Your Dog Probably Isn't as Troubled as That Study Made You Think

A paper published in 2020 in Scientific Reports became something of a landmark in dog behaviour circles. It surveyed 13,715 Finnish dogs and found that 72.5% showed at least one problematic behaviour, with fear and aggression cropping up at rates that genuinely alarmed owners and vets alike. That figure got cited everywhere. It shaped how a lot of people talked about whether their dog was "normal".

New analysis published on October 5, 2026 and covered by dvm360 is now pushing back on those conclusions, and the challenge is worth understanding if you've ever been told, or told yourself, that your dog has a behaviour problem.

What the original research actually claimed

The Finnish study leaned on owner-reported data. Owners filled in questionnaires about whether their dogs showed fear, aggression, compulsive behaviours, and so on. That's a reasonable method, but it comes with a structural problem: owners don't have a calibrated baseline. If your dog barks at the postman, is that "aggression" or just... a dog being a dog? The threshold for what counts as a problem varies enormously from person to person.

The new analysis challenges whether those reported rates reflect genuinely problematic behaviour or whether they capture a much broader range of normal canine responses that owners, without clinical training, interpreted as concerning.

This matters more than it might seem. If vets and behaviourists are working from inflated prevalence figures, treatment thresholds shift. Owners get worried. Medicating or intensively managing behaviours that fall within the normal spectrum becomes more common. And the dog at the centre of all this anxiety is often fine.

The gap between "my dog does this" and "my dog has a problem"

A dog that freezes around strangers, barks when startled, or resource-guards a favourite toy is not automatically anxious, aggressive, or compulsive. Those are responses. Whether they cross into something clinically significant depends on frequency, intensity, how much they interfere with the dog's daily functioning, and how quickly the dog recovers.

That last one, recovery time, is underrated. A dog who barks at a loud noise and then goes back to sleeping in thirty seconds is very different from a dog who remains visibly stressed for hours. Both owners might tick the same box on a questionnaire.

The researchers challenging the Finnish data are essentially asking: were owners measuring the behaviour, or measuring their own discomfort with it?

Honest answer: probably both, in unknown proportions.

What this means if you're worried about your dog

The practical takeaway isn't "ignore concerning behaviour". Genuine anxiety, fear-based aggression, and compulsive disorders are real, they do affect a meaningful minority of dogs, and they do warrant attention.

The takeaway is that a single incident, or even a pattern of mild incidents, doesn't automatically mean your dog is broken. Context and frequency matter. One growl in a specific high-stress situation is not the same signal as repeated, unprovoked aggression.

A few things worth actually doing:

• Keep a log. Note what happened, when, what preceded it, and how long it lasted. A week of actual observations is worth more than a vague sense that "he's been weird lately".
• Distinguish discomfort from distress. A dog who pulls away from something unfamiliar is showing a preference. A dog who shakes, pants excessively, and can't settle is showing distress. The clinical interventions for these are different.
• Talk to a vet before assuming a behavioural diagnosis. Some behaviours that look like anxiety are rooted in pain or a physical condition, and no amount of desensitisation work will fix a sore hip.

If you're tracking your dog's behaviour over time and want something more structured than a notebook, Tailo's episode tracking is designed for exactly this, logging specific incidents with context so patterns become visible rather than staying as a general worry in the back of your head.

Why this kind of research correction is a good thing

Science self-correcting is how it's supposed to work. The 72.5% figure from the Finnish study travelled far beyond its original context, got simplified in transit, and ended up as a kind of background anxiety for a lot of dog owners who read about it secondhand.

The researchers publishing this challenge aren't saying dog behaviour problems don't exist. They're saying the measurement was probably off, and that we should be more careful about what we count as a problem before we start treating it as one.

That's a calibration, not a dismissal. And for owners who've spent real money and real worry on a dog who might just be a slightly dramatic, perfectly healthy animal, it might be something of a relief.

The uncomfortable version of this is also true, though: some genuinely distressed dogs have owners who don't fill in the questionnaire boxes, because the owners have normalised what they're seeing. Measurement errors run in both directions.