The method

The test never asks whether your dog is fearful, sociable or stubborn. It asks what they do in a named situation: when the doorbell goes, when an unknown dog walks straight at them, when you close the door, when a game stops dead.

The reason is simple. An owner asked for a judgement answers with the picture they hold of their dog. An owner asked for a frequency answers with what they have seen. The two answers do not look alike, and only the second one can be used.

Twenty eight situations are asked, spread over eight dimensions, three or four each. Five of them are worded the other way round, so a high answer lowers the score instead of raising it. That is what stops a hurried owner ticking the same column from start to finish without it showing.

Each answer is worth 0 to 4 on a five point frequency scale. A dimension's score is the mean of its items, rescaled to 100. Nothing else: no hidden weighting, no correction, no curve.

The never observed option is not a zero. It removes the item from its dimension's mean. A dog who has never met a child must not be counted as bad with children, they must be counted as unrecorded on that point.

When every situation in a dimension is marked never observed, the dimension comes back as not assessed, and the report says so. We would rather show an empty box than a number we made up.

Two instruments serve as landmarks for the structure. The C-BARQ, developed at the University of Pennsylvania, describes dog behaviour across a set of factors and is widely used in research. The MCPQ-R, developed at Monash University, describes dog personality across five factors.

What MyDogType borrows from them is an idea about carving: a dog's behaviour is better described by several independent axes than by a single label, and those axes are measured through situations the owner reports.

What MyDogType does not borrow: none of their items, none of their scales, none of their thresholds. The twenty eight situations were written for this tool, the eight dimensions are MyDogType's own carving, and the test has not been through the validation process those instruments went through. MyDogType is not a short C-BARQ and must not be presented as one.

Each archetype is defined by a set of weights over the eight dimensions: what has to be high, what has to be low, and how much it matters. Your dog's eight scores are matched against the ten sets of weights, and the closest fit wins.

One correction is applied before the comparison. Some archetypes would mechanically score better than others simply because their combination is more common. Each raw fit is therefore measured against the distribution of its own archetype before being compared with the other nine. Without that correction, one archetype was taking fifteen times more dogs than the rarest one.

The archetype is a summary, not the result. Two dogs of the same archetype can have very different scores, and it is the eight dimension profile that carries the information.

The report shows where your dog sits against others on each dimension. While the database holds fewer than five hundred profiles, that comparison is computed against a documented reference distribution, and it is labelled an estimate everywhere it appears.

Past five hundred profiles the calculation switches automatically to the dogs actually assessed, and the label changes. We will never present an estimate as a measurement.