Perception, 2007, volume 36, pages 525 ^ 536
DOI:10.1068/p5532
Reproducibility of distance and direction errors associated with forward, backward, and sideway walking in the context of blind navigation Nicole Paquetô½, Constant Rainville½#, Yves LajoieÁ, Franc°ois Tremblayô
ô School of Rehabilitation Sciences, University of Ottawa, 451 Smyth Road, Ottawa, Ontario K1H 8M5, Canada; ½ Montre¨al Interdisciplinary Rehabilitation Research Center, Jewish Rehabilitation Hospital, 3205 Alton Goldbloom, Laval, Que¨bec H7V 1R2, Canada; # Research Center, Montre¨al Geriatric University Institute, 4565 chemin de la Reine-Marie, Montre¨al, Que¨bec H3W 1W4, Canada; Á School of Human Kinetics, University of Ottawa, 125 University Avenue, Ottawa, Ontario K1N 6N5, Canada; e-mail:
[email protected] Received 10 October 2005, in revised form 16 May 2006; published online 14 March 2007
Abstract. The ability to navigate without vision towards a previously seen target has been extensively studied, but its reliability over time has yet to be established. Our aims were to determine distance and direction errors made during blind navigation across four different directions involving three different gait patterns (stepping forward, stepping sideway, and stepping backward), and to establish the test ^ retest reproducibility of these errors. Twenty young healthy adults participated in two testing sessions separated by 7 days. They were shown targets located, respectively, 8 m ahead, 8 m behind, and 8 m to their right and left. With vision occluded by opaque goggles, they walked forward (target ahead), backward (target behind), and sideway (right and left targets) until they perceived to be on the target. Subjects were not provided with feedback about their performance. Walked distance, angular deviation, and body rotation were measured. The mean estimated distance error was similar across the four walking directions and ranged from 16 to 80 cm with respect to the 8 m target. In contrast, direction errors were significantly larger during sideway navigation (walking in the frontal plane: leftward, 108 158 deviation; rightward, 188 138) than during forward and backward navigation (walking in the sagittal plane). In general, distance and direction errors were only moderately reproducible between the two sessions [intraclass correlation coefficients (ICCs) ranging from 0.682 to 0.705]. Among the four directions, rightward navigation showed the best reproducibility with ICCs ranging from 0.607 to 0.726, and backward navigation had the worst reliability with ICCs ranging from 0.094 to 0.554. These findings indicate that errors associated with blind navigation across different walking directions and involving different gait patterns are only moderately to poorly reproducible on repeated testing, especially for walking backward. The biomechanical constraints and increased cognitive loading imposed by changing the walking pattern to backward stepping may underlie the poor performance in this direction.
1 Introduction Spatial navigation is the process of determining and maintaining a course or trajectory from one place to another (Gallistel 1990). Two types of navigation are recognised, according to the scale in which the navigation is performed (Trullier et al 1997). First, local navigation includes displacements in the immediate environment. Second, wayfinding is moving around in a large-scale environment and, usually, the destination is not in the immediate environment. Navigation activities require the integration of sensory, motor, and cognitive functions (Trullier et al 1997). Visual inputs are largely used to successfully negotiate obstacles during walking. However, navigation is sometimes executed in the absence of vision, for example during displacements in darkness or with low visual acuity. Occasionally, involuntary contacts with structures and objects occur, which are the result of navigation errors. In this paper, navigation relates to displacement in the immediate surrounding to reach a nearby destination. Specifically, our experiment pertains to the classic paradigm of blind navigation.
526
N Paquet, C Rainville, Y Lajoie, F Tremblay
1.1 Blind navigation The ability to reach a destination, or previously seen target, in the absence of vision has been studied extensively with the blind navigation task (eg Thompson 1983; Steenhuis and Goodale 1988; Rieser et al 1990; Loomis et al 1992; Mittelstaedt and Mittelstaedt 2001). The general conclusion is that young, healthy individuals are reasonably accurate in reaching a previously seen target located 2 to 24 m ahead while walking forward without vision. However, they rarely stop at the exact location of the target, especially when the walking distance is longer than 10 m. Errors made during blind navigation may have several origins. First, subjects may misperceive the distance between them and the target. Rieser and his colleagues (1990) have shown, however, that this contributes very little to navigation errors. Second, when subjects walk towards the target, they may inaccurately perceive their own body displacement from their idiothetic information (Bo«o«k and Ga«rling 1981), ie the sensory feedback associated to body movements and displacement, and efferent copies. Third, subjects may make mistakes in updating their current position during navigation. The neural process of continuous updating of the position of an individual relative to one or more locations is path integration (Potegal 1982). This is known to produce cumulative navigation errors that bring the subject further from his/her destination, especially when several changes in direction are executed without vision (Glasauer et al 2002). Fourth, insufficient working memory to remember the position of the target while updating one's own position during displacements could account for navigation errors. However, no significant effect of time to reach the target has been found for blind navigation over distances up to 12 m (Rieser et al 1990). 1.1.1 Distance error. Past studies on blind navigation showed that distance error varies as a function of the distance to be walked. First, the precision of the performance, or constant error, increases with increasing target distance. For example, the mean constant error was found to be 0 cm for a 4 m target, 20 cm for an 8 m target, and 45 cm for a 12 m target (Elliott 1987). Subjects overshot the target when it was at a short distanceö2 m or less (Loomis et al 1993; Philbeck et al 1997) öand undershot the target at distances of more than 10 m (Fukusima et al 1997; Sun et al 2004). Second, the standard deviation of subject's final positions, or variable error, also increases with increasing target distance. The mean variable error was found to be 24, 51, and 67 cm for the 4, 8, and 12 m targets, respectively (Elliott 1987). Performance variability was also shown by subjects undershooting the 6 and 8 m targets in half of the trials and overshooting them in the other half (Rieser et al 1990). 1.1.2 Direction error. It is well known that people walk in circles when they are lost and that visual cues are of no help. This tendency was confirmed by Vuillerme et al (2002) who showed that subjects make an 88 deviation on average relative to a straight line after a 15 m walk during blind navigation. Direction error was larger while navigating sideway. For example, body rotation relative to a vertical axis was on average 88 88 during side-stepping towards a target located only 1.25 m to the right or left (Paquet et al 2003). Thus, it appears that perception and control of direction are not as accurate as perception and control of distance during locomotion without vision. 1.2 Stability over time of blind navigation performance Several studies on blind navigation have been conducted, but little information on the stability of navigation errors over time is available. Since a simple task of walking without vision towards a previously seen target involves several sensorimotor and cognitive processes, one may suspect that blind navigation performance would vary from day to day. Bonanni and Newton (1998) investigated the test ^ retest reliability of the end position of subjects who were instructed to step in place without vision.
Distance and direction errors in blind navigation
527
The ideal performance in this task is to remain at the start position. However, subjects moved forward 86 43 cm, deviated 48 338 and turned 148 508 after 50 steps. The intraclass correlation coefficients (ICCs) of those variables obtained on two consecutive days ranged from 0.66 to 0.69, indicating only moderate reproducibility. We found similar moderate to good levels of reproducibility in a pilot study on blind navigation during sideway walking (Kaldas et al 2003). ICCs of body rotation and total distance travelled were 0.62 and 0.79, respectively, on a 7 day interval. Altogether, these results suggest that the ability to perceive and control body displacement without vision may not be stable over time, as subjects may have a good performance on one day and a less good one on another day. This raises a serious concern regarding blind navigation protocols that are conducted over 2 or more days. 1.3 Objectives of the study In the past, blind navigation was tested during walking in the forward direction almost exclusively. Our question is whether blind navigation performance is influenced by the direction of walking, ie in the frontal versus the sagittal plane, and by adaptation of the gait pattern, ie walking sideway versus backward versus forward. Our first objective is to determine distance and direction errors during blind navigation towards 8 m targets in four directions: forward, backward, rightward, and leftward. Little is known about test ^ retest reproducibility of distance and direction errors during blind navigation. The second objective of our study is to quantify the consistency of navigation errors over time by establishing the test ^ retest reliability with a 7 day interval. 2 Methods 2.1 Subjects Twenty young, healthy subjects (nine males and eleven females) participated in this study. Their ages ranged from 18 to 40 years, and were 29 8 years on average. They were all right-handed with a score of 80 or more on the Edinburgh Handedness Inventory (Oldfield 1971). Subjects were without a history of musculoskeletal injury to the lower limbs or a neurological condition. They all signed an informed consent form approved by the local Ethics Committee. 2.2 Procedure Subjects participated in two testing sessions exactly 7 days apart. Testing was done in a standard-size gymnasium (19.0 m613.5 m). Lighting was uniform and noise kept to a minimum. Figure 1 illustrates the experimental set-up. Two targets (2 cm610 cm pink paper bands) were fixed on the floor 8 m away from the starting line for each target. Both targets were 3.5 m from the lateral gymnasium walls and 5.8 m from the far wall. Target 1 was used for forward and backward navigation trials, and target 2 for sideway trials. The two start line-to-target paths were 6.5 m apart. We used two paths instead of only one to prevent overlap of too many footmarks drawn on the floor (see section 2.3). Subjects did not practice the task before testing. For forward navigation, subjects were positioned on start line 1 and looked at target 1 located 8 m ahead. Then, wearing opaque goggles that completely occluded central and peripheral vision, they walked forward at a comfortable pace and stopped when they believed their feet were on target 1. For backward navigation, subjects were positioned on start line 1 with target 1 behind them. They turned the upper body to look at target 1; then, wearing the opaque goggles and repositioned with target 1 behind, walked backward toward target 1. For sideway navigation, subjects were positioned on start line 2 with target 2 located 8 m away to their right or left. They walked sideway (sidestepping without crossing the legs) toward target 2. Subjects were asked to concentrate on reaching the target without counting their steps. To avoid subjects knowing their navigation error and correcting their performance in the next trials, they were instructed
528
N Paquet, C Rainville, Y Lajoie, F Tremblay
3.5 m start line 1
target 1
8 m 6.5 m
start line 2
8 m
5.8 m target 2 3.5 m
Figure 1. Top ^ down view of the gymnasium showing the position of start lines 1 and 2 and targets 1 and 2. Start line 1 and target 1 were used for forward and backward navigation trials, and start line 2 and target 2 for rightward and leftward navigation trials.
to keep their goggles on after their blind walk and were guided back to the starting line by an assistant through a complex pathway. Five trials per direction were done, which took approximately 20 min. The testing order was always forward, backward, rightward, and leftward. 2.3 Measurements and data collection The final feet positions were marked on the floor with an erasable marker. When on their final position, subjects were asked to put their feet together, and the assistant drew a 10 cm line from the tip of the right shoe to the left one with the use of a ruler. A code indicating the trial number was written beside the mark. These marks were not visible from the start line. After the departure of the subject from the gymnasium, at the end of the testing session, three measures were taken from the marks on the floor, as illustrated in figure 2. 2.3.1 Linear distance travelled. The linear distance between the starting line and the central point of the final position mark was measured with a measuring tape. 2.3.2 Angular deviation. The angle between a line drawn from the centre of the starting line to the centre of the final position mark and the ideal, straight trajectory was measured with a laser level. For forward and backward navigation, deviation towards the subject's right side was positive, while deviation towards the left side was negative. For sideway navigation, the typical deviation associated with rightward and leftward navigation illustrated in figure 3 was positive, while the opposite was negative. 2.3.3 Body rotation. The angle between the starting line and the final position line was measured with a goniometer. Clockwise rotation was positive; counterclockwise rotation was negative. 2.4 Data transformation and analysis In order to better estimate the distance travelled by subjects when their trajectory was curved, mainly during sideway navigation, an estimated distance travelled (EDT) was calculated with the formula: EDT
linear distance travelled . 26sin
body rotation=26body rotation in radians
Distance and direction errors in blind navigation
529
target BR
ideal trajectory 8 m distance
footmark
linear distance travelled AD
start line
Figure 2. Example of one footmark and measures of linear distance travelled, angular deviation (AD), and body rotation (BR) taken.
The following example illustrates the purpose of the formula. In subject 4 navigating sideway in the leftward direction, body rotation at the end of the trajectory (trial 3) was 808. The total distance travelled of this curved trajectory, which in fact is the shortcut between the start and end points, was 684 cm. The EDT obtained after using the formula was 743 cm. The extra 59 cm provided by the EDT better represents the real distance walked by subject 4 along a curve. This formula assumes that the curvature was constant throughout the trajectory, and that body rotation determined the curvature radius. This assumption was supported by visual observation of subjects during their navigation trials. The effect of walking direction on the dependent variables was determined with a one-way ANOVA. A posteriori comparisons with Tukey tests were used to establish significant differences. A probability of 0.05 or less was accepted as significant. The test ^ retest reproducibility of the dependent variables was quantified with ICCs. 3 Results 3.1 Distance and direction errors 3.1.1 Forward direction. The mean (1 SD) estimated distance travelled was 784 123 cm, and the mean angular deviation was 08 48 (see table 1). Angular deviation was less than 108 in every subject and body rotation was less than 108 in sixteen subjects. The distance travelled was more variable among subjects with values ranging from 570 to 1094 cm. About half of the subjects undershot the target (11/20) and nine overshot the target. 3.1.2 Backward direction. On average, subjects walked 762 130 cm. A slight majority of the subjects (13/20) undershot the target. The mean angular deviation and body rotation were ÿ38 98 and 98 178, respectively. In twelve subjects, body rotation was 108 or more. 3.1.3 Sideway direction. Rightward and leftward navigations were characterised by curved trajectories (see figure 3 for examples). On average, subjects deviated by 108 158 and 188 138 while navigating rightward and leftward, respectively. These deviations
530
N Paquet, C Rainville, Y Lajoie, F Tremblay
Table 1. Mean 1 SD estimated distance travelled, angular deviation, and body rotation obtained at test and retest, and intraclass correlation coefficients (ICCs). All directions
Forward
Backward
Rightward
Leftward
753 138 791 167 0.682
784 123 820 141 0.603
762 130 790 138 0.554
720 134 764 190 0.726
744 158 789 189 0.748
Angular deviation=8 test retest ICC
6 14 9 14 0.687
04 15 0.570
10 15 16 14 0.607
18 13 19 14 0.564
Body rotation=8 test retest ICC
6 24 6 27 0.705
29 4 12 0.554
ÿ12 22 ÿ16 29 0.675
26 25 27 28 0.493
Distance=cm test retest ICC
ÿ3 9 ÿ2 7 0.094 9 17 8 14 0.499
target
forward
sideway leftward
sideway rightward
left target
right target start position
backward target
Figure 3. Typical trajectories obtained during forward, backward, rightward, and leftward navigation. For both rightward and leftward trajectories, angular deviation was positive. For rightward trajectory, body rotation was negative (counterclockwise). For leftward trajectory, body rotation was positive (clockwise).
were accompanied by ÿ128 228 and 268 258 of body rotation. Only four subjects had a straight trajectory with less than 108 of deviation and body rotation. The mean estimated distance travelled was 720 134 cm and 744 158 cm while navigating rightward and leftward, respectively. In both directions, a majority of subjects undershot the 8 m target (17/20 in the rightward direction and 15/20 in the leftward direction). 3.2 Effect of walking direction and gait patterns on navigation errors 3.2.1 Distance travelled. Figure 4a illustrates the mean estimated distance travelled obtained at test and retest. At test, the mean estimated distance travelled was similar across the four navigation directions, except for a difference between forward and rightward
531
12 10
45
8 6 4 2 0
(a)
F
B
R
L
Angular deviation=8
Estimated distance travelled=m
Distance and direction errors in blind navigation
35 25 15 5
ÿ5
F
B
R
L
ÿ15
(b) 60 test
Body rotation=8
40
retest
20 0
R F
B
L
ÿ20 ÿ40 ÿ60
(c) Figure 4. Impact of walking direction and pattern on dependent variables obtained at test (white bars) and retest (grey bars). (a) Mean estimated distance travelled; (b) mean angular deviation; and (c) mean body rotation obtained during forward (F), backward (B), rightward (R), and leftward (L) navigation. Statistical comparisons: continuous lines compare results obtained at test, and dotted lines at retest. * p 5 0:05, ** p 5 0:01.
navigation (784 123 cm versus 720 143 cm, p 5 0:01). At retest, however, there was no difference in the mean estimated distance travelled across the four navigation directions. 3.2.2 Angular deviation. Figure 4b shows that angular deviation was significantly influenced by walking direction. At test, the mean deviation during sideway navigation (rightward 108 158 and leftward 188 138) was significantly larger than during forward navigation (08 48, p 5 0:01) and backward navigation (ÿ38 98, p 5 0:01). The mean deviation was similar during forward and backward navigation. During rightward navigation, angular deviation was smaller than during leftward navigation, but only at test. 3.2.3 Body rotation. A significant effect of walking direction was found on body rotation, as shown in figure 4c. At test, the mean body rotation during rightward navigation was ÿ128 228 (a counterclockwise rotation, on average) and 268 258 during leftward navigation (a clockwise rotation). These values were significantly larger than those during forward navigation (28 98, p 5 0:01) and backward navigation (98 178, p 5 0:01). The mean body rotation was less during forward than backward navigation, but only at test.
532
N Paquet, C Rainville, Y Lajoie, F Tremblay
3.3 Test ^ retest reproducibility 3.3.1 Group data. Table 1 lists the ICCs calculated for estimated distance travelled, angular deviation, and body rotation obtained at test and retest. With the data from the four navigation directions pooled together, the three dependent variables show only moderate reproducibility (ICCs between 0.682 and 0.705). Among the four directions, rightward navigation shows the best reproducibility of navigation errors with ICCs ranging from 0.607 to 0.726, and backward navigation has the worse reliability with ICCs ranging from 0.094 to 0.554. 3.3.2 Individual data. Differences between test and retest on estimated distance travelled, angular deviation, and body rotation were calculated individually, for each subject. Table 2 presents the mean differences obtained in the twenty subjects. These means were obtained from the absolute values of the test ^ retest differences to better illustrate the extent of the discrepancy between the test and the retest. On average, the difference in estimated distance travelled between test and retest is 73 to 96 cm in the four navigation directions. Mean differences in angular deviation and body rotation at test and retest are between 38 and 108, except for body rotation during sideway navigation Table 2. Mean 1 SD difference in estimated distance travelled, angular deviation, and body rotation between test and retest (n 20).
Distance=cm Angular deviation=8 Body rotation=8
Forward
Backward
Rightward
Leftward
96 59 32 63
81 70 67 64
83 72 99 13 10
73 79 7 10 14 17
Table 3. Individual differences in estimated distance travelled (cm) between test and retest. Values have been regrouped for individuals with small, intermediate, and large differences, respectively. Difference
Forward
Backward
Rightward
Leftward
Small subject 7 subject 9 subject 12
ÿ35 ÿ55 ÿ8
47 ÿ2 ÿ1
ÿ1 ÿ46 100
14 ÿ13 31
Intermediate subject 1 subject 2 subject 4 subject 5 subject 6 subject 8 subject 11 subject 13 subject 14 subject 15 subject 17 subject 18 subject 19 subject 20
21 ÿ28 ÿ136 48 ÿ151 168 130 ÿ140 61 12 67 81 139 134
40 ÿ144 ÿ96 ÿ82 ÿ10 76 64 ÿ99 58 41 ÿ69 75 ÿ8 ÿ53
91 ÿ103 ÿ113 16 ÿ88 62 149 ÿ40 111 112 3 ÿ15 29 109
13 ÿ93 ÿ48 ÿ36 ÿ42 87 107 ÿ6 ÿ6 79 ÿ52 86 47 ÿ17
Large subject 3 subject 10 subject 16
144 183 174
162 164 301
ÿ19 318 138
216 133 324
Distance and direction errors in blind navigation
533
with mean differences up to 148. Inspection of individual data revealed that some subjects had reproducible navigation errors in the 7 day interval, while others manifest very low reliability. Table 3 lists differences in estimated distance travelled between test and retest for every subject to illustrate the range in reproducibility from excellent to poor from subject to subject. The three subgroups of subjects (smallest, intermediate, and largest differences) were similar in gender and age. 4 Discussion We found that the mean estimated distance error while navigating without vision towards an 8 m target ranged from 16 to 80 cm during walking forward, backward, rightward, and leftward. Varying the walking directions and altering the gait pattern did not modify the distance travelled. In contrast, direction errors were significantly larger during sideway navigation (walking in the frontal plane) than during forward and backward navigation (walking in the sagittal plane). During rightward and leftward blind navigation, subjects deviated on average 108 to 188, and rotated in space on average ÿ128 to 268. These navigation errors, however, are not very stable over time. We found only modest levels of test ^ retest reproducibility within a 7 day interval. 4.1 Distance and direction errors During forward navigation, subjects were accurate in reaching the 8 m target, as found in previous studies (Elliott 1987; Rieser et al 1990; Steenhuis and Goodale 1988). On average, they slightly undershot the target by 16 cm. However, the distance travelled varied considerably within subjects and between subjects, as the standard deviation was 123 cm. The extent of this distribution is more than half the one found by Elliott (1987). One explanation is that our subjects did not practice walking towards the target before test trials. Elliott (1987) has found that practice walking with eyes closed tends to decrease target undershooting. On average, our subjects did not veer when they navigated forward. The mean angular deviation was 08 and the mean body rotation was 28. A similar minimal direction error over an 8 m walk without vision has previously been reported (Rieser et al 1990), and indicates that, over this short distance, subjects controlled their trajectory direction well. In contrast, subjects had difficulty controlling their body rotation in space during backward navigation. In more than half the subjects (12/20), body rotation was 108 or more. This may be due to a larger attentional demand during backward than during forward walking, which could affect the perception or control of navigation direction. Our findings indicate that navigating in the frontal plane produces large direction errors. Instead of side-stepping straight towards the target, subjects made curved trajectories, as if they were executing a circle on the floor, always facing the centre of the circle during both rightward and leftward walking (see figure 3). This feature of sideway blind navigation has been noted in a previous study (Paquet et al 2003). In observing subjects executing side-stepping, we could see that at each step they made a slight hip flexion instead of only hip abduction. This is what made them veer. Subjects seemed unaware of this hip flexion, and apparently did not perceive or control their body veering. In the present study, the curve was less pronounced during rightward than leftward navigation, as shown by the mean angular deviation and body rotation of 108 and ÿ128 during rightward navigation and 188 and 268 during leftward navigation. This asymmetry in the extent of direction error could be due to subject's laterality, as all subjects were right-handed. However, no scientific evidence exists to support this explanation, and the asymmetry we found could be due to chance. Since we have found similar asymmetry at retest for body rotation, this other explanation is also unlikely. The investigation of
534
N Paquet, C Rainville, Y Lajoie, F Tremblay
a larger group of subjects composed of left-handed and right-handed individuals would help solve this issue.
Estimated ÿ measured distance=cm
4.2 Effect of walking direction and altered gait pattern on navigation errors We found that the distance travelled was not influenced by the change in the walking direction and the associated alteration in the gait pattern during blind navigation towards an 8 m target. However, this result refers to estimated distance travelled rather than real distance, as we did not directly measure the trajectory length when the trajectory was curved. Our goal in calculating an estimated distance travelled was to adjust the length of our measured distance travelled to better represent the length of curved trajectories. We verified the relationship between trajectory curvature, expressed by body rotation, and the impact of our calculation on distance travelled, expressed as the difference between estimated distance travelled and the linear distance we measured. Figure 5 illustrates that trajectory length (estimated distance travelled) increases with body rotation (which represents trajectory curvature). 80
60
40
20
0 20
40
60 Body rotation=8
80
100
Figure 5. Relationship between body rotation and the difference between estimated distance travelled and linear distance travelled.
From the estimated distance travelled, we found that, on average, subjects walked less than 8 m with the three types of gait pattern. Such slight undershooting at this distance has previously been reported during forward navigation (Corlett et al 1990), but not for backward and sideway navigation, as we show here. This observation suggests that the ability to perceive body displacement and to update the distance travelled during the course of blind navigation is likely unaffected by changing the navigation direction or the gait pattern, at least on short-distance targets. In contrast, walking forward, backward, or sideway had a major impact on the control of trajectory direction. We found that the gait pattern (forward and backward versus sideway) significantly modified angular deviation and body rotation. In addition, the direction of navigation in the same plane (forward versus backward in the sagittal plane, rightward versus leftward in the frontal plane) had a significant effect on body rotation. Thus, when the gait pattern is unusual, ie other than forward walking, the perception and control of path direction is highly imperfect. It suggests that the biomechanical constraints and/or unfamiliar afferents of the unusual gait pattern induce navigation errors. In addition, higher cognitive demand may be associated with less automatic walking patterns. For instance, reaction time to respond to an auditory cue was significantly longer in healthy subjects walking at very slow speed than at the preferred speed (Lajoie et al 1999).
Distance and direction errors in blind navigation
535
4.3 Test ^ retest reproducibility Over a 7 day interval, navigation errors obtained during our blind navigation protocol are not stable. Not only the ICCs for distance travelled, angular deviation, and body rotation were below 0.7 in most cases (see table 1), but the mean difference in estimated distance travelled between test and retest was up to 96 cm across the three gait patterns (see table 2). The reproducibility level of our navigation errors corresponds to those obtained in a previous pilot study (ICC 0:62 for body rotation; Kaldas et al 2003) and during stepping in place without vision (ICCs between 0.66 and 0.69; Bonanni and Newton 1998). Altogether, these results suggest that the day-to-day variability of navigation errors in young adults can be quite large when vision is absent. In spite of this variability, our study did reveal that some individuals are particularly apt at estimating travelled distance (see table 3), while others perform very poorly in such a task. This may not be so surprising, given that navigation involves complex neural processing (Lindberg and Ga«rling 1983) and the integration of sensorimotor and cognitive functions (Trullier et al 1997). We can imagine that a simple change in mental set or concentration level could lead to quite different performance on two different testing sessions. The reasons underlying the large interindividual variations in performance level remain a major issue for future studies. 5 Conclusions The navigation direction and gait pattern during blind navigation had a significant impact on direction error, but not on distance error. The biomechanical constraints of walking sideway likely lead to the large direction error obtained with this mode of walking. The test ^ retest reproducibility of navigation errors is not more than modest over a 7 day interval. It suggests that cognitive factors such as concentration and mental set have a major influence on blind navigation performance from day to day. If this is the case, navigation performance should be impaired by the execution of a dual cognitive task. We are at present conducting new series of experiments to test this hypothesis. The new information gathered from our navigation tests in a small-scale environment will form a scientific basis on which to plan experiments in larger environments. Ultimately, navigation research should be targeted at helping populations with impaired mobility due to difficulty to orient and navigate in their community Acknowledgments. This project was financed by the Canadian Institutes of Health Research. We are thankful to the Institute for Rehabilitation Research and Development in Ottawa for support. We gratefully acknowledge the assistance of Fatemeh Sabagh-Yazdi, Cheryl Beaulieu, and Peyman Ghorbani. References Bonanni M, Newton R, 1998 ``Test ^ retest reliability of the Fukuda Stepping Test'' Physiotherapy Research International 3 58 ^ 68 Bo«o«k A, Ga«rling T, 1981 ``Maintenance of orientation during locomotion in unfamiliar environments'' Journal of Experimental Psychology: Human Perception and Performance 7 995 ^ 1006 Corlett J T, Byblow W, Taylor B, 1990 ``The effect of perceived locomotor constraints on distance estimation'' Journal of Motor Behavior 22 347 ^ 360 Elliott D, 1987 ``The influence of walking speed and prior practice on locomotor distance estimation'' Journal of Motor Behavior 19 476 ^ 485 Fukusima S S, Loomis J M, Da Silva J A, 1997 ``Visual perception of egocentric distance as assessed by triangulation'' Journal of Experimental Psychology: Human Perception and Performance 23 86 ^ 100 Gallistel C R, 1990 The Organization of Learning (Boston, MA: MIT Press/Bradford Books) Glasauer S, Amorim M A, Viaud-Delmon I, Berthoz A, 2002 ``Differential effects of labyrinthine dysfunction on distance and direction during blindfolded walking of a triangular path'' Experimental Brain Research 145 489 ^ 497 Hartley T, Burgess N, 2005 ``Complementary memory systems: Competition, cooperation and compensation'' Trends in Neurosciences 28 169 ^ 170
536
N Paquet, C Rainville, Y Lajoie, F Tremblay
Kaldas J, Paquet N, Dannenbaum E, Fung J, 2003 ``A functional measure of navigation'' Proceedings of the World Confederation for Physical Therapy page 137 Lajoie Y, Barbeau H, Hamelin M, 1999 ``Attentional requirements of walking in spinal cord injured patients compared to normal subjects'' Spinal Cord 37 245 ^ 250 Lindberg E, Ga«rling T, 1983 ``Acquisition of different types of locational information in cognitive maps: Automatic or effortful processing?'' Psychological Research 45 19 ^ 38 Loomis J M, Da Silva J A, Fujita N, Fukusima S S, 1992 ``Visual space perception and visually directed action'' Journal of Experimental Psychology: Human Perception and Performance 18 906 ^ 921 Loomis J M, Klatzky R L, Golledge R G, Cicinelli J G, Pellegrino J W, Fry P A, 1993 ``Nonvisual navigation by blind and sighted: Assessment of path integration ability'' Journal of Experimental Psychology: General 122 73 ^ 91 Mittelstaedt M L, Mittelstaedt H, 2001 ``Idiothetic navigation in humans: estimation of path length'' Experimental Brain Research 139 318 ^ 332 Oldfield R C, 1971 ``The assessment and analysis of handedness: the Edinburgh Inventory'' Neuropsychologia 9 97 ^ 113 Paquet N, Kulkarni K, Fung J, Watt D G D, 2003 ``Spatial navigation after surgical resection of an acoustic neuroma: a pilot study'' Journal of Otolaryngology 32 180 ^ 184 Philbeck J W, Loomis J M, Beall A C, 1997 ``Visually perceived location is an invariant in the control of action'' Perception & Psychophysics 59 601 ^ 612 Potegal M, 1982 ``Vestibular and neostriatal contributions to spatial orientation'', in Spatial Abilities Ed. M Potegal (London: Academic Press) pp 361 ^ 387 Rieser J J, Ashmead D H, Talor C R, Youngquist G A, 1990 ``Visual perception and the guidance of locomotion without vision to previously seen targets'' Perception 19 675 ^ 689 Steenhuis R E, Goodale M A, 1988 ``The effect of time and distance on accuracy of target-directed locomotion: Does an accurate short-term memory for spatial location exist?'' Journal of Motor Behavior 20 399 ^ 415 Sun H J, Campos J L, Young M, Chan G S W, Ellard C G, 2004 ``The contributions of static visual cues, nonvisual cues, and optic flow in distance estimation'' Perception 33 49 ^ 65 Thompson J A, 1983 ``Is continuous visual monitoring necessary in visually guided locomotion?'' Journal of Experimental Psychology: Human Perception and Performance 9 427 ^ 443 Trullier O, Wiener S I, Berthoz A, Meyer J A, 1997 ``Biologically based artificial navigation systems: review and prospects'' Progress in Neurobiology 51 483 ^ 544 Vuillerme N, Nougier V, Camicioli R, 2002 ``Veering in human locomotion: modulatory effect of attention'' Neuroscience Letters 331 175 ^ 178
ß 2007 a Pion publication
ISSN 0301-0066 (print)
ISSN 1468-4233 (electronic)
www.perceptionweb.com
Conditions of use. This article may be downloaded from the Perception website for personal research by members of subscribing organisations. Authors are entitled to distribute their own article (in printed form or by e-mail) to up to 50 people. This PDF may not be placed on any website (or other online distribution system) without permission of the publisher.
