strongly defined by stimulus, d/cc addresses underlying motivation and thus generalizes. that’s bullshit, though.
cat works for all dogs that fit conditions for treatment. different training is used for dogs not meeting these conditions. duh?
(dogs who don’t recover well have *broken systems* and need to be looked at differently in order to fix.)
cat is not dependent on skill or knowledge. yeesh. even beginners get results, apparently. sigh. counterpoint is that that’s bs, of course trainer needs skills. uh, yeah.
cat isn’t flooding, dogs stay under threshold and dogs have choices. flooding is obviously unethical, and not what is happening in cat. fine. if you screw it up, you’re flooding.
it has taken 30 minutes!!!! for her to get to the point of her fucking talk! gah!
references rosales ruiz and snider, their insanely long and boring ass dvd, and pat miller’s article (she thinks the generalization thing is an issue), which is cool.
why it’s called constructional - new behavior is required for things to change for the dog, creating more options. my evaluation of the thing was that what dog is doing with body changes, emotions change, solely because dog is limbic system with legs, and physical process are hardwired to emotion.
6 steps
- arrange environment
- know threshold
- boundary approach
- switch over behavior
- interaction
- generalization
- secure subject dog, less mobility preferred, stationary during protocol
- position owner at other end of leash, heh.
- figure out a marking system.
- mark a safe boundary
- set up at location where there is reactivity.
- stim dog brought into dog’s awareness, no real reactivity
- once id’d, estimate locality of sensitization point and then stay under that.
stim dog approaches under threshold marker
if you go over threshold, you have to wait the dog out and let the dog react. and then hit yourself with a brick cause you fucked up. wait til dog does something else
alternative behavior reinforced by dog moving away. because, you know, that’s what this is about. grrr.
if subject reacts while dog leaves, dog returns to previous threshold... she didn’t do this, she had the dog stop.
if subject reacts prior to stim reaching marker, walk 1 or more steps up then wait for new behavior until going away.
both of those are uncomfortable for most r+ folks. with good reason.
as training continues, aversive stim should become less aversive, subject should start to express new behaviors, incl affiliative ones, and you work to shape the good ones
she didn’t go through dog in conflict point. :-(
eventually allow interaction, not appropriate for all dogs, keep safety in mind
interaction allows for friendliness, pass by and parallel walking is useful. light play is last on the list. retreats are still happening, though. even without interaction you can get results.
generalization - yeah. we know what that is. uh. she says “this is where we introduce a 2nd dog” they did one dog at a time going through whole process each time, also can do multiple dogs in parallel. practicing in different locations important. goal is friendly behavior to majority of dogs
train generalization
set up in actual location of reactivity, train where it makes sense to train, duh. train with the actual stimuli. this seems like intuitive stuff there.
6hr session. ouch. first thing after cat seminar for her. she said it sucked. client was happy though. next day 4hrs. yeesh. client was still happy. next day, 1.5hrs... started to try with other dogs... she read from the email lists. yay for her. seriously, though, if we get to the end of this talk and she tells us to do this like the bat people are doing, i’m going to hurt someone.
study - holy shit, finally, oh my god. all breeds, s/n, 1-7yrs, no people rx, dog rx, less that 50 ft, at least 5 previous instances of rxn to dogs on leash. worked in a warehouse in portland. outside there, it rains.
contract fr subjects had informed consent, liability, permission to film, opt-out. they were not paid for any of the stuff. studies will not make you $!
warehouse poorly lit, carpeted, grody, covered windows for no outside stim, there was still some noise. marked out distances on the floor with tape measure
stim dogs - m/l sized dogs, 4-12yrs, nonreactive at distance, some good at dog-dog interactions, mostly f/s dogs.
sessions 1-4hr/day/session, several sessions a week, no longer than a week between sessions. originially, cat was all day affair, working towards interactions. it seems like a *bad idea* to do that to a dog. or yourself.
owners asked not to expose dogs to stimuli outside of sessions.
procedure - set threshold, stimulus dog approached dog, distance decreased gradually
no new behavior within 30sec = shutdown, usually looked like dog going to sleep. if shutdown, roles reverse and stimulus dog stationary while sub dog moves. *this changes more things than they think.*
owners not interacted with dogs during trials.
as stim dogs neared, they’d curve on boundary, to banana and then they’d parallel walk.
next they’d intro new stim dog and repeat.
time did not allow for significant generalization work.
after 10 min of work, dogs get 2 min breaks. this continued throughout training work. they’d go back to their own corners and get loves, snacks and water.
set threshold at precursors, not rxns
the approach - stim dogs moves towards sub dogs until pre-rxns.
alternative behaviors are calming signals and relaxation, all prosocials are rewarded with retreat and indicate switch over.
shows vids of alt beh
when to decrease distance - changeable parameters are distance, response variability, duration. you are supposed to construct a repertoire of behavior, not just one. means that this is a creative teaching process, which is kinda good.
what we measured - training time per dog was hoped to be 12hrs
change in dist between dogs per trial
and then real life happened, and the ideals were blown out of the water
dogs who in spite of pre-screening had behaviors that made them unsuitable
rx just at home, rx prev but not now, rx to people in test situations, rx too great at max dist, escalation of rx in the home
difficult keeping methodology due to different responses to training. attrition due to injury and adoption. even though it would be nice to adapt, they had to stick to the plan.
sub 1 - rx all non household dogs bigger than him. barks. wow a chi that barks. no way.
day 1 of like 50 trials, shows progress from 30ft-2ft. pretty cool.
she says the big mistake was moving distance closer. they were being pushy, and that was a problem. ending with success and pushing for less what what they should’ve done. nice graphs!
sub 2 5yo jrt m/n, rx to all other dogs. post neuter was rx to all dogs, no bite history.
more video! i love this woman!
jrt’s baseline is 25ft. started at 50ft. 55 trials later, 6ft. 130 trials - 4 ft. later on you see more relaxed body movement and more behavior. on day 3 they reached interaction. amazing! jrt was bouncy and joyful. there was some human error involved, but they eventually fixed it all up. jrt graphs are interesting too. hasn’t generalized, is still in progress.
this not generalizing thing is a problem, though.
dog - 3 apbt mix. rx since 6wks! that is insane. has had incidents with household dogs and cats. 14 dogs in 2 houses. illegal at best. owners have tried clicker training, dog does big damage.
a rednose pit, actually, looks like was abandoned, potentially fighting lines, along with early weaning... bad news.
she notes she is focusing on mistake and trying to express that we should not make these mistakes.
started at 36 ft. squeaky dog. and yes video!
she sez: when you are doing this right, it’s really boring.
reason the cat people don’t use clicker is it’s considered a stimulus condition. is problemy for owner compliance in original protocol
big point: set foundation of many behaviors, screw distances.
results - despite history and advice not to interact w/other dogs, still interacts and just did some big damage recently.
dog 4 - m/n gsd adopted @14wks, rx to all dogs except one in household. no bite history, no medical issues or meds (true for all dogs)
videos are interesting.
when you see the switch overs start, you see a lot of conflict - also you’ve been staring at the dog for hrs, and figuring out the approximations can be difficult over long stretches. difficult to change what you use to mark beh.
shut down occurs for a combination of reasons, stress plus time, hefty cortisol and tired = shutdown. reversal of protocol seems to change behavior a great deal, but it seems to me that you should start in reversed protocol in order to get better results. that’s why bat seems like the better choice.
skipping ahead
oops:
- accidentally going over threshold - this is difficult to avoid as many behaviors can occur in a short time, slow things down, dog fixation can be dog thinking about things.
- shut downs due to over multiple threshold exposures - so back off and slow down! have to figure out the reality of the dog recovering from stress. she accidentally shaped shutdowns by shaping toward laying down and relaxing. oops! intertrial playtime important for avoiding that. do other things!
- stim dog fatigue - look out for it, can result in different behaviors due to the inherent training of the protocol. her stim dogs tensed up in the environment... that does not sound ok to me.....
- environmental factors - outside folks, outside distractions. superstitious fear-based learning can sensitize dogs to external stimuli
she doesn’t use the cat protocol with clients, but does use some of the techniques. she thinks it’s valid and can apply to some dogs, but she thinks it’s easier to go over threshold as we are using r- and there’s a learning curve, and we are not used to this.
she tells us about bat - ahimsa.org, check out the blog, it’s pretty fucking awesome. it is more user friendly, it is better for the dogs, it is better for the stim dogs.
conclusions - d/cc and cat are similarly time intensive. cat can help dogs who can’t use food or toys, who are frequently exposed to negative environs that reinforce behavior.