Map Design for Complex Architecture: A User Study of Maps & Wayfinding

Abstract

The following study seeks to determine if a printed, paper map can aid visitors in navigating through complex architectural environments. Specifically, we report on the design and testing of two different paper maps intended to help patients find dental clinics and related offices within a large medical and health sciences center. As part of an iterative design process, we first identified a variety of design factors that influence the cognitive aspects of using maps during wayfinding, and redesigned an existing map of the environment based on those principles. We then conducted user testing to further determine what information should be included or excluded on the map and to see if changes in format enhanced or detracted from communication goals. The results show that maps can indeed assist visitors in finding their way through complex buildings, but that there are limitations to their ability to overcome architectural barriers.

Full text

The following study seeks to determine if a printed, paper map can aid visi-
tors in navigating through complex architectural environments. Specically,
we report on the design and testing of two dierent paper maps intended to
help patients nd dental clinics and related oces within a large medical and
health sciences center. As part of an iterative design process, we rst identi-
ed a variety of design factors that inuence the cognitive aspects of using
maps during waynding, and redesigned an existing map of the environ-
ment based on those principles. We then conducted user testing to further
determine what information should be included or excluded on the map
and to see if changes in format enhanced or detracted from communication
goals. The results show that maps can indeed assist visitors in nding their
way through complex buildings, but that there are limitations to their ability
to overcome architectural barriers.
Map Design for Complex Architecture:
A User Study of Maps & Waynding
Karen Cheng & Sarah Pérez-Kriz
7
Map Design
Cheng & Pérez-Kriz

Healthcare facilities are often housed in complex, confusing, and dicult to
navigate environments. Their spatial problems typically develop over time
when existing units are renovated or relocated, or when new wings are built
to contain advanced technology and new areas of practice. Unfortunately,
given limitations of both time and budget, most medical centers build
where they can, with little consideration of the most logical routes for public
circulation. As a result, many healthcare facilities are decentralized and disor-
ganized—adding waynding diculties to the burden of patients who are
already anxious, pressed for time, and physically impaired.
Ideally the structure of a building itself should facilitate the movement of
its occupants. Numerous studies have examined the architectural factors
that can inuence spatial legibility. Research has shown that waynding is
facilitated when: locations are visually dierentiated (an area or zone looks
dierent from others); travelers can see views inside or outside the structure,
including interior, exterior and global landmarks; the oor plan and spatial
conguration are simple and regular; and there are fewer decision points
and corridor intersections (Best, 1970; McKean, 1972; Passini, 1980; Weisman,
1981; Bronzaft and Dobrow, 1984; Peponis, Zimring and Choi, 1990).
In addition to the architectural factors inherent in a built structure, direc-
tional signs and room numbers have been proven to help people nd their
way (Weisman, 1981). Researchers have determined that people made fewer
wrong turns in settings with signage than in those without (O’Neill, 1991),
and the presence of signs signicantly reduced perceived crowding, discom-
fort, anger, and confusion amongst visitors to a correctional center (Wener
and Kamino, 1983).
However, signage alone has not been able to overcome the navigational
problems caused by an illegible building (Arthur and Passini, 1992). Similarly,
Carpman, Grant and Simmons (1984) found that in a specic hospital corri-
dor, waynding performance decreased as the number of signs increased.
Perhaps when signs are not enough, maps might be able to provide the nec-
essary assistance to travelers. In a 1993 study by Wright, Hull and Lickorish,
a paper sketch map was given to participants to supplement an inadequate
signage system; this map was found to signicantly reduce backtracking in
a hospital setting. Additionally, maps have proven to be eective in other
ways. Subjects who used a schematic map were able to nd the most e-
cient routes to a destination (Bartram 1980), and handout maps in museums
were able to successfully orientate visitors (Hayward and Brydon-Miller,
1984). Furthermore, people view maps as eective and desirable sources of
information when traveling to new destinations (Devlin and Bernstein, 1995).
introduction
8
Visible Language
48.2
,

There is signicant evidence that maps are a unique mental aid to navigation.
Numerous environmental and spatial cognition studies have examined how
maps work in the mind, and it is generally accepted that maps can help peo-
ple to create or develop a spatial mental model known as a “cognitive map.”
A cognitive map (as originally dened by Tolman, 1948) is an internal, mental
representation of an environment. The term should not be taken as a literal
metaphor; cognitive maps often contain errors and are more likely to be
coarse, not-to-scale representations of general spatial relations rather than
precise, map-like constructs that can be mentally examined at will (Tversky,
1993). Nonetheless, there is broad agreement on the three types of spatial
knowledge that are the basis of a cognitive map: landmark knowledge,
route knowledge, and survey knowledge (Siegel and White, 1975).
Landmark knowledge allows people to determine where they are by using
natural or man-made features that are easy to see and recognize (for exam-
ple, a building, monument, lake, etc.) Route knowledge is the next level of
understanding when landmarks can be mentally connected into a sequence
(for example, “turn left at the church and go past the cemetery until you see
the lake.”) Survey knowledge is the highest form of spatial cognition when a
combined understanding of how landmarks and routes are connected en-
ables individuals to infer new paths they have not previously traveled (such
as shortcuts and detours). Survey-level knowledge is often described as
the ability to envision an environment from a bird’s-eye view—the vantage
point of most cartographic maps.
Given the unique characteristics of maps, the following study seeks to deter-
mine if a printed, paper map can aid visitors in navigating through a large,
complex medical center. From a nancial point of view, the design, produc-
tion, and distribution of a paper-based map represents a signicant cost
savings when compared to the revision of physical signs or alterations in the
built environment.
A paper map can easily be mailed to patients prior to their appointment or
downloaded online and printed at home. Travelers can use the map to plan
their trip in a relatively calm and unhurried manner, perhaps with the assis-
tance of family or friends who will accompany them. Paper maps are portable,
and can be consulted en-route, reducing the need for users to memorize a
lengthy sequence of actions. A paper map can even be highlighted and anno-
tated by the patient or healthcare sta to clarify or provide specic directions.
However, although paper maps oer the advantages described above, it
may be overly optimistic to assume that a single building map, no matter
how well designed, can eliminate the navigation errors that stem from ar-
chitectural barriers. Maps are a single component of an integrated program
9
Map Design
Cheng & Pérez-Kriz

of waynding. An eective waynding system should take advantage of
multiple elements that reinforce one another—including building maps,
oor plans, signs, information desks, and in-person volunteer guides/es-
corts. Additionally, maps are not equally helpful to all travelers, because map
reading is an acquired skill that varies in the population (as summarized by
Carlson, Hölscher, Shipley, and Dalton, 2010), and may be declining due to
the increased use of GPS devices (Klippel, Hirtle, and Davies, 2010).
To investigate the potential benets and limitations of handout maps
in assisting visitor navigation in a complex building, we examined two
dierent paper maps intended to help patients nd dental clinics and
related oces within a large medical and health sciences complex at the
University of Washington.
The dental clinics are located within the single largest university building
in the United States, the Magnuson Health Sciences Building—a complex
with a total oor area of approximately 5,740,000 square feet. The building
houses the University of Washington Medical Center, a teaching hospital,
and ve academic units—the School of Dentistry, School of Medicine,
School of Nursing, School of Pharmacy, and School of Public Health.
Due to its large size and complexity, the building is dicult for visitors to
navigate. There are more than 20 wings that were built during a period of
over 50 years. When additions were made, administrators opted to fully
connect all interior hallways, resulting in an irregular, asymmetric congura-
tion—a spatial form known to present waynding diculties.
The building complex is generally divided into two zones—the hospital (the
University of Washington Medical Center) and the academic area. Both of
these zones are further subdivided into a series of wings. The hospital names
their wings with double letters (such as AA, BB, CC, etc.), while the academic
units use single letters (A, B, C, etc.). Letter sequences do not necessarily
follow alphabetical order—for example, the T-Wing stands for the “Teaching
Wing” and the RR-Wing indicates the “Research Wing.” Additionally, some
wings are named after compass directions—for example, in the hospital, the
NE-Wing is the name of the “North East Wing.”
The building does not have a unied signage program. There are three sets
of signs that dental patients must sift through in order to nd their way. The
rst signs seen by most dental patients are for the hospital because the ma-
jority of visitors enter the building through the main hospital entrance. The
hospital signs are, in general, well designed; they are typographically legible,
visually consistent, logically placed, and properly maintained. However, the
the research setting
10
Visible Language
48.2

hospital signs do not include information about the dental clinics because
the dental clinics are not ocially part of the hospital but instead are man-
aged by the School of Dentistry, an academic unit.
Therefore, in order to guide dental patients and other academic visitors
through the hospital, an additional set of signs with a competing visual
identity has been designed and installed. These signs tend to be located
in somewhat out-of-the-way positions because more prominent areas are
taken up with ocial hospital signage.
As patients pass through the hospital and move toward the dental clinics,
the secondary set of signs disappears, and a third signage system emerges.
This last set of signs is part of an older system that is used by all academic
units in the building, including the dental clinics and the School of Dentistry.
These older signs are less well maintained and their positioning is, at times,
questionable. In places the text within the signage system is inconsistent
(for example, a covered exterior path is labeled as an “outdoor walkway” on
the second oor, but as a “bridge” on the third oor).
The administrators of the School of Dentistry have consistently received
complaints from patients who have diculties navigating to the dental
clinics. Therefore, in 2011 the School of Dentistry commissioned our design
team to develop a new building map. This new map would replace an
existing map (gure 1, next page) that they felt was too cluttered and too
complex. Additionally, sta members believed that the current map did not
adequately help visitors to understand: 1) that the building is organized
in a series of wings; 2) that the dental clinics are located in the B-Wing and
D-Wing; and 3) that the building is built on a sloped site, with the entry oor
on Level 3.
Research studies from several elds (including environmental and spatial
psychology, cognitive psychology, human geography, architecture and
urban planning) provide some guidance for how to best construct a map
that facilitates waynding. The following considerations governed our initial
choices for both the form and content for new dental clinic map addressing
the concerns identied above (gure 2, next page).
11
Map Design
Cheng & Pérez-Kriz

ORIGINAL MAP (double-sided 8.5 x 11”)
B Wing D Wing
D WingB Wing
School of Dentistry
B Wing D Wing
Section Left of Line
Reduced In Size
CLINIC
D2
CLINIC
D3
Hospital
UW
UNDERGROUND
Parking
Husky
Stadium
P A T I E N T G U I D E
University of Washington
Walkway
Endodontics
Periodontics
Faculty Practice
& Research Center
Grad Pros
471
BDevelopment
& CDE
4th Floor
E wingC wing
T wing
452
B
470
B
469
B
403
B
D458
D453
A wing
Walkway
251
258
232
BD
D
Patient
Accounts
Urgent Care
DECOD
Court
Cafe
221B
and
2nd Floor
Hospital
Dentistry
Oral Medicine
A wing E wing
C wing
D 209
229
B
Oral Medicine,B221
Oral Radiology (x-ray),B307
Oral Surgery, B350
Orthodontics, B338
Patient Accounts, B232
Patient Services, B452
Pediatric Dentistry,B323
Research Center, D453
Student Services, D323
Continuing Dental
Education (CDE), B471
D2 Clinic, D251
D3 Clinic, D351
Dean’s Office, D322
DECOD Clinic, B229
& Patient Intake
Urgent Care Clinic, B229
Endodontics (root canals), B470
Faculty Practice, D453
Fears Clinic, D458
Grad Periodontics, B403
Elevators Stairs
Note: The letter at the beginning of a room number
indicates the wing. The first number indicates the
floor. Example, D358 is in the D wing on the 3rd floor.
Walkway
351
338
Ortho
350
Oral
Surgery
Pediatric
Dentistry
307
New Patient
Registration & X-ray
323
B
B
C wing
D
E wing
322
B
B
D
323
D
3rd Floor
Hogness
Up To
A
Lobby AABBBB
Triangle
Tunnel
NE Pacific St
NE Pacific St
Cafeteria
to 1st floor
take elevator
Pacific
Elevators
Cascade
Elevators
BUS
BUS
T wing
Lobby
Hogness
DS Patient Map 7-08 twade
Entrance
Garage
Pacific Place NE
Please note: Our buildings are built
upon a slope. The front entrances are at
the 3rd floor level. The rear entrances are
at the 1st floor level.
It takes about 10 minutes to walk from
the Triangle Parking Garage to the Dental
Clinics. Please allow that extra time in
getting to your appointment.
Montlake Blvd NE
(Restorative, Prosthodontics,
Endodontics, Periodontics)
Development & Hospital Dentistry, D258
New Patients &
B307
Records Request,
B469Grad Prosthodontics,
Patient
Services
Fears Clinic
8794002
2
C
Reception
Student
Services Dean’s
Office
Reception
U W
Campus
I-5 or SR-520 to Triangle Parking Garage
Disabled Parking in Back - Area S1
Husky
Triangle
Parking
Garage
SR-520
NE 45th ST
NE Pacific
NE Pacific
Montlake
Blvd NE
15th
Ave NE
Boat St
Lake
Washington
Montlake
Blvd NE
From I-5
1)Take I-5 to the NE 45th St exit.
2) Go east on NE 45th St, about
eight blocks.
3)Turn right on 15th Ave. NE.
4) Go south about 6 blocks.
5)Turn left at light onto Pacific.
6)Turn left at the first left turn
off Pacific, there is a light.
7)Turn right into the Triangle
Garage.
From SR-520
1)Take the Montlake exit.
2) Go north on Montlake over
the bridge.
3)At the first light after the bridge
turn left onto Pacific.
4)Turn right at the next corner
and around into the Triangle
Garage.
Disabled/Visitor Parking
1)At Pacific and 15th Ave NE
continue south on 15th until it
merges with Boat St.
2) Continue left on Boat St and
through the gate and then into the
S1 parking area.
Portage
Bay
Gate
Disabled
Parking
Gate
Disabled Parking
NE Pacific
15thAve
NE
Boat St
Montlake
FD
H
T
RR
AA BB Hospital
South Campus
Center
CEG
K
I Court
3rd Floor is Ground Level on Pacific - 1st Floor is Ground Level in Back
S1
S1 to B Wing
You may enter the D wing, and
take the elevator either to the 2nd, 3rd,
or 4th floor, then cross the walkway
to the B wing at each level.
BUS
S1 to -D1 Below Ground
1) On the middle level of the
parking garage take the tunnel
going into F wing. There are signs
to the Dental School.
2) Follow the signs to the Dental
School and the D wing elevators.
You are now on level D minus 1, not D1!
S1 to D1 Above Ground
1) Enter D wing on the right side
by the covered bus stop.
2) Go straight through two sets
of doors, then turn left.
3)At the next hallway turn right
to the elevators.
J
University
District
Montlake
Capitol
Hill
Triangle Garage
B
Stadium
INTERSTATE
5
S1
ORIGINAL MAP (double-sided 8.5 x 11”)
B Wing D Wing
D WingB Wing
School of Dentistry
B Wing D Wing
Section Left of Line
Reduced In Size
CLINIC
D2
CLINIC
D3
Hospital
UW
UNDERGROUND
Parking
Husky
Stadium
P A T I E N T G U I D E
University of Washington
Walkway
Endodontics
Periodontics
Faculty Practice
& Research Center
Grad Pros
471
BDevelopment
& CDE
4th Floor
E wingC wing
T wing
452
B
470
B
469
B
403
B
D458
D453
A wing
Walkway
251
258
232
BD
D
Patient
Accounts
Urgent Care
DECOD
Court
Cafe
221B
and
2nd Floor
Hospital
Dentistry
Oral Medicine
A wing E wing
C wing
D 209
229
B
Oral Medicine,B221
Oral Radiology (x-ray),B307
Oral Surgery, B350
Orthodontics, B338
Patient Accounts, B232
Patient Services, B452
Pediatric Dentistry,B323
Research Center, D453
Student Services, D323
Continuing Dental
Education (CDE), B471
D2 Clinic, D251
D3 Clinic, D351
Dean’s Office, D322
DECOD Clinic, B229
& Patient Intake
Urgent Care Clinic, B229
Endodontics (root canals), B470
Faculty Practice, D453
Fears Clinic, D458
Grad Periodontics, B403
Elevators Stairs
Note: The letter at the beginning of a room number
indicates the wing. The first number indicates the
floor. Example, D358 is in the D wing on the 3rd floor.
Walkway
351
338
Ortho
350
Oral
Surgery
Pediatric
Dentistry
307
New Patient
Registration & X-ray
323
B
B
C wing
D
E wing
322
B
B
D
323
D
3rd Floor
Hogness
Up To
A
Lobby AABBBB
Triangle
Tunnel
NE Pacific St
NE Pacific St
Cafeteria
to 1st floor
take elevator
Pacific
Elevators
Cascade
Elevators
BUS
BUS
T wing
Lobby
Hogness
DS Patient Map 7-08 twade
Entrance
Garage
Pacific Place NE
Please note: Our buildings are built
upon a slope. The front entrances are at
the 3rd floor level. The rear entrances are
at the 1st floor level.
It takes about 10 minutes to walk from
the Triangle Parking Garage to the Dental
Clinics. Please allow that extra time in
getting to your appointment.
Montlake Blvd NE
(Restorative, Prosthodontics,
Endodontics, Periodontics)
Development & Hospital Dentistry, D258
New Patients &
B307
Records Request,
B469Grad Prosthodontics,
Patient
Services
Fears Clinic
8794002
2
C
Reception
Student
Services Dean’s
Office
Reception
U W
Campus
I-5 or SR-520 to Triangle Parking Garage
Disabled Parking in Back - Area S1
Husky
Triangle
Parking
Garage
SR-520
NE 45th ST
NE Pacific
NE Pacific
Montlake
Blvd NE
15th
Ave NE
Boat St
Lake
Washington
Montlake
Blvd NE
From I-5
1)Take I-5 to the NE 45th St exit.
2) Go east on NE 45th St, about
eight blocks.
3)Turn right on 15th Ave. NE.
4) Go south about 6 blocks.
5)Turn left at light onto Pacific.
6)Turn left at the first left turn
off Pacific, there is a light.
7)Turn right into the Triangle
Garage.
From SR-520
1)Take the Montlake exit.
2) Go north on Montlake over
the bridge.
3)At the first light after the bridge
turn left onto Pacific.
4)Turn right at the next corner
and around into the Triangle
Garage.
Disabled/Visitor Parking
1)At Pacific and 15th Ave NE
continue south on 15th until it
merges with Boat St.
2) Continue left on Boat St and
through the gate and then into the
S1 parking area.
Portage
Bay
Gate
Disabled
Parking
Gate
Disabled Parking
NE Pacific
15thAve
NE
Boat St
Montlake
FD
H
T
RR
AA BB Hospital
South Campus
Center
CEG
K
I Court
3rd Floor is Ground Level on Pacific - 1st Floor is Ground Level in Back
S1
S1 to B Wing
You may enter the D wing, and
take the elevator either to the 2nd, 3rd,
or 4th floor, then cross the walkway
to the B wing at each level.
BUS
S1 to -D1 Below Ground
1) On the middle level of the
parking garage take the tunnel
going into F wing. There are signs
to the Dental School.
2) Follow the signs to the Dental
School and the D wing elevators.
You are now on level D minus 1, not D1!
S1 to D1 Above Ground
1) Enter D wing on the right side
by the covered bus stop.
2) Go straight through two sets
of doors, then turn left.
3)At the next hallway turn right
to the elevators.
J
University
District
Montlake
Capitol
Hill
Triangle Garage
B
Stadium
INTERSTATE
5
S1
Figure 1: Original Map (double-sided 8.5" x 11")
Figure 2: First Revision Map (single-sided 11" x 17")
12
Visible Language
48.2

Figure 1 (left top)
The original map for patients
of the University of Washington
Dental Clinics. The map
printed in two colors on both
sides. Administrators felt that
the original map was too clut-
tered and too complex, and that
the map did not adequately
help visitors to understand: 1)
that the building is organized
in a series of wings; 2) that the
dental clinics are located in
the B-Wing and D-Wing; and 3)
that the building is built on a
sloped site, with the entry oor
on Level 3.
Figure 2 (left bottom)
Initial redesign of the dental
and is oset printed in two col-
ors on one-side. We increased
white space on the map by re-
placing text with symbols, and
by reducing overall information
density. We also attempted to
clarify the wing system by in-
troducing a simplied building
diagram with color-coding (of-
ces in the D-Wing are shown in
blue, while those in the B-Wing
are in black).
1___Reducing Visual Clutter
Typographers and legibility researchers generally agree that “white space”
aids in the readability of documents. Surrounding certain elements with
negative space helps to establish a visual hierarchy of information where the
reader’s eye is directed to specic information in a logical sequence. White
space can be also be used to lessen the density of large blocks of text, which
makes long passages less intimidating to readers and easier to scan.
In order to recover white space from the original map, we replaced text with
symbols wherever possible because icons generally take up less space than
their corresponding words.
We selected the symbols designed by the AIGA (American Institute of
Graphic Arts) for the US Department of Transportation in 1974–79, a system
widely used both nationally and internationally and likely to be familiar to
dental patients.
2___Reducing Information Density
To simplify a map, designers must consider how much information to in-
clude. The concepts of “route knowledge” and “survey knowledge” described
earlier can be used to dene the poles of an information continuum. A
minimalist approach involves designing a “route map,” which shows only the
relevant landmarks and the path segments that connect them. A maximalist
approach favors the design of a “survey map,” which provides a broad over-
view of an environment by depicting all information in a selected area.
The main advantage of a route map is the direct focus on the path to be
taken. Each decision point on the route can be shown clearly, with turn-by-
turn instructions. Because route maps are simplied in form and content,
providing only the smallest possible set of information that is needed for
people to nd their way, they require relatively low cognitive eort—they
are easy to understand and use.
However, route maps have a signicant disadvantage. If a person using a
route map deviates from the path by mistake, there is no information to help
them re-orientate and get back on the route because this style of map does
not show the surrounding environment. Therefore, once the travelers are o
the route, they are lost.
The alternative to a route map is a survey map, which does show an entire
environment in detail. The broader scope of a survey map makes it more
ecient than route map because the same survey map can show multiple
routes (a single survey map can direct many individuals who come from
dierent origins to dierent endpoints).
13
Map Design
Cheng & Pérez-Kriz
INITIAL DESIGN DECISIONS
clinic map. The map is 11"x17,"
measures 8.5"x11," and is oset

In principle, survey maps should enable a better understanding of a place
than route maps; a viewer builds survey knowledge by seeing routes in the
context of their larger surroundings. However, researchers believe there
are limits to how much graphic information can be absorbed from a map.
Once a critical point of visual density has been reached, comprehension may
decline (Dobson, 1980). This is the main disadvantage of survey maps—ex-
tracting, understanding and keeping track of a route within a dense and
visually cluttered map can be cognitively demanding (Phillips, 1979; Rosen-
holtz, Li, and Nakano, 2007).
In the design of the new dental clinic map, our goal was to compromise and
create a hybrid “schematic map” that would blend the two map archetypes
described above. Schematic maps are simplied versions of survey maps
that may distort certain spatial characteristics in order to emphasize and
clarify specic aspects of a route and its environment. In this case, we hoped
to focus attention on a central route that begins at the patient parking
garage and continues through the hospital toward the dental registration
oce (typically the rst destination for new patients). Several additional
destinations would need to be shown on the map (such as specic clinics,
patient services, etc.), but these would be subordinate in importance.
3___Communicating the Wing System
To facilitate the understanding of building’s wing system and to clarify the
location of the dental clinics, we created a simplied building diagram that
highlighted the position of the B and D-wings. Because the map is oset
printed in two colors, we were able to color-code these spaces (oces in the
B-wing were shown in black, while those in the D-wing were shown in blue).
We wanted viewers to progress from this simplied diagram to a more
detailed oor plan that showed the locations of all dental facilities on three
oors. By rst showing the simplied diagram, we hoped for a controlled,
progressive disclosure of information that would avoid overwhelming the
navigator. This strategy follows the broader psychological phenomenon of
“chunking,” where limits on the human capacity for processing information
are extended by grouping similar items together in a “chunk” (Miller, 1956).
The technique also makes sense given that people have been shown to
mentally divide maps for “conquering”—by partitioning a map into sections,
learners create manageable “chunks” that can be further examined individu-
ally for more information (Thorndyke and Stasz, 1980).
To encourage users to view the simple and detailed diagrams in a linear
sequence, we placed them side-by-side. This composition required revising
the original, double-sided 8.5x11” format to a single-sided tabloid sheet
(11x17”). We were concerned that patients would nd this tabloid map
awkward to hold; we tried to alleviate this issue by folding the map in half
14
Visible Language
48.2

to letter size, and positioning elements around the fold. We anticipated
that this larger format could be printed at home on letter size paper by
using “shrink to t.”
To further clarify the wing system, we also created a small diagram that
explained the room numbering scheme used in the building. We hoped that
this diagram would be easier to scan and understand than the explanatory
text caption currently used on the original map.
4___Communicating Floor Levels
Because the building is constructed on a sloped site, visitors entering from
the front-facing street arrive on Level 3. At the back of building, street level
entrances are on Level 2.
On the original map, the text label identifying the 3rd Floor is relatively small
and easily overlooked. On the new map, we enlarged all oor labels and
emphasized them via enclosure in a ruled box.
In any design project, there are questions that can only be answered
through empirical evaluation. In this case, we hoped to resolve the following
issues in map content by observing individuals using a map to perform
waynding tasks in the actual environment.
1___Entry Points
It would be helpful to know which entrances to include and/or emphasize
on the new map. There are three main entrances that could be used by
visitors: a hospital entrance, an academic entrance, and an intermediate
wing entrance. Sta members suspect that the hospital entrance is the most
popular, and that the intermediate wing entrance is not used by visitors.
However, the administration had no data to conrm this assumption.
2___Landmarks
In a ve-year study of three American cities, urban planner Kevin Lynch
found that people understand their environment through ve elements:
paths, edges, nodes, districts, and landmarks (Lynch, 1960). All of these
elements are typically visualized on maps, but the last element, landmarks, is
particularly important when navigating with maps.
A number of studies (as summarized by Davies and Peebles, 2007) suggest
that people prefer to use a landmark strategy when using a map. Speci-
cally, to orient themselves, travelers look for conspicuous landmarks in the
environment that can be matched to visual representations on the map. In
fact, this procedure is recommended by several instructional guides on “how
design decisions requiring user testing
15
Map Design
Cheng & Pérez-Kriz

to use a map” such as the Boy Scout Handbook (2009). Therefore, to design
a successful map, one must include the landmarks that people are likely
to notice.
Certain features of our particular setting provide obvious and eective land-
marks, but unfortunately, this building (and many other health care settings)
consists mainly of long, featureless corridors. In such an environment, em-
pirical evaluation may help to determine what objects or structures travelers
use as landmarks.
3___O-Route Information
It is not clear how much information should be included on the new map to
help travelers who make wrong turns at decision points. Extensive context
might be necessary to help them re-orientate and recover from mistakes—
but adding excess detail may complicate and clutter the map.
Researchers, cartographers and design practitioners (Arnheim, 1976;
Southworth and Southworth, 1982; Tufte , 1983; Arthur and Passini, 1992)
generally suggest that maps should be as simple as possible to avoid cog-
nitive overload. However, this advice fails to address the specics of content
inclusion in this particular context.
To better understand how paper maps are used during waynding in a com-
plex building and to answer specic questions regarding content inclusion
on a patient map, our team recruited 24 paid adult volunteers who were
unfamiliar with the building and its surrounding area. Participants were met
by a researcher at a campus café and escorted to the pedestrian tunnel
in the main parking garage of the medical center. Here participants were
asked to wear a head-mounted camera that would record video from their
lines of sight.
Using a head-mounted camera to capture user behavior was a novel aspect
of this study. The camera tracked participants’ gazes and visual focuses
as they navigated through the environment, allowing the research team
to analyze how and when participants consulted the map and environ-
mental signage during their waynding experiences. Additionally, the
head-mounted camera allowed the participants to navigate through the
environment independently, without having to be accompanied or followed
by an observer.
After helping the participants put on the camera, the researcher gave the
participants a dental clinic map. Twelve participants received the original
map, and twelve participants received the revised map.
TEST PROCEDURE
16
Visible Language
48.2

Each participant was then asked to independently proceed from the parking
tunnel to the Dental Registration Oce (B-307). At the entrance to the
Dental Registration Oce, the participant was greeted by a second member
of the research team. This second researcher then asked the participant to
walk to the D-2 Dental Clinic. The researcher provided verbal directions that
were considered typical from a script that was written by the Director of the
Dental Patient Registration Oce .
After participants reached the D-2 Dental Clinic, the researcher escorted
them to a nearby waiting area, and asked them to complete a questionnaire
that: 1) rated their familiarity with the building and area prior to the exper-
iment; 2) rated their overall sense of being lost during the experiment; 3)
estimated their overall sense of direction (based on the Santa Barbara Sense
of Direction Scale, as developed by Hegarty, Richardson, Motello, Lovelace,
and Subbiah, 2002), 4) rated the map that they received; and 5) provided
demographic information.
After completing the questionnaire, participants were asked to look at their
map again and draw the two routes that they had traveled on the map (from
the parking garage to the Dental Registration Oce and from the Dental
Registration Oce to the D-2 Dental Clinic). Participants were also asked if
there was anything that they would change about the map and if they had
any suggestions for improving the map.
The recorded video from each participant was analyzed and coded to
determine the frequency and length of time that each participant looked at
either the map or a sign. Gazes had to last at least two seconds in order to be
coded. Timestamps on the video were used to calculate the duration of each
gaze. The total travel time for each route was also calculated from the video
timestamps, as well as the duration of errors and stops.
Twelve participants received the original map, and twelve participants re-
ceived the revised map. The age range of those who received the original map
was 21–54, with a mean age of 28; this group contained four men and eight
women. The age range of those who received the revised map was 18–59,
with a mean age of 30; this group contained two men and ten women.
On a scale of 1-7 (with higher scores indicating a better sense of direction),
those who received the original map had a mean score of 5.0; those who
received the revised map had a mean score of 4.2. Neither group was familiar
with the building or area prior to the experiment. The group that received
the original map had a mean score of 1.0 (on a scale of 1-7, with 1 being “Not
Familiar” and 7 being “Very Familiar”). The group with the revised map had a
participant demographics
17
Map Design
Cheng & Pérez-Kriz

mean score of 1.5 for familiarity. Statistical testing revealed no signicant dif-
ferences between the two groups in terms of age, gender, sense of direction,
or building familiarity.
All participants were able to reach the two target destinations (the Dental
Registration Oce and the D-2 Clinic). However, 22 of the 24 participants
experienced navigation diculties during the course of the experiment,
stemming from omissions of content in the map, architectural barriers, illog-
ical site organization, and inadequate signage, as described below.
No signicant dierences were found between the two groups of map
users. Specically, there was no statistical dierence between participants
who used the original or revised map in terms of their overall time to the
destinations, the number or length of waynding errors, the number or
length of pauses en-route, or overall satisfaction ratings of the map. Given
these results, our discussion will focus on the general waynding behaviors
observed in the study and the issues of content inclusion, with map type
noted where relevant.
1___Building Entrances
As described earlier, there are three main entrances to the building complex:
an entrance to the hospital, an entrance to the academic area, and an
intermediate entrance between the other two, which leads to the BB-Wing
(gures 3A and 3B).
The hospital entrance is the most popular; 16 out of 24 participants (67%)
used this entrance. The remaining eight participants were equally divided
between the academic entrance and the intermediate BB-Wing entrance
(each of these entrances were used by four participants, or 17% of the test
group). These eight participants were all individuals who had chosen to exit
the parking garage and continue on foot via the surface street. They later
explained that they wished to avoid navigating inside the building complex;
they thought it would be easier to avoid getting lost if they followed a street
that runs parallel to the medical complex.
Unfortunately, on the new map we had removed the label to the intermedi-
ate BB-wing entrance because sta members believed that it was rarely used
by visitors. As shown by the behavioral data, this assumption was not true.
Although stakeholder input and opinions can serve as a valuable source of
information, assumptions can sometimes be false. The waynding results
underscore the importance of using a combination of user opinions and
behavioral studies to determine how a space is actually used.
18
Visible Language
48.2
RESULTS

All of the walking participants who had the original map used the BB-Wing
entrance—possibly because this entrance is located directly opposite the
main crosswalk near the parking garage. Participants did not continue
towards the third academic entrance—perhaps because this entrance is
set back from the street, and the building sign is not clearly visible from
the crosswalk.
However, three out of four of the walking participants with new map did
use the academic entrance. The remaining participant used the BB-Wing
entrance. However, because the BB-Wing entrance was not shown on her
map, she was confused and deliberated for ~1 minute before entry.
Figure 3a
The intermediate BB-Wing
entrance.
Figure 3B
The rst crosswalk by the main
parking garage leads into the
entrance to the BB-Wing. The
academic entrance is not visible
from this crosswalk.
19
Map Design
Cheng & Pérez-Kriz

Similarly, of the three participants with the new map who used the academ-
ic entrance, there was one individual who rst navigated to the BB-Wing en-
trance. This participant traversed the area for ~4 minutes due to uncertainty
regarding her location. She eventually backtracked to the original crosswalk
area and traveled to the academic entrance.
Clearly, in future iterations of the map, we need to include the BB-Wing
entrance. This entrance also seems likely to be popular with patients who
arrive by bus, or use the valet parking service since the entrance is closest to
the bus stop and valet station.
2___Environmental Assumptions
Most people assume that when entering a building, they are arriving on the
rst oor, or Level 1. However, as described earlier, because this complex is
built on a sloped site, participants entering from the street level arrive on
Level 3 (at front of the building) or Level 2 (at the back of building). Level 1
is the lowest (basement) level on the hospital side of the complex. On the
academic side, there are two additional lower levels, and these oors are
designated as Levels “-1” and “-2.”
Shortly after entering the building, four participants tried to take an elevator
to Level 3. Two additional participants tried to use stairwells. Therefore, 25%
of all participants (six out of 24) were not aware of the oor numbering
system, despite oor number labeling on the maps. The six test participants
who made these errors were evenly divided between those who had the
original map, and those who had the revised map.
Three of the participants did not seem particularly stressed by their oor
level error, perhaps because the time lost was relatively minimal (26 sec-
onds spent climbing stairs and 7-35 seconds spent waiting for an elevator,
entering the elevator, pushing a oor button, then exiting the elevator
after realizing the mistake). The fourth participant seemed more irritated,
possibly because she had spent longer (~2 minutes) repeatedly traversing
stairs between Levels 1–3. Of course, such incidents may be more upsetting
to actual patients, who are often pressed for time, anxious about upcoming
procedures, and/or experiencing physical discomfort, as opposed to paid
volunteer study participants.
For similar buildings constructed on a grade, Arthur and Passini have
recommended the terms “Level 1” for the ground level, and “1 Below” and “2
Below” for basement levels (Arthur and Passini, 1992). A dierent solution
was veried by a research study conducted at the University of Michigan
General Hospital (Carpman, Reizenstein, Grant, and Simmons, 1983), where
ve oor-numbering options were tested with patients and visitors. In this
study, the terms “Sub 1” and “Sub 2” were interpreted correctly most often.
20
Visible Language
48.2

To improve waynding, these or other naming options could be evaluated
for this building. At minimum, perhaps the term “Minus” should be spelled
out, because the negative sign is relatively small and easily missed.
It is possible that oor level information can be made more prominent on fu-
ture design iterations of the map. However, this information may continue to
be overlooked because people have strong pre-existing assumptions about
oor levels based on their previous experiences in other buildings.
3___Architectural Barriers
There are several areas in the building where the architecture is in conict with
the desired waynding behavior. For example, when entering from the main
hospital entrance, visitors need to walk through a small enclosure in order to
continue towards the dental clinics. This enclosure contains a set of elevators
(the Pacic Elevators). However, from the main route, the enclosure appears to
be a dead end—the opening to the subsequent hallway is not visible from the
outside (gures 4A and 4B).
Six participants (25% of all participants passing through this area) stopped
or paused at this decision point and consulted their map. Two participants
had the original map, and four participants had the revised map. Of the six
participants who hesitated, ve participants eventually proceeded through
the enclosure (after 3-27 seconds of deliberation).
The remaining participant also paused at the decision point but made a
wrong turn that continued into the maternity area of the hospital. In the
maternity ward, the participant searched her map (neither map shows the ma-
ternity department) and decided to backtrack to the original position outside
of the elevator enclosure. Back at the decision point, she again deliberated
before correctly passing through. The participant spent approximately three
minutes on this error loop.
Figure 4A
To reach the dental clinics,
patients must walk through an
enclosed area that contains a
set of elevators. From the main
hallway, the enclosure appears
to be a dead end.
21
Map Design
Cheng & Pérez-Kriz

visitors often perceive these walkway doors to be locked or alarmed exits.
This seems to be a common misapprehension, as several of these metal
walkway doors have ad-hoc, occupant-created signs that state: “This door
is NOT alarmed” and “This door does NOT lock behind you.” Of the seven
participants with walkway door issues, three had the original map, and four
had the new map.
The problems described above occur because the architecture discourages
users from the correct waynding behavior. It is also possible for the built
environment to encourage an incorrect navigational response. For example,
the main academic entrance opens into a lobby and waiting area with a high-
ly visible staircase (gure 6) that leads to a single location—an auditorium
used for special events. One test participant attempted to use this staircase
to reach other areas of the building, despite passing a posted sign that states
“To auditorium only.”
In future iterations of the map, we need to include the location of the ma-
ternity ward, as it is a destination that is commonly reached by users. It also
seems that neither map adequately shows the way through these architec-
tural barriers (the elevator enclosure, the walkway doors and the auditorium
stairs). We may be able to better explain the architectural elements by an-
notating the map with adjacent text. However, each annotation adds clutter
and visual noise to what is already a complex layout.
Figure 4B
Once inside the enclosed area,
travelers can see the subsequent
hallway that was not visible
from the main hallway.
In a similar fashion, outdoor walkways (gure 5) were also problematic for
seven out of ten participants. The short hallway that leads to the walkway
is small and easily overlooked, as evidenced by the six participants who
missed the turning point and had to backtrack. In addition to those who
missed the turn, there were two participants who reached the walkway
door but did not go through. One participant turned away from the door
immediately; the other participant looked through a small window set in
the door before turning away. Several sta members later explained that

4___Preference for Straight Routes
Participants who used the intermediate entrance—the entrance that leads
to the BB-Wing—made a fairly consistent error. Instead of turning right
towards the A-Wing and the rst waynding target, the Dental Registration
Oce, three of the four participants (75%) continued traveling straight
down into the RR-Wing (gure 7, next page).
In the RR-Wing, these three participants searched their map (All three
participants had the original map, which did not show the RR-Wing). There
they deliberated for some time before backtracking to the original BB-Wing
entrance and deciding to move down the correct path. This error loop con-
sumed 33 seconds to ~2 minutes.
Interestingly, none of the other participants who reached this same decision
point made the same error. When using the hospital entrance, participants
must pass by the BB-Wing entrance on their way to the Dental Registration
Oce. Of the 16 participants who used the hospital entrance, none made a
wrong left turn down to the RR-Wing.
A similar preference for straight routes was observed during the second
waynding task, when participants were asked to navigate from the Dental
Registration Oce to the D-2 Clinic. Participants were directed by the
research associate to make an immediate left into the B-Wing. (Signage also
instructs visitors to travel down the B-Wing to reach the D-2 Clinic). How-
ever, half of the 20 participants who had been traveling straight (from the
Hospital and BB-Wing entrances) ignored these directions and proceeded
straight ahead into the C-Wing. Of the ten participants who moved straight
into the C-Wing, six had the original map, and four had the revised map.
Figure 5 bottom left
Visitors often perceive the doors
to the walkways to be locked or
alarmed exits.
Figure 6 bottom right
The main academic entrance
opens into a lobby and waiting
area with a highly visible
staircase. This staircase leads to a
single location—an auditorium
used for special events.
23
Map Design
Cheng & Pérez-Kriz

Based on this analysis, the map should be modied to include the RR-Wing
since this is a common error destination. It is less clear how the map could
be revised to prevent travelers from continuing straight and missing turns.
It may be necessary to instead alter signage or the environment to make
entrances to A-Wing and B-Wing more prominent.
5___Room Numbering
For the four participants who used the academic entrance, a preference for
straight routes should have neatly aligned with the initial waynding task.
The rst target destination (the Dental Patient Registration Oce) is located
in the B-Wing, and the entrance to the B-Wing is straight-ahead and visible
from the academic entrance.
However, there is an addressing problem with the Dental Patient Regis-
tration Oce, which is situated at the corner of A-Wing and B-Wing. The
oce has two doors; the address that patients are directed to (B-307, in the
B-Wing) is permanently locked and no longer in use, and instead, the second
door in the A-Wing serves as the main oce entrance. The administrators
have not provided the A-Wing door with a room number, although there are
overhead text signs that point to this second door location.
Unsurprisingly, this addressing issue caused problems for two of the four
participants who used the academic entrance. Upon entry, these two partic-
ipants, both with the new map, proceeded straight down into the B-Wing.
There, they spent 1-3 minutes traversing the wing (and trying the locked
door to B-307) before deciding to backtrack to the entrance. After scanning
the lobby area a second time, they discovered the signs that point to the
oce door in the A-Wing.
Figure 7
Three participants (75%) who
used the intermediate BB-Wing
entrance continued straight
down into the RR Wing. These
travelers missed the secondary
set of signs to the dental clinics
posted by the elevators.
24
Visible Language
48.2

The location of B-307 was not an issue for participants who used the
hospital and intermediate BB-Wing entrances, because these participants
pass through the A-Wing on their way to the B-Wing. The oce door in the
A-Wing is visually prominent; if participants are passing through, they are
unlikely to miss the oce (gure 8).
It is not clear how the map could be altered to prevent the confusion caused
by the existing naming scheme. Instead, we recommend that all room num-
bers be made consistent with their actual location.
6___Wing Names
Three participants (one with the original map, and two with new map)
spent time backtracking in the BB-Wing, because they failed to distinguish
between the B-Wing and the BB-Wing. This appears to be a common prob-
lem because the door to BB-306 is posted with occupant-created signs that
direct dental patients to B-307. These unocial signs are clear evidence of
waynding failure.
We are unable to envision how the map could be altered to avoid confusion
between wing names with the same letters (i.e., A-Wing/AA-Wing, B-Wing/
BB-Wing, etc.). Instead, it may be desirable to give wings more distinctive
and memorable names. Designer Paul Mijksenaar has discussed the similar
problem of coding oors in car parks; he admires a system that exists at
Disneyland, where oors are named after Disney characters, such as Pluto,
Mickey Mouse, etc. The oor names are further reinforced in the environment
by the use of character illustrations as wall graphics (Kinross, 1993).
Figure 8
The main oce door to B-307
is visually prominent. However,
the door is located in the
A-Wing rather than the B-Wing.
25
Map Design
Cheng & Pérez-Kriz

7___Route/Wing Selection
During the second half of the test, while navigating from the Patient Reg-
istration Oce to the D-2 Clinic, ve participants (21%) selected the wrong
wing at a decision point—a wing that led away from their desired destina-
tion. These errors all occurred on Level 2, after the participant had walked
down one set of stairs from Level 3 (the entry oor).
Specically, at the intersection of the C-Wing, D-Wing, and E-Wing, three
participants continued straight ahead to the E-Wing (instead of turning
left to the D-Wing). As described earlier, this error may be the result of a
preference for straight routes. Two of these participants had the original
map, and one had the revised map.
The other two errors are more dicult to explain. One participant made
a wrong turn after going down the B2 stairwell, where it is possible to go
left (correct) or right (a dead end). The other participant made an incorrect
right turn at the intersection of the C-Wing, B-Wing, and A-Wing—and
traveled back towards the hospital, rather than to the D-2 Dental Clinic.
Before making these errors, neither participant looked at their maps for
a substantive length of time (both participants had the original map).
The rst participant looked at the map for seven seconds as he began the
second route, then glanced at the map for two seconds while descending
the stairs. The second participant is remarkable for not looking at the map
at all during the entire experiment (he relied completely on signage).
There are several possible explanations for these wrong turns. Partici-
pants may have become disorientated after traveling down one level—
research studies have shown that level changes can negatively impact
navigation (Hölscher, Meilinger, Vrachliotis, Brösamle, and Knau, 2006).
Alternatively, it is possible that their route choice could be the result of a
preference for right-hand turns—a number of studies have found a right-
ward turning bias in pedestrian behavior (as described by Bitgood and
Dukes, 2006). In either case, it is not clear how the map could be modied
to prevent these errors.
8___Map Orientation & Alignment
Two of the participants with the revised map made a wrong turn upon
entry to the hospital (both of these participants missed the secondary set of
signs to the dental clinic). Instead of turning right towards the dental clinics,
they turned left and traveled deeper into the hospital area. One participant
corrected his route after reaching the hospital elevators. The other partici-
pant traversed several hospital areas (including the hospital elevators, the
hospital information desk and the hospital pharmacy) before self-correcting.
These error locations were noted for inclusion in future map iterations (none
of these hospital facilities were shown on the new map being tested).
26
Visible Language
48.2

The initial wrong turn of these two participants could be the result of the
map alignment errors. The original map had a “forward-up” orientation from
the parking garage. Dental sta felt this orientation was incorrect, since this
conguration places South at the top of the map, counter to the “North-Up”
convention that is standard in most modern-day maps. Therefore, adminis-
trators requested that the revised map be “North-Up” even though several
studies of xed “You-Are-Here” ( YAH) maps (Levine, 1982, and Levine, Mar-
chon, and Hanley, 1984) have shown that people make signicantly fewer
errors when their map is aligned to be “forward-up”—when the map directly
corresponds to the traveler (i.e., left and right on the map are also left and
right for the traveler).
Of course, unlike a xed wall-mounted YAH map, a hand-held map can be
freely rotated in any direction. The only indication of the “correct” orien-
tation is the positioning of text labels. It is not certain if aligning the main
route and text labels in a forward-up orientation on a hand-held map will
necessarily facilitate better waynding, especially considering the violation
of the classic North-up convention in this particular circumstance. However,
both participants were observed examining their maps prior to making the
wrong turn, and neither had rotated the map to match their forward move-
ment. Also, this wrong turn error did not occur for any of the original map
participants, who had the forward-up orientation.
Perhaps one solution to a possible orientation problem is to install wall
mounted YAH maps at key decision points, such as the hospital entrance.
While there are a few wall-mounted YAH maps in the wings where the dental
clinics are located, these YAH maps only indicate position, not alignment.
Additionally, these YAH maps vary widely in style and content. It would be
desirable for all building maps and hand-held maps to be part of a unied
system, so that patients can more easily relate the map that they receive and
study in advance of their appointment to maps that identify their position
and orientation in the actual environment.
9___map size
One of the twelve participants who received the new map failed to notice
that the map was folded, and could be extended to a tabloid (11x17”)
format. Therefore, the participant navigated during the test using only the
basic building overview—she did not view or use the area that shows the
B-Wing and D-Wing in greater detail with specic dental clinics and oces.
There are a few possible explanations for this occurrence. People may not
normally receive tabloid size documents, so it may not be immediately
obvious to unfold the map. It is also possible that the participant focused
on the rst side of the map immediately after receiving it because it allowed
her to see her initial starting point. Researchers have hypothesized that
27
Map Design
Cheng & Pérez-Kriz

travelers mentally divide their route into sub-segments that are bounded by
landmarks or areas of relative familiarity (Tverksy, 1992). The rst segment of
the test route begins at the parking garage, so travelers may have wished to
keep this known anchor point in view as they navigated toward an unknown
destination. The parking garage was not shown on the second side of the
map, so that side may have seemed irrelevant.
Given this analysis, it seems best to include the parking garage on the de-
tailed side of the map. It may also be advantageous to reformat the map to
t on a single letter-size document.
10___Misunderstood/Missing Information
During the nal stage of the test procedure, participants oered their
suggestions for improving the map. These suggestions largely concerned
areas that participants felt should be easier to anticipate and understand.
Specically, two participants asked if building entrances could be claried
and emphasized (see #1, Building Entrances). Three participants asked if the
enclosure containing the elevators could be claried (see #3, Architectural
Barriers). One participant suggested that the sign to the Dental Patient
Registration Oce, B-307, be shown more prominently (see #5, Room Num-
bering). One participant asked if the dierence between the B-Wing and
BB-Wing could be shown more clearly (see #6, Wing Names).
In addition, participants suggested that certain elements be added to the
map—these were either error locations (such as the RR-Wing, as described
above in #4, Preference for Straight Routes), or landmarks (such as the
escalators/stairs from the parking garage to the hospital entrance, the hos-
pital information desk, and interim rooms.) This last suggestion came from
participants who used room numbers to orientate themselves on the map.
However, because of confusion with wing names (i.e., B-Wing vs. BB-Wing),
door numbers were not always eective landmarks for participants.
On the original map, four participants commented that they did not
understand the symbol used for stairs. This issue appears to be resolved on
the revised map, where we used an enclosed form rather than the original
shbone” icon (there were no participant comments regarding the new
stair symbol).
28
Visible Language
48.2
Original ‘shbone’ icon for stairs Revised enclosed stairs icon
“

The goal of this study was to determine if a printed, paper map could aid
visitors in navigating through a large, complex medical center. Given that all
study participants were able to arrive at both target destinations by using
a map, this artifact does appear to be of benet to users. However, because
almost all participants (22 out of 24) experienced waynding errors, there is
clearly room for improvement.
Certainly, in this particular building there are a number of issues in spatial
organization that make waynding with a map dicult. These factors are
signicant negatives for visitors to overcome when navigating in an unfamil-
iar environment. Clearly, the plan of this building needs to be re-examined.
Potential solutions could involve operational changes (such as directing
people to the most ecient building entrance or renaming specic wings),
alterations to the built structure, and/or the adoption of new technologies
for visitor communication and guidance.
Most importantly, what is needed is a waynding program that systemat-
ically integrates multiple waynding elements (oor plans, signs, verbal/
written directions, information/help desks, and maps) into a coherent and
consistently applied design program. Such an endeavor will require substan-
tial, long-term eort, and signicant dedication of nancial resources. In the
interim, a printed, paper map can function as an inexpensive supplementary
measure while support is gathered for a larger initiative.
A secondary goal of this study was to determine was how to best design a
paper map for a specic context. The research described above reveals the
complexity of maps as information design problems. Even though maps are
commonplace objects, and their purpose is well understood, determining
the ideal form and content of an eective map is not trivial.
As part of an iterative design process, we began by identifying a variety of
design factors that inuence the cognitive aspects of understanding and
using maps during waynding. For some cases, it was obvious how research
ndings could be applied to our design problem; in others, the research was
descriptive rather than prescriptive, and empirical testing was required to
determine what information should be included or excluded on the map
and if changes in format enhanced or detracted from communication goals.
Playback and analysis of participant videos did allow our team to identify
specic information that was missing and should be included on future
iterations of the map, including key landmarks and the locations of common
route errors. Following this eort, we plan to revise the map accordingly, and
conduct additional user testing. In this way, the combination of the design
principles, client/stakeholder input, and behavioral results works to inform a
continuous cycle of design development and renement.
conclusion

30
Visible Language
48.2
AUTHORS’ BIOS
Karen Cheng is Professor of Visual Communication
Design at the University of Washington, where she
teaches information design and data visualization.
Karen received her Master’s degree from the University
of Cincinnati College of Design, Art, Architecture and
Planning; she also holds an Honors Bachelor’s degree
in Chemical Engineering from Penn State University.
Prior to joining the faculty at the University of Wash-
ington in 1997, she worked in Brand Management at
the Procter and Gamble Company. Her book, Designing
Type, was published by Yale University Press in Spring
2006, and has been translated into French, German,
Spanish, Korean and Chinese.
kcheng@uw.edu
Sarah Pérez-Kriz is Associate Director of Institutional
Assessment at George Mason University, where she
manages academic program review and oversees the
assessment of the university’s learning spaces. Before
joining the Office of Institutional Assessment, Sarah
served as the Founding Director of the Office of
Undergraduate Research at the University of San
Diego. She also spent several years as a faculty
member in the Department of Human Centered
Design + Engineering at the University of Washington,
where she taught courses in human-computer inter-
action, visual communication, and research methods.
Sarah earned her PhD in cognitive psychology from
the University of California, Santa Barbara in 2006.

BiBliographY
Arnheim, Rudolf. 1976. “The Perception of Maps.” Cartography and Geographic
Information Science. 3 (1): 5-10.
Arthur, Paul, and Romedi Passini. 1992. Wayfinding: People, Signs, and Architecture.
New York: McGraw-Hill Book Co.
Bartram, D. J. 1980. “Comprehending Spatial Information: The Relative Efficiency of
Different Methods of Presenting Information about Bus Routes.” The Journal of
Applied Psychology. 65 (1): 103-10.
Best, G. A. 1970. “Direction-Finding in Large Buildings.” A rchitectural Psychology:
Proceedings of the Conference Held at Dalandhui, University of Strathclyde,
28 February-2 March 1969. D. Canter (Ed.), London: RIBA Publications, 72-91.
Bitgood, Stephen, and Stephany Dukes. 2006. “Not Another Step!” Environment and
Behavior. 38 (3): 394-405.
Boy Scouts of America. 2009. Boy Scout Handbook. Irving, Texas: Boys Scouts of America.
Bronzaft, Arline L., and Stephen B. Dobrow. 1983. “Improving Transit Information
Systems.” Journal of Environmental Systems. 13 (4): 365-376.
Carlson, Laura, Christoph Hölscher, Thomas Shipley, and Ruth Conroy Dalton. 2010.
“Getting Lost in Buildings.” Current Directions in Psychological Science.
19 (5): 284-289.
Carpman, Janet Reizenstein, Myron A. Grant, and Deborah A. Simmons. 1983.
“Wayfinding in the Hospital Environment: The Impact of Various Floor Numbering
Alternatives.” Journal of Environmental Systems. 13 (4): 353-364.
Davies C., & Peebles, D. (2007). Strategies for Orientation: The Role of 3D Landmark
Salience and Map Alignment. Cognitive Science Society, Inc.
Devlin, Ann Sloan, and Jason Bernstein. 1995. “Interactive Wayfinding: Use of Cues by
Men and Women.” Journal of Environmental Psychology. 15 (1): 23-38.
Dobson, Michael W. 1980. “The Influence of the Amount of Graphic Information on Visual
Matching.” The Cartographic Journal. 17 (1): 26-32.
Hayward, D. G., and Brydon-Miner, M. 1984. “Spatial and Conceptual Aspects of
Orientation: Visitor Experiences at an Outdoor History Museum.” Journal of
Environmental Systems, 13 (4), 317-332.
Hegarty, Mary, Anthony E. Richardson, Daniel R. Motello, Kristin Lovelace, and Ilavanil
Subbiah. 2002. “Development of a Self-Report Measure of Environmental Spatial
Ability.” Intelligence. 30 (5): 425-447.
Hochmair, Hartwig H. 2005. “Investigating the Effectiveness of the Least- Angle Strategy
for Wayfinding in Unknown Street Networks.” Environment & Planning B: Planning
& Design. 32 (5): 673-692.
Hölscher, Christoph, Tobias Meilinger, Georg Vrachliotis, Martin Brösamle, and Markus
Knauff. 2006. “Up the Down Staircase: Wayfinding Strategies in Multi-Level
Buildings.” Journal of Environmental Psychology. 26 (4): 284-299.
Kinross, Robin. 1993. “Paul Mijksenaar: A Conversation and a Lecture.” Information Design
Journal. 7 (2): 105-114.

Klippel, Alexander, Stephen Hirtle, and Clare Davies. 2010. “You-Are-Here Maps: Creating
Spatial Awareness through Map-like Representations.” Spatial Cognition &
Computation. 10 (2-3): 83-93.
Levine, Marvin. 1982. “You-Are-Here Maps: Psychological Considerations.” Environment
and Behavior. 14 (2): 221-237.
Levine, Marvin, Iris Marchon, and Gerard Hanley. 1984. “The Placement and
Misplacement of You-Are-Here Maps.” Environment and Behavior. 16 (2): 139-157.
Lynch, Kevin. 1960. The Image of the City. Cambridge, Mass: MIT Press.
McKean, J. 1972. “University of Essex: Case Study.” Architects Journal,156: 645-667.
Miller, George A. 1994. “The Magical Number Seven, Plus or Minus Two: Some Limits on
Our Capacity for Processing Information.” Psychological Review. 101 (2): 343-52.
Passini, R. 1980. “Wayfinding in Complex Buildings: An Environmental Analysis.” Man-
Environment Systems, 10 (1), 31-40.
Peponis, John, Craig Zimring, and Yoon Choi. 1990. “Finding the Building in Wayfinding.”
Environment and Behavior. 22 (5): 555-590.
Phillips, Richard. 1979. “Making Maps Easy to Read: A Summary of Research.” Processing
of Visible Language 1, Kolers, P.A., Wrolstad, H.E. and Bouma, H. (Eds.) New York:
Plenum, 165-174.
Rosenholtz, Ruth, Yuanzhen Li, and Lisa Nakano. 2007. “Measuring Visual Clutter.”
Journal of Vision. 7 (2): 1-22.
Siegel A. W., and S. H. White. 1975. “The Development of Spatial Representations of
Large-Scale Environments.” Advances in Child Development and Behavior. 10: 9-55.
Southworth, Michael, and Susan Southworth. 1982. Maps: A Visual Survey and
Design Guide. Boston: Little Brown.
Thorndyke, Perry and Cathleen Stasz. 1980. “Individual Differences in Procedures
for Knowledge Acquisition from Maps.” Cognitive Psychology. 12 (1): 137-175.
Tolman, Edward C. 1948. “Cognitive Maps in Rats and Men.” Psychological
Review. 55 (4): 189-208.
Tufte, Edward R. 2001. The Visual Display of Quantitative Information. Cheshire,
Conn: Graphics Press.
Tversky, Barbara. 1993. “Cognitive Maps, Cognitive Collages, and Spatial Mental Models.
”Spatial Information Theory: A Theoretical Basis for GIS, COSIT’ 93.
A. U. Frank and I. Campari (Eds.) Berlin: Springer-Verlag, 14-24.
Weisman, Jerry. 1981. “Evaluating Architectural Legibility: Wayfinding in the
Built Environment.” Environment and Behavior. 13 (2): 189-204.
Wener, Richard, and Robert Kaminoff. 1983. “Improving Environmental Information.”
Environment and Behavior. 15 (1): 3-20.
Wright, Patricia, Audrey J. Hull, and Ann Lickorish. 1993. “Navigating in a Hospital
Outpatients’ Department: The Merits of Maps and Wall Signs.” Journal of
Architectural and Planning Research. 10 (1): 76.

33
Map Design
Cheng & Pérez-Kriz
authors’ acknowledgements
Special thanks to Tanya Test and Ronald Viernes, two students
of the Visual Communication Design program at the University
of Washington, for their work on the design of the new dental
clinic map. Also, many thanks for the four students who
worked with us to recruit study participants and conduct the
user tests: Brennen Birch, Joy Palludan, Caryn Carlson Rothe,
and Calder Thami. We are grateful for the useful suggestions
of Linda Norlen, Aliate Assistant Professor at the University
of Washington, as well as the three anonymous reviewers. We
would also like to thank the Wycko Milliman Endowed Chair
in Art from the School of Art at the University of Washington,
for their nancial support of payments to test subjects.

CopyrightofVisibleLanguageisthepropertyofUniversityofCincinnati,onbehalfof
VisibleLanguageanditscontentmaynotbecopiedoremailedtomultiplesitesorpostedtoa
listservwithoutthecopyrightholder'sexpresswrittenpermission.However,usersmayprint,
download,oremailarticlesforindividualuse.