Table 2. Repeatability estimates for behavioural responses in the attention bias tests, across all trials and between pairs of adjacent trials using raw data and across all trials using ranked data.

Repeatability was estimated using the package rptR. Breed was retained as a fixed effect in the models for eating frequency and latency to eat only, based on AIC and BIC.

Behaviour	All trials			Trials 1 and 2 only			Trials 2 and 3 only			All trials (ranked data)
	r	s.e.	CI	r	s.e.	CI	r	s.e.	CI	r	s.e.	CI
Vigilance (s)	0.58	0.06	[0.44–0.68]	0.55	0.08	[0.38–0.68]	0.70	0.06	[0.57–0.79]	0.56	0.06	[0.43–0.67]
Zones crossed (n)	0.47	0.07	[0.34–0.59]	0.43	0.09	[0.23–0.59]	0.59	0.07	[0.44–0.72]	0.46	0.07	[0.32–0.59]
Eating frequency (n)1	0.36	0.12	[0.11–0.56]	0.50	0.17	[0.13–0.76]	0.21	0.13	[0.00–0.48]	0.35	0.07	[0.19–0.47]
Time at door (s)	0.29	0.07	[0.14–0.43]	0.46	0.09	[0.27–0.61]	0.35	0.09	[0.15–0.52]	0.36	0.07	[0.21–0.49]
Latency to eat (s)	0.20	0.07	[0.07–0.34]	0.31	0.10	[0.10–0.50]	0.12	0.10	[0.00–0.34]	0.23	0.07	[0.09–0.37]
Attention to dog wall (s)	0.08	0.06	[0.00–0.22]	0.05	0.08	[0.00–0.27]	0.28	0.10	[0.10–0.46]	0.33	0.07	[0.17–0.46]

Note:

¹Original scale approximations for repeatability estimates are presented.