Skip to main content
Account
Fig. 4 | Animal Cognition

Fig. 4

From: Testing cognitive models of decision-making: selected studies with starlings

Fig. 4

Response time as a measure of preference. a Starlings faced a fixed (dark) and a variable (light) option, where variability was in reward size or delay. Consistent with predicted risk aversion for amounts and risk proneness for delays, latencies to peck in no-choice trials (shown) were shorter for the fixed reward and for the variable delay option. Choice proportions in the same experiment showed risk proneness for delays but did not differ significantly from random for amounts (not shown). Adapted with permission from Reboreda and Kacelnik, Behavioral Ecology (1991)—Fig. 3—https://doi.org/10.1093/beheco/2.4.301. b Starlings were trained with 15 pairs of options differing in amount and/or delay to food, and then given choices between members of each training pair. Choice percentage (ordinate) was predicted using relative value of latencies (latency A /(latency A + latency B) × 100) to respond in probe trials (abscissa). Each treatment pair is represented by a different shading. Also included are the identity diagonal and two linear regressions. One of them (----) is unconstrained and the other (—) is constrained to pass through the origin. The R2 values are generated by the average obtained results. The fit of SCM predictions was the best among a set of candidate predictive models. Adapted with permission from Shapiro, Siller and Kacelnik, Journal of Experimental Psychology: Animal Behavior Processes (2008)—Fig. 8f—https://doi.org/10.1037/0097-7403.34.1.75. c Difference of response times in single-option trials minus response times in choice trials of the same option against all other options (n = 9, mean ± SEM). Inset shows individual results. For illustration, in the rare trials in which birds chose one of the two poorest option in the environment (1 pellet / 10 s) when that option was paired with any of the other 5 options, response time was ~ 25 s shorter than when the same option was met in a sequential (single-option) trial. Adapted with permission from Monteiro et al. (2020)—Fig. 3c—https://doi.org/10.1371/journal.pbio.3000841.g001

Back to article page