Assess a dog for warning work by scoring alertness, confidence, recovery after a startle, bonding to people, reaction to a sudden noise and tolerance of handling, on a lead and in repeated sessions on different days, without provoking it. Any dog that shows fear-biting, guards food against people or never recovers from a startle fails the assessment however high it scores elsewhere.
Part of Guard Dog Training
Temperament assessment is a structured set of observations of a dog’s reactions to people, noise, novelty and handling, used to decide whether it will make a safe, reliable household alarm.
Safe Dogs Fail Gently And Recover Fast
The parent article lists what to evaluate in any dog:
- Alertness — does it notice environmental changes? A dog that ignores a stranger walking into the room is a pet, not a sentinel.
- Confidence — does it investigate novel things or cower? A fearful dog is an unreliable alarm; it may freeze or flee instead of alerting.
- Recovery — when startled, how fast does it return to normal? Fast recovery means a stable temperament.
- Bonding — does it orient to humans? A dog that bonds to its handler or family will naturally guard them; one indifferent to humans will not.
- Sound sensitivity — test with a sudden loud noise such as a clap or a dropped pan. An ideal candidate startles, orients to the source, then investigates. A poor candidate panics, runs or shows prolonged distress.
The assessment below adds two more checks that matter for safety: how the dog tolerates handling and how it behaves around food. A warning dog that is steady but cannot be handled cannot be treated when injured and may bite a child who touches it; a dog that guards food will bite someone who walks past its bowl.
What You Need
- A fenced or enclosed space the dog does not know well, and a second familiar space
- A lead and a harness, and a muzzle or cloth strip for safety
- Two adults, one to hold the lead and one to run the tests
- A watch to time recovery
- A notebook to score each test
- No children as test subjects, ever
Method: Six Observations, Repeated
Step 1: Orientation and bonding. Let the dog settle with the handler. Does it check in, follow, look to the handler when something new happens? Score 0 for indifferent, 1 for some interest, 2 for clear bonding.
Step 2: Novelty. Open an umbrella or lay a tarp on the ground at a distance while the dog is held. Score 0 for panic or flight, 1 for cautious avoidance, 2 for approach and investigation.
Step 3: Sudden noise. Clap or drop a pan from a distance behind the dog. Score 0 for panic, 1 for startle with slow recovery, 2 for startle, orientation and investigation. Time the recovery, the interval from the noise to normal behaviour, and compare dogs on the same test.
Step 4: Stranger. A calm person the dog does not know walks into view at a distance and stands. Watch the dog’s body: relaxed alertness and a bark are good; stiffness with a fixed stare and a lunge, or a dog hiding behind the handler, are not. No stranger approaches the dog or makes eye contact; no one threatens it.
Step 5: Handling. The handler touches paws, ears, tail and gently lifts a lip. Score 0 for snapping, 1 for tolerating with stress, 2 for relaxed acceptance.
Step 6: Food. Put the bowl down and watch from a distance. A dog that freezes, stiffens, growls or eats faster when a person walks past at a distance is guarding. Never reach into the bowl; asking about its history is safer than finding out with a hand.
Step 7: Repeat on another day, in another place. A dog that is quiet in a new place may be only stunned. Score again.
Worked Example: Scoring And What To Reject
Six observations scored 0-2 give a total of 0-12. For one dog, 2 + 2 + 2 + 1 + 2 + 2 = 11, a good result; for another, 2 + 2 + 2 + 2 + 2 + 0 = 10, a high total with a 0 for food. The second dog is rejected because a 0 on a safety test cannot be bought back with points elsewhere.
The rule used here is mine and not from any source: any 0 on handling, food or the stranger test, or any score of 0 on the noise test that does not recover by the end of the session, rejects the dog. Keep the scoring table, repeat the tests after a week, and compare. A dog whose scores swing from 2 to 0 between sessions is not stable.
Common Mistakes
- Provoking the dog to see what it will do. Teasing, striking or threatening a dog is cruel, unsafe and tells you nothing about how it behaves normally.
- Testing a hungry, exhausted or frightened dog. It shows you stress, not temperament.
- Judging a puppy as if it were an adult. I recall that puppy tests predict adult behaviour only weakly, though I cannot cite a study.
- Ignoring history. A bite to a person is more informative than any test.
- Testing alone. A second adult controls the dog and watches the dog’s body.
Safety
Test on a lead, keep children out, and never provoke
A frightened or aggressive dog bites when it is cornered or surprised. Run every test with the dog on a lead and harness in an enclosed area, with a muzzle available, and with only adults present. Children are most often bitten by dogs they know, and a child who is made part of a test can be hurt; no child takes part. If a dog bites during testing, irrigate the wound at least 5 minutes and leave it open; see Wound Management. A dog of unknown vaccination status is also a rabies concern.
What’s Next
- Learn next: Alert Training — training a dog that has passed
- Combine with: Stranger Differentiation — the harder version of the stranger observation