Conor Smithson¹, Nurit Gronau², Isabel Gauthier¹; ¹Vanderbilt University, ²The Open University of Israel

Individuals differ in the level of detail they can visually encode. Prior work (Gronau et al., submitted) suggested that a domain-general perceptual ability (Op) for discriminating among object identities predicts memory for contextual detail, specifically, where an object was seen. Additionally, Op predicted the 'meaningfulness effect', wherein the locations of objects that were independently rated as highly familiar were better remembered than the locations of objects rated as rare or difficult to recognize. Here, we extend this work by experimentally manipulating meaningfulness by changing the level of processing of the images. Storybook images were encoded during a repetition detection task in which participants attended to either the artist style or the deeper semantic content. We tested whether Op predicted the meaningfulness effect on memory for location as well as for temporal position within a sequence. We controlled for low-level visual ability, working memory, and intelligence. Although preregistered structural equation models were inconclusive due to imprecise parameter estimates, multiple regression analyses replicated the predictive effect of Op on memory for location, as well as the meaningfulness effect on location memory. Op did not, however, predict memory for sequence position, nor the meaningfulness effect on this type of memory. Domain-general scene perception ability was an even stronger predictor, capturing all that was accounted for by Op, as well as additional variance, likely because it also captures variance in the ability to use color, texture, and spatial relations between objects. Scene perception also predicted memory for sequence position, but like Op, it did not predict the meaningfulness effect on memory for sequence position. High-level visual discrimination abilities appear to predict the meaningfulness effect on where stimuli were seen, but not on when they were seen. This is consistent with previous research showing that object location and object identity are not perceived entirely independently.

Acknowledgements: This work was supported by the David K. Wilson Chair Research Fund and NSF BCS Award 2316474