Introduction

In dog Agility, handlers and dogs compete in a race to get around an obstacle course in the fastest possible time without making any mistakes (knocking bars over, going the wrong way, and so on). It’s pretty fun, but some people take it a bit more seriously than others! Recently, I was sucked into a debate about whether the taller dogs (‘500’ dogs, who are in the height category that encompasses most Border Collies, as well as my own little Kelpie rescue dog Dash) run faster than dogs of other heights (200, 300, 400, and 600). As it turns out, somewhat unsurprisingly, they do. You wouldn’t have really thought you’d need statistics to show that, but some people are hard to persuade, so since I had access to some data on the topic, I wrote an R script to show that.

A small dog doing Agility as fast as it can

This analysis is based on data extracted from K9 Entries (https://www.k9entries.com/), for both Victorian and Queensland Agility competitions from 2016 to 2018. Thanks to Alison Muddle for extracting the data and to Judy Kloeden for initial analyses. There are almost 18,000 individual entries in this analysis.

All of the data analyses were conducted in R Statistical Software and compiled using R Markdown in the R Studio package. Note that where there was more than one entry per dog, entries were averaged for the analysis. In addition, speeds faster than 10 m/s and slower than .8 m/s were cropped from the analysis to reduce the effect of outliers. Pairwise comparisons are corrected with the Tukey method, which is fairly conservative.

Results for Novice Agility

Overall, there was a strongly significant difference between rates of travel for the different heights, F(4, 594) = 25.98, p < .001. The 500 height dogs were significantly faster than all the other heights, all p-values < .01, corrected (see tables below for estimated marginal means and details).

Results for Excellent Agility

Overall, there was a strongly significant difference between rates of travel for the different heights, F(4,346) = 11.08, p < .001. The 500 height dogs were significantly faster than 300 and 200 but not 400 and 600 dogs, with p-values < .01, corrected (see tables below for estimated marginal means and details).

Results for Masters Agility

Overall, there was a strongly significant difference between rates of travel for the different heights, F(4,407) =23.43, p < .001. The 500 height dogs were significantly faster than all other heights, with p-values < .001, except for the difference between 500 and 600 (p = .011), corrected (see tables below for estimated marginal means and details).

Results for Open Agility

Overall, there was a strongly significant difference between rates of travel for the different heights, F(4,392) = 14.68, p < .001. The 500 height dogs were significantly faster than all other dogs, with p-values < .01, corrected (see tables below for estimated marginal means and details).

Jumping Results