Xueyi Huang¹, Angela Shen², Emil Olsson², Michaela Klimova¹, Ali Ekhlasi², Oğul Can Yurdakul², Megan A. K. Peters^2,3, Jorge Morales¹; ¹Northeastern University, ²University of California, Irvine, ³University College London

The vividness of visual mental images is conventionally measured on a 1–5 scale. However, because mental imagery is inherently subjective, vividness alone provides limited information about the underlying mental images. Can vividness ratings be linked to specific visual properties on an individual basis? Are people accurate at reporting visual properties from imagination? Is there a “shared” vividness concept across people? We investigated these foundational questions across four experiments leveraging a novel method that captures trial-by-trial links between vividness and visual properties: Subjects imagine real-life objects as vividly as possible, rate their vividness, and either reconstruct or classify their imagery’s visual properties—sharpness, opacity, and saturation—on Voronoi tessellations. Experiments 1 & 2 (n=200 each, preregistered) revealed that while subjects associate specific visual-property combinations with particular vividness degrees, there are large individual differences. Experiment 3 (n=50) validated the reconstruction method: Subjects viewed realistic images altered with random property values and accurately reconstructed them on Voronois. Experiment 4 (n=100) replicated Experiment 1 in a new sample. Subjects then rated the vividness of realistic images whose visual properties were either taken from the subject's own reconstructions or from the averages of all subjects during the imagery task. Consistent with Experiment 2 which used a double-pass procedure to show within-subject reliability, ratings of subjects’ reconstructions were highly correlated with subjects’ original vividness ratings. Surprisingly, subjects’ responses were even more strongly correlated with the original ratings when using group-average visual properties, making them an even more stable index of vividness. Together, our findings demonstrate that despite their subjective nature, mental images and their experienced vividness can be described accurately and objectively. Crucially, our novel paradigm captures both individual differences and converging interpretations in how visual properties map onto vividness. It appears that we do know what we’re talking about when we talk about vividness.

Acknowledgements: This work was supported by the Templeton World Charity Foundation, Inc. (funder DOI 501100011730) under the grant https://doi.org/10.54224/22032 to (JM & MAKP)” and in part by the Canadian Institute for Advanced Research (CIFAR, Fellowship in Brain, Mind, & Consciousness; to MAKP).