In this episode of K9 Conservationists, Kayla speaks with Lucia Lazarowski about long-term working dog success.
Science Highlight: None
Links Mentioned in the Episode:
Behavioral Characteristics Associated with Detection Dog Success
Olfactory Research in Dogs by Lucia
Where to find Lucia: Website | Research Papers
You can support the K9 Conservationists Podcast by joining our Patreon at patreon.com/k9conservationists.
K9 Conservationists Website | Course Waitlist | Merch | Support Our Work | Facebook | Instagram | TikTok
Transcript (AI-Generated)
Kayla Fratt 00:09
Hello and welcome to the K9Conservationists podcast, where we are positively obsessed with conservation detection dogs. Join us every other week to discuss detection, training, welfare, conservation, biology and everything in between. I’m Kayla Fratt, one of the cofounders of K9Conservationists, where we train dogs to detect data for land managers, researchers, agencies and NGOs. Today, I have the joy of talking to Lucia Lazarowski from Auburn University about kind of long term predictors of working dogs success. This is a very recent paper that has just come out, we’re super excited to talk about it. So we’re not going to do a science highlight today, because the entire episode is a science highlight. We also don’t have any new reviews to highlight. So we’ll just jump right into it with Lucia. Welcome to the podcast. We’re so excited to have you!
Lucia Lazarowski 00:55
Thank you so much for having me. I’m excited to be here.
Kayla Fratt 00:58
Yeah, so why don’t you start out with giving us a very brief overview, I think we’ve done it as a science highlight in the past, kind of the original study where you had a cohort of puppies, and you were doing all these tests at different ages. Talk us through what that looks like. And then how that led into this, this more recent study?
Lucia Lazarowski 01:15
Sure. So um, that original study, the purpose was, we it’s our program at Auburn, just to give a brief overview is a breeding program for dogs, primarily for explosive detection. But generally for different detection roles. And so we we breed our dogs, here, they go through early development and socialization and training, they’re part of the program until about a year old, at which point they are presented to various customers for single purpose detection. And then hopefully, they go out in the field. And so part of our process is that we conduct behavioral assessments at different ages to kind of track the puppies progression, get ideas of which dogs seem to have potential, which don’t. And we had just a sort of internal behavioral assessment that we had been using for many years.
Lucia Lazarowski 02:08
So we got to the point where we decided it was a good time to try and validate our tests. So anytime you have a novel, behavioral test, it’s a good idea to validate what you’re doing to essentially, see if what you think you’re measuring is what you’re really measuring, is it reliable, over time, reliable over context. And so our goal was to take our existing tests, run a validation on it, and make sure that we were getting, you know, reliable and valid results. And so we did that by taking a just a cohort of puppies within our program, during a particular year, about 60 Puppies, and looking at their performance on these behavioral test at different time points, so three months, five months, 10 months, and then around 12 to 14 months at the end of training, and then seeing if how they performed on that test was predictive of their selection for a detection role.
Lucia Lazarowski 03:06
And so in the field of behavioral assessments, there’s lots of different ways that you can look at the validity of a test. And so we looked at kind of three main things. One was, is the test reliable? And you can look at reliability in different ways. What that generally means, do you get the same results? Across at, you know, when you perform the test at different contexts? Or does is it repeatable? Can different people perform the same tests and get the same result? So for that goal, we that’s really getting to whether our protocol for running the test was objective, and reliable. And if the definitions that we were using were objective, and so one way to do that is, Can different people take the same list of definitions and of tests that we’re using? And can they come up when they look at the same dog and get the same results so that your, your test is as objective as possible, which is important with behavioral tests, which usually do have a lot of subjectivity.
Lucia Lazarowski 04:08
To accomplish that goal, we had multiple different people, different trainers, as well as research assistants observe and score the same dog using the same definitions but independently assessed the dog. And mostly they matched in their scoring. So we found for that, that we did have really good agreement between different people, which tells us that our protocol for running the tests and for scoring the dogs was pretty, pretty consistent across different people. We also looked at convergent validity. So this is looking at we all you know, we assume that we design a test to measure the thing that we’re interested in measuring. But you it’s important to compare that to some other measure to tell you if what you’re measuring is what you think you’re measuring and that can be difficult to do, because you have to compare it to some other methods, some other gold standard. And there’s currently not a whole lot out there for detection dog behavioral assessments.
Lucia Lazarowski 05:11
So we use the C-BARQ, which is the the canine behavior assessment, research questionnaire, something along those lines. It was developed by Dr. Serpell at the University of Pennsylvania. And it is used a lot. It’s basically a survey that assesses different aspects of dog’s temperament. And it’s used by lots of different research programs, mainly for pet dogs, it’s also been used for assistance in guide dogs. And although it’s not the time, that was not meant as an assessment of detection, dog behavior, there are things that it measures that aren’t important for detection dogs like fearfulness and boldness and trainability that we were trying to measure in our test.
Lucia Lazarowski 05:55
And what we did was we compared how dogs scored on our behavioral test to, we had the trainers fill out that questionnaire that was asking questions about those similar behaviors, and looked at how that compared and again, we found consistency between how the trainers were ranking their dogs on this previously validated questionnaire, and how they were scoring them on the actual behavioral test. So that was for convergent validity. And then finally, the last one was predictive validity, which was probably the most important for us. And that’s whether the way that the dog behaves on our test is predictive of how the dog is going to perform in the real world, or how they’re going to be their likelihood of being selected for an operational role.
Lucia Lazarowski 06:41
And we did find that the performance on the test, generally, the better the dog scored on different items, the more likely it was to be selected around a year old for a detection role. So that was the original study was just to validate, okay, our test is measuring what we want to be measuring. And it’s reliable, and it’s consistent. And it’s, you know, a useful tool for us to gather data on our puppies and make decisions about their training and their outcomes.
Kayla Fratt 07:10
Yeah, pause here for a second because I think, you know, when we’ve done a couple different episodes about, you know, Puppy assessments and puppy selection and dog selection, and, you know, I’ve done these before, I used to host a podcast called pandemic puppy podcast, that was when I was raising my puppy, niffler. And we just did, like, you know, crate training and like all of these, like basic puppy things, but we did a puppy assessment episode as well. And I think, you know, as a potential buyer, you like, for us, we think about like, is this test predictive, as, like, kind of the only thing we’re really, really worried about in a lot of cases. And I love that you laid out, you know, it’s not just whether or not this test predicts it, but do multiple people get the same result over and over, because, you know, if I could see situations where if you’ve got someone who’s just raised hundreds and hundreds of dogs, and they’ve got this test in front of them, and they know, that confidence and boldness are supposed to be correlated, and then, but they’ve just got an eye for like, what you know, whatever it is, that is going on with that puppy that actually is not necessarily objectively in that like syllabus.
Kayla Fratt 08:25
You know, you could just end up with all sorts of screw results where like one, one judge, so to speak, would be really good at predicting it, and someone else who has much less experience would be a far less experienced. So yeah, I just I want to call out, like how complicated this is, and how hard this is. And I think one of the things we talked about, in one of the episodes we’ve done on this in the past is just, you know, to me, sometimes it feels like when we’re looking at these assessments of like, like niffler, my puppy, who’s three years old now, but when I first got him, they did their temperament tests and assessments at seven weeks. And it’s like trying to take a room of like three year olds and decide which one you want to hire as a lawyer, like, you know, it’s way you know, the LSAT isn’t even that predictive, but at least at that point, you’re so many more years into the situation and so much closer to actually being a lawyer. And, you know, particularly with puppies, you know, we want something that’s going to be productive before they leave mom, because that’s when we decide to take them home in most situations, you know, you at Auburn and you know, guide dog programs, it’s a little different, because you’re holding on to those dogs for a year or two. But you know, I wasn’t gonna ask the breeder to hold on to Niffler for two years.
Lucia Lazarowski 09:44
Yeah, that’s generally what what we found and what others have found is that the, the older the puppy is when you do the assessment, the the higher you know, the predictability is going to be obviously because there’s a lot that changes over there the first year Five, but the ultimate goal is the earlier the better, because then you can, you can save time and resources, if you have a puppy that, you know, is just strong indicators that it’s not going to work out and needs a career change or making alterations to your trading plans. So, you know, we did find that the older, you know, 510 12 months, the closer you get to the end, obviously, the stronger the predictive ability is, but we did find indicators at three months, that data that did show how likely a dog would be to be selected.
Kayla Fratt 10:33
Yeah, yeah. Which is, you know, and that’s just kind of how like, almost all predictions get like, I’m a little bit of a political polling nerd. I listen to the 538 podcast, and they do all sorts of polling. And, you know, they just had an episode a couple of weeks ago about like, how predictive our polls about the presidential election a year out before the election, and they were like, they’re not, like 25 point swing, which, like, you know, we, you know, you’re looking at something and it says that, so such and such candidate is 10 points up. But if you’re a year out that actually has ZERO predictive value, and, you know, we’re all anxious about stuff or excited or whatever, about something that just really doesn’t have predictive value. And like, I find listening to that podcast, like, when I sat down in my first graduate level stats class, I was like, wow, listening to this podcast for the last couple of years really has me like understanding a lot of the like, limitations better, like, you know, the math is still really hard.
Kayla Fratt 11:29
But anyway, that’s a side note. So now, okay, so we’ve got this, this earlier test, or this earlier paper, where we looked at puppies that, you know, up into up until placement, basically, and we found what was predictive up until placement. But now the question is, okay, but what if the dog washes out of training? Or what out of placement or what if, you know, like you mentioned, you might have a measure that you catch in a puppy that’s three months old in the new adjustment program, and then that could help them succeed. But the same thing can happen with like a 12 month old dog, you could hold them back and work on something, and then they could still get placed. So that’s basically the question for this next paper, right?
Lucia Lazarowski 12:09
Yeah. So generally, most, most research that’s out there on working dog selection tests, is looking at, you know, how how different assessments at different ages predict what can be called, you know, graduation success, or training success, basically, their placement at the completion of training, if they get selected for a role placed in an operational role, but there’s kind of this black box after that, where there’s generally not a lot of feedback once the dog is in the field of how the dog is performing.
Lucia Lazarowski 12:42
And there’s a lot that happened after the dog is selected changes that can take place, because usually, a lot of cases, they go to another program where they get more rigorous, more advanced, more specific training, and they go through a whole training course with a specific handler, but they actually go out and they’re placed in the field in a specific setting or environment. And there are dogs, you know, there’s, there’s attrition that we see early on during the puppy development period. But then there’s also attrition that happens once in the field. And so there’s not, you know, previously been a lot of information about how reliable these tests are in predicting long term success in your career beyond just that placement. And so, because we have the luxury of being able to track where a lot of our puppies go, because we have good relationships with the customers to get them, we were able to get a lot of feedback on the dogs.
Lucia Lazarowski 13:34
And so that was our question was okay, it’s great to know that our puppy tests predict if they get selected or not, but we want to know, if that selection has, you know, long term, the longevity of that. And so what we did was we followed up with those same 60 Puppies, three years later, we contacted the the, in some cases, the handlers or the trainers, where the puppies went, but in other cases, if the dogs had, if they were washed out, and they were adopted, and we were able to contact people that adopted them, and we just wanted to know, you know, what’s the current status if the dog had been selected for for a job? Is it still working? And for the dogs that just across the board, looking at what their behavior looks like now? And you know, is this behavior stable across development into adulthood? And is their placement their selection? Is that consistent with their working success and just kind of more more long term longitudinal information on the dogs how they’re doing now beyond that, that your old time point?
Kayla Fratt 14:40
Yeah, yeah, definitely. So, you know, I guess one of the things this isn’t necessarily addressed in the paper, but it just popped into my head, um, given how expensive it is to purchase a one of these dogs, what do we know about kind of quote unquote, how bad It has to be for a dog to be dropped from a program post selection. Because I can just imagine if I’ve spent tens of thousands of dollars on a dog, just how much loss aversion might come into that decision. And, you know, obviously, you know, we also just can’t know in this sort of setup, like, the quality of the training that these dogs received, or the type of training or intensity of training. Because you can imagine, different trainers in different programs have vastly different or maybe not vastly different, I guess they’re all kind of going into similar roles, but different levels of skill sets. So anyway, that’s a couple of different like questions and things, I just want to pull on a little bit with you if that’s okay.
Lucia Lazarowski 15:44
Yeah, and that’s definitely a challenge to this is that, in a, in an ideal world, we would have like the same, the same customer assess all of our dogs using the same standards and same decision making. But there’s obviously, there is changes in the supply and demand of dogs. And that can affect their decision about whether they’re going to select a dog or not. And like you said, different trainers are going to vary in how they trained dogs, and you know how good of a trainer they are. And so there’s lots of different factors that come into play in, you know, in regards to the dog’s long term success. And so it’s not definitely not perfect.
Lucia Lazarowski 16:23
But in general, everybody wants the dogs to succeed, if somebody has invested, you know, in, in selecting and purchasing the dog and putting it through training, they’re going to want that dog to succeed. And again, there’s going to be things that come into play, like large organizations have tons of dogs coming through, they have a little bit more flexibility to just watch the dog that is going to require more work, whereas smaller programs that, you know, every dog is very important, because they’ve already put a lot into it, they might be more willing to invest more into making sure that dogs disease, but for the most part, you know, we were just looking at whether, you know, we’re kind of assuming that the dogs have to have some minimum level of behavioral characteristics that allow them to perform their job, and that they’re only going to wash if the if something happens, or they exhibit behaviors that are incompatible with performing their job.
Kayla Fratt 17:18
Totally. Yeah, yeah, yeah. So let’s cut to the chase, how many of these or kind of roughly what proportion of the dogs that were placed or purchased ended up still being in the job a couple years later?
Lucia Lazarowski 17:31
So I guess it’s kind of two different answers to that question. So of this cohort of 60 dogs that are kind of primary question was, how accurate was the selection designation, how that matched up to there three years later, still working in the field. And it was actually almost perfectly consistent that the percentage of dogs that graduated was the same as the percentage of dogs that were still working in the field three years later. However, there is a little bit of nuance there, because which ended up being sort of surprising and probably uncommon. But for one, only two dogs from that cohort of 60, that was selected, ended up washing from the field after selection. But we also had two dogs that were initially not selected that received additional training a little, you know, they matured a little bit ended up getting selected later on. And so on balance, it ended up being equal, you know, to wash and then to replace that originally got categorized as washed out. So ended up being the same, but in reality, for dogs changed. Their status. Really was Yeah.
Kayla Fratt 18:39
Yeah, that’s a good way. Yeah, that’s a good way of putting it. And I mean, again, I think that makes that I hope makes sense to most of our listeners, like if I look at, like, so many dogs at one year old. That’s actually it’s much more predictive than I would have expected, I think. Yeah, absolutely.
Lucia Lazarowski 18:57
So so that was exciting, because that’s pretty, that’s very good numbers in terms of, if we look at just a general success rate from that original cohort, I want to say it was around 60% of the dogs got selected, which is, that’s pretty consistent across the board for working dog programs, they tend to average anywhere from like 30 to 50, or 60, not as high as we would like, our program usually boast better numbers closer to 80. This particular cohort was a COVID cohort. And so they’re not as ideal as we would like. And so the first study of the the dogs that got selected was a lower success rate than we typically see, you know, across several years. We were just looking at like one one particular year, which happened to be right during COVID. So we I mean, we have no way to know for sure, but we did know that there were multiple letters in there that didn’t get the same exposure to different people. Just because we had a lot of restrictions on Who could come in but but then in terms of the consistency of outcome from selection to working for years later, that was extremely stable. And so that that was a grading just that, we can be confident that the decision that’s made around a year old is going to be pretty reliable and consistent years later.
Kayla Fratt 20:20
Yeah, that’s really cool. So what and this might not be something that you’re as involved in, so it’s okay to say you don’t know, what is kind of the process like for buying one of these dogs. So like, when I when I acquire a dog, most of our dogs are shelter dogs, so we get, you know, as much information as we can, but usually, you know, we don’t even have a pedigree. You know, we don’t know parentage, we have no idea like, what this pup was what this dog was like, as a teenager, or maybe the dog is still as a teenager, but we don’t really know like, what their fear periods were like, or whether or not fear periods actually exist. But you know, whether they were like a weird teenager, like, Are people handed like, you know, like a file folder of like biographical information that with his dog, and then like videos of them, like, do they assess them hands on? How does it actually work?
Lucia Lazarowski 21:11
So our program, because because we are your research program, and because we breed the puppies here and raise them here, we have a wealth of information that I want to say most programs probably don’t have that much information on their dogs or just play depends on the organization. But generally, customers that are coming to select dogs, you can offer them all this information. And they generally don’t want it because they don’t want it to bias their decision. Or they might, you know, they might take it and look at it after they’ve made their decision. But we’ve we’ve kind of run into a lot of that, like, we can give you all kinds of information about how they were as puppies and all these things, but they want to look at how that dog performs on that day, because that’s what matters to them. And they don’t want to, they don’t want us to bias their decision.
Lucia Lazarowski 22:00
But typically, what it looks like is, you know, we have relationships with different customers, you know, locally or around the country, that they will go on their procurement trips, when it’s time for them to procure new dogs, I’ll usually visit several different programs looking for dogs. TSA is a good example, they were just here recently, they will visit different programs, they will run their own version of their test. So the test that’s described in our papers is our internal tests that we use to get data on the puppies and to make, you know, get a good idea of their chances. But the customers that are coming to select dogs, they’re gonna run their own test. Sometimes it’s similar in some ways, sometimes it’s different. Generally, they, they’ll typically have I mean, they’re consistent what they’re looking for, but they don’t want people to be able to, you know, train to the test. And so they’re gonna introduce some some novel aspects so that, you know, the dog does it look fluffed up for the test, but it’s not representative. And so they’ll run their own their own assessment. And that will vary depending on who the customer is, and what they’re looking for in terms of what kind of environment the dogs and we working in some, some might emphasize a lot more of how the dog is around people and crowds if the dog will be working in a place like that. Whereas others may not care about that, if it’s going to be you know, a dog that’s going to be working screening cargo or something like that doesn’t matter. So everyone kind of has their own test.
Lucia Lazarowski 23:30
The medical information is definitely something that’s important that we do, you know, hand them electronic files or folders full of certain screenings. So TSA has specific medical requirements, the dog has to meet certain standards, medically to even be considered. And again, different, different customers will vary in what things they may or may not accept, depending on like the physical demands of the job and how something might be. We are, our program is currently part of a national breeding consortium. It’s being funded by DHS, and it’s through John Hopkins University. And they’re basically working with several different detection dog breeding organizations around the country, some small kennels, some bigger groups like us to try and address this issue of not having enough quality detection dogs domestically. So a lot of times, groups like TSA have to go overseas to look for dogs. When you’re doing that, you’re usually not getting as much information on the dogs or might not be as accurate as more medical issues. And so it’s kind of this incentive to help domestic breeders increase the supply of dogs.
Lucia Lazarowski 24:52
So anyways, we’re part of this consortium, and part of that is gathering lots of data on our puppies And then actually providing the customer like TSA with a profile of how the dog has performed and their prediction of being selected for that particular role. And so that’s a good example where when they, when they come, they don’t look at that until after they’ve done their assessment to not bias them, but then they will use that information to look at how accurate it was and fill in any gaps that maybe they didn’t have.
Kayla Fratt 25:25
Yeah, yeah, definitely. So yeah, you you do your assessment, you get your own, you make your own judgment. And then you can see if it’s then based on the paperwork, if it looks like that dog was having an unusual day, or Yeah, something like that. Cool. Yeah, that’s really helpful. So okay, so four of the 60 dogs changed status a couple of years out? What, what were what are some of these tests? Like, you know, do you what can we Yeah, can walk through what that was?
Lucia Lazarowski 25:59
Sure. So the tests are designed to, to assess, kind of, I guess, I would say three broad categories of what we consider to be important behavioral characteristics. One of those obviously, being the specific detection, skills and abilities. So how good the dog is at using it to nose in different different olfactory behaviors. Another being what a lot of people call environmental soundness or emotional stability, kind of everything that has to do with their confidence, or fearfulness or lack thereof. And then the other being kind of motivational characteristics.
Lucia Lazarowski 26:41
And so the the test is broken up kind of into a series of sub tests that like multiple mini tests that are all in the then kind of address those those bigger domains. So the first part is looking generally at the detection behaviors. And that’s, that’s typically set up as a series of searches. And all of these tests across the different domains, they they generally get more challenging, as the puppies are older. So when they’re three months old, the test is shorter, because the dogs have as much endurance and they haven’t had as much training. And they get kind of more longer, more challenging as they get older. But generally, a series of searches, when they’re young, they’re just, we’re just hiding their toys, and they’re looking for their toy, because they haven’t actually been trained to specific odors. And we’re just looking at, you know, how, how enthusiastic they are in searching different areas, how, how efficiently they’re using their nose versus using visual targets. You know, how the level of arousal while they’re searching, so just basically anything with anything to do with your actual hunting abilities. Then we through that, we’re also looking at the dogs motivation for their reward, we use toy reward for promotion by dogs, to throughout the search, as they’re getting rewarded, we’re looking at the dogs possession, so how much they want to physically possess the ball in their mouth and play tug, and their kind of interest and persistence and motivation for the reward. Because obviously, that’s what’s used in training, and the higher their motivation for that toy. The more trainable they’re going to be, the more efficient training is going to be. And then the third part of the test being the environmental soundness. And that usually looks strange, because it looks very artificial and far removed from what the dogs are going to be doing. But what we’ve found in our analyses is that you can do these kind of battery of, of exposures to strange objects and stimuli. And that’s going to be a good predictor of the dogs, fearfulness and confidence in more natural environment. So that’s usually kind of fronting the dog with a novel object we’ll use. We have like a dinosaur statue, you know, different kinds of like, statue, object type things that the dogs probably haven’t seen before. And we’re looking at how, you know, Are they scared of it? Do they approach it, we will do a visual of like a surprise element. So with an umbrella opening or a bag being thrown in front of a dog, something that suddenly appears in front of a dog and their reaction to an acoustic startle, so loud noise and then animated objects or something kind of robotic, in motion, usually like a race car, a robotic toy, and looking at their reaction to that. Interestingly, kind of what I was getting at a second ago, we used to have another component that was also looking at this environmental soundness domain was we would take the dogs to a high stimulus environment like downtown or campus and we’re looking at how they react to crowds of people and cars. Are the noises in the environment and stairs, that’s a very typical assessment for working dogs. But there are issues with it because well, it’s time consuming to do that. And to find the areas, that’s another problem we had during COVID is that there was we couldn’t find places with people to expose the dogs to there wasn’t a lot of traffic, the the environment can be very different from one day to the next. So we’re on a college campus, we like to use campus areas, Student Center a lot to test the dogs. But if it happens to be during, you know, holiday break, we’re doing this and there’s their students around that the dining room, and that they will be very different than a busier day.
Lucia Lazarowski 30:40
So there’s a lot of just factors that can vary, we found is that the dogs performance on this battery of, you know, presentations to these novel unusual objects, that was just as predictive of how to perform in these different operational environments as actually exposing them to the environment. And so what that tells us is that we don’t have to do that longer, more complicated, more variable tests, we can just do the simple battery and get the information that we need.
Kayla Fratt 31:09
That’s really interesting. That actually is kind of surprising to me. Do you know if that has been validated a crossbreed groups at all? Or is this just with your group?
Lucia Lazarowski 31:20
We so I don’t know that, that that’s the test that we use as far as like the battery of objects that is used a lot by different groups looking at dog behavior, both for working and working dogs. As far as I know, it has never been compared to like operational exposures. But generally, that kind of test is has shown really good results because it’s much more standardized and structured as you totally
Kayla Fratt 31:49
Yeah. Yeah, I was just thinking, so like my book, both of my dogs are Border Collies, and they’re notorious, but you know, like, Where does this come from? How much of this is just kind of what we all say about them, but they’re kind of notorious for being bad with Southern environmental contrast. And I can imagine both of my dogs doing much better with like walking into an airport or a student center or a festival than they would with like, an air horn going off behind their heads. So yeah, that’s kind of the question, I guess, is like, it’s actually surprising to me that those things are that tightly correlated, because I’m not confident it would be for my dogs. But yeah, like that also could just be, I could just be biased in some way in my own.
Lucia Lazarowski 32:41
I mean, that brings up a good point for our studies in general is that we have a very homogenous population of purpose read Labradors. And so, you know, we always caveat with, this needs to be replicated with other breeds, other programs, you know, our dogs were all raised the exact same environment, same protocol for training, very closely related dogs. And so this information is great for us. But whether and that’s actually something that we’re currently working on, in collaborations with other groups, like the pen networking dog center, is we want to see if the same tests are predicted for other breeds, or just the same breeds that are raised differently or have different, you know, different tasks that they’re being trained for. So that is, yeah, that’s a really great point.
Kayla Fratt 33:29
Yeah, cuz they have, like, Penn vet has a lot more of the pointers. And they do a lot more dual apprehension stuff. And the I could see, yeah, I would also honestly be a little surprised if like a German shepherd had that same reaction to like something appearing suddenly, versus just kind of like being in our crowd. But again, like, that would be interesting. If that is the case, and I’m just wrong.
Lucia Lazarowski 33:52
There’s also the issue of dogs, you know, different roles. The dogs need different types of kind of behavioral profiles. So all of our dogs, obviously single purpose textures, and we don’t do you know, we don’t have the opposite our dual purpose dogs. So for like a for dual purpose dog that’s doing detection and apprehension, patrol, whatever. There is some level of I don’t know if the right word is reactivity, but they Yeah, they they actually need to react in a way that we probably wouldn’t want our dogs to react to the role that they’re going into and so it may be that the tests are more or less predictive because of the specific job that the dog is needed.
Kayla Fratt 34:42
Yeah, yeah, absolutely. Yeah. I’ve never worked with dual purpose dogs. But like, I can imagine, I might want a dog that can I would need a dog that can handle a crowd if need be, but also is going to at least respond to me If something has come up like, yeah, I don’t know, you might want that dog to be able to notice it. Environmental contrasts on its own. I think we come up with in conservation all the time, we like we do a lot of work with Search Dog Foundation, where they’re, you know, they’ll send us some of their potential career change dogs. And, you know, our first question is always like, well, have we seen him around wildlife? You know, that’s like our first, our first wonder, that is just not, it’s nowhere near as important for most other fields where like, prey drive and toy drive often, like correlate, and a lot of times when I’m talking to like sport dog people, they call toy drive prey drive. And like, we actually really try hard to pull those things apart and find the dogs that will choose a ball over a squirrel and the dog views those as different reinforcers. And that’s not a given.
Lucia Lazarowski 35:57
Yeah, and we often joke that we are selecting for the Darwin or dogs, because our TAs, people often say, like, well, that’s not fair. Well, you’re asking the dog to do, like, I don’t want my dog. And one of our tests, we literally have a taxidermy bear, that we, you know, the dog walks into a room, and there’s just this giant bear right there. And the better scores are the dogs that walk right up to it, and sniff it and leave it and they, you know, they don’t show any fear or apprehension to it. You know, it begs the question of like, well, do you want your dog to survive in certain environments, but the reality of like, what these dogs are, how they’re being selected, the level of, you know, extreme things that they’re being faced with. That’s, you know, that’s what we deal with.
Kayla Fratt 36:43
But, right, yeah, like, I might not want my dog to choose to approach. But I also would certainly not want a dog that’s going to like bark and charge at a bear. Like, choose what you’ve described. I like I guess, yeah, having not seen like a whole suite of potential behaviors, but like, that, like, not worrying about it, and still probably a pretty good thing for us. But yeah, I’d say, you know, it’s, it’s a huge question. And, you know, I wouldn’t necessarily ask our dogs to go through that right away.
Lucia Lazarowski 37:18
Yeah, we’re not looking for dogs that are just completely, you know, if they’re not reacting at all, you’re asking like, okay, is this their, their eyes working? Like, is there something you want them to notice things, but what we kind of emphasize is their ability to recover from it. We don’t want dogs that pancake and want to flee and leave the area. And they’re so scared, like, they need to be able to, if they’re in the real world working, and they’re confronted with a loud explosion, a loud noise, or, you know, a random object in our scene before they need to be able to work through it. We don’t want them to be here, but it is okay if they show some apprehension, but they take it upon themselves to then explore and investigate. They don’t need a lot of handler, encouragement and support and they can overcome it. And it’s not impeding their ability to work. So yeah,
Kayla Fratt 38:07
And that is such an important point. I remember, gosh, this was like years ago, listening to some podcasts with Ian Donmar on it where he said something about, you know, well, if you open up an umbrella and a puppies face and it doesn’t react, all you’ve confirmed is that puppy isn’t scared of umbrellas, you actually know nothing about their ability to recover. And for most of us, that’s actually what we need. I mean, it’s good to know that you’ve got a puppy that isn’t nervous, etc. But you also as much if not more, want that response. And I’m so sorry about Norbert. He’s just wandering around screaming. This is life with that cat.
Rachel Hamre 38:48
Rachel here from K9Conservationists. One of the ways you can support K9Conservationists is by checking out our online store at K9Conservationists.org/shop. You can find hats, stickers, pet food mats, reusable grocery bags, mugs, and all sorts of shirts and jackets. I have one of the hoodies and I think it’s so comfortable and cozy. I also really love the pride stickers. If you really don’t need anything new you can also make a tax deductible donation to K9Conservationists at k9conservationists.org/support-our-work
Kayla Fratt 39:27
Yeah, it’s so cool. I’m so I’m trying to think there any other like specific questions I had about the study. Was there anything that like surprised you as you were going in or anything that you were like? Yes, that’s what we were hoping to say.
Lucia Lazarowski 39:39
It was. It was really cool with this follow up study. Like I said, we not only tracked dogs that have been placed in the field and I guess I don’t know if I finished explaining this but we not only follow it up with dogs to find out their status, but we also for as many of the dogs in a cohort that we that were generally local that we could we brought Got them back in. And we repeated the test to see. So we’re interested in not only does the puppy test predict selection and working status three years later, but we also wanted to know if their actual behavior was consistent over time. And so we we repeated the test with the dogs now at three years old. And we brought in both dogs that were in the field that their handlers were willing and able to bring them in. And that washed out from early training that had been adopted by someone in the community, we had them bring their dogs in, and we repeated the test.
Lucia Lazarowski 40:32
And so that was really cool to see both sides of it, and to get that confirmation, because a lot of the decisions are driven by the customer selecting the dog or not, but a lot of them are also made internally as a program, we don’t necessarily present every single dog because if there’s a dog that is just obviously not, you know, not cut out, we’re not going to put them through that we’re not going to waste people’s time. And so we do make some decisions of, we’re gonna go ahead and adopt this dog out, go ahead and wash them out. And they’re not actually getting that third party assessment. And so that was really good to have those dogs come back in, run the test those dogs that we decided to wash out and then confirm, like, wow, they really, you know, these dogs were scared to even walk in the door with them, we didn’t even complete the task, because, you know, probably would have broken them. But so as you know, as sad as it is, obviously, to see the dogs that are scared, it was validating for us to see that we had made the right decisions about dogs like that and ticket to compare a dog that had been adopted out because it was showing signs of fearfulness that was, you know, scared to walk in the door and very nervous, versus one of the dogs that was in the field working that had come in, that just, you know, was a cakewalk for them just night and day though screams. So that was cool to see.
Kayla Fratt 41:51
Yeah, that is really neat to see. And that’s such a good point, too. You know, like in the conservation world, I think we’re, you know, those of us that go with shelter dogs I’ve heard, you know, like maybe 60% of the dogs that we identify actually go on to succeed. And that’s kind of like the equivalent of us just getting to do that assessment at 14 months. But also looking at populations that are probably a while, definitely a ton more variable. But yeah, we also don’t know how much like, I think our stats would feel, or look somehow more impressive. If we actually went into a shelter and actually just assessed all of the dogs there, they’d be like, Oh, we’re really good at rejecting dogs that aren’t going to work. Because we also like, we choose to actually assess, like 10, 5%. Even the dogs that come to us via email and people think are going to work for us. And often I’d like to be totally honest, a huge part of part of that is like we’re a small program, we’ve got five dogs, like a huge portion of it is just like, we don’t have anyone looking or like, we’ve got a couple females in our program that are not necessarily the best fit to live with other females, that are also a little dog selective.
Kayla Fratt 43:10
So we’ve got, you know, there’s, and that’s a huge part of it, though, like you can’t ignore that when you’re looking at dog selection is that like, we might turn down dog just because like the only person on our team who’s looking has a female that’s not super tolerant of other females or like, you know, I we get a ton of requests or a ton of emails from people with like, really nice looking German shepherds or males, and we just all of us are currently prefer smaller dogs. So it’s just like, yeah, that dog probably could do the job, and probably would be great. And like, we’re happy to forward those dogs along and help them out. But, you know, there’s just so much personal choice that comes into I’m like rambling a little bit right now.
43:53
Yeah, that ties into what I was saying earlier about, you know, these numbers are not perfect, because, you know, people even from the same organization, depending on who shows up that day, they may, they may run the test differently or make a different decision about the dog or like I said earlier, they the contract that they’re trying to fill may call for 40 dogs on one visit and they may call for two dogs a different visit and so getting selected or not for these kinds of external variables, not necessarily the dog himself, but that’s.
44:26
Yeah, well, Mike I know and this kind of goes back to that inter-rater reliability or injuries over reliability. We talked right up top, but I used to work at a large animal shelter. And one of the things that we would do is I think we used it wasn’t C-BARQ. It was match-up we use match-up 2 as our test. And I remember especially early on I just had to talk to my teammates. I was like hey, if we get like female chocolate labs coming through like I can’t be the one to assess them like I had a had a chocolate lab but that was my one of my heart dogs and like, now, if I got hired again, you know, if it was like Border Collies like I can’t, I can’t be the one who’s assessing them because I will like, make excuses or whatever for it. And that might be a little less common when you know, a someone’s a professional dog assessor and be, you’ve got a more homogenous population. But that was definitely something that we really contended with in the shelter is kind of being like, well, but he’s only been in the shelter for two days. So like, should we just do the test again in a couple days, but now he’s been exposed to it twice, like, you could desensitize, you know, you don’t know which way it’s gonna go. But still, it’s like, it’s just really hard and very, very non scientific on the shelter side.
Lucia Lazarowski 45:46
But while we’ve I mean, we have seen some biases in preferences for what people look for a pretty consistent finding of males being more often selected than the females. And there was some, there was some thought that it’s hard to say if it’s because the males actually outperform the females or because in this industry, there is a preference or a bias for males. And so there are things like that that we have.
Kayla Fratt 46:17
That’s interesting, have you actually been able to look at like, are the males outperforming the females?
Lucia Lazarowski 46:21
We have found that on, on some measures, and at certain ages, they are and we think, from the limited data that we have looked at on this, it seems like it probably has to do with the hormonal phases that the females are going through where they kind of go through more sensitive periods when they’re having these, you know, when they’re approaching maturation puberty, that the males aren’t as affected, by the way, yeah, we have seen a little bit of like, higher arousal and higher levels of boldness in the males compared to females at some ages. But at the end of the day, there doesn’t seem to be any meaningful difference between the two.
Kayla Fratt 47:06
That’s super interesting, because actually, it’s funny in the sport dog world, there’s like the opposite perception and a lot of at least in agility, like when I talked to sport dog people and you know, again, because I’ve got Border Collies, I talked to agility people a lot. They’re all like, I just don’t understand why you would go with a boy dog. They mature so slowly, they’re so dorky. They’re so goofy, you know, and they’re all of these perceptions. And it’s all women telling me this. Males, and then I would imagine, explosive handlers are like, Is that is that a male heavy field?
Lucia Lazarowski 47:38
Definitely. Yeah. The question I have is, is it the preference of? Or is it just that who the person is and yeah, even
Kayla Fratt 47:47
Like, I’m in a search and rescue group, and a lot of people in there will talk about like, oh, well, we wouldn’t want anyone on our team to have an intact female, because then during all of the trainings that would like, throw off the males and like, that is kind of a real thing. You know, especially when people in search and rescue say that I’m like, but what happens if there’s an intact female in your search area? And there’s a missing person?
Kayla Fratt 48:09
Yeah, okay. So actually, you might be the right person to ask this question. And now we’re, like, totally off the rails, just asking you about anything you may have seen. But in sport dogs, again, there’s this really common perception or statement that like females mature faster than males? Is that anything that you’ve seen? Because it actually sounds like what you’re saying is like, in kind of, like, teenager dumb, they might actually be less bold than the males but like, yeah.
Lucia Lazarowski 48:38
Yeah, I don’t know. I don’t know if I can speak to how them maturing faster or slower. It seems like what we’ve seen is just more that, around that, around that period of adolescence, the females are more affected by it because they’re going through more hormonal changes. One thing that we have actually seen in our population consistently which has been something that’s been just a anomaly, I guess I would say is that our dogs seem to the females seem to mature much later than the tip the breed or than even like other programs that we’ve seen, and we think it could have to do with kind of like the the athlete effect you know, like female gymnasts. Maybe just because they’re there they you know, the physical activity they’re engaging in and your diet and things like that, but I remember reading a paper that talked about this adolescence effect the teenage dog syndrome with a population of Guide Dogs, and how they saw like big changes in their behavior around six months, which they said was right around the their first heat, and we were all like, Whoa, our female average first heat is like 14 months or later. Oh, It’s nothing. We are. I’m not as involved in this research program is trying to get a feel for why that is what the cause is about. And yeah, it’s only been interesting and, and something that we have to deal with because we can’t, you know, we have to hold on to the females longer to bring them into heat later.
Kayla Fratt 50:23
Yeah, one. I guess there’s also always the possibility when you’re looking at like smallish breeding populations or closed up books or whatever, that you just have founder effects like, yeah, you know, they’re like, yeah, like niffler is letter, for example, there was seven puppies, I think. And like five of them went to experienced homes, including one that stayed with the breeder. And everyone reported like, these guys are taking a long time to potty train. And they puppy culture, there was like, nothing wrong with them. They were all like, lovely, lovely puppies. But all of them except for niffler. And one of the others were like, really unusually slow to potty train. You know, it’s just like, I don’t know, like, it could be like, it could be something in the environment for sure. Or like, maybe there’s just something about that pairing that like, created something. And when you’ve got these small, close breeding populations, like you could just have.
Lucia Lazarowski 51:23
Yeah, we actually have a good example of that. So like the Auburn dog, like our poster child, dog is the Labrador. But we crossed in a several years back, we crossed in the, the German Wirehaired Pointer, or technically the the jar to capture like the air scenting behavior. And so our dogs, we don’t really have a lot of the wirehairs anymore, but most of our labs have like a little bit of a little percentage still of the Wirehair. Because the first, so we bred a male German Wirehaired Pointer to a female lab. And that first cross was amazing. They all went out and they were working. That was like 10 years ago, and they’re just not retiring or close retirement. That was a phenomenal letter. But then every litter after that had a lot of behavioral issues, especially like emotional reactivity, fearfulness, so we kind of stopped. That’s why we don’t really have a lot of the Wirehairs anymore. Because it just time and time again, the litters were
Kayla Fratt 52:27
Interesting, so like, the F1s were good. And then the F2s were weird.
Lucia Lazarowski 52:31
Yeah. And then, again, looks like if you look at our population, and this is sometimes like struggled to put in our analyses to look at breed effects, because it is that a breed effect? Or is it because it was literally one sire, that we brought into our population? So there’s something that he you know, he didn’t exhibit any of that maybe something that he carried? Or maybe it was the particular pairing that we’ve matched them up with? So, you know, it’s impossible to say when you have such a small column, yeah, that’s, that’s exactly what we run into to.
Kayla Fratt 53:06
Yeah, gosh, yeah. That’s so fascinating. It’s so, so interesting. And it really makes you wish that like, I don’t know, I guess, I guess I don’t know that this isn’t happening, but. And it wouldn’t necessarily be applicable to dogs. But like, if people were doing that sort of behavioral crossing, breeding sort of stuff with like mice, where you can just get so many more answers so much more quickly, or probably, and I’m sure that’s happening in psychology somewhere. Those are not the sorts of papers I’ve been following. And again, then it’s like, okay, but how do you translate that over to working dogs and like breeds that are probably a lot more different than a lot of rat, or mouse like laboratory strains are but.
Lucia Lazarowski 53:56
Yeah, and there’s, I feel like, I’ve heard of something referred to as like, hybridization where hybrid vigor with the first cross, but then the, any subsequent crosses are not as good. So yeah, interesting.
Kayla Fratt 54:12
Yeah, I wonder I look, I know, I’m not really in the doodle community. But I listen to the Functional Breeding podcast and they’ve got some really good doodle breeders who come on and, you know, sometimes I listen to them talking about Yeah, these differences between F1 and F2 crosses. I’m like, whether you’ve got maybe it’s your third generation, but instead of breeding a golden doodle to a golden doodle, you’re actually bringing breeding the golden doodle back to a lab. So you’re getting three quarters lab. Like it’s just so complicated, and even like the very, very best breeding programs are still so small, and you can only have so much information. Yeah, it’s great. I mean, I’ve been thinking a lot about like, I think my dream next dog that I would love to try would be a sprollie. So a springer Border Collie cross. Um, and frankly, I think I’m just gonna have to make them myself if that’s something I want to do, because I can’t find anyone doing it also would just be you have no idea like we might get like that first letter might be total like firecrackers like exactly what we’re looking for. But that doesn’t mean that you can just, I mean, I guess you could try to do the same thing with the, with the same letter again. But like, then you can’t just keep making more and more of them unnecessarily get that like homogeneity that we like out of our purebred dogs. Yeah. Yeah, thanks for letting us go down that rabbit hole. Is there anything else that I didn’t ask you about the paper? Or that you wanted to expand on to circle back to other caveats and things should be that people should be aware of?
Lucia Lazarowski 55:46
No, I think we hit on on everything. Yeah, I guess like the the main kind of limitation is just again, the generalizability to other breeds. And so those are things that we’re working on that we’ll hopefully be able to address in the future. Something else that we are currently looking at is so we have focused really on the traits that are important for detection dogs, but specifically explosive detection dogs, because that’s just where That’s where our dogs go. But right now we’re looking at are there? They’re pretty sure there are but what are the specific traits that differentiate like the ideal explosive detection dog versus a conservation dog versus search and rescue? So you know, how, like, what are the specific constellations of behavioral traits that differ between the comments detection dogs of any differ between those specific disciplines? Absolutely,
Kayla Fratt 56:41
Yeah, that would be so neat to see. And gosh, I’m so excited to see more and more of this stuff coming out. Yeah, like I again, this is something I’ve said previously on the podcast, but it’s also I guess, if you’re looking at raw scores across breeds, this might still be consistent, but like when I was looking at niffler as litter, one of the things that I was really looking for is I kind of wanted the most spaniel-y of the Border Collies or the retriever of the Border Collies and that litter. So like, I got to actually be there the day that they went outside for their very first time. And he was the one who was just like, nose to the ground, tail up, like off ready to be red tail hawk food, like 50 meters away from the rest of the puppies while they were all like paying attention to mom and, and breeder, and I was like, that seems like what I think I want, but if I was looking at a litter of Spaniels, I probably wouldn’t want the most independent and the most like, I’m just listening to my nose screw you have the spaniels? I know what the Border Collies I’m gonna get the responsiveness. Yeah, like, I know, that’s something that just, that is maybe the most guaranteed thing you get out of that. And then with the Spaniels, it’s like, I feel pretty confident, I’ll be able to get the like the cornering and the sniffing and those sorts of things. But like, can I get that dog to respond? Well, to me around birds, it’s more like my fabric. Anyway, look, I guess, if you’re looking at raw numbers, that’s not as much of a concern, but certainly for people who are like looking at a litter, or looking at a suite of dogs, like, if you’re comparing within the group, instead of looking at raw numbers like that always has to be a concern.
Lucia Lazarowski 58:15
Yeah, and I guess that brings up another important kind of caveat about all of this research that we’ve done is that we are definitely not at the point where we’re gonna test a puppy and say with 100% confidence, it totally, this is all about like, trends. And typically, the better they score, the higher their likelihood, but there’s a gray area where our we’ve gotten to where our test is, is pretty good at differentiating, like your best dogs and your worst dogs. But there is a lot of overlap in those middle ground dogs that could go either way. And that’s just gonna come down to personal or programmatic pros and cons and then weighing the risks and benefits of keeping a dog that could you know, go on a few ways versus watching the dog. Let me do what I’ve been fine. So yeah, there’s still there’s definitely room for improvement.
Kayla Fratt 59:01
Oh, yeah, absolutely. And again, that comes back so much I would imagine to those programs, because like, so far, you know, I say conservation dogs, it’s like, what I’ve heard from industry experts is like, yes, 60% or so we at K9Conservationists are out 100% of the dogs that we’ve selected are still in our program. But we also like, our handlers has a max of two dogs, who are all really dedicated people, like trainers who have been professional trainers in the past, like, and with, because we all are co-founders, like we can decide which dogs go on which projects and who’s suited to that. Like if we needed to be able to take a dog and say This dog can do A, B, C and D just as well as this next dog. We would probably have closer to that like 60% success rate within our team. But there’s just we’ve just got a lot more programmatic flexibility, and also like, really dedicated creative trainers to get them through whatever they need or to work on, work on whatever their weak points are. Yeah. And, you know, part of that is just that we’re not the TSA. Like, we’re, we’re not government agencies where we have these, like, super strict requirements, or you know, that like the dog at 14 months must not have any of that. Cool. Well, I had a blast. I feel like I could keep picking your brain about this for ages. And, yeah, maybe we’ll get to do that someday at some conference. I hope. So. For anyone who’s interested in learning more about Auburn’s program or keeping up with your work, is there anywhere that people should be finding those things online?
Lucia Lazarowski 1:00:48
Sure. Um, so our programs website, if you just Google Auburn University, Canine Performance Sciences, that is our programs website, I don’t know how up to date it is with our research lab. But if you Google my name, people oftentimes email asking for PDFs of papers and things like that, and I am just kind of general PSA, the scientists are usually more than happy to share their papers for free. So before you, you know, go and purchase it through, you know, the journal, we don’t make any money off of that. And we have no issue with it. As long as we have the right to share the paper, which a lot of times, we have a version we can share. I’m happy to share it. So if you either just email me or if you just Google my name, I do have ever ResearchGate page and I have a Wix website where all my papers can be accessed. And then through that, you can usually find our programs other papers, too. And so we try and make all of our research as accessible as possible.
Kayla Fratt 1:01:51
Yeah, you were very responsive with that. And I think that’s how we ended up doing this episode. I think I just emailed you for the full text. And I might have gotten halfway through writing that email and then was like, Wait, what am I doing? I need to ask her on the podcast. Originally, I was just going to ask you for the full text because I’m a PhD student. I’m not paying $70. I haven’t figured out how to use the university access. Yes. Because I’m sure I have access to it. Now. I just haven’t figured it out. So well. Yeah. Cool. Again, thank you so much.
Kayla Fratt 1:02:19
And for everyone at home. I hope that you enjoyed this episode as much as we did. And maybe you’re feeling inspired to get outside and be a canine conservationist in whatever way suits your passions and skill set. You can find our website at k9conservationists.org where we have Patreon and a course and you can also just look at cute photos of our dogs. We just updated the website. So the homepage, the aspiring page, and the partnership page are all brand new. And I worked very hard on those, so go look at them and then tell me how much you like them on social media, will make me very happy. And yeah, we’ll be back not next week but the week after. Maybe we’ll get back to weekly episodes. I hope so because I really miss doing this but it’s busy. Again, thanks so much.