LIBRARY Mlchlgan State Unlverslty PLACE IN RETURN BOX to remove this checkout from your record. TO AVOID FINES return on or before date due. I DATE DUE DATE DUE DATE DUE MSU Is An Afflrmotivo Action/Equal Opportunity Institution anyhow-dam”.- THE EFFICACY OF N ORTH-AT-THE—TOP AND ALIGNED YOU-ARE-HERE MAPS FOR TASKS WITH DIRECTIONAL CONTENT by Margaret Livingston A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of MASTER OF ARTS Department of Geography 1992 ABSTRACT THE EFFICACY OF NORTH-AT-THE-TOP AND ALIGNED YOU-ARE-HERE MAPS FOR TASKS WITH DIRECTIONAL CONTENT by Margaret Livingston Map users have various and conflicting assumptions about map orientation. Familiarity with conventional north-at-the-top orientation makes it most useful for some map use tasks. You-Are-Here maps aligned with the terrain have been shown to be more effective than misaligned ones for navigational tasks. This study examines the effectiveness of north-at-the-top and aligned You-Are-Here map orientations for tasks involving geographical directions and for tasks involving relative directions. You-Are—Here maps of two fictitious college campuses were created in both orientations, for tests with two groups of subjects. Each group viewed slides of two maps of unique campus-orientation combinations and answered map use questions. North-at-the-top and aligned map orientations were compared for geographical direction questions and for relative direction questions. Results for the two test groups were equivocal: one group performed significantly better on the aligned map for both types of questions, while the other group’s performance did not significantly differ between the map orientations for either question type. . To all those who have helped me get to this point and to remembering that life is a journey, not a destination. ACKNOWLEDGMENTS I would like to thank my advisor, Dr. Richard Groop, for his insight, direction, and patience. I am also thankful for Dr. Judy Olson’s careful reading and suggestions through the course of this work. My fellow graduate students have provided invaluable encouragement and assistance. iv TABLE OF CONTENTS LIST OF TABLES ........................................................................................................... vii LIST OF FIGURES ........................................................................................................ viii CHAPTER: I. INTRODUCTION AND LITERATURE REVIEW ............................................ 1 Map Axes ........................................................................................................... 1 North-at—the-Top Convention 3 Map Orientation Research .................................................................................. 4 Route Maps ........................................................................................................ 6 Strip Maps .......................................................................................................... 8 You-Are-Here Maps ........................................................................................ 10' Geographical Directions and You-Are-Here Maps .......................................... 13 Research Problem ............................................................................................. 15 II. THE EXPERIMENT .......................................................................................... 16 Test Maps ......................................................................................................... 16 Test Questions .................................................................................................. 17 Test Procedure .................................................................................................. 22 III. RESULTS AND DISCUSSION ......................................................................... 28 Approach of the Analysis ................................................................................. 28 Comparing Map Orientations ........................................................................... 30 Testing the Research Hypotheses ..................................................................... 33 Geographical Direction Questions: Results ...................................................... 33 Geographical Direction Questions: Discussion of Results ................................ 35 V Relative Direction Questions: Results ............................................................... 37 Relative Direction Questions: Discussion of Results ........................................ 37 Question-by-Question Analysis: Results .......................................................... 39 Question-by-Question Analysis: Discussion of Results ................................... 44 Summary of Results ......................................................................................... 45 IV. SUMMARY AND CONCLUSIONS ................................................................. 47 Appendix: A. Test Questions ...................................................................................................... 50 B. Oral Instructions ................................................................................................... 56 LIST OF REFERENCES ................................................................................................ 59 LIST OF TABLES Table 1. Mann Whitney U Test for Difference between Test Groups ................................ 29 2. Chi Square Tests for Differences between Map Orientations (Geographical Direction Questions) ......................................................... 34 3. Chi Square Tests for Differences between Map Orientations (Relative Direction Questions) ................................................................. A . 38 4. Chi Square Tests by Question for Differences between Map Orientations (Group 1) ............................................................................. 40 5. Chi Square Tests by Question for Differences between Map Orientations (Group 2) ............................................................................. 42 :9. 9.0099999er h—I .9 LIST OF FIGURES Example of an AAA T riptik strip map ................................................................... 9 Diagram of “Forward = Up” Equivalence ............................................................ 11 Aligned Version of Fresno State University Test Map .......................................... 18 North-at-the-top Version of Fresno State University Test Map ............................. 19 Aligned Version of Santa Maria University Test Map . ........................................ A . 20 N orth-at-the-top Version of Santa Maria University Test Map ............................. 21 Sample of Questions for Fresno State University Map ......................................... 23 Sample of Questions for Santa Maria University Map .......................................... 24 “Student” Viewing You-Are-Here Map ............................................................... 26 Campus and Map Orientation Matrix .................................................................... 31 viii CHAPTER ONE IN 'I'RODUCI‘ION AND LITERATURE REVIEW finding our way around on the earth’s surface is of critical importance. Whether in a local environment or a wider-reaching setting, we are constantly faced with deciding where to go and determining the route that will take us there. Maps assist us in achieving a sense of place and thereby knowing how to proceed; the axes of the map play a role in that function. Map users have various and conflicting assumptions about map orientation, or about what is assigned to the vertical axis. This study is concerned with the orientation of You-Are-Here maps, a specific type of fixed-location navigational map. Orientation is integral to You-Are-Here map design as it affects our comprehension of location and our ability to navigate. Mum Perceptual literature suggests a significant role in the formation of a figure’s primary axis, with a secondary axis perpendicular to the primary one. Rock (1973) argued that assigning an orientation to a figure is an inseparable, intrinsic part of perceiving it. The interpretation of a figure, of course, depends on its orientation within a frame of reference as well as its natural orientation. Considerable research in perception and sensation has revealed a special status for vertical and horizontal axes (e. g., Clark, 1974; Howard and Templeton, 1966). Gravity defines vertical, the perpendicular to gravity defines horizontal; we ourselves are vertically extended and move around on the horizontal plane. Rock further showed that vertical and horizontal symmetry are more salient than diagonal symmetry. Because the up-down dimension is so important and ubiquitous, it is extensively used in structuring our perceptions, our verbal descriptions and our maps. 2 A map sheet (which contains figures on backgrounds) has axes that correspond to the vertical and horizontal. Such a coordinate system aids orientation and location within the frame of reference (the map), and in relating to the environment. The vertical dimension is of particular relevance to the present study. Braine (197 8) presented evidence that subjects across cultures and ages agree on the best orientation of simple figures, preferring vertical spatial extension and focal features at the top. The present study investigates just what direction might best be located at the top of a map. The special status of the “up” on the map has long been recognized and observed in cartography. Practices regarding map orientation, i.e., which direction or feature should be assigned to upward direction on the map, have developed along with the growth of geographical knowledge and of cartography itself. In early maps, the salient feature at the top of the map was often a culturally significant one. For example, one type of circular map from the Middle Ages, the “T in 0” map, symbolically portrayed the world in three parts divided by the Don and Nile Rivers and the Mediterranean Sea. The Orient (and, symbolically, Paradise), the least known and most difficult-to-reach area, was at the top of the map, and thus farthest from the map reader. From this practice is derived the term “to orient” a map, in the sense of turning it such that the directions make sense to the reader. Subsequently, knowledge of the properties of the earth sphere increased and related inventions such as the spherical coordinate system and the magnetic compass were developed. This expanded knowledge allowed wide acceptance of the geographical directions as a useful and applicable system, which came to replace feature-based reference systems on most maps. The unique poles conferred by the earth’s rotation and the magnetic compass (pointing in predictable direction if less than the true direction of the poles) provide more or less invariant and accessible reference directions. Other locational reference systems which may use cultural references points (such as Mecca) or allow determination of location in greater detail (such as latitude/longitude or the Universal Transverse Mercator system) also incorporate aspects of the geographical reference 3 system: one must know the correct geographical direction in order to face Mecca while praying; latitude/longitude and UTM designate location as north or south of the equator and east or west of a designated meridian. The geographical reference system is applicable over the entire earth’s surface and on any map, unlike a location- or culture-specific reference point such as Mecca. It requires little technical knowledge or measurement (unlike UTM). The geographical reference system is doubtless the most widely used and best-understood earth-based system. Useful in determining location and direction, geographical directions are commonly indicated or implied on maps. In a study by Adeyemi (1982), subjects indicated that geographical directions on a map help in wayfinding and even requested that eight, not merely four, geographical directions be shown. - - - v n' n “One of the strongest conventions in cartography”, that of orienting maps with north at the top, has developed over centuries (Robinson, 1952). Arguably, while any of the geographical directions might have prevailed as the most salient direction and therefore been placed at the top of the map, north was so designated because global navigators and map makers were predominantly from the northern hemisphere. The convention of designing maps with north at the top (i.e., at the edge farthest from the map viewer) has for some time dominated to such a degree that if geographical directions are not indicated on the map, it is assumed that north is at the top. The strength of the convention is further evidenced by the fact that other map orientations (particularly south-at-the-top) are often considered aberrations or jokes. Robinson states that “[t]he convention is so well established that when a map is not oriented ‘properly’ it is considered unorthodox” (Robinson, 1952, p. 62). Others have pointed out that there is a long history of maps with “orientations other than north and many of them are remembered principally 4 for that characteristic. . . . Even after we have looked at such maps for some time, they tend to have a strangeness about them” (Cemy and Wilson, 1976, p. 132). Our language provides further evidence of the strength of this convention and the ways in which we adjust to it: “up no ” and “down south”; “upper Michigan” and “lower East side” (Robinson et al., 1969). A brief clarification of the various uses and meanings of the term “orientation” is necessary. To “orient oneself” generally means to get one’s bearings or to fix one’s location relative to a reference point or within a given setting. In field use, to “orient the map” is to lay it flat and turn it so that it parallels the terrain. The “orientation of the map”, as used in the context of this paper, however, refers to the design of the map and designates what is located at the top of the map, be it north or some other direction or a particular feature. llQ"Rl The predominance of north-at-the-top maps causes map readers to be most familiar with that arrangement. Very little research has been done on map orientation, and only a small part has focused on map design or on the effect of that design on wayfmding. The field of psychology has shown more interest in the subject of orientation (and related cognitive processes) than has cartography. Downs and Stea (1977) have noted that the first step in wayfinding is orientation, using that term in the sense of knowing where one is in relation to selected places on the earth’s surface. As Cemy and Wilson (1976) point out, the question of “how readily...we recognize ...[non-conventionally oriented] maps and orient ourselves on them” has been generally ignored in cartographic research (p.132). Blades and Spencer (1987) contend that there has been “no research at all concerned with how the design of a navigational map might influence the way a person uses it” (p. 71). 5 Some research has suggested that map readers do have more difficulty performing map recognition tasks when maps are oriented other than conventionally. Cemy and Wilson (1976) investigated the recognition of rotated and normal maps. The results revealed significantly poorer performance on the rotated maps. However, the “maps” were merely simple outlines of states and countries and, as the authors note, “[i]t would be desirable to repeat the experiment with more maps, richer in detail and information” (p. 137). A study by Lloyd and Steinke (1984) investigated the cognitive processes used in viewing maps in non-conventional orientation. Dot maps and graduated circle maps of the lower peninsula of Michigan were displayed in various degrees of rotation, in both correct 8 and mirror representations. The results clearly indicate that the time required to identify ‘ normal or mirror image maps was strongly related to map orientation, increasing with degree of rotation from north-at-the-top. The studies on orientation have fundamentally investigated shapes, not real-life maps, and for the most part involved only recognition tasks. To the extent that map use entails shape recognition, these studies are of some use to cartography. There is no apparent reason, however, why such maps (particularly the outline thematic maps used in the Lloyd and Steinke study) would need to be oriented other than conventionally. Nevertheless, the results of these studies, in revealing that geographical shapes are recognized most easily in the conventional orientation, do support the notion that people are most familiar with, and presumably create their mental representations of these geographical areas from, north-at-the-top maps. The north-at—the-t0p map design convention appears not to be the only force at work in this respect. In a study of distortion of memory in maps, Tversky (1981) presented evidence that cognitive processes may influence map readers’ association of north with the top of the map. She investigated the “rotation” heuristic as adopted to facilitate encoding and retrieval of spatial orientation of maps. “Rotation” is the tendency to 6 remember the axes induced by a figure as convergent with the axes of the fiame (e.g., the map sheet). Tversky found that when a figure (such as South America or the San Francisco Bay) has a natural orientation (axis) that does not quite correspond to that of the frame of reference, subjects “rotated” the figure in memory to converge with the vertical (north-south) or horizontal (east-west) axis of the frame of reference. Thus, the San Francisco Bay, which runs northwest to southeast, is remembered as running north-south, i.e., in convergence with the map axis. This evidence that people distort geographical figures to allow the figure’s axis to converge with the north-south axis of the page indicates further the strength of the association in map readers’ minds between north and the top of the map. BMW Map use involves more than mere shape and name recognition in navigation and route-findin g tasks. For these uses, the map reader must go beyond recognition of the area represented on the map, and must identify his or her present and target locations on the map. Further, the user must determine how to proceed in the environment to reach the goal. It is in regard to maps designed and/or used for navigational purposes that the question of map orientation (i.e., what direction or feature is at the top of the map sheet) has most often been raised. Clearly, map orientations other than north-at-the-top are possible. “Needless to say, since there is no up or down on a spherical surface, there is no reason why a map cannot be oriented in some other way” (Robinson et al., 1984, p. 74). It is the element of movement which would seem to warrant consideration of non-conventional map orientations. “[Tlhe practice of [non-conventional orientations of] route maps might be extended‘to all maps so that if there is a significant amount of ‘direction’ in the function of a map it could be so oriented” (Robinson, 1952, pp. 62-63). The content and purpose of the map (whether the 7 map involves an aspect of direction or movement) are the critical factors in considering whether a non-conventional orientation is appropriate. Robinson et a1. (1984) add further insight to the reasons for considering various orientations in map design: Since we think of the top of the maps as “away” from us, it is apparent that in some instances orienting a map in the direction of interest or movement, if any, may well be more useful. On the other hand, having most maps oriented to north probably helps most users because the orientation is familiar (p. 74). This points out an essential question: on route and navigation maps, is conventional orientation warranted due to map users’ familiarity with north-at-the-top maps or are such maps more usefully oriented in the direction of movement? Map users generally turn hand-held maps such that the directions on the map (held horizontally) parallel the directions in the environment. In this way, the direction of the target in the terrain is determined by projecting out from the target symbol on the map. This technique is advised in training manuals and has been observed to be common (Levine, 1982; Robinson, 1952). A study by Adeyemi (1982) confirms the usefulness of this technique. That study revealed that when using a north-at-the-top map, subjects who were facing north (i.e., who held the map so that map directions paralleled ground directions) answered directional questions correctly more often than did subjects holding the map with north at the top but facing some direction other than north. However, employing this technique with a north-at-the-top map while facing in any direction other than north, causes the text to vary from right-reading; this is the crux of designing maps oriented in the direction of movement Because text can be right-reading when a map is held in only one position relative to the map user, a map that would primarily be used along a direction other than south-to-north might arguably be more useful if oriented along that direction; the text, printed parallel to the “bottom” of the map sheet would thus be 8 right-reading and, as such, a clue to the best way to hold the map when using it. Such maps ’ allow alignment parallel with the terrain and perpendicular to right-reading text, but they imply an inherent direction of movement. Two types of route maps that can be designed to be oriented in the direction of movement of direction of viewing with inherent direction are strip maps and You-Are- Here (Y AH) maps. These two types of maps differ in that YAH maps are usually in a fixed location while strip maps can be carried along and used while travelling. Both are used for navigation within a specific, defined environment. $21:ng A strip map (generally rectangular) focuses a map user’s attention on an individual, linear feature or route, presented as a sequence of points. While early strip maps often appeared in strip form due to lack of information about the area beyond the limits of the map, the most common use of strip maps in modern times has been for travel. Strip maps often have non-conventional orientation for convenience in positioning several strips on a page and/or to allow parallel alignment of map and ground directions as well as right- reading text. In a series of such strips, each representing one section of a complete route, location of north would thus vary from strip to strip as necessary. As in the case of the TripTik strip maps produced by the American Automobile Association (Figure 1), the orientation of the strip is often based on the direction of intended travel along that section. That is, the strip can be oriented such that the user travels from the “bottom” of the page (the edge of the map nearest the user when held horizontally) to the “top” of the map (the point farthest from the user when held horizontally), thus providing a correspondence with travel from a present (nearby) location to a distant location. The strip map differs from a more standard reference map in that it focuses on the features of a single route and eliminates other, more general information. In this way, it can SIOUX FALLS ------- CHAMBERLAIN. S. DAK. $04 For Emergency —— RoldServrcn see "SUPEQ‘VUMBER" Insure —— _ -..L“- '. O L SIOUX FALLS I “(an turn: 0 131 27$— m . Stanley Como" 4-343" . 65 .1 65 ./ crosses level to gently rolling, fertile. diversified rmland, producing corn. sorghum and small 6r—— gram. (2:25) 3 Q? 9500‘ 0 C-ufirfl‘“ll 'va-l-UIA 504 Figure 1. Example of an AAA T riptik strip map. The reproduction here is at 75% of the original. (© AAA Reproduced by permimion) 10 be seen as a set of instructions for moving from one place to another. It is ideally suited to route following but poorly suited to route planning (MatEachren, 1986). It seems reasonable that strip maps created for route following possess a sufficient inherent direction to warrant orientation in the direction of movement. - -reM Another type of map that does not consistently follow the north-at-the-top convention is the You-Are-Here (Y AH) map. YAH maps are found in various locales such as shopping centers, hospitals, libraries, and college campuses. These terrains are complex enough that the first-time (or subsequent) visitor requires the map information to navigate successfully. A YAH map indicates the location of the map on the map itself and therefore (either implicitly or explicitly and precisely or irnprecisely, depending on the design) of a person viewing the map. Levine (1982) specifically addressed the design of YAH maps. He focused on improvement of the map through clear and complete YAH symbols and on placement of the map to best allow the map user to relate the map to the terrain. YAH maps are generally in a fixed, vertical position (and therefore in a fixed orientation relative to the viewer), against a wall or attached to a post. Levine proposed an equivalence between “forward” in the horizontal plane of the terrain and “up” in the vertical plane of the YAH map. This equivalence is due to a 900° forward rotation of a horizontal map on which directions parallel the terrain. This correspondence has also been described and illustrated by Shepard and Hurwitz (1984) (Figure 2). Based on this “Forward = Up” equivalence, Levine suggests “aligned” design and placement of the YAH map, such that the upper part of the map corresponds to the area forward in the terrain. With such “aligned” placement, there also exists direct agreement between right and left on the map and right and left in the tenain. 11 Point B is further ahead of Point A on the horizontal ground and “higher" than Point A' on the vertical “map". A right turn from Point B to Point C on the ground also projects to Figure 2. Diagram of “Forward = Up” Equivalence. 12 Levine et a1. ( 1984) investigated the relative ease of use of aligned and misaligned YAH maps. “Misaligned” maps are those placed out of alignment by 900° or more; “contraligned” refers to the specific case of misalignment by 18000. In a laboratory situation, subjects viewing slides of YAH maps were asked to indicate in which direction they would move toward a target location. Even when extensive instructions about the YAH map and YAH symbol were provided, correct responses were given substantially more often on the aligned than the misaligned maps. Subjects tended to ignore the YAH symbol on the misaligned maps and respond as though the map were aligned. Field testing confirmed these results. Subjects were asked to navigate through an unfamiliar complex floor of a university library to a target office, having viewed a YAH map (aligned or misaligned) of the floor. Subjects who viewed an aligned map reached the target significantly more often and more quickly than did those who viewed a misaligned map. Levine et a1. (1984) have demonstrated that aligned YAH maps are easier to use than misaligned maps, thereby supporting the “Forward = Up” equivalence. That is, they have shown that YAH map users tend to assume that what is “up” on the vertical map is “ahead” in the environment, and therefore that the position of the map relative to the terrain makes a difference. These results are in agreement with the findings of Shepard and Hurwitz (1984) whose study investigated subjects’ interpretations of right and left turns from aligned and misaligned paths. They found that identification of a turn as right or left takes increasingly longer as the direction of the line going into the turn departs farther from alignment. Tversky (1981) found that subjects’ sketches of local environments were drawn from subjective constructions rather than from memory of previously learned maps. Only a small minority of subjects drew maps with north at the top of the page; the vast majority drew aligned maps, “where the ego is location at the bottom center” such that a right turn on the map is a right turn in the real world. Research on axes and orientation has shown that maps with conventional, north-at- the-top orientation are most helpful in some map use tasks (Cerny and Wilson, 1976; 13 Lloyd and Steinke, 1984), whereas maps with inherent direction of movement may be more useful when oriented other than conventionally. You-Are-Here maps oriented with the terrain allow map users .to head off more successfully toward their target than do misaligned maps (Levine et al., 1984). fiegmphical Direcg'gns and You-Are-Here Maps The study by Levine et a1. did not involve geographical direction navigation. Rather, the setting portrayed by the maps was a “closed” one in which subjects indicated movement forward, to the right, etc., and no mention of any outside reference system was made. All directional references were either absolute (a line drawn on a page) or relative to entities on the map (“facing building J”). The use of a closed system with no references to geographical direction avoids the recognition that there exist two sets of assumptions about the orientation of a map. The dominance of the north-at-the-top convention causes map readers to assume that north is at the top, and correspondingly, that east is to the right, south to the bottom, and west to the left on the map. In conflict with this is the finding that YAH map users tend to assume that maps are aligned with the terrain (Levine et al., 1984). In contrast to the experimental setting of the study by Levine et a1. (1984), real- world settings in which YAH maps are found or used often involve “geographical direction content” or “geographical direction processing” on the part of the user. Names of entities commonly contain geographical directions (e. g., “Old South Hall”; “the west concourse”), and geographical directions are frequently used in giving directions (e.g., “Go north on State Street”) or to indicate relative location (e.g., “It’s the building just east of the fountain”). The stacks in the Main Library at Michigan State University are divided between the east and west wings. References to “north campus” and “south campus” to designate the parts on either side of the river are common. And the local shopping mall contains Meridian West and Meridian East cinema complexes. 14 YAH maps are often found in enclosed, indoor settings which offer few or no environmental clues as to where the geographical directions lie. YAH maps users are generally unfamiliar with the terrain (hence they consult the map) and therefore even in the case of an outdoor setting such as a college campus, the environmental geographical directions may not be known. Also relatively inaccessible are the points of the compass. While these may be learned in relation to fixed local landmarks such as bodies of water, mountains, etc., or may be roughly inferred (if prOper account is also taken of time of day and of season) from the direction of the sun or shadows, during the day, or of constellations, during the night, the points of the compass are not available to humans in the absence of familiar landmarks, clear sky and knowledge of time of day and year (Shepard and Hurwitz, 1984, p. 163). Thus, the frequent and everyday use of geographical directions combined with map users’ frequent unawareness of location and geographical directions in YAH map settings raise the question of whether north-at-the-top orientation might be more effective for YAH maps. It is logical that the design of the map should encourage the most effective and appropriate map use strategies (Blades and Spencer, 1987). Studies have shown that the axes of the map sheet play a major role in map reading and map use. Map readers are accustomed to the vertical axis representing north, and this conventional map orientation has been shown to be most effective in some map reading tasks. In the specific case of YAH maps, orientation in alignment with the terrain has been shown to allow most successful navigation. These two apparently conflicting findings can co-exist because pe0ple use different maps differently. The setting and purpose of a map are critical to its design and use; the mapmaker should use the orientation that is most effective under the circumstances in which a map is to be employed. To this end, further research into specific 15 map uses and settings is needed to provide insight as to which map orientation is most effective. Ideally, map designers could then design maps with the most helpful orientation. W This research is designed to investigate the orientation of YAH maps in relation to geographical directions to discover the most effective design for YAH maps. The purpose of this research is to evaluate the usefulness of a conventionally oriented YAH map when geographical directions are involved in the use of the map. Maps oriented with north at the top can help readers recognize familiar geographic shapes such as lower Michigan, but how effective that orientation is in wayfinding tasks, and particularly in the YAH map setting, has not been conclusively determined. The goal of the present study was to determine whether north-oriented YAH maps are more useful in situations involving geographical directions. An experiment was devised to compare north-at-the-top YAH maps to environmentally aligned YAH maps. To determine the effectiveness of north-at-the-top YAH maps, the following hypotheses were proposed: Hypothesis 1: Map users can perform map reading tasks that involve mm direction references more accurately with north-at-the-top You-Are-Here maps than with aligned You-Are-Here maps. Hypothesis 2: Map users can perform map reading tasks that involve rel_atiye direction references more accurately with aligned You-Are-Here maps than with north-at-the-top You-Are-Here maps. To test these hypotheses, a methodology was devised involving map user response to map reading problems using north-at-the-top and aligned You-Are-Here Maps. CHAPTER TWO THE EXPERIMENT To test which YAH map orientation allows better performance of tasks involving geographical direction terms and of tasks involving relative direction tasks, test maps of both orientations were produced for two locales. Two groups of subjects viewed slides of these YAH maps and answered questions based on the maps. IcstMam Two sets of YAH maps were included in the test. From among the settings in which YAH maps are appropriate, it was decided to use university campus settings because campuses are sufficiently large and complex to allow a significant number of questions to be derived. The maps were based on real university campuses (San Francisco State University and University of California, Santa Barbara). Feattn'e names and configuration were changed somewhat to reduce the likelihood of a subject identifying the locales and possibly biasing the results. The test campuses were named Fresno State University and Santa Maria University. Two maps were prepared for each campus: one with aligned and one with north-at-the-top orientation. A distinctly visible YAH indication included the words “You Are Here”, a short straight line representing the map itself, and an arrow pointing toward the map-line. It followed Levine’s(1982) recommendations on improved YAH symbols. He noted that the relationship between the map-line and the arrow allows the map user to determine his or her location relative to the map and thus in the terrain. The YAH symbol was located in the appropriate orientation and location on each map. On the aligned maps, the arrow is at the bottom of the map with the arrow pointing up (Levine, 1982). The YAH symbol on the north-at-the-top test maps is located at one side of the map, pointing toward the middle of the map. 16 17 In addition to the YAH symbol as described above, a line drawing of a map user (referred to as “the student” in this study) was shown with each map, to allow subjects to grasp that the map portrayed is being viewed from the YAH symbol. This was intended to avoid, in the subjects’ minds, the confounding issue of whether or not the map was aligned. Except in the unique case of a map which is aligned and has north-at-the-top (which is rarely feasible in real-world YAH maps and did not occur in this study), north- at-the-top maps are misaligned. As this study did not directly address the alignment- misalignment issue, but rather the effect of north-at-the-top (and inherently misaligned) YAH maps, the pictorial “map user” was provided to reduce the influence of that issue. Cardinal geographical directions were used in some feature names on the test maps (e. g., South Campus Drive; West Gate) and a prominent north arrow was included on all maps. Copies of the maps are presented in Figures 3, 4, 5, and 6. The maps were prepared in black and white using Aldus Freehand“ on an Apple Macintosh“ computer in the Center for Cartographic Research and Spatial Analysis at Michigan State University. Maps of the real campuses were scanned and these images enhanced (and altered somewhat) with the illustration software. Prints of the maps were produced on an Apple LaserWriter“; these were then photographed, using Kodak EPY 135 film, and slides were commercially produced. flies; Questions To measure the effectiveness of different orientations, sixteen questions were developed for each campus. The corresponding questions were parallel between the two sets; thus, question No. 5 for the Santa Maria University map (“The student will walk in which direction to go to Parking Lot A?”), corresponds to question No. 5 for the Fresno State University map (“The student will walk in which direction to go to the Student Health Center?”). l8 Fresno State University Campus Map ——————————— ‘t " T '— \ N(— 'l \ \ / " _ —' _ 'J ‘ // I i I \ i Cox i Stadium M \ Student Umon Francisc - Bldg. Science \‘chcnnu Theater ' \ I I Ro rt \ I I M I Strong Hall\ ‘ W / I"s II ' v / Q / ~ Malone \Student la? / HST—é) Field y I § I Child Care / \lélealth ll 15 / Ill, ‘6 Center // On Q I h§l '/F // Figure 3. Aligned Version of Fresno State University Test Map (Center / ’— \ simian? ’ P”°"°'°3V\ Fresno State University Health Campus Map \ \ ¢\ 0‘ \ \ \ .Chm Care - Mary Park Center Hall Robert Strong Hall McKt-mra 'l'hcutcr Figure 4. North-at-the—top Version of Fresno State University Test Map 20 Receiving Santa Maria University —>N All campus buildings are (.1 accessible to wheelchairs. East Gate ‘ Here Figure 5. Aligned Version of Santa Maria University Test Map 21 CentralR ‘ ' ' _- Stores and Santa Maria Unrversrty 1' Receiving . All campus buildings are West a accessible to wheelchairs. Gate Goleta ' d. 97213 (W) ~ ~ \§§ \ Stk 9 Lot 5 - ‘3 \ 55 Q \ \ xx - N at: a.) T \\ ,g ~Z‘ ~=s \ . .4 - v . South . Y7 Hall _%torke 9' ower do -o "m _0 Faculty Club , _, , Figure 6. North-at-the-top Version of Santa Maria University Test Map 22 Of the sixteen questions for each campus (see samples in Figures 7 and 8), the eight odd-numbered questions involved geographical directions (north, south, east, west, and intermediate ones) and the eight even-numbered questions involved relative directional references (right, left, in front, behind). Twelve questions (Nos. 1—8, 10, 12, 14, 16) were ’9 relative to the pictorial “student’s location and four questions (Nos. 9,11,13,15) were relative to the location of campus features. Fourteen questions were multiple choice, and two questions (Nos. 7, 10) asked the subjects to draw a line from a given point. A copy of the actual test booklet used in the experiment is reproduced in Appendix A. W You-Are-Here maps are generally in a fixed vertical position, and thus in a fixed orientation to the map viewer. Slides of YAH maps projected on a vertical screen were used in this study to retain that specific relationship; providing a subject with a hand-held map would allow the map to be turned and thus allow the orientation between map and viewer to vary. Testing individual subjects on more than one map raises the possibility of subjects’ improving their performance over the course of the test, or “learning”; such learning could bias the results. To reduce the effects of any such learning and any other possible effects of order, two groups were tested and viewed the maps in reverse order: aligned first and north-at-the-top second; north-at-the-top first and aligned second. Both test groups viewed a Santa Maria University campus map before seeing a Fresno State University map. A pilot test determined conditions necessary for clear visibility of the map slides by the subjects. Two projectors displayed identical images on screens in the two front comers of the room. Blinds were drawn and overhead dimmer lights were adjusted to allow subjects to read the test booklets. Subjects in the pilot test were allowed unlimited time to view the slides and answer the questions. The map slide was not changed until everyone 1) 2) 3) 4) 5) 6) 7) Figure 7. 23 PLEASE CIRCIE THE CORRECT ANSWER OR FOLLOW OTHER DIRECIIONS Which entrance of Cox Stadium is farthest from the student: east or west? The Psychology Building is: to the student's left, to the student's right, ahead of the student, or behind the student. West is: to the student's left, straight ahead of the student, to his right, or behind him? Which is nearer to the student: the McKenna Theater or Cox Stadium? The Student will walk in which direction to go to the Student Health Center: north, south, east, or west? The majority of the campus is: in front of the student, to the student's left, to the student's right, or behind the student. Place your pencil on the dot below. Assuming that your pencil is the student looking at the map. draw the direction in which the student would walk if he decides to walk southwest. PLEASE CONTINUE ON THE NFX'I‘ PAGE Sample of Questions for Fresno State University Test Map l) 2) 3) 4) 5) 6) 7) Figure 8. 24 PLEASE CIRCLE THE CORRECT ANSWER OR FOLLOW OTHER DIRECTIONS Which entrance of Buchanan Hall is closer to the student: the north entrance or the south entrance? The Ag. Hall is: to the student's left, to the student’s right, ahead of the student, or behind the student ? East is: to the student's left, straight ahead of the student, to the Student's right, or behind the Student? Which is nearer to the student: Cheadle Hall or South Hall? The student will walk in which direction to go to Parking Lot A: north, south, can, or west? The majority of the campus is: in front of the student, to the Student's left, to the student's right, or behind the student? Place your pencil on the dot below. Assuming that your pencil is the student looking at the map, draw the direction the student would walk if he decides to walk southwest. PLEASE CONTINUE ON THE NEXT PAGE Sample of Questions for Santa Maria University Test Map 25 had answered all the questions. Results revealed an overwhelming proportion of correct answers; it was decided to limit the time for each map slide to six minutes, to induce distinction between the maps. The test booklets consisted of a consent form on top and three pages (pink) of questions about the Santa Maria University map followed by three pages (blue) of questions about the Fresno State University map. The sets of questions were color coded to allow clear distinction between and reference to the two sets during the test. Students in two undergraduate geography courses were tested. University students were deemed suitable subjects, as they must frequently navigate in the sorts of complex and often unfamiliar settings in which YAH maps are appropriate. The test lasted approximately fifteen minutes. In one case, the test was administered midway through the class session; all students in that course were tested The test in the other course was administered at the beginning of the class session; students who arrived late were not tested. Seventy-one subjects were tested in one course and fifty-seven in the other. A brief introduction (Appendix B) informed the subjects that the experiment concerned YAH maps and provided a general description of YAH map use. A slide of the pictorial “student” viewing a YAH map was shown (Figure 9) and was described as a student looking at a YAH map of an unfamiliar campus. The next slide was a campus/orientation combination on which that group would not be tested. It was explained as a blow-up of the map the picrorial “student” was looking at. Based on the findings of Levine (1982) and Levine et a1. (1984) that subjects have difficulty grasping the components and function of the YAH symbol, these were described. Test booklets were then distributed and subjects were asked to read and sign the consent form. It was then explained that a map slide would be shown for six minutes during which they were to answer the sixteen questions on the pink pages; the first test map slide was then shown. The subjects were told when three minutes and five minutes had passed. 26 Figure 9. “Student” Viewing You-Are-Here Map 27 At the six minute mark a blank screen was projected while subjects were asked to turn to the blue section of the test booklet They were told they would also have six minutes to answer the sixteen questions based on the second test map, which was then shown. The three minute and five minute marks were again noted. After six minutes a blank screen was projected and the booklets collected. The focus of the study on the orientation of the maps in relation to geographical directions was explained. A transcript of all oral instructions is provided in Appendix B. CHAPTER THREE RESULTS AND DISCUSSION Correctness of response to the multiple-choice questions was straightforward; subjects either chose the correct answer or chose an incorrect answer. Responses to the questions in which subjects were asked to draw a line in a specified direction were deemed correct if the line drawn was in the correct quadrant, following Levine et al. (1984). A score for each subject was determined by the total number of correct answers out of the thirty- two questions asked. To create test groups of equal numbers of subjects, as necessitated by the statistical test to be performed, the responses of fourteen randomly selected subjects were eliminated from Group 1, the larger group; thus each test group comprised fifty- seven subjects. Wigwam On examination, a large proportion of high scores was revealed. The fifty-seven subjects in each test group were each asked thirty-two questions; thus each test group answered 1,824 questions. Group 1 subjects answered incorrectly only 264 questions (14.5%). Group 2 subjects gave incorrect answers to 324 questions (17.8%). Overall, subjects answered only 16.1% of the questions incorrectly. Normal probability plots confirmed that the scores were not normally distributed. Transformations of the data failed to produce normal distributions. I decided, therefore, to use non-parametric tests, which do not rely on assumptions about the distributions of the scores. Since the sample size was large (fifty-seven in each group), parametric tests should also be valid and I have used them as confirmatory tests. A Mann-Whitney U test was used to determine whether the two test groups were from the same population. This test analyzes the differences of mean between samples by comparing the ranks of the observed values. All scores for each group were ranked; the rankings for each group were summed and the U statistic was calculated. The value of U 28 29 can be used to estimate the probability of the observed differences between groups occurring by chance. The mean ranks, U statistic, and significance are presented in Table l. The probability value, 0.079, was low but above the chosen rejection level of 0.05. Since the difference between the two groups could be attributed to chance, the two groups could have been aggregated; however, because of significant differences in results between the campuses, the two test groups were examined separately . Table 1. Mann-Whitney U Test for Difference between Test Groups H0: Samples are from populations with identical distributions (or = 0.05) 1251919212 2 __L__a_Mc n R nk 1.1 2 Decision 1 57 69.63 2388 0.079 acceptHo 2 57 58.11 Given the variables of this study, an analysis of variance would appear to be the appropriate test. However, conventional analysis of variance does not allow for the non- parametric, multivariate, matched pair character of this study. Thus, in order to maintain some distinction among the multiple variables, the test groups were analyzed separate1y. For each of the two test groups, differences in performance on the north-at-the-top and aligned map orientations were analyzed for the geographical direction questions; the results for the two groups were then compared. A similar analysis by test group was used for the relative direction questions. A more detailed, question-by-question analysis followed. 30 'n M ' 'n The primary objective of this study was to determine whether the presence of geographical directions in map use tasks influences the effectiveness of YAH map orientations. Effectiveness was measured by map reader performance on questions about the map viewed. Since the subjects answered parallel questions on the two maps, their performance on the two maps can be directly compared. Figure 10 indicates on which maps, by campus and orientation, each group was tested. Subjects in Group I viewed and answered questions about the aligned Santa Maria University map (Map C) and then answered parallel questions about the north-at-the-top Fresno State University Map (Map B). Subjects in Group 2 saw the north-at-the-top Santa Maria University map (Map A) and then the aligned Fresno State University map (Map D). Since eight of the sixteen questions for each map were geographical direction questions and eight were relative direction questions, each subject could have a maximum score of eight on the geographical direction questions and eight on the relative direction questions for each map. A subject who scored seven on the geographical direction questions on Map C and scored five on the geographical direction questions on Map D performed better on Map C for those questions. Tabulations were made, by groups and by question type, of the number of subjects who performed better on one of the orientations or equally on both. The tabulations were examined to determine whether one map orientation was more effective for each of the two question types. For either of the types of question, if the two orientations were equally effective, one would expect the number of subjects performing better on the aligned map to approximately equal the number performing better on the north-at-the-top map; that is, one would expect the distribution of the scores between the two orientations to be even. By the same reasoning, a more effective map orientation would cause more subjects to perform better on that map. The observed numbers can be compared to the expected numbers, and the differences can be tested for significance. 31 . Santa Maria Fresno State UHIVCTSII)’ Campus University Campus North-at-the-top Group 2 Group 1 onentatron (Map A) (Map B) Aligned orientation Group 1 Group 2 (Map C) (Map D) Figure 10. Campus and Map Orientation Matrix 32 Chi square tests with one degree of freedom were used to test the probability that the observed frequencies were so different from the expected frequencies that they were unlikely to be due to chance variation. X2 is a measure of this difference: the greater the difference between the expected and observed values, the larger the value of X2 and the greater the probability, that the distribution is not random. The formula for the X2 statistic is: x2 = Z ( O - E )2 E where O is the frequencies observed and E is the expected frequencies. In this study, the number of subjects who performed the same on both maps was split and half attributed to each of the map orientations. This procedure allowed the performance of the total sample of fifty-seven to be retained. Logically, the performance of those subjects who scored the same on both map orientations reflects directly on the issue of the influence of map orientation and must be considered. To eliminate these “ties” and evaluate only subjects who performed differentially between the map orientations would be to ignore relevant data and possibly bias the results. The observed frequencies thus comprised the number of subjects who performed better on that map orientation and half the number of Subjects who performed the same on both orientations. The expected frequencies are half the total number of subjects (fifty- seven), or twenty-eight and one-half. After 12 is calculated, the probability of such a distribution occurring by chance can be estimated. The calculated value of x2 can then be compared with the critical value of X2 at the 0.05 level of significance. A calculated value of X2 less than the critical value indicates such a distribution could be expected to occur by chance more than five times in one hundred. In such a case the null hypothesis could not be rejected. 33 WW Both of the research hypotheses involved comparisons of different map orientations. Since the test groups viewed different campus-orientation combinations, the performances of the two test groups were analyzed separately in order to hold this variable constant. The first research hypothesis, which addresses geographical direction questions, was first tested for each group and the results were compared; a similar approach was then applied to the second research hypothesis, concerning relative direction questions. hi lDir ti n uestions: R sults To test the first hypothesis, the differences in subjects’ performances on geographical direction questions on the two maps were compared to the expected even distribution. The results of the differences between the two map orientations, by test group, are presented in Table 2. In Group 1, fifteen subjects performed better on the north-at-the-top map, twenty- one performed better on the aligned map, and twenty-one performed the same on both. When the “ties” are split between the two map orientations, the adjusted results indicate that twenty-five and one-half subjects performed better on the north-at-the-top map and thirty-one and one-half subjects did so on the aligned map. The difference between these frequencies is not significant at the 0.05 rejection level; the null hypothesis for this comparison cannot be rejected. In Group 2, nine subjects performed better on the north-at-the-top map and twenty- eight performed better on the aligned map. When half of the twenty subjects whose performance did not differ between the map orientations were allotted to each of the map orientations, nineteen subjects counted as performing better on the north-at-the—top map and thirty-eight on the aligned map. The difference between these frequencies is significant; the null hypothesis for this comparison is rejected. 34 £883. a: E82. 36 as a 303.0 vwfi ax g .326 and chd on 33% 53 :0 088 3.58.2. 9.3 38.33 .8 2:32 a 33.3% Soc .5 2:8 35.8.2— 2; 38.33 a 83.52 Gad I 3 an N 5.3.3 36.8.03 23 98......» a 83.52 nAm g mad—dug ..o .023 Senora 9.3 98.38 e 83.52 235:3... 92 05 503.3 9335.6 333 a 855.88.. .88.. "a: 3838.6 8:85 833.289 82.8:on 9.2 :00an mooeobba .8 38,—. 0.33m 20 a 32.3.. @ 3a.— 33333.3. 8 .28.. 35.8.2. 23 38.33 e 83.52 “.830 nfiN 3.8.8.. m — 3a.. 83.33% 8 .88.. 38.8.2— 2; 98.3.... .o 8252 135 .u use. 35 in'Di Si fR 1 Map orientation did not make a significant difference in Group 1 subjects’ ability to correctly answer questions involving geographical directions. Subjects in Group 2 did , perform significantly better on the aligned map. Results for both groups fail to confirm the first research hypothesis; the results for Group 2 in fact point to the opposite conclusion, i.e., that aligned YAH maps are more effective for map tasks involving geographical directions. However, the results for the two test groups must be taken together, and the difference in the findings addressed. This discrepancy between the two groups was unforeseen and in fact every effort was made to achieve uniformity between the groups. Factors that might have caused the difference in results are now examined. While some aspect of testing conditions may have differed between the groups, both groups were tested in the same room with the same lighting conditions and arrangement of projectors and screens. However, Group 2 was tested in mid-morning and Group 1 in the evening. It is possible, although unlikely, that the Geography courses from which the subjects were drawn attract students with significantly different map reading experience, spatial abilities, or analytical skills. Another variable which must be considered is that two campuses were used in the test. Both test groups saw one map of each campus, but the two groups did not see the same campus-orientation combinations. It is possible that a particular campus viewed in a particular orientation may affect map reading tasks in such a way as to cause the discrepancy in findings for the two test groups. Statistical tests confirm a significant study- wide difference between the two campuses. For each subject a score was calculated for each campus. All scores on each campus were ranked. A Mann-Whitney U test was used to test the difference of rank means. The calculated probability of 0.008 is less than the rejection level of 0.05; the null hypothesis of no significant difference is rejected. To 36 examine this with a different approach, the difference between performance on the two campuses, by subject, was tested for independence with a matched pairs t-test. The resulting probability of less than 0.001 confirms a significant difference between the two campuses. The experiment included maps of two locales in an effort to broaden the applicability of the findings, that is, to generalize beyond map-specific results. Since two campuses were used as the map settings, it was hoped that consistent results would apply at least to campus YAH maps generally. The maps were designed to be similar and were so considered by reviewers before the tests were conducted. Again, both locales were university campuses; they contained comparable numbers of buildings and other features; a uniform north arrow was evident on each; the buildings were filled with the same gray tones on both campuses, providing like contrast and visibility; the “student” map viewer appeared with the YAH symbol on both campuses; and parallel text was of the same size on both campuses. There are, however, conceivable differences between the campus maps which may account for the.dissimilar test results. The map of Santa Maria University portrays the coastline of the Pacific Ocean which may, in a particular orientation, function as a frame of reference and influence map readers’ sense of direction. For one of the campuses, the alignment of the buildings to the map frame may interact with one of the orientations in a way that doesn’t occur with the other orientation. However, this study did not foresee, and the experimental design did not address or test, the influence of any such variables. In sum, the results for the two test groups differ. In Group 1, map orientation had no significant effect on performance of map use tasks involving geographical direction; in Group 2, aligned orientation allowed significantly better performance on those map use tasks. These inconsistent results may be attributable to variation in test conditions, to effect of uncontrolled variables that differed between these maps, or to a type I error. 37 The second research hypothesis, which follows on the findings of Levine et a1. (1984), posited that aligned YAH map orientation would be more effective for relative direction map use tasks. It was tested by comparing the differences in subjects’ performance on the relative direction questions on the two map orientations to the expected even distribution. The results, by test group, are presented in Table 3. In Group 1, twenty-five subjects performed better on the north-at-the-top map, sixteen performed better on the aligned map, and sixteen performed the same on both maps. With the “ties” split between the two map orientations, thirty-three subjects counted as performing better on the north-at-the-top orientation and twenty-four as better on the aligned map. The difference between these frequencies is not significant at the 0.05 rejection level; the null hypothesis for this comparison cannot be rejected. In Group 2, four subjects performed better on the north-at-the-top map, forty-three performed better on the aligned map, and ten performed the same on both maps. Splitting the “ties” results in nine subjects counted as performing better on the north-at-the-top map and forty-eight on the aligned map. The difference between these frequencies is significant at the 0.05 rejection level; the null hypothesis is rejected. l'v Dire in esinzDic sinofR lt These dissimilar results between test groups are consistent with the findings on the geographical direction questions. In Group 1, map orientation did not affect performance on relative direction questions. In Group 2, the aligned map again allowed better performance on the relative direction questions. The Group 2 results support the second research hypothesis; the Group 1 results do not. For the reasons discussed regarding the results of the geographical direction questions, the dissimilar results between the test groups were unforeseen and puzzling; the above discussion of possible variants between 38 o: .00? a: 5000.. 3.3 8.8 NR g NR .336 $3 a... «X g «X .826 on 53 ..o 083 35.8.0.— 2; 58.3.... .o .0952 2 30:35.... 53 :0 088 38.8.2. 2.3 08.3.... .o 83.52 89¢ a 3 3. Q. .8 .033 35.8.0.— 2; «.83... .0 .0952 en 3 38.30 .6 .0203 35.8.0.— 23 3003:. ..o .352 28.8.85 03. 05 503.3 3.58.... 2:33 a. 08.8.8.0.— .Bom "an 3:87.30 5.80.5 03.205 38.8.8.0 9.2 503.09 32.0.0me .8 «30,—. 0.33m 20 a 8.3.3.. 8 .033 35.8.2— 033 900.38 .0 2:52 mm mu .5 .093 35.8.0.— 3... 800.33 ..o .0952 .m 055. 39— 32...? 3a.— 39 groups applies to this analysis of relative direction questions. Some aspect of testing conditions and/or the effect of uncontrolled variables is likely to be behind the dissimilar findings. The above analysis, aggregated by question type (geographical direction or relative direction), gave an overall view of performance. A more detailed, question-by-question analysis was undertaken to investigate differences in individual questions. Because of the dissimilar results between test groups in the aggregated analyses, quesfion-byquesfion tests were also done by test group. For each question, the numbers of subjects who performed better on the north-at-the-top map, performed equally well on the two maps, or performed better on the aligned map were tabulated. The “ties” were split between the two map orientations. Chi square tests were used to test the distribution of these performance differences against the expected even distribution. The result of all tests of differences between the map orientations, by question, are presented for Group 1 and Group 2 in Tables 4 and 5, respectively. The analysis of Group 1 reveals significant differences between map orientations for one geographical direction question (No. 7) and for one relative direction question (No. 10). For both, subjects performed significantly better on the aligned map. These questions are linked in that they are the two questions which asked subjects to draw a line in a given direction, and which are like those in the Levine et al. (1984) study. As in that study, the aligned orientation proved more effective for this task; here, it was more effective even for the tasks that involved geographical directions. The other questions revealed no significant difference in performance between map types. For Group 2 subjects, neither question No. 7 nor question No. 10 revealed a significant difference in performance between orientations (Table 5). However, A093 50.. ..e 3:558. $2.88 a... a... .v , 68 m 68 m. n. 0.3.08. a... 82. a. 63 a 68 e m. £388 3.... 82 a. a . 5.... m. .. cm .38. a... E. a a .8 . as N a a: .8? a... 5.... a .3. mm 68 N .. fags. x... 82. .n 3 . 68 n n a: 5.88 3.. as d a. 68 n 55. . m £2.08. an E... e. 5.. v .03 0 . 553 $.00...» N. 3553.... 353...... 33...... 302.0 53 so 053 :0 .033 .5 .33 5.30:0 35.8.3 35.8.3 35.8.3 9.3 38......» 9.3 98.3... 9.3 38.3... 3.25.... .o 3.52 .o .852 38.33.... 05 5003.3 5% .3? .003. 080.383 3 35:2 Go... n .8 2.2.8503 03. 05 5003.3 3.3.5.... 3.9.90 0. 00:05:83 .053 "a: A. 95.9 058.8508 9.2 5003.3 32.0.8.5 .8 .5330 .3 9.0... 0.03m 20 .v 03.... 41 c: 2.000.. c... 2.000.. c... 2.000.. a: .00.... 0: 2.0000 c: 2.000.. o: 2.000.. c: 2.000.. van end 3% can van end cw...” 0mm «x; .856 08d 0n...— 8nd to... R. 6 $86 95d Bed «8 Rfiflfi8$998 20.33.... 50.. :0 0.5... 00.58.00 9.; 28.3... ..o .252 6.2.08 .0 030... .8. 68. .20. .208 .8. .20. .20. .00.. :0 .0300 008.88.. 9.3 0.00.33 .0 .0952 3mm... 5.. 60.0 6.5 Mv-tMV'O 8 .33.. 00:28.2. 9.... 3.0.3... .o .252 c. v. N. o. viooo 5.80:0 42 0.. .003. 0m. .0080 0: .0080 0.. .00000 0.. .00000 0... .0080 0... .0080 0... 2.0000 .0000 .00.. 00 00.0.0000. 00m 3.0 .0 mm 35 .N 00.. N. N 00 0.0 N. 00m mm. .0 mm 8 .. 3...” 8N .0 .0 g m 00m NN. .. 0m 6.00 m. 3...” wk .. 0m .80 0 00m 90 .0 00 g 0 00m 000 .0 00 .8 N «0.0.0:...» 00 500.0000 00.0.0003... .3000 .000 00 0.000 00 .0000 00008.00 00008.00 003 0.00.000 003 0.00.0... ..0 .0052 .0 .0057. 000000.000 0... 0003.00 .00. ..000: .0000. 0800.00.00 0. 0.00.52 60.0 a 0. 000000.000 03. 0... 083.00 00.00.05 0:00.00 0. 000000800 .0000 .0: .0 95.0. 000080000 002 0003.00 0800.805 .8 0000000 00 0.00... 0.0000 0.0 .n 0.00... 0 0. n m . 03.0.5 0000000 43 0.. .00_.0.. 0.. 2.0000 0: 2.0000 0: 2.0000 0: 2.0000 0.. .0080 0.. 2.0000 0.. 2.0000 00..“ 0»... 00m can 00.... SM 006 Ed 050.00; .8000 080 000.0 «no.0 «SN NSN 000.0 0.0.0 «no.0 x @XQSSQRQ .00.. 00 00.00 020.0000 003 800......» .0 .3052 80000. .n 0.00... GAG 0. 0.9 .. A00 .. 00 6.0 ..o .0309 00008.00 003 0.8.300 .0 80.002 0. 0. N. 0. 010000 0000000 44 considerably more Group 2 subjects did perform better on the aligned than the north-at-the- top orientation for those questions: thirteen vs. five for question No. 7; eleven vs. none for question No. 10. Group 2 subjects performed significantly better on the aligned map for one geographical direction question (No. 15) and one relative direction question (No. 16). These are the last two questions asked for each map; since the aligned map was the second map these subjects viewed, the better performance on these last two questions of their test may be a result of subjects’ ‘learning’ and improving over the course of the test. While the last two questions asked of Group 1 (Nos. 15 and 16 on the north-at-the—top map) did not show significant improvement, a pattern similar to that of Group 2 is revealed. A considerable number of Group 1 subjects did perform better on those two questions on the second map they viewed (nonh—at-the-top) than on the first (aligned): thirteen vs. three for ' question No. 15; sixteen vs. three for question No. 16. Thus subject ‘learning’ may have had some effect on performance in this study. Other questions revealed no significant difference in performance between map types for Group 2. in- - inAnal i'Dicu in fR l The fact that both groups showed significance on the same number of questions (two) but on different questions, is puzzling. However, in both cases, the trend for those questions is supported by the other test group, although not at the significant level. Taken together, the results for the two test groups thus reveal four questions which show significance. The relative magnitude of the number of “ties”, i.e., subjects who performed equally on both map orientations, influences the results. The large number of “ties” can mask certain phenomena; while the overall test for Group 1 did not show significance between map orientations, the question-by-question analysis diminishes the impact of the 45 “ties” and reveals more detailed results. For most of the questions, the number 0 “ties” (Tables 4 and 5) represents a large proportion of the subjects (n=57) and can be seen as support for the notion that map orientation does not affect performance; that is, that either map orientation would be as helpful (or unhelpful) as the other. The question-by-question analysis for Group 1 revealed that, apart from the two significantly different questions (which favored the aligned map), the majority of questions favored the north-at-the-top map. More subjects performed better on the north-at-the-top map on seven of the eight geographical direction questions and on five of the eight relative direction questions (Table 4). The analysis for Group 2 (Table 5) revealed a different pattern. More Group 2 subjects performed better on the aligned map on five of the geographical direction questions and on eight (all) of the relative direction questions, including the two significantly different questions. These results indicate strikingly different trends for the two test groups. While Group 1 did not perform significantly differently on either map orientation overall, the north-at-the—top map was more helpful to subjects on twelve of the sixteen questions. On the other hand, Group 2 subjects clearly found the aligned map more helpful on thirteen of the sixteen questions. These equivocal findings again point to some fundamental difference between the groups themselves or the campuses, such as the subjects’ abilities, the effects of the campus-orientation combination seen, or some other factor. Summary of Results In this experiment, subjects viewed an aligned map and a north-at-the-top map and were asked to answer geographical direction questions and relative direction questions about both. Since the questions for the two maps were parallel, a subject’s performance on one map could be compared with his or her performance on the other map. To determine 46 whether one map orientation was more effective for one type of question, the subjects’ performances on the two maps were compared. Two research hypotheses were tested by comparing the disu'ibution of subjects’ performance to a hypothetical even distribution. Two groups of subjects were tested, and separate analyses performed to allow other variables to be held constant. Findings differed between the two groups. To test the first hypothesis, the subjects’ performance on geographical direction questions was compared between the aligned and north-at-the-top orientations. This hypothesis was not confirmed, as Group 1 subjects did not perform . ..— significantly differently on either map orientation and Group 2 subjects performed better on the aligned map. The second hypothesis could not be unequivocably accepted either. Map orientation did not make a significant difference on Group 1 performance on relative direction questions. Aligned orientation was more effective for Group 2 subjects in answering these questions. Subjects’ performance on the two orientations was also examined on a question- by-question basis. For Group 1 subjects, the aligned orientation was more effective for the two questions which asked the subjects to draw a line in a given direction. For Group 2 subjects the aligned orientation was more effective for the last two questions of the test, which may be attributable to subject ‘learning’. Subjects appeared interested in and expressed enthusiasm about this experiment. After the test several subjects mentioned that they had never thought about why some navigational maps are easier to use than others. Other comments indicated that map users are not always aware of the orientation of a map. The north arrow was revealed as an important map element by the comment of several subjects that “the north arrow was really clear on all the maps”; the prominence of a north arrow in map use may dictate its prominence in map design. CHAPTER FOUR SUMMARY AND CONCLUSIONS This study investigated the issue of orientation in You-Are-Here maps. Traditionally, some salient feature or direction has been placed at the “top” of the map (the edge most distant from the reader); for whatever reasons, north is the feature to which maps are conventionally “oriented”. Map readers are quite familiar with this orientation, which is so strong as to be assumed if no north arrow is depicted A second useful orientation is alignment with the terrain such that what is “up” on a vertical YAH map is “ahead” in the environment. This study addressed YAH map orientation in settings involving geographical directions. ‘ . Two research hypotheses were tested. Based on map readers’ familiarity with conventionally oriented maps, the first hypothesis posited that map reading tasks involving . geographical directions would be performed more accurately with north-at-the-top YAH maps. The findings did not support this hypothesis: in one test group, the orientation did not significantly affect performance and in the other test group, the aligned orientation proved more useful. The second hypothesis followed the findings of Levine et a1. (1984). Their study, which did not involve geographical directions, found that subjects would set off in the correct direction from a YAH map significantly more often when the YAH map was aligned than misaligned. The present study thus tested the hypothesis that tasks involving relative directions would be performed more accurately with aligned YAH maps. This hypothesis was partially supported by the findings: one test group did not perform significantly differently between orientations and the other group performed significantly better on the aligned maps. The findings, while inconsistent in one sense, do indicate, in all cases where one map orientation was shown to be more effective, that the aligned orientation prevailed. North-at-the-top orientation was never shown to be significantly more helpful. This would 47 48 indicate that aligned orientation of YAH maps is likely not detrimental and may be preferred and more effective. The experiment was designed to simulate YAH map use as realistically as possible in other than field conditions and to allow testing of large numbers of subjects. While viewing a vertical map slide is not dissimilar to reading a posted YAH map, answering questions on paper does differ from the real-life task of moving away from a map in the conect direction. While Levine et a1. (1984) found results from field testing to support laboratory tests, it would be useful for further research on the effect of geographical directions to be done under real—life conditions. The specific maps used in this experiment raise several questions. A campus setting for the map was chosen because campuses are generally of an appropriate size and complexity to allow a reasonable number of navigational tasks. However, the subjects in this experiment were university students and as such may be quite familiar with maps of and navigation around a campus setting. Use of a different locale (e.g., airport, shopping complex, amusement park) with which students might have more difficulty might have provided more variation. The experiment included two different campuses in order to derive more general results than those based on a single map (campus) would allow. The campuses were designed to be similar enough to avoid a disparity in difficulty. The difference in results for the two test groups, however, points to the campuses as a possible influence. It is unclear from the findings whether the effect of map orientation is map-dependent; that is, whether some aspect of one of the test campuses is responsible for the dissimilar findings. Further research that tests more than two maps (locales) is recommended to investigate this issue. Other factors of interest to future research on this topic include field of study, gender and time limitation. The present study used as subjects students enrolled in geography courses; using students from other fields of study (e. g., biology, music, history) could provide insight into differential spatial abilities among people in various disciplines. 49 Gender was not addressed in this present study but could also be investigated for differential effects. Determining the amount of time needed to answer individual questions could prove quite useful. In the present study, subjects were allowed a block of six minutes to answer sixteen questions and it is therefore not possible to determine the response times for specific questions, or to analyze the degree of difficulty of the questions beyond “correct/incorrect”. It is possible, for example, that a subject may have answered quickly those questions he or she found easiest and spent the bulk of the six minutes answering one or two difficult questions. Limiting subjects to the same amount of time for each question, or alternatively, recording response times using maps displayed on a computer monitor, would allow far greater discernment of the difficulty of specific questions. A related issue of interest is the feasibility of YAH maps which are both aligned and north-at-the-top. Such maps would seem to provide advantages of both orientations and avoid any conflict map readers might experience in their assumptions about map orientation. Since aligned maps are designed for and specific to a given location, and because what is “up” on the map (north) would have to be ahead in the environment, however, the use of such maps is somewhat limited, for instance to southern entrances to YAH map-appropriate settings. But where possible, such maps might be doubly effective and less confusing, and therefore warrant investigation. APPENDICES Appendix A Test Questions 1) 2) 3) 4) 5) 6) 7) APPENDIX A (FRESNO STATE UNIVERSITY) PLEASE CIRCLE THE CORRECT ANSWER OR FOLLOW OTHER DIRECTIONS Which entrance of Cox Stadium is farthest from the student: east or west? The Psychology Building is: to the student's left, to the student's right, ahead of the student, or behind the student. West is: to the student's left, straight ahead of the student, to his right, or behind him? Which is nearer to the student: the McKenna Theater or Cox Stadium? The student will walk in which direction to go to the Student Health Center: north, south, east, or west? The majority of the campus is: in front of the student, to the student's left, to the student's right, or behind the student. Place your pencil on the dot below. Assuming that your pencil is the student looking at the map, draw the direction in which the student would walk if he decides to walk southwest. PLEASE CONTINUE ON THE NEXT PAGE 50 8) 9) 10) ll) 12) 13) 51 If the student were to walk from Mary Park Hall to the Residence Dining hall, what building would be on his right: Child Care Center, Robert Strong Hall, Verducci Hall, or none of these? Which building on campus extends farthest north: Humanities, Cox Stadium, Creative Arts, Psychology, or none of these? Put your pencil on the dot below. Assuming that your pencil is the student looking at the map, draw the direction in which the student would walk from the map to the library. Which corner of the Humanities Building is closest to the Admin. Building: northeast, northwest, southeast, or southwest? After walking around campus for a while, the student finds the Education Building is on his right and the Creative Arts Building on his left. What building is the student facing: Student Union, Science, Residence Dining, Parking Garage, or none of these? Which bleachers at Maloney Field are closer to Verducci Hall: the north bleachers or the west bleachers? PLEASE CONTINUE ON THE NEXT PAGE 14) 15) 16) 52 If the student were to walk from the Gymnasium to the Admin. Building, what building would be on his right: Science, Humanities, Psychology, Business or none of these? The Franciscan Building is in which direction from the library: north, south, east, or west? The student gets in his car in the Parking Garage and leaves campus via North State Drive. What will be the last building on his left as he drives off campus: Verducci Hall, Student Health Center, Mary Park Hall, or none of these? S T O P! DO NOT TURN THE PAGE PLEASE SIT QUIETLY UNTIL FURTHER INSTRUCTIONS ARE GIVEN 1) 2) 3) 4) 5) 6) 7) 53 (SANTA MARIA UNIVERSITY) PLEASE CIRCLE THE CORRECT ANSWER OR FOLLOW OTHER DIRECTIONS Which entrance of Buchanan Hall is closer to the student: the north entrance or the south entrance? The Ag. Hall is: to the student's left, to the student's right, ahead of the student, or behind the student ? East is: to the student's left, straight ahead of the student, to the student's right, or behind the student? Which is nearer to the student: Cheadle Hall or South Hall? The student will walk in which direction to go to Parking Lot A: north, south, east, or west? The majority of the campus is: in front of the student, to the student's left, to the student's right, or behind the student? Place your pencil on the dot below. Assuming that your pencil is the student looking at the map, draw the direction the student would walk if he decides to walk southwest. PLEASE CONTINUE ON THE NEXT PAGE 8) 9) 10) 11) 12) l3) 14) 54 After walking around campus for a while, the student finds Gervetz Hall on his right and the Arts Building on his left. What is the student facing: the bus stop, South Hall, the Library, Biological Sciences, or none of these? Which building on campus extends farthest south: South Hall, Kerr Hall, Ag Hall, Faculty Club, or none of these? Put your pencil on the dot below. Assuming that your pencil is the student looking at the map, draw the direction in which the student would walk from the map to the Library. Which end of the lagoon is closer to Parking Lot B: the northwest end or the southeast end? If the student were to walk from South Hall to the Arts Building, what would be on his right? Library, Gervetz Hall, the bus stop, Parking Lot B, or none of these? Which corner of Kerr Hall is closest to Cheadle Hall: northeast, northwest, southeast, or southwest? If the student were to walk from the Event Center to Parking Lot C, what would be on his left: Biological Sciences, Cheadle Hall, the West Gate, Centennial House, or none of these ? PLEASE CONTINUE ON THE NEXT PAGE 55 15) The Library is in which direction from the Biological Sciences Building: northeast, northwest, southeast, or southwest? 16) The student gets in his car in Parking Lot H and leaves campus via the East Gate. What will be the last feature on his right as he drives off campus: Events Center, Ag, Hall, Pauley Track, Centennial House, or none of these? STO P! DO NOT TURN THE PAGE. PLEASE SIT QUIETLY UNTIL FURTHER INSTRUCTIONS ARE GIVEN. Appendix B Oral Instructions APPENDIX B Oral Instructions “Hello. My name is Margaret Livingston and I’m a graduate student in Geography, studying cartography. Your instructor has agreed to let me use some of your class time, and I’m going to ask you to participate in an experiment. You will be viewing slides on the two screens in the front of the room; the same slide will be shown on both screens, so you can look at whichever one you can see best.” “I’m going to show you some maps and ask you to answer some questions about them. The maps are You-Are-Here maps, the kind you find at shopping centers, airports, large buildings, and other complex settings. In fact, there’s one just outside this building, which you may have seen. You’ve all used these maps at some time -- you arrive in one of these settings, and don’t know your way around, or don’t know how to get to a specific store or building, so you look at the map, which shows you where you are in relation to everything else. The questions you answer in this experiment will involve information on the map and you will decide on the answers based on the map.” (slide of the “student” looking at the map is shown) “Now I’ll first show you this example of a student looking at a You-Are-Here map. This student has arrived at a campus for the first time, and takes some time to look at the map to figure out where he is and where other things are.” (slide of example test map is shown) “This second slide shows the same student in the same spot looking at the same map, but also shows a blow-up of the map he’s looking at. The You-Are-Here symbol (indicated on slide) indicates the location of the map itself with this straight line (indicated on slide) and of the viewer with the arrow (indicated on slide).” 56 57 (blank screen is projected; test booklets are distributed) “You each have a booklet. I’ll ask you to read and sign the consent form, which is the white sheet on the top of your booklet. If you prefer not to participate, please just sit quietly.” (when all subjects have signed the consent form) “Next, you will be looking at a map and answering the questions on the pink pages. There are sixteen questions and you will have six minutes to answer them.” (first test map slide is shown) “Here’s the first map. You may begin. You have six minutes.” (at the three minute mark) “Three minutes have passed. You have three minutes left.” (at the five minute mark) “Five minutes have passed. You have one minute left.” (at the six minute mark; blank screen is projected) “Time’s up. Now please turn to the blue section. You will again have six minutes to answer these sixteen questions about this next map.” (second test map slide is shown) (at the three minute mark) “Three minutes have passed. You have three minutes left.” (at the five minute mark) 58 “Five minutes have passed. You have one minute left.” (at the six minute mark; blank screen is projected) “Time’s up. Please pass your booklets to the right, and they’ll be collected. Briefly, what I’m looking at in this research is the orientation of the You-Are-Herc maps. You may have noticed that one of the maps you looked at had north at the top and the other one didn’t. Also, some of the questions had geographical directions (what we think of as “compass” directions) in them and some didn’t I’m investigating which map orientation is more helpful when geographcal directions are used. Thank you all very much for your help.” LIST OF REFERENCES LIST OF REFERENCES Adeyemi, E.O., 1982, “The Effect of Map Orientation on Human Spatial Orientation Performance.” The Cartographic Journal, Vol. 19, No. 1, pp. 28-33. Blades, M. and Spencer, C., 1987, “How do People Use Maps to Navigate Through the World?” Cartographica, Vol. 24, No. 3, pp. 64-75. Braine, LG, 1978, “A New Slant on Orientation Perception.” American Psychologist, Vol. 33, No. 1, pp. 10-22. Cemy, J., and Wilson, J ., 1976, “The Effect of Orientation on the Recognition of Simple Maps.” Canadian Cartographer, Vol. 13, No. 2, pp. 132-138. Clark, H. H., 1973, “Space, Time, Semantics, and the Child” in Moore, T.E., (Ed.), Cognitive Development and the Acquisition of Language, New York: Academic ' Press. Downs, RM. and Stea, D., 1977, Maps in Minds: Reflections on Cognitive Mapping, New York: Harper and Row. Howard, 1. and Templeton, W., 1966, Human Spatial Orientation, New York: Wiley and Sons. Levine, M., 1982, “You-Are-Here Maps: Psychological Considerations.” Environment and Behavior, Vol 14, No. 2, pp. 221-237. Levine, M., Marchon, 1., and Hanley, G., 1984, “The Placement and Misplacement of You-Are-Hcre Maps.” Environment and Behavior, Vol 16, No. 2, pp. 139-157. Lloyd, R. and Steinke, T., 1984, “Recognition or Disoriented Maps: the Cognitive Process.” The Cartgraphic Journal, Vol. 21, No. 1, pp. 55-59. MacEachren, A., 1986, “A Linear View of the World: Strip Maps as a Unique Form of Cartographic Representation.” The American Cartographer, Vol. 13, No. 1, pp. 7-25. Robinson, A., 1952, The Look of Maps, Madison: University of Wisconsin Press. 59 60 Robinson, A., Sale, R., Morrison, 1., and Muehrcke, P., 1984, Elements of Cartography, 5th edition, New York: John Wiley and Sons. Robinson, A., and Sale, R., 1969, Elements of Cartography, 3rd edition, New York: John Wiley and Sons. ' Rock, 1., 1973, Orientation and Form, New York: Academic Press. Shephard, RN. and Hurwitz, S.,_ 1984, “Upward Direction, Mental Rotation, and Discrimination of Left and Right Turns in Maps.” Cognition, Vol 18, pp. 161- 193. ' Tversky, B., 1981, “Distortion in Memory for Maps.” Cognitive Psychology, Vol. 13, pp. 407-433. SEQHEQI Refemnces Calkins, RD, 1909, “Orientation of Maps.” Journal of Geography, Vol. 8, No. 1, pp. 15- 17. Gulliver, F.P., 1908, “Orientation of Maps.” Journal of Geography, Vol. 7, No. 3, pp. 55- 59. Hintzman, D.L., O’Dell, GS, and Amdt, DR, 1981, “Orientation in Cognitive Maps.” Cognitive Psychology, Vol 13, pp. 149-206. Kozlowski, L., and Bryant, K., 1977, “Sense of Direction, Spatial Orientation, and Cognitive Maps.” Journal of Experimental Psychology, Vol. 3, No. 4, pp. 590- 598. Loftus, G., 1978, “Comprehending Compass Directions.” Memory and Cognition, Vol 6, No. 4, pp. 416-422. MacEachren, A., and Johnson, G., 1987, “The Evolution, Application and Implications of Strip Format Travel Maps.” The Cartographic Journal, Vol. 24, No. 2, pp. 147— 158. 61 Maki, R., 1979, “Processing Relative Locations in a Natural Space.” Bulletin of Psychonomic Society, Vol. 14, No. 1, pp. 25-28. Maki, R., Maki W., and Marsh, L., 1977, “Processing Locational and Orientational Information.” Memory and Cognition, Vol 5, No. 5, pp. 602-612. Rock, 1., 1974, “The Perception of Disoriented Figures.” Scientific American, Vol. 230, NO. 19 pp- 78'85. Schone, H., 1984, Spatial Orientation, Princeton: Princeton University Press. Shepard, R. and Cooper, L., 1982, Mental Images and Their Tranjonnations, Cambridge: MIT Press. Shimron, J ., 1978, “Learning Positional Information from Maps.” The American Cartographer, Vol. 5, No. 1, pp. 9—19. Stevens, A. and Coupe, P., 1978, “Distortions in Judged Spatial Relations.” Cognitive Psychology, Vol 10, No.4, pp. 422-437. Trowbridge, C. C., 1913, “On Fundamental Methods of Orientation and ‘Imaginary Maps’.” Science, Vol. 38, No. 990, pp. 888-897. "7|llllllllllllllll