A false positive indication is one where your dog indicates, or alerts, on something that is NOT her target scent source (Lehnert and Weeks, 2016). To all intents and purposes, it looks as though she’s correctly located the target odour but … she hasn’t! You’ve been misinformed.
At Scent : Detect : Find Ltd, false positive indications are an extremely rare occurrence. If they do happen, they’re easy to identify and quick to remedy. In large part this is because …
a) All dogs are trained to perform a highly reliable, passive indication with duration – a sustained nose touch, or point. This indication is easy to distinguish from any passing investigatory sniff – which the handler can confidently ignore. In contrast, the correct indication is clear and unambiguous and is always generously rewarded.
b) Importantly, all training is broken down – split – into its, tiny, constituent parts rather than being lumped together into one, large, unmanageable whole (Berg, 2018. Zerubavel, 1996). Splitting allows any training difficulties to be addressed immediately, including any false positive indications, before the problem has the chance to escalate further or become entrenched.
If you’d like to know a little more about Splitting and Lumping, simply follow this link to one of my previous Blog Posts – Smoothing out the Lumpy Bits; Why it might be wise to teach the indication before the search and all that back chaining stuff.
Strangely, false positive indications seem to be an accepted part of Scent Detecting practice – “Just one of those things“. “Nothing to worry about“. “To be expected“. “Par for the course“. “It happens“.
Well, I’m here to tell you otherwise …
False positive indications are not just one of those things.
False positive indications are something to “worry” about.
False positive indications should never be an expected, or usual, part of Scent Detecting training or practice.
False positive indications are a symptom of one, or more, Scent Detecting training problems which need to be addressed rather than accepted!
What are the likely problems with false positive indications?
Well, if you’re “just having a bit of fun” with your dog, perhaps a false positive indication doesn’t matter too much to you. That said, it’s worth asking yourself why “just having a bit of fun” is any less deserving of careful training than any other activity you choose to participate in? Would you only half train a Sit? Are you happy with a Recall that only happens sometimes? Is it OK for your dog to choose her own route around an Agility Course? …
A dog who doesn’t understand what’s required of her, who hasn’t benefitted from careful guidance, who gets rewarded on some occasions but not on others (dependent on her handler’s whimsy on a particular day) … well … that’s NOT a dog who’s having “fun“. It’s far more likely that she’s confused and anxious. If you want to use Scent Detecting as a fun activity for your dog, there are plenty of ways to achieve this including scent trails, scatter feeding, treat searches, hiding toys and using Scent ‘n Snack mats. These type of scent-based activities can provide numerous behavioural benefits for your dog and require little, if any, training – your dog is a natural scavenger and comes pre-programmed to locate food!
If you’d like some more ideas on how to entertain your dog’s nose, look out for one of my next Blog Posts – Scent-Based Nose Games – Just a bit of fun!
In contrast, if you’re the handler of a working dog, or take part in competition style Scent Detecting, then the reliability of your dog is key to your success. Be under no illusion, a dog who gives false positive indications is an unreliable Scent Detecting dog. If you cannot trust her indication you’ll only know if she’s correct if you can see the target scent source yourself or watched where it was hidden at some earlier point in time. When faced with a truly blind search – one where the location of the target scent source is unknown to you – well … I wish you all the luck in the world!
What are the potential causes of false positive indications?
There can be numerous potential causes for false positive indications but, very sadly, the vast majority are handler and training related. The root cause of your problems may be just one, very particular, issue or could be the result of a combination of different issues.
- Contamination / Residual Odour
This is probably the number one rationale offered by handlers to “explain” away their dog’s false positive indication – “The area must have been contaminated”. “The scent wasn’t handled correctly”. “My dog’s nose is just so good that she picks up on every little bit of residual odour”. Sound familiar? Well, what if I were to tell you that although a contaminated search area is a real thing – it does happen and it happens all the time – the root cause of your problem is not odour related, it’s training and handler related?
Your dog is what’s referred to as a macrosmat or, ‘keen smeller’. Her olfactory threshold, the level at which she can first start to detect an odour, is reportedly as low as 2 – 3 parts per billion (ppb) and, quite possibly, for some odours, as low as 2 – 3 parts per trillion (ppt) (Helton, 2009). Given this incredible degree of sensitivity to odour, it seems a little surprising that your dog doesn’t give you a false positive indication more often than she does. So, what’s going on?
As an aside, keep in mind that if your dog indicates on any residual odour or contaminated area, she’s actually correct! It isn’t a false positive indication. What she’s doing is indicating on a very low concentration of the target odour … something you may not have wanted or planned … nevertheless, your dog is correct. For clarity, I consider residual odour to be any odour “left behind” by the target odour source when it has been removed from a search area or container. In contrast, contamination occurs when the target odour source leaves a residual odour in an otherwise “clean” area or when the target odour source becomes contaminated itself by another odour. Given the “catch all” nature of the definition of contamination, this is what I’ll be referring to throughout the following discussion.
As Gadbois (2016) has pointed out, your dog’s olfactory abilities are so good that no amount of cleaning, or careful handling of the target scent source, or use of gloves, will reduce contamination sufficiently for her not to be able to detect it. In addition, those gloves that you’ve been advised to wear are a recognised contaminant in themselves (Learning Centre Valutek Blog, 2016). Dogs whose handlers use Nitrile gloves have been found to indicate on anything that the gloves have come into contact with! By now you’re probably starting to appreciate that, to a very large extent, attempting to avoid contamination is a waste of your time. Far better to acknowledge that contamination is a real issue and work with it.
Scent : Detect : Find Ltd adopts a number of different strategies to overcome any so-called false positive indications as a result of odour contamination.
1. Initial training is conducted using a large concentration of target odour
Dogs are rewarded generously for indicating the location of this “large” odour source. Although they are likely to notice contaminated areas in their search environment – dogs may hesitate and investigate areas of the room where the scent source has previously been – importantly, they aren’t rewarded for this behaviour. It doesn’t pay off. In short, they’re learning to ignore any contamination as they discriminate between the different concentrations of odour. Dogs will work with this highly concentrated odour source until they’ve gained considerably more experience. At this point, the odour concentration will be reduced – Diminished or Diluted.
Interestingly, according to Laing et al (2003.), a “change in concentration can actually modify the quality of … [an] … odor“. As Wilson et al (2006) point out, “for many odours, as the concentration increases, the perceived quality of the odour changes“. With this in mind, Goldblatt et al (2009) suggest that Scent Detecting dogs should be “trained and maintained on a range of intensities of [odours] that they may encounter in [their Scent Detecting work]”.
A recent pilot study by DeChant et al (2021), investigating the ability of dogs to both generalise across, and discriminate between, different odour concentrations, has provided further support for Scent : Detect : Find Ltd’s training approach to odour concentration. As described previously, all dogs begin their training with a very high concentration of their target odour and only after some considerable amount of Scent Detecting experience are they then introduced to the same target odour at much lower concentrations. This approach applies to all target odours used.
According to DeChant et al (2021), “dogs can be trained to actively not respond to concentrations of a trained odorant below a desired threshold“. Put simply, your dog can be trained to ignore contaminated search areas. In DeChant et als (2021) study, “false alarms decreased after Concentration Discrimination training. This may result from training dogs not to respond to the low, but detectable concentration odor”. Clearly, on this account, false positive indications cannot be attributed to issues of contamination. There’s a definite need to examine the training protocols in use and pro-actively train using varying concentrations of odorant.
2. On locating their target odour, all dogs are trained to give an unambiguous passive indication with duration
As already discussed, all dogs at Scent : Detect : Find Ltd begin their training by working with a large concentration of target odour. This is what they learn to indicate, or alert, on. I favour the passive indication and train my own, and other handlers, dogs to place their nose as close to the target scent source as possible and hold it there for a minimum of 5 seconds or until their handler asks them to move away. There can be no mistaking the precise location of the target odour source when the dog’s nose is either touching or pointing toward it. Adding duration to the indication adds even more certainty. A passive indication with duration allows the handler to “ignore” any small acknowledgements of contamination within a search area and simply wait for the added commitment – the duration – that comes when the dog is convinced that she’s located the highly concentrated target odour source. The dog is free to check out all areas of possible contamination because the handler can be confident that once investigated, she’ll dismiss these contaminated areas and move on.
Recently, I’ve noticed a growing trend toward relying on a dog’s change of behaviour (COB) to help the handler recognise that their dog has located the target scent source. Typical changes in behaviour might include slowing down, speeding up, heightened interest in a particular area of a room, walking on tip-toes with head held high … the list of possibilities is almost endless. Each of these behaviours may help you narrow down the location of the target scent source but may also act to totally mislead you. The dog may simply be checking out an area of contamination.
For further discussion of indications, you might like to follow this link to one of my previous Blog Posts – The Indication. It’s not an optional extra.
3. In very specific circumstances, it’s worth taking care to avoid any possible contamination
Although, generally, I don’t worry about contamination, there are a couple of very specific circumstances where target odour handling needs to be a little more thoughtful.
a) Storage – Keep your target odour in the most secure container you can get your hands on. Your dog needs “down-time”. Remember, if you’ve taken the time to familiarise your dog to the target odour it will have become a “motivational magnet” for her. If she can detect the odour she’ll be in “work mode”. Give her a break!
b) Frequently used containers – If you’re likely to use a particular container on repeated occasions, make sure you can identify it – label it in some way – and only use it to hold the target odour. Frequent contact with the target odour source will, over time, result in a heavily contaminated container closely matching the concentration of the target odour source itself.
2. Handler Pressure and the Clever Hans Phenomenon
As Galambos et al (2021) point out, “dogs are unmatched amongst non-human animals regarding their inter-species social skills … thousands of years of domestication [have] enabled them to excel at the reading of and responding to human social-communication signals” … and here lies the problem … our dogs, and their performance, are affected by us!
Two potential Scent Detecting problems arise from this.
1. Our dogs will be aware of our mood and emotions and respond accordingly
As Galambos et al (2021) states, our dogs have a “human-like susceptibility to social influence” and demonstrate a “spectacular level of social sensitivity“. This is likely to be a particular problem within competition-style Scent Detecting circles where handlers can place enormous “pressure” on themselves to do well and gain rosettes. Unintentionally, the handler’s “pressure” is then “transmitted” to their dog.
Scent Detecting videos, available freely on various social media platforms, are well worth checking out. Make your focus of attention the dog and his body language rather than the actual Scent Detecting task itself. Watch for any Calming Signals (Rugaas, 2006) – lip-licking, yawning, turning away from the task, hesitancy, slowing down – and link these behaviours to the handling skills that you observe. Mariti et al (2017) consider these signals as having “a communicative role” – they can help you identify how your dog is feeling. More often than not, this will be a reflection of how you’re feeling! Custance and Mayer (2012) refer to this phenomenon as “emotional contagion“.
And the consequences for your dog’s Scent Detecting performance? If she’s becoming uncomfortable in the search area because of the way you’re feeling and the demands you may be placing on her, the quickest way to end the experience is … you guessed it … to indicate. Unfortunately, more often than not, this will be a false positive indication. Again, without a doubt, this is reflective of a training issue. Are you and your dog ready for this Scent Detecting challenge? Have you put in the necessary preparatory work? Do you both have the requisite level of experience? Do the expectations you have of your dog’s performance outstrip their current level of ability?
2. The Clever Hans Phenomenon
The Clever Hans Phenomenon, or Clever Hans Effect, refers to any situation where a person or animal can be influenced by the subtle and unintentional cueing of another (Jackson, 2005). Your dog is able to recognise and respond to the cues delivered by those around her … including you, her handler! Problems arise when those cues misinform her. Importantly here, cues that misinform her about the location of the target scent source resulting in a false positive indication.
In a study by Szetei and Miklosi (2003), almost 50% of the study dogs would go to an empty bowl indicated to them by a person pointing at it rather than to a bowl in which the dog had previously seen and smelt food! The dogs appeared to place greater weight on the human cues than their own senses and experience.
In another study by Lit et al (2011), 18 drug / explosive detection dog and handler teams were given a number of detection tasks to complete. None of the tasks contained drug / explosive scent. In other words, they were all blank searches. Any alerts / indications that the dogs gave would be incorrect; false positive indications. Two of the detection tasks included a small piece of paper that falsely marked a scent location and two other detection tasks contained a decoy scent (toy / food).
In the detection tasks that included the false paper markers the handlers reported that their dogs alerted more at the marked locations than in other parts of the search area. According to Lit and her co-authors, when the handlers believed that there was scent present in the search area (paper markers) they were more inclined to claim that their dogs had indicated on a target scent; a false positive indication. As Lit et al (2011) state, this would seem to confirm that “handler beliefs affect outcomes of scent detection dog deployments“.
As Udell et al (2010) point out, with “increasing evidence for an acute sensitivity to human gestures and attentional states“, guarding against the Clever Hans Phenomenon in our dogs can be highly problematic but an essential consideration in the development of a reliable scent detecting dog. Again, any false positive indication in these circumstances is a training issue rather than simply “one of those things“. Suggestions for reducing the Clever Hans Effect include;
Practice more blind searches – If you don’t know where the target scent source is located you’re less likely to influence your dog.
Keep an OPEN mind – Try not to “second-guess” where the target scent source might, or might not, be. Chances are you’ll be wrong and, in the process, misinform your dog resulting in a false positive indication.
Allow your dog to work independently – Giving your dog the space and freedom to work by herself helps to reduce any tendency she might have to watch for cues from you rather than search the environment carefully herself. If you start to show interest in a particular part of the search area it’s highly likely that your dog will too. As soon as this happens, believing that your dog has located the target scent source, you’ll start to look harder yourself. A vicious circle with the almost inevitable result – a false positive indication from your dog prompted by inadvertent cues from you.
Beware of your audience – Keep in mind that your dog will be sensitive to cues from your audience as well as from you, her handler. Encouraging your audience to talk to one another can be helpful here. The more attention they’re giving one another the less attention they’ll be paying to your dog as she searches the area. Be particularly alert to ‘audience silence’ and ‘breath-holding’ as your dog comes close to her target scent source. Without a doubt, your dog will notice it even if you don’t!
Want to know more about this subject? Just follow this link to False indications, Clever Hans and You.
3. Target Odour is not a “Motivational Magnet” for your dog
Familiarising your dog to the target odour refers to the process of Classical or Pavlovian conditioning that helps her form positive associations with the scent. What starts out as a completely irrelevant odour, something that your dog would not normally be attracted to, will, after conditioning, start to become highly relevant. The presence of the odour will predict that good things will follow. According to Hall et al (2014), the process of familiarisation enhances odour discrimination training and can reduce overall training time.
During the familiarisation process the target scent becomes what Berridge et al (2009) describe as a “motivational magnet” – something that has to be approached, often compulsively. This fits well with the notion of “scent obedience” – the idea that the target scent itself will act as a cue to your dog to search it out and stay with it.
For more information on the odour familiarisation process, you might like to follow this link to Smoothing out the Lumpy Bits: Why it might be wise to teach the indication before the search and all that back chaining stuff .
Self-evidently, if the target odour is not important to your dog – if you haven’t turned it into a motivational magnet for her – then she has no real reason to search it out or indicate to you that she has found it. Pair this situation with a reliance on a change in your dog’s behaviour (COB) – rather than a fully trained indication – to show you where the target odour might be located and you have the recipe for a Scent Detecting disaster. To clarify, any passing sniff at the “olfactory messages” left by another dog, or the odour of food crumbs dropped on the floor, will look like (and is) a COB and, as such, will be considered an indication. Unfortunately, it’s a false positive indication.
The simple take-home message – review your training. Take the time to familiarise your dog to the target odour and train a reliable, unambiguous, indication!
4. Your dog has not learnt what you thought you had taught her
Recently, I came across a video clip of a dog working a “Line-Up” comprising a number of identical containers. Apart from considerable hesitancy on the dog’s part to both start, and continue, working (which points strongly to issues of Handler Pressure and Incomplete / Inadequate Training), it was very obvious that the dog did not fully understand the task in hand. On approaching the Line-Up, the dog gave false positive indications on multiple containers before, eventually, arriving at the correct container – the one that actually held the target odour source. If the handler had not placed the target scent source there herself, her dog’s multiple false positive indications would have completely misled her. Unfortunately, on the evidence provided by this video clip, this is an unreliable Scent Detecting dog.
Although this type of issue can become evident during many different types of Scent Detecting challenge, it’s probably most obviously identifiable in Line-Up tasks. In this particular case, the handler obviously believed she had taught her dog to deliver a passive indication, with duration, when she located the target scent source. In reality, her dog had learnt to deliver a passive indication, with duration, when it encountered any of the containers. The dog had learnt a visual cue – the containers – rather than the intended target odour cue.
Again, in common with other false positive indications, this is a training issue. One that needs to be addressed by the handler if her dog is to progress any further.
5. Incomplete / Inadequate Training
A few things to consider when training your dog …
1. Do you have a plan?
Do you have a clear picture of what you want to achieve? Have you thought about the steps you might need to take to achieve this end goal? Do you have realistic expectations of just how long this goal might take to achieve? Have you assessed your dog’s current level of ability and used this as the starting point for your training? Are you prepared to take one, or, possibly more steps back in your training plan to ensure that any new training is building on firm foundations? Have you considered your dog’s current level of health, fitness and emotional well-being? Have you considered how long your training sessions should be to optimise learning? What about rewards? Is your training plan Evidence Based? Without a doubt, these are only some of the questions you should ask yourself before starting to formulate your training plan or schedule.
2. Have you split your training?
As discussed earlier, when it comes to Scent Detecting training, I’m a big advocate of splitting rather than lumping. Splitting requires you to break your training down into tiny, incremental, steps. Speed of progress is dictated by your dog rather than any pre-determined time-frame decided by you, her handler. This forces you to consider your dog’s understanding of the particular training task in hand, and her ability to perform the element reliably, before attempting to move on with your training. If you move on too quickly, poor understanding, or performance, of a particular training element will result in false positive indications as your dog struggles to work out what is expected of her.
3. Are you training or testing?
If you ever find yourself saying something like, “let’s just see if she can do this” then it’s highly likely that you’re testing your dog rather than training her. If your dog does happen to succeed at some challenge you’ve set her, this will have far more to do with luck than any carefully considered application of training skills on your part. Why should that matter? Well, it matters because, next time, your luck might just run out. As the saying goes, “It takes more than one Swallow to make a Summer” and it takes more than one lucky success to make a RELIABLE Scent Detecting dog.
and in conclusion …
False positive indications should never be an accepted part of Scent Detecting practice. Without a doubt, the most likely culprit will be a training issue. If you want a reliable Scent Detecting dog … STOP. ANALYSE and ADDRESS the issues even if that requires you taking many steps back in your training.
As with all of my blogs, I include a reference list. This allows you to investigate the topic a little further, check out the sources of my information and decide for yourself whether my interpretations of the literature are an accurate reflection of the author’s work. Happy reading.
© Scent : Detect : Find Ltd – 2021
References / Further Reading
- Berg J (2018) Lumping and splitting. Science. 359. 6382. 1309
- Berridge KC, Robinson TE and Aldridge JW (2009) Dissecting components of reward; ‘Liking’, ‘Wanting’ and Learning’. Current Opinion in Pharmacology. Feb. 9(1): 65-73
- DeChant MT, Bunker PC and Hall NJ (2021) Stimulus Control of Odorant Concentration: Pilot Study of Generalization and Discrimination of Odor Concentration in Canines. Animals. 11. 362
- Custance D and Mayer J (2012) Empathic-like responding by domestic dogs (Canis familiaris) to distress in humans; An exploratory study. Animal Cognition. September. Vol 15. Issue 5. 851-859.
- Gadbois S (2016) Modern Dog Training and Science: Are We Doing it Right? Common Misconceptions and Pitfalls. The Science of Sniffer Dogs. Oxford: Two Day Conference
- Galambos A, Gergely A, Kovacs AB and Topal J (2021) Affect Matters: positive and negative social stimulation influences dogs’ behaviour in a subsequent situation involving an out-of-reach object. Applied Animal Behaviour Science. Journal Pre-Proof
- Goldblatt A, Gazit I and Terkel J (2009) Olfaction and Explosive Detector Dogs. IN: Helton WS (Ed) Canine Ergonomics. The Science of Working Dogs. London: Taylor and Francis Group
- Hall NJ, Smith DW and Wynne CDL (2014) Effect of odor pre-exposure on acquisition of an odor discrimination in dogs. Learning and Behavior. Jan
- Helton WS (2009) Overview of Scent Detection Work. Issues and Opportunities. IN: Helton WS (Ed) Canine Ergonomics. The Science of Working Dogs. London: Taylor and Francis Group
- Jackson J (2005) The Clever Hans effect – a horse’s tale. Critical Thinking. http://www.critical-thinking.org.uk/pdf/clever-hans.pdf
- Laing DG, Legha PK, Jinks AL and Hutchinson I (2003) Relationship between molecular structure, concentration and odor qualities of oxygenated aliphatic molecules. Chemical Senses. 28. 57-69
- Learning Centre Valutek Blog (2016) Odor and Color Determine the Quality of Your Cleanroom Gloves. https://blog.valutek.com/home/odor-color-determine-the=quality-of-your-cleanroom-gloves# Accessed: 27.8.2020
- Lehnert M P and Weeks E N I (2016) Trained Dogs in Insect Detection. IN: Jezierski T, Ensminger J and Papet L E (Eds) Canine Olfaction Science and Law. Advances in Forensic Science, Medicine, Conservation, and Environmental Remediation. London: Taylor and Francis Group.
- Lit L, Schweitzer JB and Oberbauer AM (2011) Handler beliefs affect scent detection dog outcomes. Animal Cognition. 14: 387-394
- Mariti C, Falaschi C, Zilocchi M, Fatjo J, Sighieri C, Ogi A and Gazzano A (2017) Analysis of the intraspecific visual communication in the domestic dog (Canis familiaris): a pilot study on the case of calming signals. Journal of Veterinary Behavior Clinical Applications and Research. Volume 18. March-April. 49-55.
- Rugaas T (2006) On Talking Terms with Dogs: Calming Signals. 2nd Ed. Washington: Dogwise Publishing
- Szeiti V and Miklosi A (2003) When dogs seem to lose their nose: an investigation on the use of visual and olfactory cues in communicative context between dog and owner. Applied Animal Behaviour Science. Sept 5. Vol 83. Issue 2. 141-152
- Udell MAR, Dorey NR and Wynne CDL (2010) What did domestication do to dogs? A new account of dogs’ sensitivity to human actions. Biological Reviews. 85. 327-345
- Wilson DA and Stevenson RJ (2006) Learning to Smell. Olfactory Perception from Neurobiology to Behavior. Baltimore: The Johns Hopkins University Press
- Zerubavel E (1996) Lumping and Splitting: Notes on Social Classification. Sociological Forum. 11.3. 421-423.