The Problem of Testimony
The talk will detail work drawing on behavioural results, formal analysis, and computational modelling with agent-based simulations to unpack the scale of the challenge humans face when trying to work out and factor in the reliability of their sources. In particular, it is shown how and why this task admits of no easy solution in the context of wider communication networks, and how this will affect the accuracy of our beliefs. The implications of this for the shift in the size and topology of our communication networks through the uncontrolled rise of social media are discussed.
Commonly used face cognition tests yield low reliability and inconsistent performance: Implications for test design, analysis, and interpretation of individual differences data
Unfamiliar face processing (face cognition) ability varies considerably in the general population. However, the means of its assessment are not standardised, and selected laboratory tests vary between studies. It is also unclear whether 1) the most commonly employed tests are reliable, 2) participants show a degree of consistency in their performance, 3) and the face cognition tests broadly measure one underlying ability, akin to general intelligence. In this study, we asked participants to perform eight tests frequently employed in the individual differences literature. We examined the reliability of these tests, relationships between them, consistency in participants’ performance, and used data driven approaches to determine factors underpinning performance. Overall, our findings suggest that the reliability of these tests is poor to moderate, the correlations between them are weak, the consistency in participant performance across tasks is low and that performance can be broadly split into two factors: telling faces together, and telling faces apart. We recommend that future studies adjust analyses to account for stimuli (face images) and participants as random factors, routinely assess reliability, and that newly developed tests of face cognition are examined in the context of convergent validity with other commonly used measures of face cognition ability.
Error correction and reliability timescale in converging cortical networks
Rapidly changing inputs such as visual scenes and auditory landscapes are transmitted over several synaptic interfaces and perceived with little loss of detail, but individual neurons are typically “noisy” and cortico-cortical connections are typically “weak”. To understand how information embodied in spike train is transmitted in a lossless manner, we focus on a single synaptic interface: between pyramidal cells and putative interneurons. Using arbitrary white noise patterns injected intra-cortically as photocurrents to freely-moving mice, we find that directly-activated cells exhibit precision of several milliseconds, but post-synaptic, indirectly-activated cells exhibit higher precision. Considering multiple identical messages, the reliability of directly-activated cells peaks at a timescale of dozens of milliseconds, whereas indirectly-activated cells exhibit an order-of-magnitude faster timescale. Using data-driven modelling, we find that error correction is consistent with non-linear amplification of coincident spikes.