Jeffery et al.'s target article explores behavioural and neurobiological studies from many species and draws attention to the gaps in the literature on human three-dimensional navigation. Our own work addresses this gap by using a novel paradigm to examine human searching behaviour in large-scale space. We have used this setup in a series of adult and developmental studies (Pellicano et al. Reference Pellicano, Smith, Cristino, Hood, Briscoe and Gilchrist2011; Smith et al. Reference Smith, Gilchrist and Hood2005; Reference Smith, Hood and Gilchrist2008; Reference Smith, Hood and Gilchrist2010).

Our large-scale search laboratory contains an array of green light-emitting diode (LED) light switches embedded in the floor. Participants press illuminated switches to discover targets that change colour when pressed. Potential targets can be manipulated by colour, saliency, and probability of distribution. Our paradigm preserves the experimental rigor of classic visual search, by recording button presses to millisecond accuracy, while scaling up to large-scale search in which the human participant has to move through the search array. This provides a first step towards investigating the processes involved in human navigation, which addresses a concern that has been raised regarding the validity of modelling human search solely via two-dimensional computer monitor–based search (Dalrymple & Kingstone Reference Dalrymple and Kingstone2010; Foulsham et al. Reference Foulsham, Walker and Kingstone2011; Foulsham & Kingstone Reference Foulsham, Kingstone, Morimoto, Istance, Mulligan, Qvarfordt and Spencer2012; Ingram & Wolpert Reference Ingram and Wolpert2011).

Work comparing visually guided versus hidden target conditions revealed increased search latencies when the target is not visually available (Smith et al. Reference Smith, Hood and Gilchrist2008). In both visually and non-visually guided conditions there were fewer revisits to previously explored locations than are present in classic computer screen visual search studies, emphasizing a key distinction between visual search and large-scale search (Smith et al. Reference Smith, Hood and Gilchrist2008).

Jeffery et al. also highlight a gap in human developmental work on navigation and spatial cognition, which is crucial to building an understanding of how humans develop complex spatial representations. Research in our paradigm has investigated the role of short-term memory in preventing revisits to previously inspected locations (Smith et al. Reference Smith, Gilchrist and Hood2005). Children made significantly more revisits to previously examined search locations when performing a more effortful search with the non-dominant hand, suggesting that when the task is more effortful, individuals are less able to optimize search. In addition, although children's general fluid intelligence was unrelated to search time, there was a significant relationship between search latency and visuo-spatial short-term memory (measured via Corsi blocks). These data highlight the role of spatial working memory in the development of efficient exploration of large-scale space (Smith et al. Reference Smith, Gilchrist and Hood2005).

Further work focused on the learning of likely target locations (Smith et al. Reference Smith, Hood and Gilchrist2010). Here, targets were located on one side of the room in 80% of trials. Participants were able to learn likely targets when both the start location and search locations were fixed throughout the task. However, when room-based and reference frames were disrupted, there was no evidence of learning, suggesting that encoding of likely locations depends on a combination of egocentric and allocentric cues (Smith et al. Reference Smith, Hood and Gilchrist2010).

This same paradigm has been used in individuals with autism, following the suggestion by Baron-Cohen (e.g., Baron-Cohen Reference Baron-Cohen2008) that the exceptional visual search skills of children with autism are due to an increased aptitude at systemizing spatial cognition. However, children with autism were less efficient, less systematic, less optimal, and overall less likely to follow probabilistic cues than ability-matched typical children. This provides further evidence that visual search and large-scale search do not always share the same cognitive mechanisms (Pellicano et al. Reference Pellicano, Smith, Cristino, Hood, Briscoe and Gilchrist2011).

Recent work has investigated the role of saliency in building spatial representations. In work with adult participants, the perceptual salience of search locations was manipulated by having some locations flashing and some static, a paradigm adapted from established visual search literature (Theeuwes Reference Theeuwes1994; Yantis & Jonides Reference Yantis and Jonides1984). Adults were more likely to search at flashing locations, even when explicitly informed that the target was equally likely to be at any location (Longstaffe et al. Reference Longstaffe, Hood and Gilchrist2012). We propose that attention was captured by perceptual salience, leading to an automatic bias to explore these targets. When this work was extended to a population of typically developing children, they were also more likely to search at flashing locations, and the magnitude of this effect did not vary with age. However, there was a strong developmental trend in the number of times children revisited previously examined locations. The developmental trajectory for ability to remember previously visited locations and limit revisits shows a development in spatial working memory occurring separately from perceptual inhibition. This suggests individual executive sub-processes may play different roles during search, with different developmental trajectories. Further work in this paradigm examines the effect of mediating working memory load on search behaviour. This investigation of working memory during search is crucial for a transition into real-world tasks, such as searching for lost keys or a lost car in a parking lot.

The review by Jeffery et al. commendably addresses the scope of research across many species and paradigms; here we present work in a novel paradigm that expands our understanding of human adult and child populations in the three-dimensional world. Understanding how individuals incorporate reference frames and employ cognitive mechanisms such as memory and attention is key to modelling human navigation.