UC Santa Barbara

Spatial perspective taking is the process through which judgments (usually of distance) are made from an imagined perspective. For example, you may be sitting at a desk now, and I could ask you to image standing in front of the desk, facing your chair. Then I could ask you to point to the door nearest you. Spatial perspective taking allows you to imagine the standing position described and subsequently make a judgment about where the door is from that imagined perspective. Perspective taking is related to many other spatial abilities, including mental simulation of bodily rotations (Kessler & Wang, 2012), giving directions (Hegarty & Waller, 2004), and navigation (e.g. Holmes, Marchette, & Newcombe, 2017). There are also large individual differences in perspective taking performance (Hegarty & Waller, 2004; Kozhevnikov & Hegarty, 2001). This dissertation describes two lines of research focused on perspective taking: first, examining what factors influence perspective taking performance, and second, exploring the relationship between perspective taking and navigation tasks (map use and environment learning).

Spatial perspective taking is influenced by multiple factors, one of which is the agency of cues embedded in perspective taking task arrays (Tarampi, Heydari, & Hegarty, 2016). Previous research has used the Spatial Orientation Test (SOT; Friedman, Kohler, Gunalp, Boone, & Hegarty, 2019; Hegarty & Waller, 2004; Kozhevnikov & Hegarty, 2001) to measure perspective taking ability. In this task, participants are shown an array of objects, and asked to imagine standing at one object in the array, facing a second, and then to point to a third. This is a judgement of relative direction (JRD) task (Shelton & McNamara, 1997). Work by Tarampi, Heydari, and Hegarty (2016) found that adding a human figure to the task array improves performance, and eliminates a previously robustly demonstrated sex difference in performance on this task favoring men.

In the present work, two hypotheses regarding the influence of the human figure were tested: first, if the agency of the human figure is the primary characteristic of importance, as suggested by previous research on agency and perspective taking (e.g. Shelton, Clements-Stephens, Lam, Pak, & Murray, 2012; Clements-Stephens, Vasiljevic, Murray, & Shelton, 2013; Tversky & Hard, 2009), then a cue that is agentive (a human figure) should yield improvements in performance above and beyond cues that are not agentive. Second, if the directionality of the human figure is what influenced performance on this task in previous research, then a cue that is only directional and not agentive (an arrow) should be sufficient to improve performance.

Experiments 1, 2a, and 2b of this dissertation compare how the agency and directionality of the cue could be influencing performance. Experiment 1, established that participants performed better with a human figure in the task array than without any additional cue, as in previous research. In Experiments 2a and 2b, an arrow (directional) was compared to a human figure (directional and agentive), and it was found that the arrow was sufficient to improve performance. These findings indicate that for this type of task display, directionality of an additional cue embedded in the task array is sufficient to improve performance. This suggests that cue characteristics in combination with display characteristics are both important factors to consider when measuring perspective taking performance.

In these experiments, other components of the perspective taking task were manipulated. Specifically, both magnitude of initial perspective shift and pointing direction to the target object were systematically varied across trials. Findings indicated that perspective taking was faster and generally more accurate when perspective shifts were small and when pointing to the front of the imagined heading. Target direction relative to actual heading (not imagined heading) was considered, though was not systematically varied per trial. Neither interaction between cue types and perspective shift nor pointing direction were significant. This suggests that additional cues do not make larger shifts easier to imagine or pointing to less visually accessible directions easier. Rather, findings from these experiments suggest that the benefit of the additional cue, either an arrow or human figure, is that it facilitates initial imagination of oneself in the array.

Experiments 3 and 4 of this dissertation focused on examining the connection between perspective taking and other spatial abilities: navigating using a map (Experiment 3), and learning the layout of an environment (Experiment 4). The relationship between perspective taking and various components of environment learning has been established in previous work (Allen, Kirasic, Dobson, Long, & Beck, 1996; Galati, Weisberg, Newcombe, & Avraamides, 2015; Holmes, Marchette, & Newcombe, 2017; Kozhenikov, Motes, Rasch, & Blajenkova, 2006; Muffato & Meneghetti, 2020; Weisberg, Schinazi, Newcombe, Shipley, & Epstein, 2014).

Experiment 3 examined the relationship between perspective taking and map use, specifically looking at how misalignment of the map relative to the user influences navigation performance, and how perspective taking ability might ameliorate some of this alignment effects. During perspective taking one often engages in perspective-based transformations of the egocentric frame of reference (Zacks & Michelon, 2005), and doing this well relies on good underlying spatial transformation ability. Because using a misaligned map also requires similar transformations, it was predicted that good perspective taking would be correlated with more efficient navigation and also with less pronounced alignment effects. Good perspective takers were more efficient navigators than those with poor perspective taking ability, but the predicted interaction with map alignment was not observed. Findings from this experiment overall indicate that perspective taking is related to navigation using a map, but being good at perspective taking does not make someone immune to alignment effects.

Additionally, navigation performance was examined both when a map was present and when a map was absent. Performance on these different trials indicated that more able perspective takers were more efficient navigators than less able perspective takers, but only when a map was present. When the map was absent, there were no significant differences between more and less able perspective takers, though there was a general trend for more able perspective takers to be more efficient than less able perspective takers. This suggests that when navigating with a map, perspective taking ability might influence how easy it is to use the map. When navigating without a map, however, other abilities may be more influential to navigation efficiency than perspective taking. For example, participants could have been relying on a memory of the map or a memory of navigating in the environment during the no-map trials, but because memory is imperfect and often distorted (e.g. Friedman & Brown, 2000), perspective taking ability may not have been as influential on performance as memory for the map itself.

Experiment 4 explored the connection between perspective taking and environment learning, and more specifically the development of survey knowledge of a new environment. Survey knowledge is configural knowledge or representations with metric information about distances and directions (Chrastil & Warren, 2013). Two components of perspective taking that could also be important for the development of survey knowledge are the ability to imagine changes in orientation (heading) within an environment, and knowledge of the configuration of objects in the environment. In order to examine if both of these specific abilities were supporting the correlation between perspective taking and the development of survey knowledge when learning an environment, two measures of survey knowledge were employed. These tasks tested participant’s ability to determine which headings were the same in an environment, and determine which objects were closest to each other in the environment based on Euclidean distance, which is one way of measuring participant’s configurational knowledge of the environment. Perspective taking ability was examined in relation to the development of survey knowledge, and it was predicted that good perspective takers would perform better on these measures of survey knowledge than poor perspective takers. Findings from this experiment supported this prediction, and add support to the idea that being able to flexibly and continually update one’s heading is fundamental to both perspective taking and environment learning. This in turn suggests that this ability facilitates development of survey knowledge of an environment.

This work contributes several important findings. First, it shows that perspective taking with an abstract display is influenced by cue directionality and not agency. Second it shows that additional cues facilitate initial imagination of oneself in a new position within a perspective taking task, but do not reduce the difficulty of shifting or pointing from an imagined perspective. Third, it shows that good perspective takers are better at using maps to navigate than poor perspective takers perhaps because of their fluency with perspective-based transformations. Finally it shows that both flexible updating of heading and configurational knowledge of the environment in part underlie the connection between perspective taking and survey knowledge.