Dog training statistics · 2026 · computed from library v0.3.1

0 of 17 published protocols are backed by a controlled trial.

That is the whole field, not a complaint about it. We encoded every named separation-anxiety and leash-reactivity protocol we could source into a typed library — attribution, evidence grade, cadence, sources — and these are the counts that fall out. Every number on this page is computed from that library, not written into it, and each record’s underlying citation is public on the graded protocol list.

17 protocols graded

The evidence-grade distribution

One grade per protocol, weakest phrasing bound to each: “studies show” is only ever attached to trial-backed records — which is why it appears nowhere on this site.

4 of 17

Authority-grade: 4 of 17 named protocols — a position statement or clinical text from a veterinary or behaviour body backs it.

Source · per-protocol grades and citations

13 of 17

Practitioner-grade: 13 of 17 named protocols — a named, credentialed practitioner teaches it, but no trial has ever tested it.

Source · per-protocol grades and citations

Who the protocols come from

Attribution counts

Attribution is part of what a buyer pays for. Where a protocol has no single named author, the library says so instead of inventing one.

14 of 17

14 of 17 protocols trace to a named practitioner, author, or veterinary body — Overall, DeMartini, Naismith, Donaldson, McDevitt, Stewart, ASPCA, VCA among them.

Source · the attribution on every record

3 of 17

3 of 17 protocols are trainer-community lineage with no single credentialed author identified — labelled exactly that way rather than laundered into an authority.

Source · the attribution on every record

16

16 distinct attributions across the 17 protocols, carried on 22 protocol-level source citations.

Source · the graded protocol list

The rest of the library, counted

Honesty-layer statistics

The same library records where sources disagree, what is folklore, and when the correct action is to refuse the sale.

9 of 33

9 of the library's 33 red flags are absolute refusals to sell — a bite that broke skin is refunded and referred to a veterinary behaviourist, never coached remotely.

Source · what is sold and what is refused

7 of 8

8 recorded disagreements between credentialed sources; 7 are shown to buyers with the side we take and why.

Source · the separation-anxiety guide

5

5 quarantined folklore claims — held only so they can be recognised and corrected, never emitted as instruction.

Source · the separation-anxiety guide

7

7 named places where the published field has no data at all — no separation-anxiety dropout rate, no "% of cases needing medication", no reactivity treatment-timeline research. Recorded so nobody here is tempted to invent one.

Source · the graded protocol list

Method

How these numbers are made

The library encodes published protocols from their original sources — clinical texts, position statements, named practitioners’ own programmes. Every record carries its citation; escalation and refusal copy may only rest on documents we read ourselves, enforced at build time. The counts above are computed from the library on every deploy, and this page is refreshed each January. Read the protocols themselves in the separation-anxiety guide and the leash-reactivity guide, or the full graded list at /protocols.