The stubborn location: Spatial location is more resistant to removal than non-spatial features in visual working memory.

Working memory (WM) removal serves the critical function of freeing up limited capacity to accommodate task-relevant information. While previous research has primarily focused on the removal of entire objects, it remains unclear whether all features within an object share a common fate during removal. In particular, given the privileged status of spatial location across various cognitive processes, it is an open question whether location information is treated differently during removal. To address this issue, we tested the stubborn-location hypothesis that spatial location would show lower removal effectiveness than non-spatial features. The alternative possibilities were that location would show either comparable or greater removal effectiveness relative to non-spatial features. Across two experiments (N = 197), participants first encoded both the spatial location and color of four items. Following a retro-cue, they were instructed to remove three items from memory. On valid trials, either the location or the color of the remaining item was tested. Critically, in low-prevalence trials, either feature of one of the removed items was tested. Results consistently supported the stubborn-location hypothesis, showing that spatial location had lower removal effectiveness in WM than color. Specifically, removal effectiveness - defined as the accuracy drop from valid to low-prevalence trials - was lower for location than for color. Furthermore, location memory remained above chance accuracy on the first low-prevalence (surprise) trials, whereas color did not. These findings were robust across both simultaneous (Experiment 1; position cue) and sequential (Experiment 2; order cue) presentation formats. Our results highlight the privileged status of location in WM removal.