Layout of reference points during navigation: Effects of a central reference point as anchor Kayoko Ohtsu (id-ant@moegi.waseda.jp) Faculty of Education and Integrated Arts and Sciences, Waseda University, 1-6-1, Nishi Waseda, Shinjuku-ku Tokyo, Japan Abstract nonvisible destination located east of the station, one can head to the east based on environmental reference frames or This study examined the effects of two variables on spatial learning: the objects’ array (with or without a central one can find a direction toward the destination from reference point) and the learning method (virtual walk memory by imagining oneself standing with one’s back through or seeing still images). After learning the objects’ against the station and recalling the route from there. locations and their positional relations in a virtual room in Learning spatial layouts has been a major topic of spatial egocentric reference frames, participants judged the directions cognition, and much research has been conducted with to a target from an imaginary position in the room. The results respect to the reference frames. However, fewer studies revealed that having a central reference point facilitated learning, especially in the virtual-walk condition with a have examined a layout’s own effect on spatial learning. central node. The findings are discussed in terms of the Some studies have focused on layouts of objects in learning interactions between the effect of the central reference point tasks set in middle-scale spaces. Mou and McNamara and virtual walk and the relationship between intensive (2002) and Mou, Zhao, and McNamara (2007) reported that encoding in egocentric reference frames and stored when a layout of targets has an intrinsic axis of representations of the layout in environmental reference configuration (e.g., desks in a classroom arranged vertically frames. and horizontally represent a line-column axis), the intrinsic Keywords: spatial learning; layout; reference frames; reference frame is given priority over viewing perspectives reference points; anchor points in layout learning. Kelly, Avraamides, and McNamara (2010) reported that features of a layout learned in advance Introduction affect subsequent learning. In these studies, participants In everyday life, people move about in their surrounding learned the layouts from single or multiple viewpoints. space, they get out of bed, walk into the kitchen for coffee, Therefore, one could argue that the layout features and an go into the garage, and drive to their office. Such daily alignment effect in an egocentric reference frame were spatial behavior is based on spatial cognitive functions in compared, rather than comparing multiple layouts on the which we comprehend positional relations between our acquisition of representations. bodies and objects. Because spatial locations are essentially In the present study, we examined whether differences in relative, reference systems to describe position and layout of objects affect spatial learning, using a free- orientation are important (Pani & Dupree, 1994). In the exploring task that invokes egocentric reference frames in a areas of human navigation, wayfinding, and spatial memory, middle-scale virtual space. Specifically, using post-tests, we two types of reference systems have been distinguished in compared the learning results of two different layouts: one the literature: egocentric and environmental (or allocentric) consisting of four objects arranged like spots on a die and (Werner & Schmidt, 1999). In egocentric reference systems, the other consisting of five objects in which one more object locations are defined with respect to one’s body (e.g., there was added at the center. is a table in front of me); in environmental reference Our previous research using a real labyrinth (Ohtsu & systems, locations are defined with respect to external Ouchi, 2010) suggested that a particular layout condition objects and, in some cases, an objective coordination such may facilitate spatial learning. In our experiment, as an azimuth orientation (e.g., Denmark is located to the participants explored the fylfot-shaped labyrinth (Figure 1), north of Germany). found four targets, and revisited them; they executed either When walking around a familiar town, people rely on their one of two kinds of visiting orders during the learning phase. spatial memory of the environment, including elements such The first corresponded to the Circle-Order procedure in as routes, landmarks, and their layouts. Many studies which participants revisited the targets in a clockwise and indicate that spatial layouts tend to be organized in the counter-clockwise order (i.e., visiting A→B→C→D→A, frame of environmental reference systems rather than and then A→D→C→B→A), so that at the central egocentric reference frames (e.g., Kelly, Avraamides, & intersection of the labyrinth, they constantly updated their Loomis, 2007; Montello, 1991; Shelton & McNamara, position relative to the destination targets situated in the 2001). However, when moving through a space, we also same self-to-object relation and turned to the left or right. rely heavily on egocentric reference frames, and both types The other order (i.e., visiting A→B→D→C→A, and then of frames are available for recalling the relations of spatial A→C→D→B→A) represented the Non-Circle-Order elements. For example, when walking from the station to a procedure in which, at the intersection, participants updated their position relative to the multidirectional destination 465 targets in the different self-to-object relations and turned becomes difficult when reference points increase in number, right or left or went straight ahead. In the post-tests, the but the ease of learning is not influenced solely by the participants in the Non-Circle-Order condition performed number of reference points. Lindberg and Gärling (1981) better than those in the Circle-Order condition. investigated spatial learning during locomotion with differences in the number of reference points. In their experiment, participants walked paths along reference points guided by experimenters, and at the stopping points Target D designated by the experimenters, they estimated distances Target A and directions to the reference points (one to three). The results revealed that the number of reference points did not Target C affect the accuracy of the directional estimations. If the central point acts functionally like an anchor point that organizes other spatial information into a layout Target B (Golledge, 1999), the positional relation of the reference points can be recalled more easily and more accurately with Entrance the central point than without it. In addition, considering Figure 1: Layout and Images of the Labyrinth. that people can form configurations of reference points from information acquired in egocentric reference frames as well Although our experiment was conducted to investigate the as environmental ones when learning an environment effects of the different types of directional inference through navigation or wayfinding (Sholl, 1996), it would be (unidirectional versus multidirectional), in the course of more efficient to encode the relations when one puts oneself examining the results, we conceived a hypothesis that along in the space and exerts sufficient egocentric reference with the updating, the difference in the reference points’ frames than when learning the relations from restricted layouts recognized by the participant as a consequence of viewpoints. their inferences might also have resulted in the superiority Based on the hypothesis that the fifth central reference of the Non-Circle-Order condition. In particular, we point added to the four reference points serves as an anchor presumed that the participants in the Non-Circle-Order that organizes other points into a layout and facilitate spatial condition recognized the intersection as a fifth reference learning through navigation, we conducted an experiment to point in addition to the target locations, whereas those in the examine the effect of the central reference point when Circle-Order condition recognized only four reference learning via virtual navigation and from multiple vantage points in the target locations. In our experiment, the points using still images. Via virtual navigation, one might locations of the targets ought to have been recognized as well be able to recognize the relations between the reference important reference points, but the significance of the points in egocentric and environmental reference frames, intersection might depend on the conditions. For those in the whereas via still images, one basically see the objective Non-Circle-Order condition, the intersection was more relations of the reference points shown by images. important than for those in the Circle-Order condition, because it was only in the former condition that the Experiment participants needed to pay extra attention to determine the multidirectional relation in the egocentric reference frame Factors and choose their way. In contrast, in the Circle-Order One factor was the array of objects: Square or Central Node condition, the participants could automatically turn left or (Figure 2). The other factor was the learning method: right during the revisiting after perceiving the layout. Virtual-Walk or Still-Image. The term “reference points” was originally used in categorization in cognitive science (Rosch, 1975). Subsequently, Sadalla, Burroughs, and Staplin (1980) proposed describing notable landmarks and places whose locations are relatively better known among others as reference points; the reference points serve to define the location of adjacent points. Among those reference points, Square Central Node there might be a difference in the level of importance. Figure 2: The Arrays of Objects. Golledge and Spector (1978) proposes, in the anchor point theory, that distinctive locations, features, path segments, or Environmental Setup and Materials familiar districts become an anchor of the cognitive map, A virtual circular room with a diameter of 12 m (Figure 3), and they influence encoding, storage, and decoding generated by CAD software (shade dreamhome 2.0.3) was processes used when accessing stored information in a used in all the conditions. Four common objects (Figure 4) decision-making context (Couclelis, Golledge, Gale, & and curbs indicating the front of the objects were placed, Tobler, 1987; Golledge, 1999). Learning apparently and a central node object (Figure 4) was added at the center 466 of the room only in the Central Node conditions. Six kinds Practice Phase The participants in Virtual-Walk conditions of rooms that included all combinations of the placement of explored the practice room for one minute, whereas those in common objects were prepared, and each one was assigned the Still-Image conditions saw images of the room for one to four participants in four groups. In the Virtual-Walk minute. Then, participants in both conditions took Post-test conditions, participants were seated in front of a computer 1 so that they would consciously remember the relative display and operated a keyboard for the virtual walk through positons of all objects in the exploring phase. Throughout in which continuous images from the perspective of a virtual the practice phase in both conditions, experimenters camera (height: 1500 mm, field-of-view: 80 degrees) were monitored and checked whether the participants understood shown according to the key operation. The starting points in what they needed to judge in the post-test. the Virtual-Walk conditions were in front of a common object that varied between the four participants assigned to each kind of room. In the Still-Image conditions, still images of the six kinds of rooms from 12 vantage points (three perspectives for each common object, see Figure 5) were shot by the virtual camera in the same terms as the Virtual-Walk conditions. Each participant watched the four sets of three images in front of the same display. For Square practice in all the conditions, another virtual circular room with a diameter of 5 m in which three objects and curbs were placed was used. Images of the practice room for Still- Image conditions were shot similarly as for the experimental room. Central Node Figure 5: Still Images of Three Perspectives. Exploring Phase At the beginning of the phase, the participants in the Virtual-Walk conditions were instructed to move to four positions where they could see each common object in front of them, and then allowed to move Square Central Node around the room freely. Their walking trajectories were Figure 3: Images of the Virtual Circular Room recorded by a screen capture program. The overall time of from Virtual Camera. the virtual-walk was four minutes. In contrast, those in the Still-Image conditions randomly saw the four sets of three images showing each object from three vantage points. Each set was presented four times for 15 seconds each (5 seconds per image) so that the overall viewing time would be four minutes. Four Common Objects (cue objects in the post-tests) and Node Object (the potted plant) Post Test Phase Two post-tests and a sketch map test were conducted for each group. In Post-test 1, participants were Figure 4: The Objects Used in the Experiment. asked for directions to one of the common objects from imaginary locations. A cue target was presented first, Participants followed by a fixation point, and then the target object Ninety-six undergraduate and graduate students with normal (Figure 6). The participants selected one of the keys to vision were randomly assigned to each group. The ratio of indicate directions to the target objects as if they were male to female was the same in each group (12 male and 12 standing and facing a cue object. One set of 12 randomized female). The average ages of each group were 22.5 (SD: trials, including all possible combinations of two objects, 4.65; Square-Walk), 22.4 (SD: 4.02; Square-Still), 21.0 was conducted twice. (SD: 3.60; Central Node-Walk), and 21.4 (SD: 2.74; Central In Post-test 2, participants were asked for directions to one Node-Still). of the common objects while assuming they were standing in the center of the room and seeing a cue object in a Procedure particular position. The cue object indicated which direction one was facing. After the first picture showing the cue The experiment was divided into three phases: practice, object, the target object was displayed following a fixation exploring, and post-test. At the beginning, participants were point, and the participants selected one of the directions in instructed to remember the objects, their locations, and the the same manner as in Post-test 1. One set of 24 randomized positional relations between them. 467 trials with all possible combinations of two objects was Walking Trajectory conducted twice. Except for two participants’ trajectories that could not be Finally a Sketch Map Test was conducted in which recorded due to technical difficulties, 40 recorded participants wrote down, on a circular piece of paper, the trajectories (19 in the Square-Walk and 21 in the Central names of the objects and curbs in the position they Node-Walk) were examined by the experimenters. In both experienced during the exploring phase. the Virtual-Walk and Still-Image conditions, after visiting Post-test 1 Post-test2 the common objects as instructed, some participants kept moving from one zone to another (see Figure 8) with short + + stops, while others were more likely to stay at some Cue Object 500 ms Target Cue Object 500 ms Target locations longer and overlook the room. Among the 2500 ms 2500 ms behavioral patterns common to the groups, frequencies of “if you were standing and facing an object “if you were standing in the center of the X, select the direction of a target”. room and seeing an object X as in the zone migration and staying at the center zone (turning left or picture, select the direction of the target”. The correct answers: b, d, or f The correct answers: a, c, e, or g right to overlook the room for more than 10 seconds) were d e counted (Table 1). Two-sample t-tests were performed on c f The participants selected the mean values of the zone migration and staying, and there b g one of the seven directions. was a significant difference only in the frequencies of a staying [t(38) = 2.19, p < 0.05]. Analysis of the correlations Figure 6: The Post-tests. between each value and the mean error rate in each post-test was performed, and a significant possible negative RESULTS correlation was found only between the migration frequency and the mean error rate in Post-test 2 in the Central Node- Sketch Map Test Walk condition; however, a regression analysis revealed no The sketch maps drawn by the participants were checked to significant relationship between the variables. determine whether the locations of objects were recalled 0.80 0.80 correctly. Eleven participants (three in Square-Walk, three 0.70 0.70 in Square-Still, three in Central Node-Walk, and two in 0.60 0.60 Central Node-Still) could not recall the objects and/or put them in the wrong positions. Since some of the 11 0.50 0.50 participants seemed to abandon the judgment in the post 0.40 0.40 tests (e.g. selecting the same directions in any trial), we 0.30 0.30 decided to exclude the participants from the analyses based 0.20 0.20 on the success or failure of the map. 0.10 0.10 Judgments 0.00 0.00 Square Central Square Central After the angular transformation, the mean error rates in Still each post-test were analyzed by a two-way ANOVA. Walk Results for Post-test 1 revealed a main effect of the learning Figure 7: The Mean Error Rate and method [F(1,81) = 38.69, p < 0.01] and a significant Standard Error. interaction effect [F(1,81) = 4.98, p < 0.05] (Figure 8). 6000 6000 Results for Post-test 2 revealed main effects of the array 5000 5000 [F(1,81) = 11.40, p < 0.01] and learning method [F(1,81) = 4000 4000 5.74, p < 0.05], and a significant interaction effect [F(1,81) = 5.10, p < 0.05] (Figure 8). Another two-way ANOVA was 3000 3000 performed with only the results of the Virtual-Walk 2000 2000 conditions in Post-test 1, using the following factors: the 1000 1000 array and correct directional responses (front, diagonally forward left and left-hand side). The analysis revealed only 0 0 a main effect of the layout [F(1,40) = 6.15, p < 0.05]. Square Central Still Square Central Walk Response Time Figure 8: The Mean Response Time (ms) The mean response time (Figure 9) in each post-test was and Standard Error. also analyzed by a two-way ANOVA, and a main effect of the learning method was detected in Post-test 1 [F(1,81) = 7.73, p < 0.05] and Post-test 2 [F(1,81) = 4.95, p < 0.05]. Figure 9: The Zones in Analyses of Walking Trajectory. 468 Table 1: The Mean Frequency and Standard Error. Virtual-Walk condition were more elaborated than those in the Still-Image condition. However, the most influential zone migration staying at center condition was the Central Node-Walk; participants’ Square 6.58 (.53) n=19 1.05 (.21) n=19 representations in that condition might have been the most Central 5.90 (.49) n=21 0.52 (.14) n=21 elaborated. Indeed, the ANOVA performed on the Virtual- Walk conditions using the correct directional responses as an extra factor suggests that overall positional relations of Discussion the objects as layout were learned better in the Central Node-Walk than in the Square-Walk. We hypothesized that, when learning the reference points In Post-test 2, the participants did not know that they through navigation, the central point would serve as an would be asked for directions to common objects from the anchor that organizes other points into a layout, which then center of the room, so they could not intentionally facilitates layout learning. The results supported the remember the positional relations from the center point in hypothesis and suggested that the central reference point is egocentric reference frames during the exploring phase. more effective when learning the spatial relations of objects Therefore, when they judged directions, they had no choice via virtual virtual-walk than via still images from multiple but to reorient the positional relations between one’s body viewpoints. The number of participants who failed the and the targets using the representations of the layout in sketch map test in each group indicated that, regardless of environmental reference frames. The results of the second condition, nearly equal numbers of participants could test also support the hypothesis. Although there were main remember and recall the objects’ locations. However, in effects of the two factors, the significant interaction effect Post-test 1, when participants were asked to indicate the and the finding of the highest percentage of correct answers direction to a target object, the presence of the central in the Central Node-Walk condition showed that the two reference point led to a major difference in the judgments, factors had positive effects on mainly the Central Node- which varied according to the learning method. When Walk condition. In contrast, the error rates did not differ learning occurred via still images, the central point did not greatly between the Square-Still, Square-Walk, and Central affect the judgments, as suggested by the absence of a main Node-Still conditions, although the participants in the effect of array; however when learning via virtual-walk, the Virtual-Walk conditions, including the Square-Walk, could central point seemed to affect learning. This could be answer quicker than those in the Still-Image conditions. explained by an interaction between the factors, in which Analyses of the participants’ trajectories suggest that the mean error rate in the Central Node-Walk condition was participants in Square-Walk recognized that the center zone notably lower than the rates of the other groups, and a main was an important vantage point for layout learning, as they effect of the layout in the ANOVA with only Virtual-Walk tended to stay in that zone and overlooked the room with a conditions. higher frequency than those in Central Node-Walk. It could In Post-test 1, learning method affected the judgments due be interpreted that, at the center zone, they paid attention to to the large difference between Virtual-Walk and Still- the positional relations of the four common targets in their Image conditions in the error rates and the response times. view and tried to remember the layout in environmental The difference in the error of judgments would arise from a reference frames, instead of thinking of the relations qualitative difference in layout representations between the between the targets and themselves—standing at the learning conditions. The cognitive manipulation performed center—in egocentric reference frames. In contrast, the by the participants in the first test can be speculated as participants in Central Node-Walk who thought they had to follows: recalling a layout stored in the environmental remember the relations, including the central node object, reference frames, reorienting a target object in the would have intensively encoded the relations from the egocentric reference frames from the layout, and imagining central node object to the others in environmental reference being in front of a cue object. Of course, one could provide frames at the center zone near the object. the directions using representations only in the egocentric Lastly, we would like to suggest the process whereby the reference frames by remembering 12 possible positional central reference point act like an anchor point in our relations from one target to the others, but it is rather experiment. The first factor lies in the general feature of the unlikely that many participants would have applied such a layout. When connecting the objects with a straight line, the cumbersome and uneconomic strategy. Instead, it seems that lines in both conditions overlap each other. However, the participants remembered what they were asked in the first diagonal relations in the Central Node Array are segmented test and made an effort to remember the layout for the post- by the central node object. The segmented components test. Considering that the participants in both conditions might result in elaborate layout representations and facilitate remembered the layouts equally well, it could be inferred accurate judgments because the components allow one to that those in the Virtual-Walk condition could reorient a describe and encode more diverse positional relations. The target object in egocentric reference frames from the second factor lies in the superiority of the virtual walk imaginary locations more easily and faster than those in the through over learning by still images. Learning structures Still-Image condition. This would mean that the via navigation and wayfinding involves the integration of representations in environmental reference frames in the 469 local perspectives and views that a traveler has learned Meilinger, T. (2008). The network of reference frames independently (e.g., Meilinger, 2008; Poucet, 1993). The theory: A synthesis of graphs and cognitive maps. In C. Virtual-Walk conditions provided more diverse relations Freksa, N. S. Newcombe, P. Gärdenfors, & S. Wölfl (Ed.), between the objects due to one's own movement compared Spatial Cognition VI (pp. 344-360). Berlin: Springer. to the Still-Image conditions. This would have led to the Montello, D.R. (1991). Spatial orientation and the superiority in forming the layout representations. angularity of urban routes: A field study. Environment Considering all the factors together, the Central Node-Walk and Behavior, 23, 47-69. condition might have led to the elaboration in which one can Mou, W., & McNamara, T. P. (2002). Intrinsic frames of efficiently manipulate the layout representations, reorients reference in spatial memory. Journal of Experimental oneself, and judges directions by using both environmental Psychology: Learning, Memory and Cognition, 28, 162- and egocentric reference frames. 170. The present study revealed the effect of differences in Mou, W., Zhao, M., & McNamara, T. P. (2007). Layout layouts with and without a central reference point under the geometry in the selection of intrinsic frame of reference condition of a virtual walk through. The intensive encoding from multiple viewpoints. Journal of Experimental of the positional relations from the central reference point Psychology: Learning, Memory, and Cognition, 33, 145- by virtual navigation apparently results in elaborated layout 154. representations. In our previous experiment using a real Ohtsu, K., & Ouchi, Y. (2010). The influence of route labyrinth, a process similar to that in this experiment would planning and its execution on spatial learning. have occurred, although there was no obvious central object Proceedings of the 32nd Annual Meeting of the Cognitive indicating a prominent location in the labyrinth. Future Science Society, 2494-2499. work should confirm the speculations arising from the http://mindmodeling.org/cogsci2010/papers/0610/paper06 present findings by clarifying (1) that the layout contains a 10.pdf distinctive reference point that serves as an anchor, and (2) Pani, J. R., & Dupree, D. (1994). Spatial reference systems the relationship between encoding a reference point in in the comprehension of rotational motion. Perception, 23, egocentric reference frames and stored representations of 929-946. the layout in environmental reference frames. Poucet, B. (1993). Spatial cognitive maps in animals: New hypotheses on their structure and neural mechanisms. References Psychological Review, 100, 163–182. Couclelis, H., Golledge, R. G., Gale, N., & Tobler W. Rosch, E. (1975). Cognitive reference points. Cognitive (1987). Exploring the anchor-point hypothesis of spatial Psychology, 7, 532-547. cognition. Journal of Environmental Psychology, 7, 99- Sadalla, E. K., Burroughs, W. J., & Staplin, L. J. (1980). 122. Reference points in spatial cognition. Journal of Golledge, R. G. (1999). Human cognitive maps and Experimental Psychology: Human Learning and Memory, wayfinding. In Golledge, R. (Ed.), Wayfinding behavior: 5, 516-528. Cognitive mapping and other spatial processes (pp. 125- Shelton, A. L., & McNamara, T. P. (2001). Systems of 151). Baltimore: Johns Hopkins. spatial reference in human memory. Cognitive Golledge, R. G., & Spector, A. N. (1978). Comprehending Psychology, 43, 274-310. the urban environment: Theory and practice. Sholl, M. J. (1996). From visual information to cognitive Geographical Analysis, 10, 403-426. maps. In J. Portugali (Ed.), The construction of cognitive Kelly, J. W., Avraamides, M. N., & Loomis, J. M. (2007). maps (pp. 157-186). Dordrecht: Kluwer Academic Sensorimotor alignment effects in the learning Publishers. environment and in novel environments. Journal of Werner, S., &Schmidt, K. (1999). Environmental reference Experimental Psychology: Learning, Memory, and systems for large-scale spaces. Spatial Cognition and Cognition, 33, 1092-1107. Computation, 1, 477-473. Kelly, J. W., Avraamides, M. N., & McNamara, T. P. (2010). Reference frames influence spatial memory development within and across sensory modalities. In C. Holscher, T. F. Shipley, M. O. Belardinelli, J. A.Bateman, & N. S. Newcombe (Eds.), Lecture notes in artificial intelligence. Spatial cognition VII (pp. 222-233). Berlin: Springer. Lindberg, E., & Gärling, T. (1981). Acquisition of locational information about reference points during locomotion with and without a concurrent task: Effects of number of reference points. Scandinavian Journal of Psychology, 22, 109-115. 470