Attention control contributes to ensemble perception regardless of selection demands

Ensemble perception refers to the ability to extract summary statistics (e.g., mean, variance) from groups of similar objects. In some ensemble perception tasks, participants must select specific items from a set before calculating a summary statistic (e.g., the average size of red circles among blue ones). It is generally assumed that this selection process does not alter the ability being measured. However, research in working memory suggests that introducing selection into tasks can shift the focus of what is measured-from working memory capacity to attention control. In this study, we examined whether selection alters ensemble perception's role in mean discrimination tasks using a latent variable approach. We found that ensemble perception and attention control are correlated at a construct level and that both contribute approximately equally to performance in mean discrimination tasks, irrespective of task format. We found no support for the idea that selective ensemble perception tasks are more strongly predicted by attention control than nonselective ensemble tasks. PUBLIC SIGNIFICANCE STATEMENT: We find that ensemble perception-the ability to gauge collective properties of a group of objects-relies on attention control, even in tasks designed to measure ensemble perception alone. This suggests that when people assess group characteristics, like the average size or color of items, they also use mental resources to focus selectively on specific items. Understanding interactions between perception and attention could improve the design of tasks that measure cognitive abilities, helping researchers better distinguish between different types of mental skills.