Exploring the relationship between fluid intelligence and visual working memory for simple features vs. real-world objects

Poster Presentation 36.460: Sunday, May 19, 2024, 2:45 – 6:45 pm, Pavilion
Session: Visual Memory: Working memory and objects, features

Kaira Shlipak1, Yong Hoon Chung1, Viola Störmer1; 1Dartmouth College

Visual working memory performance has been shown to strongly correlate with measures of fluid intelligence. These correlations have been observed with working memory tasks that used simple visual stimuli such as colors or simple shapes (e.g., Fukuda et al., 2010). However, recent research has demonstrated that visual working memory performance is increased when people are asked to remember meaningful stimuli such as real-world objects relative to simple feature items (e.g., Brady & Störmer, 2022). How does performance on working memory tasks using real-world objects relate to measures of fluid intelligence? Here, we examined this question by measuring working memory performance for simple stimuli and real-world objects, and correlated individuals’ performance on these tasks with measures of fluid intelligence. Participants (N = 103) completed different visual working memory tasks that used either colored circles or images of real-world objects as well as two different intelligence tests: the Raven’s Advanced Progressive Matrices Test of fluid intelligence and the C-test of crystallized intelligence. First, we replicated the real-world object advantage in visual working memory, such that memory performance was higher for real-world objects relative to colors. Second, we found reliable correlations between all working memory measures (both objects and colors) and fluid intelligence (r2 = 0.38, p < 0.0001 for object condition; r2 = 0.25, p = 0.01 for color condition), and no significant differences between these correlations across stimuli types. Notably, working memory performance was not reliably correlated with crystallized intelligence in either object or color conditions, which indicates that crystallized knowledge is not linked to visual working memory performance even for meaningful objects. These findings suggest that visual working memory performance in tasks using both simple features and complex real-world objects relies on shared processes that relate to fluid intelligence.