Zeisel John - Chs 7-10 Inquiry by Design (1984)
Zeisel John - Chs 7-10 Inquiry by Design (1984)
TOOLS FOR
ENVIRONMENT-BEHAVIOR
RESEARCH
JOHN ZEISEL
Harvard University
OBSERVING
PHYSICAL TRACES
89
OBSERVING PHYSICAL TRACES 91
90 CHAPTER 7
lmageable
of environmental change was reached, changes in behavior took place as well:
patients began to greet each other more, to converse more, and to read more, and Observing physical traces provides rich impressions and is highly illustra
staff members began a crafts program on the tables in the ward. And it all began tive. Walking through a home for older veterans in Oxford,-New York, investi
when Sommer noticed a difference between how custodians left chairs in the gators saw, for example, wheelchairs in odd places, old furniture, new medical
morning and how patients and visitors left them at the end of the day. equipment, direction signs, people in uniforms, open cans of food on window
The following discussion presents (1) significant qualities that observing sills, and patients' get-well cards taped to walls in rooms (Snyder & Ostrander,
physical traces has for use in E-B research, (2) types of devices for recording 1974). The walk gave researchers an initial picture of what life in that home was
observed traces, and (3) a classification of trace types to make visible those like: its design successes, some problems, exceptional situations, patterned wear
relations between people and environment that are useful for designing. and tear. At the beginning of a research project, such observations can be used to
spark investigators to think about what the observed objects might mean. Skillful
observers will notice even commonplace physical traces and figure out which of
Observing Physical Traces them will lead to fruitful inferences to pursue further. At Oxford, investigators
focused their attention on cans of food on windowsills-developing from this
Qualities of the Method information a central research hypothesis that residents Jived a 24-hour life-style
Imageable out of phase with the institution's 6:00 a.m. to 7:00 p.m. schedule.
Unobtrusive From a trace investigators ask questions about what caused it, what the
Durable person who created the trace intended, and what sequence of events led up to the
Easy trace. The imageable quality of physical traces makes it easy to generate hypoth
eses about causes, intent, and sequence, but from the trace alone researchers
Recording Devices cannot tell how tenable their hypotheses are; to do this, they need other methods.
Annotated diagrams For example, in a brief evaluation of a somewhat run-down housing project in
Drawings Roxbury, Massachusetts, Zeisel (1973b) found large, well-kept flowering shrubs
Photographs in residents' backyards. At first he falsely assumed that residents beautified their
Counting small yards because they cared about the appearance of the project and wanted
their own vistas more scenic. In later interviews with residents he found that
What to Look for shrubs had been planted years before in response to a management-sponsored
By-products of use competition for the best garden. A closer second look revealed that even good
Adaptations for use looking plants in the backyards had been very much neglected.
Displays of self The same potential pitfall can arise when investigators falsely infer intent.
Public messages One morning a group of architects visiting a housing project for older people in a
Context
predominantly Italian section of Boston noticed a bocce-ball court surrounded by
apartment windows. It looked as if it had never been used. They tentatively
concluded that something was wrong with the facility, that residents did not like
playing bocce ball, or that they did not like the location of the court. In fact, the
QUALITIES OF THE METHOD
court looked brand new because workmen had just completed it several days
before. In addition, it was early morning, and anyone who might have used the
Observing traces is an exceptionally useful research tool that can produce
court was still at home.
valuable insights at the beginning of a project, test hypotheses in the middle, and
be a source of ideas and new concepts throughout. If you take into account what It is also difficult to infer process. In a suburban Boston prison, cell walls
are papered from ceiling to floor with Playboy, Penthouse, and Swank center
the method can and cannot do, you can achieve the results you want; like any
folds. At first glance it seems impressive that prisoners fix up their dwelling units
tool, if used inappropriately it can be destructive. The method can be a source of
so extensively-that they mark out and personalize territory so dramatically. But
provocative images, is unobtrusive, is easy to use, and deals with long-lasting
the impression the traces give is misleading. Most centerfolds have been glued to
phenomena. It provides opportunities for investigators but also sets up some
traps. the cell wall by a series of previous inmates. Walls are not stripped when a new
92 CHAPTER 7 OBSERVING PHYSICAL TRACES 93
inmate moves in, every 6 to 12 months. The wallpapered surroundings that beforehand, putting away such physical traces as toys in the living room, which
inmates move into offer them many diversions but little chance to personalize. might indicate how different rooms are used.
Visual trace records can be used as illustrations of research concepts. This Observing or measuring traces does not require being present when the
can prove useful to investigators who want to follow up on trace observations with traces are created. The method is therefore particularly useful to find out about
interviews to test their hypotheses. In studies of property damage in parks rare events, hard-to-see events, private behaviors, and behavior of groups who
(Welch, Zeisel, & Ladd, 1978) and in schools (Zeisel, 1976a), for example, cannot be interviewed. Zeisel's school study (1976a) provides an example of
investigators showed slides of damaged property to groups of teenagers, park using physical traces to document private behavior that is hard to observe direct
personnel, and persons living next to the property in order to focus discussion on ly. During the day teenagers can be seen hanging out around schools, playing
what these people thought about property damage. stickball against walls, and sometimes climbing onto rooftops. At night they
In lectures and reports, pictures of vivid traces can help viewers and sometimes find out-of-the-way places around back to sit together, drink, and
readers understand physical settings in which projects were carried out. Lenihan smoke. Boston teenagers treat these half-hidden settings as clubhouses where
(1966), in his report evaluating the VISTA program in the 1960s, wanted readers outsiders are not allowed. The first hint of such nighttime clubhouse activity
to understand the wide variety of volunteers' assignments: Appalachian mountain came from physical traces: empty beer cans, discarded playing cards, cigarette
villages, Southwestern desert towns, urban slums. He used photos of physical butts, graffiti, and broken lights.
traces to augment the poetry of his writing.
The force of concrete visual impressions can be a pitfall for careless re Durable
searchers. The visual impact of even low-frequency observations can be so
great-flowering bushes, nearly new facilities, vandalized windows-that they Many traces have the advantage for researchers that they do not quickly
dominate a researcher's mind. To a person walking through a well-kept housing disappear. Investigators can return to a research site for more observations or
development, the beauty of a few flowering bushes can give the impression that counting and can document traces with photographs or drawings. Of course, the
there are flowers in bloom everywhere, even though few residents have bushes more permanent a trace is, the greater its chance of being observed at all. For
and only some are flowering. When such traces are photographed and presented example, rock gardens and paving stones in someone's garden will be visible for
out of context, they can mislead-a problem of false emphasis the visual com years, long after grass and flowers have virtually disappeared.
munications media face every day. It is important that observers also train them There is, however, the problem of selective deposit. Some activities are
selves to see traces that do not stand out, such as the scarcity of certain expected more likely to leave traces than others. The extent of beer drinking in back of a
objects or the absence of wear and tear. If you ask yourself "What traces are school can be detected the next day by the number of cans. Playing poker or
missing?" in addition to "What traces do I see?" you are more likely not to be smoking nonfilter cigarett�s may leave no traces at all.
seduced by visually impressive traces. You will begin to see what is not there. Another consequence of the durability of traces is their cumulative quality;
earlier traces can encourage later ones. A large number of people may feel free to
Unobtrusive
cross a lawn because people who did so before left a path, whereas few people
would do so were there no path. This cumulative quality can cause problems for
Observing traces is an unobtrusive method (Webb et al., 1966). It does not investigators who overlook it, who think each act is independent of earlier ones.
influence the behavior that caused the trace. But if traces are not taken out of context, their cumulative character can provide
Unobtrusiveness is particularly valuable when gathering data about which insights for data gathering and analysis. The finding, for example, that litter
respondents are sensitive or when respondents have a stake in a certain answer. tends to beget litter (Finnie, 1973) is particularly useful if you want to arrange
For example, an investigator who wants to know how strictly hospital attendants maintenance schedules in parks and around schools.
follow fire-safety rules will learn more from counting the fire exits blocked by
stretchers than from interviewing attendants, who may want to paint a rosier Easy
picture than actually exists. School principals who want to avoid showing they
are not doing a good job may report less damage to school property than a Physical-trace observation is generally inexpensive and quick to yield in
researcher might observe directly. And principals who want the school commit teresting information. The inexpensiveness of a brief physical-trace survey
tee to increase the budget for maintenance may magnify the damage. If a respon makes it possible in most research projects not only to discover but also to
dent at home knows a researcher is coming, she may neaten up the house explore in greater depth a host of initial hypotheses. Using more costly methods
94 CHAPTER 7 OBSERVING PHYSICAL TRACES 95
would mean discarding possibly fruitful but implausible hypotheses without are. In settings where cameras are out of place or lighting is difficult and the
looking at them closely. This same quality means, however, that researchers can researcher does not want to use flash attachments, written trace notation is
waste their energy because time and money do not force them to think through appropriate. Annotated diagrams are also well suited when traces can be recorded
each initial proposition rigorously before going into the field. on two-dimensional plans and then studied. The arrangement of chairs Sommer
The speed and ease with which physical traces can be recorded-in still (1969) observed in the patient dayroom could perhaps be represented in plan
photographs, sketches, notations-make the method useful for collecting a great more effectively than in photographs.
many data for speedy review. An initial site visit can yield enough recorded When annotated diagrams are chosen as one of the recording devices,
observations for weeks of review and analysis. This is helpful in generating a several rules of thumb can be helpful. Agreement among researchers on a set of
range of testable propositions and hypotheses. Yet the harvest can be so rich that standard symbols will increase comparability of the data within a project. For a
it seduces a research team not to look further: "We already have so much residential floor plan, for example, a team might use traditional architectural
information. Why do we need more?" symbols for furniture. When researchers on several projects use such standard
and easily understood symbols, their data can be more easily compared and
In sum, observing physical traces is imageable and unobtrusive, deals with shared.
durable data, and is easy to do. The following sections of this chapter discuss
ways to record trace observations and a classification of traces particularly rel
D
0
evant to questions of design.
RECORDING DEVICES
Investigators save energy and time by deciding before going into the field
Bed TV Dining table
[u I 11 Couch
how and when they will record trace observations: annotated diagrams, draw
ings, photographs, precoded counting lists, or a combination of these. If photo Architectural furniture symbols
graphs are chosen, researchers decide such issues as whether prints or slides will
be more useful for the purposes of the study or whether both are needed. Each
decision affects how trace observations can be analyzed, how they can be used in Outdoors and in special settings, investigators may have to be more inven
conjunction with other research methods, and how findings will be presented. tive about the symbols they use. In their study of peddlers and pedestrians on
Observations ought also to be timed to avoid possible systematic effects of Rome's Spanish Steps, Gunter, Reinink, and Giinter (1978) developed a set of
maintenance schedules or predictable activity cycles on the data-for instance, symbols for recording how peddlers arranged their wares (see next page).
If you want your observation notes not to be confused with your reacti9ns
early morning cleanups that obliterate signs of teenagers' night life around
to what you saw, you must not analyze them in the field. Provisions need to be
schools.
made to facilitate subsequent analysis.
A simple device can facilitate preliminary analysis of field notes with a
Annotated Diagrams minimum of fuss: Original notes and diagrams are made on the left half of the
notepaper, leaving the right half open for recording hunches and preliminary
Recording traces verbally and diagrammatically, as a rule, requires little hypotheses (see the illustration "Furniture Layout in El Barrio Apartment," p.
preparation and no special skills. Except for a notepad, the recording method is 97). A wide margin can be made on any notepaper simply by creasing it.
unobtrusive; to make it still less obtrusive, trained observers may memorize If investigators know the floor plans of the places to be observed, and if
major traces in a setting and record them later. This is especially possible when more than one similar place is to be observed or the same place is to be looked at
the setting is simple and the objective standardized, as when making diagrams of several times, their notepaper can have a floor plan printed on it. This facilitates
furniture layouts in people's living rooms for a study of what furniture people making notes and ensures comparability of diagrams. This method can be used
own and how they arrange it. equally well for interiors, such as offices, waiting rooms, or dwelling units
During a two-person interview one interviewer can inconspicuously draw a (Zeise}, 1973a) and for exteriors, such as playgrounds, street comers, or plazas
plan of the setting and note where objects are located and where physical traces (Giinter et al., 1978).
96 CHAPTER 7 OBSERVING PHYSICAL TRACES 97
OBSERVATIONS COMMENTS
Drawings Arrayed in this way, photographs enable all members of a research group to
participate in initial visits to the site.
If observers have the skill to make sketches of the traces they see, the time
Color slides have other benefits. In addition to being convenient and capti
it takes may well be worthwhile. Drawings can be extremely useful in final
vating during oral presentations, slides can be easily grouped and regrouped for
reports because they are highly imageable and inexpensive to reproduce.
analysis on light tables or in projectors. Some slide films can be developed
commercially in just a few hours. When it is essential to know that you have all
Photographs your data before leaving a site, or when you want to make a presentation shortly
after making observations, slide film that can be quickly developed or even
Photographs of physical traces taken at the beginning of a research project
instant-print film may be a lifesaver.
can give all parties working on it an initial overview of the types of things they
are likely to see in the field. Discussion of photographs among team members
can quickly generate hypotheses about possible fruitful issues for further study. Counting
A group can leisurely discuss what behavior a trace might reflect and what intent
might be behind it. For these reasons, it is generally valuable to document both Certain traces yield their full value only when their quantity is taken into
easily photographed outdoor traces and indoor traces, although indoor ones may account. In such situations it will suffice to record in detail one or two examples
be harder to photograph. Photographs are particularly valuable if the research site and count the rest. For example, in a housing project where some families have
is not easily accessible because it is too far away, requires special permission to fenced in their backyards and some not, photographs of a few along with a
visit, or is altogether temporary (for example, a circus). careful count will do the job.
When investigators expect to count traces, they can first analyze photo If you know what you want to count beforehand, precoded counting pads
graphs of observations to decide on categories for counting. Photographs can be or checklists can be arranged-possibly linked to the site plan for accurate
used as stimuli in focused interviews, to determine the categories respondents use location data.
when they see such things. At the end of a project, photographs are excellent to As important as choosing appropriate categories is intersubjectivity of the
illustrate verbal presentations of findings. Many of these qualities hold for photo categories among observers. Each member of a team of observers faced with the
graphs in research, whether they are of physical traces or of behavior. same physical trace ought to record it as a trace in the same category if data are to
In the field several rules of thumb and a few tricks can possibly save time, be comparable. To achieve a degree of intersubjectivity, observers in the U.S.
money, and embarrassment. Expensive cameras are seldom more useful as re housing census are shown photographs representing distinct levels-and there
search tools than inexpensive ones. Researchers need to take some photographs fore categories-of housing deterioration. On the basis of these "exemplars" this
themselves because they know what to record for analysis-what to include in very large group is expected to develop a shared way of looking, at least to some
the picture and what to leave out. For illustrative photographs, one can always extent.
hire a professional photographer (or choose the most skilled researcher). Even Another practical way to develop intersubjectivity among investigators is
then one will have to tell the professional precisely what to photograph. When to take them on a site visit to settings similar to those at the research site. Through
extra equipment is needed-for instance, flash attachments or tripods for interior group discussion they can learn from one another and arrive at a consensus of
photographs-it must be selected with consideration of both research require how items they see would be recorded.
ments and respondents' sensitivity. Each way of recording traces catches another dimension of the trace and
A researcher's choice of film has perhaps the greatest consequ�nces for the provides researchers with new data.
rest of the study. Black-and-white photographs, useful as illustrations, can also
be made useful as objects for group discussion. Color photographs are expensive
and difficult to print. From contact sheets or directly from negatives, researchers WHAT TO LOOK FOR
can choose a number of photographs which seem to cover the range of concerns
they are aware of, which seem to be most interesting, or which require more What an investigator chooses to observe depends on what he wants to do
discussion and analysis to understand. These photographs can be inexpensively with the data he gathers. If I want to identify my mother in a crowd, I will try to
printed as large blowups on a microfilm printing machine, available at most notice only women whose hair is brown with a gray streak. If you want a police
libraries. Although such prints cannot be used as permanent records because they officer in New York City, you will look for and "see" only people in dark blue
fade after several months, they can be put on a wall for analysis and discussion. uniforms.
100 CHAPTER 7 OBSERVING PHYSICAL TRACES 101
The following categories for looking at and gathering data about physical some place as their own, to make it express who they are personally: a flag or a
traces are organized to increase designers' control over the behavioral effects and religious symbol on front lawns; mementos of trips on windowsills. Public
side effects of their decisions and to increase people's own control over their messages are changes such as wall posters and graffiti, by which people use en
relation to the environment. Both these purposes are means to another end: to vironments to communicate with a large public audience, sometimes anonymously.
increase everyone's ability to intervene through design to make settings better What you look for depends on what you want to do with the data. Ruesch
suited to what people actually do. These purposes translate into such questions as and Kees, in their perceptive book Nonverbal Communication (1970), describe
the following: How do environments create opportunities for people? Where do using data on facial expressions, body movement, and physical traces to under
people and their surroundings impinge on each other? Where do they limit each stand how people communicate without words. Their emphasis on communica
other? How do people use the environment as means to an end? And to what tion leads them to underplay traces in the categories of adaptations for use and
ends? What design skills do people have? How do they manipulate their sur by-products of use but provide a more detailed analytic scheme for displays of
roundings? How do people change environments to meet their needs? What takes self. Another important description of how to observe physical traces is included
place in particular settings? To answer such questions, the following organization in Webb et al., Unobtrusive Measures (1966). Webb et al. describe the useful
for observing physical traces is useful. ness of a range of measures-for example, counting bottles in garbage cans to
see how much people drink, observing litter in the park, and analyzing suicide
notes. The categories they develop are not all equally suited to solving E-B
Physical Traces to Look for questions. For example, they use the term accretion to describe any type of
physical trace left behind, without specifying the manner in which it was left
By-products of Vse -the actor's environmental intent. All but one of the categories discussed in the
following pages and several discussed in Chapter 12, on archival methods, are
Erosions
Leftovers examples of accretion. For clarity I have, therefore, scrupulously avoided the use
Missing traces of this important but broad term.
Some erosion traces, such as the scars in the butcher's table, indicate to the
interested researcher that planned and predicted activities have taken place;
others that the environment is being used in a new way, such as the path across
the lawn. Because most environments sustain some wear and tear, observers
must be careful to distinguish between erosion traces that signify bad design,
those that reflect uses designers planned for, and traces left when new and
appropriate activities took place. Erosion traces, and in fact all by-products
of use, can be the first step in finding out what those who use the setting feel
about it.
Leftovers. Physical objects as the result of some activities get left behind: Missing traces
cigarettes in ashtrays after a party, dishtowels hung on kitchen-cabinet knobs
next to a sink, open cans of food stored on windowsills in a veterans' residence. may lead to not very useful answers: "The apartment is vacant because tenants
just moved out." But it may also lead quickly to fruitful insights, because not to
use an available space is quite a strange thing to do.
Connections. Physical adaptations for use may connect two places, enabl
ing people to interact in new ways: holes that teenage�s strategically cut in a
playground fence to enable players to get in without walking around to a distant
gate; pass-throughs cut in walls between living rooms and windowless kitchens to
Props provide a view out when residents eat in the kitchen. Buildings converted to
restaurants often have windows cut into swinging kitchen doors so that people
New props may have been added because users or uses have changed or serving can avoid bumping into each other when coming from opposite direc
because certain activities were overlooked or considered unaffordable in original tions.
designs. Props added for either reason may reflect a particular user's idiosyncrat Connections that users of a facility make can indicate that the original
ic wants, such as the living-room stove, or they may reflect more normative designer overlooked a common behavior that requires being able to move, see,
behavior common to a larger group. hear, or talk between one space and another or that such activity developed since
the place was designed (as with the window in a swinging restaurant door). Of
Separations. Changes may separate spaces formerly together, increasing course, sometimes users may want a connection that setting managers do not. An
such qualities as privacy, control, and darkness or more sharply dividing territo example would be hacksawed bars on a prison-cell window after a jailbreak.
ries: ground-floor apartments with covered-over windows, stones along some
one's property line, "Keep Out" signs on back doors of buildings.
Separations can be particularly informative about side effects of design
decisions. The parking areas in the interior of Castle Square, a housing project in
Boston's South End, were deeded officially to the city so that it would maintain
them, plow them, and pick up garbage on them. But as an unanticipated side
effect, people who work in the surrounding neighborhood park there during the
day and sometimes all weekend. Residents feel that this infringes on their infor
mal right to park their cars just in front of their houses, and so they place wooden
sawhorses across the parking places in front of their doors to stop other people
from parking there.
Connections
Displays of Self
Residents change environments to put their stamp on them-to say "This is
mine and it says something about me." Displays of self may be directed toward
other people, but just as often the changes mean something mainly to the person
who makes them: mementos of trips, family portraits, doll collections. Displays
may help others identify a person's environment-name plaques on the front
door-or may tell people about the person by announcing what groups she is a
Separations member of.
106 CHAPTER 7 OBSERVING PHYSICAL TRACES 107
Personalization. People use environments to express their uniqueness and Did he do this on purpose to leave his mark for the next student? Would he use a
individuality: a style of furniture in the living room, trinkets on the windowsill, name tag provided by the administration? If so, what would he feel about it?
silly signs on businesspeople's desks. Each such use shows how someone is More important, what would this indicate about the relationship between students
different from his neighbor-in taste, in personality, in habits. and administrators?
How permanent a trace is may also be significant. Does the name of a
family etched into the wood of their front door mean they hold different attitudes
toward the neighborhood than their neighbors whose name is spelled out with
store-bought plastic letters in the lawn? The family with plastic letters may feel
no less permanent, but rather have greater respect for wooden doors.
Public Messages
Official. Probably the most frequently seen public messages are official such messages. But the usual traces left from unofficial public communications
ones erected by institutions, which may even pay for the right to do so: advertis are shreds of paper stuck to lampposts, brick walls, and newspaper stands.
ing signs, names of commercial establishments, place names. They reflect offi
cial uses of settings-the behavior of paying clients. Illegitimate. Messages to the general public which are _ not planned for, for
which environmental adaptive changes are not made, and which, although some
times expected, are seldom if ever approved of, are considered by man to _ be
� _
illegitimate uses of public environments. The mo�t f�equen� exam�le of illegiti
mate public messages is graffiti. Political graffit i with antiauthonty slogans or
antiethnic slogans often appear in prominent public places. Members of teenage
gangs in large American cities stake out their turf by writing their name and street
number on walls.
.
Official public messages
' .. •
�-..�-�
�--- -- �- -�-·-
Official public messages usually appear in environments designed for that
�-==-,:::-- ���- -
purpose. The private right to display official public messages is increasingly �-- ~__..__ --
,._..,_��---"-' H- - - -
-
.,.__~ � _�:
being challenged by the public, asserting its right not to see them. -- -- --
��? �--�
' ' ...,.- �- -
Unofficial. Individuals and groups also communicate publicly by means
�"'- __,.�-
of settings not designed specifically for that purpose. Unofficial messages usually
Illegitimate public messages
announce short- term events and are often accepted and even expected on surfaces
in public places: theater placards on wooden walls surrounding construction sites,
political posters stapled to telephone poles, and "Lost Cat" announcements taped Ille8itimate as I am using it here does not imply a value judgment. It
to laundromat windows. merely r�fers to official disapproval of the activity. Those who engage in the
activity may find it completely legitimate. For example, almost everywhere
students paint lines on walls of schools to enable themselves to play games: a
hockey goal to play street hockey or a square strike zone to play stickball. They
consider such lines as legitimate as the neatly painted official lines on the basket
ball court (Zeisel, 1976a). Others may consider the lines attacks on society.
Such "illegitimate" expression may have useful social side effects. Gang
graffiti, establishing territorial boundaries, possibly reduce gang conflict. Politi
cal slogans give minority political groups visibility.
Context
Unofficial public messages Traces clarify their context and are clarified by them. A square painted on a
wall may mean nothing. Near a school it is a stickball strike zone and signifies
that the area is used for street games. When looking at physical traces, research
Informal public messages tell investigators about such things as types of ers must keep in mind that they are trying to look beyond the trace itself to
cultural events taking place in an area, proportion of students living there, and understand a larger picture. That larger picture can emerge only if you see the
political activity. Some bookstores and supermarkets establish tack boards for context of what you observe.
110 CHAPTER 7
OVERVIEW Chapter 8
A good way to begin almost any E-B research project is to walk around the
research site looking for physical traces of behavior. It is easy to do, can be done
un�btrusively, and provides investigators with rich imagery to build on in solving
OBSERVING
therr problem. Trace observation can be carried out both qualitatively and quanti ENVIRONMENTAL
tatively.
This chapter has discussed categories of traces particularly appropriate for BEHAVIOR
E-B observations: by-products of use, adaptations for use, displays of self, public
messages. The frrst category represents remnants of what people do in an en
vironment, the others of what people do to it. This way of looking is aimed at
increasing our ability to intervene through design to make settings better suited to
Observing behavior means systematically watching people use their en
what people actually do.
vironments: individuals, pairs of people, small groups, and large groups. What
The next chapter discusses how to observe the other half of the E-B
do they do? How do activities relate to one another spatially? And how do spatial
equation: behavior.
relations affect participants? At the same time, observers of environmental be
havior look at how a physical environment supports or interferes with behaviors
taking place within it, especially the side effects the setting has on relationships
between individuals or groups. In a park, for example, an observer sees a child
playing, watched over by her father, who anxiously jumps up every time the
child moves out of his sight. The child's being hidden from view triggers a
reaction by her father. The event tells an observer something about the child's
activity and the importance for the relationship of maintaining a visual link
between father and child.
Observing behavior in physical settings generates data about people's activ
ities and the relationships needed to sustain them; about regularities of behavior;
about expected uses, new uses, and misuses of a place; and about behavioral
opportunities and constraints that environments provide.
You do not have to be an expert to observe behavior. Before entering a
party or a restaurant, you may survey the scene to see what behavior is appropri
ate there. An alert new student in a school watches who plays where in the
gymnasium, who sits where in class, and who sits with whom in the cafeteria.
Environment-behavior researchers systematically make the same types of obser
vations with different ends in mind.
Hall's classic description of how people behave in and use space, The
Hidden Dimension (1966), draws heavily on behavior observation in natural
settings. Sensitive behavior observation led Hall to discover the important spatial
dimension to human communication. He observed, for example, that how far or
how close people stand reflects their social relationship-distance generally
meaning coldness and closeness generally meaning friendliness. Further behavior
observation turned this rather simple conclusion into an exciting insight: the way
people from different cultures interpret spatial distances can lead to misunder
standing, even insult. For instance, an American might feel he is being friendly
by standing several feet from an Arab friend during a casual conversation. The
111
112 CHAPTER8 OBSERVING ENVIRONMENTAL BEHAVIOR 113
Arab, attributing meaning to space, feels the American to be cold and distant and Empathetic
moves closer. The American takes this move to be aggressive. He steps back. To
Researchers observing people soon get a feeling for the character of a
the Arab, this is clearly an attempt to be unfriendly-an insult.
situation. Observation, especially participant observation, allows researchers to
This chapter presents qualities of the research method for E-B studies,
"get into" a setting: to understand nuances that users of that setting feel. When
some practical steps observers can take to prepare for observing environmental
personal quirks of observers influence the recording of observations, their relia
behavior, and how to organize observations to learn the most about the relation
bility can be questioned. Yet personal feelings may provide essential initial 1
center are aspiring middle-class civil servants who live in buildings that reflect hospital emergency room when an ambulance arrives. As Wiseman's perceptive
modem norms to which some of the more traditional Indians do not keep. For documentary film Hospital (1970) shows, an ambulance arrival can have simul
example, some residents reported that they used the kitchen counters to prepare taneous effects on nurses, doctors, other patients, nearby staff members, police
meals, but when Brolin (1972) looked more closely, he found that they followed officers, and many others who participate, actively and passively.
the traditional Indian practice of cooking on portable stoves on the floor. One In complex situations observers begin to get a sense of chain reactions: the
resident assured Brolin that, caste distinctions being obsolete, everyone including effects of effects. No other effort gives a researcher so much of an idea of how
servants used the front door. Brolin was surprised to observe household servants people bring places to life. Ellis' (1974) explanation of "occasioning" among
using the back door. Had Brolin used only interviewing techniques, he might poorer Black people shows how they manipulate both behavior and time to cope
never have observed such rule-breaking activity. with limited space. For example, although kitchens are predominantly associated
People also tend not to report to interviewers activity they think is trivial with cooking and eating, residents might regularly use that room for other occa
and therefore not worth reporting. Nonetheless, a seemingly trivial datum may be sions: card parties, sewing bees, meeting the boys. Although "occasioning,"
central to an environmental research question. For example, if someone asked according to Ellis, is a strategy used by poorer Blacks in the United States,
you now to describe what you had been doing for the last two minutes, you would observing behavior among other groups of people could test the hypothesis that
probably say that you had been reading. You might describe as well the position they use the strategy as well.
you are in-sitting, lying. You probably would not say that you were leaning When you observe behavior, you soon become aware of repetitive activi
forward or backward and that you had just turned the page, although to design a ties in identifiable places-what Barker calls "standing patterns of behavior"
comfortable library these details may be important. (1968). Place-specific activities within such a pattern are more closely related to
Because observing behavior can be intensely personal, trained and sensi one another than to patterns of activities in other places-for example, the set of
tive researchers able to perceive relevant nuances can use the method more activities in a drugstore connected to ordering, making, drinking, and paying for
fruitfully. Being on the spot allows researchers to adjust their observations to a an ice cream soda are more closely related to one another than to those activities
particular setting and to a refined understanding of the situation. Whyte's person which constitute getting a prescription filled-although there may be no precise
al research capabilities are evident in his participant-observation study Street boundary defined between the two places where they occur. Training helps
Corner Society (1955). His day-to-day involvement with a street gang enabled observers to identify sets of activities that are closely related to one another, to
him to uncover more than ordinary evidence. identify significant patterns, and to distinguish significant patterns from unimpor
Whyte noticed, for example, that one gang member, Alec, regularly tant ones.
bowled higher scores than gang leaders during the week. But when the whole For example, in doing research on bank design, an observer might watch
gang bowled together on Saturday night, their scores paralleled the gang's hierar customers make bank transactions, from filling out slips at the desk to getting or
chy. The leaders bowled the highest scores, while Alec came in last. depositing money at the window. It is easy to overlook parts of the sequence that
When a "follower" was bowling too well, his companions would heckle occur before clients enter the bank or after they leave the teller's cage: seeing that
him, saying such things as "You're just lucky!" and "You're bowling over your documents are in their pockets and that money is safely put away into a purse.
head!" When Doc, the leader, bowled poorly, they would shout encouragement, Does the security guard standing watch consider himself part of every transac
telling him he could do better. Whyte noticed that gang members exerted subtle tion? To look carefully at events, observers continually question whether they see
-and not so subtle-social pressures on one another to conform to the hierar the whole event, whether they see all the participants, and whether something
chy. He was able to make this insightful observation on what sociologists call significant has been missed.
"social control" because he had many opportunities to observe general and speci Observers in dynamic research situations can test their hunches on the spot.
fic gang behavior and could adjust his observations to each situation. An observer who believes she has detected a regularity can try to predict what the
next few persons will do and can revise or refine the hunch right away, depending
Dynamic on how these persons act. Instant feecback like this enables researchers in the
beginning of a study to test many hunches, quickly identifying the more fruitful
As you look at people doing things, what you see changes: activities affect research ideas.
other activities; episodes take place. You get a glimpse of the role of time in the The more explicitly predictions and tests are made in notes and reports, the
life of an environment: a mother leaning from her window calling her child to more you can use team members to check your interpretation. Writing down
supper, the child coming. More complex chains of events are exemplified by a predictions and tests also helps observers avoid the trap of thinking that false
116 CHAPTER8 OBSERVING ENVIRONMENTAL BEHAVIOR 117
starts have really been well tested and enables them to review their own work OBSERVERS' VANTAGE POINTS
later with a clearer mind.
Observers can choose to be outsiders or participants in any situation. As
Variably Intrusive outsiders, they may be secret observers or recognized ones; as participants, either
marginal or full.
Researchers have to decide how far they will intrude and from what social
and physical vantage point they want to participate in observed events. At one Secret Outsider
extreme they can choose to record and observe behavior unobtrusively from a
distance-for example, with a telephoto lens. In addition to possibly creating The distant observer unobserved by participants in a natural setting is a
ethical problems, observing in this way removes the observer from the scene of secret outsider. Moore (1973) initially chose this vantage point for a study of
action, depriving the method of a large part of its research potential. However, children's play at an elementary school in Berkeley, California. School officials
close participation increases the chance of unwittingly affecting the observed replaced half an acre of blacktop with dirt that children could dig in and objects to
situation. Choice of vantage point depends on such things as research problem, play with, such as timber, aluminum pipe, and tree stumps. For five months,
available time, and investigator skills. every week at the same time, before, during, and after the change, Moore
To offset research bias resulting from their presence, participant observers climbed to the roof of the school and recorded what the kids did, using time-lapse
adopt social positions with which people are familiar. In a hospital, this could photography. He chose this vantage point so he would not alter their behavior
mean sitting in the waiting room like a patient; in a restaurant, working as a with his recording equipment until he showed them the film and because he
waitress or being a customer. To be able to take account in data analysis of thought this would enable him later to analyze patterns of use. He found, howev
changes they themselves induce, observers record any incident in which people er, that by choosing to record only an overview of the playground, he missed
may be reacting differently because the observers are present in their adopted what individuals did over time and any indications of depth of personal involve
position. For example, patients in a waiting room may be whispering because ment in what they did. To catch some of these dynamic attributes of his topic, he
another patient (the observer) is waiting too. The more crowded a setting-for took the camera down to the ground, becoming a recognized outside observer.
example, a rush-hour subway platform-the less observers' actions affect the
situation. Recognized Outsider
Of course, intrusion may be part of the research project's design. For
example, the observer has the ability to change situations and watch results, as When Blau (1963, 1964) compared two job-placement offices, he intro
Lefkowitz et al. (1955) did in their natural experiment, mentioned earlier, on duced himself as a researcher to those who were to be observed, explained his
pedestrians' reactions to differently dressed jaywalkers. Felipe and Sommer study, and was given a desk by the department head to work at and observe from.
(1966) used themselves as both observer and stimulus to test a personal-space A pitfall of such a recognized-outsider position is what is known as the
hypothesis that people get uncomfortable enough to leave if their personal-space Hawthorne effect-that subjects who know they are being observed as part of an
norms are broken. Observers sat very close to students in a library and compared experiment often change the way they act. The Hawthorne effect derived its
the time before the students left with the time before another student across the name from the now-classic environmental experiments at the Western Electric
room moved, whose space was not invaded. The same natural-experiment Company's Hawthorne Plant in Chicago, where Roethlisberger and Dixon (1939)
approach to observing behavior can be taken by moving furniture, erecting signs, wanted to determine, among other things, how lighting levels affected workers'
or changing an environment in some other way. Natural experiments are an productivity. They carried out their studies as recognized observers. When they
example of artificial intervention made possible because observing behavior is raised light levels, production increased. When they lowered light levels, produc
such a variably intrusive research method. (The students next to whom Felipe and tion increased also. They concluded that consciously being under a microscope
Sommer sat regularly moved away first.) changes workers' behavior.
You can try to minimize the Hawthorne effect by spending enough time at
In sum, observing behavior is both empathetic and direct, deals with a your research site that people there get used to you and take you more for
dynamic subject, and allows observers to be variably intrusive. These qualities granted. Observers can develop tasks for themselves to do while observing so
make the method useful at the beginning of research to generate hunches, in the that people begin to see them as other people with something to do. Whatever
middle to document regularities, and late in a research project to locate key observers do, there will always be the danger of some Hawthorne effect, which
explanatory information. must be recognized and considered during data analysis.
118 CHAPTERS OBSERVING ENVIRONMENTAL BEHAVIOR 119
Another problem for recognized observers is that no matter how honestly before. Such an attitude dulls the observer's ability to be surprised by what she
and convincingly they present themselves, their study, and their ethical commit sees-an ability crucial if research is not merely to record the obvious.
ment to respect privacy, someone may not believe them. Observers can exacer An observer who is familiar with her vantage point can also be misled into
bate this problem by oversight. Blau obtained permission to study the placement assuming that she knows how others in a setting feel about being watched. For
offices from the department head. The staff members therefore assumed that Blau example, the marginal observer assumes when watching an informal football
would report everything he saw to their boss. This was a mistaken, but not game in the park that he is taken to be a casual spectator. Meanwhile, the football
surprising, interpretation. Observers need to avoid giving off clues that they are players think he is a park attendant about to tell them to stop playing on the grass.
partisan watchdogs. They must remain as unaffiliated as possible by being care To increase the validity of their research, observers must test their assumptions
ful about who introduces them, where they sit, whom they have lunch with, about how they are perceived by others. For example, observers can slightly
whose office they use to make phone calls from, and generally from whom they change their natural behavior to see how people in the situation respond.
accept favors. Ways to control unwanted side effects include deliberate choice of cloth
Sometimes you cannot help being a victim of natural institutional mistrust, ing, physical posture, and objects one is carrying. Researchers observing in
particularly when you are interested in informal uses of physical settings. Wel Harvard Yard will be seen very differently if they carry green bookbags than if
fare recipients with relatives staying over in the living room, students smoking in they carry leather attache cases. One useful trick is to use one's behavior
the school bathroom, teachers making private calls from an office phone, patrol recording device as a prop to indicate a familiar, yet inconsequential, participant
men resting in coffee shops between emergencies are worried about being caught position: camera for tourist, notebook for student, sketchbook for amateur artist.
by someone in authority. In such situations, subjects tend to fear that researchers In general, being a marginal participant observer requires the least amount
are spies-perhaps tax inspectors or school administrators. Subjects normally of research preparation time. But precisely for this reason it requires that ob
play along with the "spy," feeding him harmless information but not admitting servers be introspective and self-aware.
the mistrust they feel. The more researchers explain their harmlessness, the
guiltier they seem. To reduce the effects of mistrust on the validity of the
research, observers must sensitively record situations in which mistrust is likely Full Participant
to have changed behavior. They can also make a special point not to ask ques
tions about rule-breaking activity clearly irrelevant to the study problem. To observe behavior, researchers can use positions they already are in and
Secret- and recognized-outsider vantage points both have disadvantages positions they adopt central to the situation they are studying. Full participants in
along with their advantages. Secret observers are by definition distant and re a study of housing design might be residents of a neighborhood. A study to plan
moved from the action. Their position also raises ethical questions. Recognized an office might be helped by researchers taking jobs as office clerks and typists.
observers may affect action in unknown ways. Participant observation by a waitress would have been appropriate in an
E-B situation described by Whyte (1949). In 12 restaurants in which tension was
high between dining-room and kitchen staff members, he observed that when
Marginal Participant waitresses gave orders to the cooks in the kitchen, the cooks resented it. They
were higher-paid and resented taking orders from less-skilled waitresses.
Researchers who adopt the vantage point of a commonly accepted and Although they could not avoid communication flow in this direction, they could
unimportant participant want to be seen by actual participants as just another avoid taking orders directly. Tension was reduced when a clipboard was installed
patient in a hospital waiting room, another subway rider, or another art student in some restaurants on the counter between dining room and kitchen. Waitresses
drawing in a park. A marginal-participant vantage point is a comfortable one for put order slips on the clipboard. Whenever a cook decided to take the next order,
E-B researchers to adopt because observant professionals and laypersons adopt it he went to the board and picked up a slip. He put the plate back on the separating
naturally in daily situations. counter. He no longer took orders directly; the environmental change gave him
Marginal positions that observers choose are likely to be somewhat famil control over his own actions.
iar. We have all been bus passengers, members of the audience at a street In some cases researchers may not be able to choose full participation, as
concert, and restaurant patrons. Familiarity, however, can prevent observers when all participants are highly skilled professionals (doctors in a hospital) and
from looking carefully at what is actually going on. It is tempting to assume that a when membership in the setting being studied is restricted (men's athletic clubs).
quick glance will tell you everything because, after all, you have seen it all Gaining full participant-observer status by taking up residence, taking a job, or
120 CHAPTERS Table 8-1 (continued)
joining an organization usually means making a long-term commitment. Return Police arrive with stretcher. Why do they announce it? For nurses
on investment potentially comes in the form of an insightful and empathetic Announce in loud voices that they to clear a path?
position from which to gather behavioral data. have a woman who fell down and
passed out. She is lying still on
stretcher with eyes closed, covered.
RECORDING DEVICES All other patients sitting in corridor Patients looking again! ls it just some
lean forward in chairs to look. The thing to do?
stretcher will not fit through corri
Devices suited to recording behavior observations include verbal descrip dor where patients are sitting.
tions and diagrams, precoded checklists for counting, floor plans or maps, still Police struggle to maneuver stretch Hallway waiting causes traffic prob
photographs, and film or videotape. What devices to choose depends mainly on er through the crowd of nurses and lems for stretcher cases.
how much detailed information the problem demands and how much the observer doctors in the nursing station to get
already knows about the behaviors to be observed. to uncrowded corridor on other
side. Patient is put in examining
, room. Curtain pulled part-way
Notation closed by last policeman to leave.
Patients waiting in corridor have This probably bothers patients being
Recording behavior in verbal and diagrammatic notes demands that observ full view of patient in exam room. examined.
ers decide what to describe and what to overlook on the spot. For example, in A policeman wheels stretcher out
describing how people use a hotel lounge, the observer must decide whether to back door into middle of waiting
record how people meet each other and move around, how people sit and watch area, while another tells a nurse the This public discussion surely seems
details about the woman they like an invasion of privacy.
others, how they hold their newsp&pers and shift their weight, how they move brought in, leaning over counter at
their eyes and twitch their noses. Each level of analysis is useful to design nurse's station.
researchers for solving different problems. Each individual observer decides on Nurse leaves nurse's station, walks Nurse in her "station" cannot see the
and then isolates that level of analysis particularly relevant to his or her own around counter into corridor, scans informal or overflow waiting area in
study. If multiple observers work on the same research project, they must be all patients waiting there. She corridor. What are the design impli
trained and sensitized together, comparing their observations so that each knows walks up to one man who is seated, cations? Behavior of nurse in telling
what types of behaviors to note. That well-trained observers make decisions stands three feet away and tells him lab results is another type of invasion
the results of lab tests and what they of privacy.
about levels of analysis can be an opportunity to see richness in a situation and mean. Doctor walks over and asks
catch that richness in discrete notes. same patient to go into exam room
Procedures for descriptive behavioral notation are relatively simple. Notes with him.
are recorded by researchers working alone or by one team member when the Doctor's voice, shouting angrily, What acoustical control is needed in
other member is conducting an interview. As with notes of physical traces, it is comes from an exam room. exam rooms?
useful to crease a note page, creating a wide right-hand margin. When observa Doctor leaves nurse's station, Consultation in waiting areas may be
tions are written in the left-hand column, the right side is open for individual or approaches woman waiting in standard emergency-room procedure?
group analysis. Table 8-1 shows a sample of field notes. wheelchair, pulls up a chair, sits ls there a way to allow this to take
down beside her, and talks in low place but provide more privacy?
tones. Other patients sitting nearby
watch and occasionally speak to
Table 8-1. Sample field notes from site visit to hospital emergency room. (Ob each other.
servations made from nurse's station at 1:00 p.m.) Does this perhaps relax people in
Sound of friendly chatter, laughing
from one exam room. waiting area?
Observation Comment
Woman waiting in wheelchair has Does watching emergency activity
been waiting in corridor between make waiting easier?
nurse and row of examining rooms Field notes by architect/researcher Polly Welch for "Hospital Emergency Facilities: Trans
since at least 10:30. She is watch lating Behavioral Issues into Design," by P. Welch. (Graham Foundation Fellowship Report.)
ing all the activity. Cambridge, Mass.: Architecture Research Office, Harvard Graduate School of Design, 1977.
121
122 CHAPTERS OBSERVING ENVIRONMENTAL BEHAVIOR 123
Several small tricks help avoid embarrassing mistakes in descriptive be office workers eating in an open-air plaza. Looking at behavior recorded on maps
havior notes: always include yourself in observations to avoid finding out that a can give investigators a better sense of how a whole place is used at once than
crucial observed behavior actually was a response to the observer's presence; looking at statistical tables.
when sitting and taking notes in public, make a drawing on the top page of the Maps are also useful to record sequences of behavior in settings where
notepad so that anyone who looks over your shoulder will find an acceptable people have a choice of several paths: from home to bus stop, from desk to desk
sketch; never leave notes around. What to a researcher are harmless descriptions in an open-plan office. Analyzing map records in the light of an actual setting can
of the obvious, to participants can be highly insulting snooping. give an idea of the characteristics of popular paths.
If investigators want precise physical-location data, they can construct base
Precoded Checklists maps with grids corresponding to regular elements in the actual setting, such as
floor tiles or columns.
Descriptive notes provide a qualitative understanding of what is going on:
what types of behavior patterns there are, what characteristics of participants are
Photographs
salient, and what level of descriptive abstraction is appropriate to solve a prob
lem. If researchers want to know in greater detail how often an activity takes Still photographs can capture subtleties that other methods may not record:
place, they can use qualitative observation data to develop a precoded checklist the way someone sits on a chair or leans against a column; the way two persons
for counting. The qualitative approach serves in such situations as the diagnostic avoid looking at each other by adjusting their body postures. In addition, as
phase of the research project. presented in Chapter 7, photographs are useful throughout a research project
In their study of behavior on a psychiatric ward, Ittelson, Rivlin, and because of their illustrative quality. The same procedures hold for deciding on
Proshansky (1970) recorded over 300 descriptions of behaviors during extended photographs to record behavior as were described for using photographs to record
periods of time. For example, patient reclines on bench, hand over face, but not physical traces.
asleep; patient cleans table with sponge; patient plays soccer in corridor; patient
sits on cans in hall watching people go by. For counting purposes, they coded the
Videotapes and Movies
descriptions into categories representing types of activities observed, such as
lying awake, housekeeping, games, and watching an activity. Whenever time is a significant element in an E-B problem, motion photog
For each activity on a checklist, observers record characteristics of partici raphy-videotape or movies-ought to be considered. For example, urban
pants (alone or in groups), place, time, and other relevant conditions, such as the design of streets for handicapped and older people demands understanding their
weather. Perhaps the most significant task in developing a checklist is specifying pace: how fast do they move, how long can they move before resting, how fast
the descriptive level of abstraction to record. Ittelson et al. decided, for example, can they move out of other people's way? To design a safe escalator, it is
that activity types (housekeeping, personal hygiene) were more relevant to their essential to know how different types of people approach it, prepare to get onto
problem than activities were (cleaning a table with a sponge, setting one's hair). it, and embark (Davis & Ayers, 1975).
Rather than describe subjects in terms of approximate age, sex, weight, and
height, which might be relevant to a study of children's play equipment, observ
ers in the psychiatric ward coded sex of subject, whether acting alone or in a WHAT TO OBSERVE
group, and, if in a group, of what size and sex mix.
To set up a checklist demands previous diagnostic observation, a thorough Observing behavior looks like a simple E-B research technique. Everyone
understanding of how the data will be used, and an understanding of how to watches people every day. Doesn't everyone know how to do it? In a way, yes;
develop coding categories. Once a precoded checklist is set up, it provides but few know what to look for and how to analyze what they see so that it is
relatively comparable quantifiable data with only a moderate amount of training useful to design.
for observers. Designers make places for people to do things in-either alone or together
with other people. A structure for looking at environmental behavior useful to
designers results in data to help physical designers make decisions that improve
Maps
places for people. The better information designers have about how the people
Recording activities on floor plans, diagrams, or maps is particularly con they design for behave in physical settings and how those people relate to or
venient if researchers want to observe and analyze several people in one general exclude other people, the better they can control the behavioral side effects of the
area at the same time: groups at a cocktail party, patients in a waiting room, design decisions they make.
1 24 CHAPTER 8
But that is not enough. Designers mw,t also know how the contexts of
observed activities affect the activities, because in different sociocultural and
physical settings the same behavior can have different design implications. For
example, children may do homework at the kitchen table for different reasons in
a house with several available rooms to study in than in a one-bedroom apartment
where four people are living. In some groups people react to neighbors sitting on
the front stoop with disdain, while for others the front rather than the back is
where everyone sits.
When you structure the way you look at something, you replace complex
reality with a simpler version to guide your reactions and action. To increase our
control over the behavioral side effects of design decisions, we can describe
behavior in terms of actor, act, significant others, relationships, context, and
setting (see box).
Who is Actor
The following illustrations are verbally annotated to show how you can use
these observation categories to describe environmental behavior in actual situa
tions.
Each observation comprises a relationship between an actor and a signifi
cant other to which the physical setting in some way contributes.
125
126 CHAPTERS OBSERVING ENVIRONMENTAL BEHAVIOR 127
Who: Actor the front doors of a school while five others look on from a bench nearby and one
gets ready under a tree to play, it would be misleading to write in one's field
The subject of a behavioral observation, the "actor," may be described in notes: "A group of ten boys are playing street hockey at the school entrance." To
numerous ways, depending on the purpose of the description. Designers can use design a place to play street hockey, the relationship among players, spectators,
research in large design projects to better understand similarities and variations and reserve players is relevant.
among types of people. For example, instead of designing a school for 273 A group of two also raises problems for observers: are they a group acting
unique individuals, a designer can use research to differentiate the needs of together with common significant others, or do they themselves represent actor
students, teachers, principals, and maintenance workers. Nursing homes can be and significant other for each other? If they are very similar and are doing the
planned for patients, nurses, doctors, maintenance crews, and visitors; furniture same thing, it may be appropriate to describe them together: two boys playing
can be designed for the range of people who work in offices. In a sense, indi street hockey with each other, two elderly men playing chess, two women
viduals in observations are treated as representatives of a social group. walking down the street together. However, when the couple is made up of two
We can use individuals as such representatives by describing a person's different types of individuals interacting, it may be useful to describe them
social position or status: age status, marital status, educational status, profession separately, seeing one of the two as the actor in the observation: parent and child
al status, and so on. It helps to be complete in observations if we describe both a in the park, nurse and patient in a hospital. But even here, as with all descriptive
person's ascribed statuses (the characteristics that a person has automatically, observation techniques, the researcher's judgment is the most significant deter
such as sex and age) and his or her achieved statuses (those that the person had to minant of what is important to describe.
do something to get, such as finding a job, graduating from college, getting
married, or inviting people to a party). Many positions are defined as part of a
relationship to others: party hostess (guests), wife (husband), teacher (student),
nurse (patient), salesperson (customer). Doing What: Act
An observer unable in field notes to describe statuses accurately can de
scribe clues from which he and other researchers reading the notes may be able to The people you observe will be doing something. An observer needs to
infer status. For example, Snyder and Ostrander, in their Oxford nursing-home decide the level of abstraction he will use to describe behavior and how he will
study (1974), observed people who were patients, family members, visitors, and distinguish individual acts from a connected sequence of acts.
staff members. After a few days they knew most individuals personally or could The level of description observers choose depends mainly on the design
infer their status from such things as dress (uniform means nurse; bathrobe means and research problem facing them. Let us take as an example an observational
patient) and tools (stethoscope means doctor; sitting in wheelchair means pa study to write a behavioral program for a shopping-center design. Observers
tient). But when they were not sure, they described in their field notes whatever could describe very generally that some people there are "shopping" and others
clues they had and whether they were guessing about the person's status. It is are just hanging around. More precisely, they can describe that some shoppers
better to record "It could be a nurse's aide resting in the wheelchair" than to write browse, while others buy something. Or observers might record where and count
"It is a patient asleep in the comer," so that other researchers can help evaluate how often a supermarket patron stops in the aisles. Observers might record how
the data. high patrons reach and how low they stoop when getting items off the shelves. Or
Sometimes relevant descriptions of actors in behavioral observations are observers might go to the trouble to observe and record in what direction patrons
names of groups-teens, teachers, girls-not individuals. In Zeisel's property tum their heads and focus their eyes while walking down the aisle. Each observa
damage study (1976a) researchers observed groups of boys playing street hockey tion is either interesting or useless, depending on the problem researchers are
and stickball in open spaces around schools. It was not important for their trying to solve. The series of design questions in Table 8-2 shows how each level
research and design problem to identify each street-hockey participant as an actor of described activity might be useful.
in a separate act. Researchers treated the group as the actor, describing the Along with deciding on appropriate levels of analysis, researchers must
group's size and composition. Groups can be described in the same status terms explain how the acts they describe relate to one another. In the sequence of acts
as individuals. For example, the psychiatric ward study by Ittelson et al. (1970) called "shopping," a person prepares a shopping list, leaves home, goes to the
identified groups by the number of male and female patients, doctors, and visi store, looks at items in the store, reaches for them, examines them, walks down
tors they contained. the aisle, pays at the cash register, returns home, and unpacks. Each of these can
One pitfall for observers to avoid is subsuming significant individuals be seen as a discrete act linked to the others as part of a larger "shopping"
under general group descriptions. If four teenagers are shooting hockey pucks at sequence. If researchers observing behavior maintain clarity of descriptive level
128 CHAPTERS OBSERVING ENVIRONMENTAL BEHAVIOR 129
Table 8-2. Behavior descriptions and corresponding questions for a shopping With Whom: Significant Others
center design, by level of detail
Acts people engage in are partly defined by how other people are or are not
Behavior Observation
included. Other people whose presence or absence is significant in this way can
Design Question
be seen as participants in the act itself. Girls for whom boys playing street hockey
General "Shopping" as opposed In a shopping-center plan, show off make the activity what it is. If they were not there, it would be another
Description to "hanging around" how many places are needed situation. The same is true in reverse for studying alone in the library. Those who
for people to hang around, are not there-friends, roommates, strangers-contribute to the situation by
and how can they be designed
to augment rather than inter- their absence. To understand and present what is going on, descriptions of girls
fere with shopping? watching the boys and of absent roommates must be included in research obser
vations of behavior.
Shoppers browsing as How should items be dis- "Significant others" are especially important in environmental design re
opposed to buying played so that browsers and search because so many design decisions about adjacencies, connections, and
something buyers can see them but
buyers have greater access separations have side effects for relationships. To continue one of our earlier
to them? examples, boys playing street hockey need a hard, flat surface to play on. If this
surface is provided for them in the middle of a deserted field far from other
Where and how often How can flooring materials, activity, it is unlikely to be used, because the "significant others," the girls and
shoppers stop in lighting, and aisle length be passers-by, have not been taken into account. A tot lot with no places for parents
supermarket aisles designed for maximum conven-
ience to customers, maximum to sit and watch may go unused in favor of a more convenient one or will be used
exposure of sales items, and in a different way than the designer had hoped.
minimum maintenance? The positions or statuses by which actors are described often have standard
role relationships associated with them. In a family, for example, one finds role
How high patrons will What shelf design and what relationships between parent and child, sister and brother, husband and wife,
reach and how low product placement (what size
they will stoop container on what shelf) will grandparent and grandchild. In hospitals there are role relationships between
ensure that customers have the doctor and patient, doctor and nurse, patient and nurse, patient and visitor, nurse
easiest time reaching items? and visitor, patient and patient. A sensitive researcher observing a doctor making
notes in a hospital will use the concept of significant other to direct attention to
Detailed Where customers' Where should standard signs be the relationship the doctor making notes has set up between herself and patients,
Description eyes focus while placed to convey the most in-
moving down an aisle formation, and where ought sale nurses, and other doctors. Does she sit among patients in the waiting room, or
signs be located to catch cus- does she retire to a private lounge? Does she discuss notes with nurses or just
tomers' glances? hand them in? To design appropriately for notetaking in hospitals, the answers to
these relational questions can be important.
and completeness in describing related acts, they will be able to analyze their data
more easily. Relationships
I have stressed the skill that observers need to decide how and what to
describe. It is equally important that they have the ability to describe what they Between actors and significant others in a situation there will be specific
see with minimum interpretation. Well-recorded observations leave ample time relationships for observers to describe. In extreme cases relationships can be
and space for analysis after data have been collected. If observers try to interpret described simply: "together" (two lovers on a park bench at night) or "apart" (a
what they see before writing it down, they run the risk of recording interpreta prisoner in solitary confinement).
tions rather than description, losing the data for good. The data cannot be re Most E-B relationships, however, are not so simple. Are two persons
trieved to be analyzed by others or reviewed later. If data on behavior are to be talking to each other through a fence together or apart? What about two persons
sharable, it is vital that observers record "a smiling person," not "a happy sitting back-to-back in adjacent restaurant booths? The problem researchers face
person," because a smile can mean many things. is to systematically describe relationships like these so that differences and simi-
130 CHAPTER8 OBSERVING ENVIRONMENTAL BEHAVIOR 131
larities between two situations are clear. Then researchers and designers can use Context
the information to develop broader strategies for design rather than continually
People react to other people differently in one situation than in another and
approaching each situation as totally new. To gather such information, research
differently in one culture than in another. It is as if they filtered what they saw
ers need to agree on a set of categories to describe connections and separations
through a series of screens-situational and cultural. The screens are usually
between people, and they must understand how the effects of relationships on
used unconsciously, as Sommer ( 1969) and Hall ( 1966) have pointed out. People
activities differ in different behavior settings.
assume that other people see things the same way they themselves do. It is the
Hall ( 1966) shows us that behavioral connections and separations between
observer's job to identify how people's situational and cultural screens are con
people in environments can be conveniently and efficiently described in terms of
structed-how they interpret their own and others' behavior.
four physiological senses and a symbolic perceptual dimension: seeing (visual),
This is particularly important in environmental design research because the
hearing (aural), touching (tactile), smelling (olfactory), and perceiving (sym
meanings people attribute to relationships determine how they react to environ
bolic).
mental features, such as walls, doors, and lights, that affect those relationships.
Describing two people as completely together, or "copresent" (Goffman,
1963:17), means that, like two children in the bathtub, they can see, hear, touch,
Situations. A person's sitting alone and apart from others, facing a wall in
and smell each other, and they feel that they are "in the same place."
a library, probably means she wants to be left alone to read or study. In a bar, this
When we move away from extreme relationships, the sensory terms we
same physical behavior can be interpreted as an invitation for conversation (Som
have for describing relationships enable us to discriminate among and compare
mer, 1969). The person might still reject the advances, but she is unlikely to be
various types and also to begin to identify the role that the physical environment
distressed and insulted, as the person disturbed in the library may be.
plays in relationships between people. A mother on the third floor calling to her
An extreme example of how a situation can influence the meaning people
child playing on the street is connected visually and aurally but is separated in
attribute to behavioral relationships can be seen if you watch people's shocked
terms of touch, smell, and perception. Two students studying at opposite ends of
reactions when you talk in a normal voice to a friend over the hush in a crowded
a long library table are separated symbolically and in terms of smell and touch but
elevator. In a department store, a market, or a crowd viewing a parade, your
are connected visually and aurally. Persons in an L-shaped living room, around
voice would not even be noticed. In an elevator, however, the definition of
the corner from someone cooking in the kitchen, are separated by sight, touch,
personal space is different, and so are the definitions of unacceptable behaviors.
and perception but are connected in terms of food smells and sound.
An observer must try to understand the situational rules being applied by partici
pants to interpret the meaning they attribute to even a simple observation such as
"Two persons stood next to each other talking."
Culture. Cultural context also influences how people interpret and react to
behavioral relationships. For example, Hall (1966) reports that in England sitting
alone reading in a room at home with the door open means "Do not disturb; do
not even knock." In the United States you would close your door to indicate you
wanted to be alone; an open door means you are available. It would not be
inappropriate for people to knock on an open door and ask whether they might
come in. An interior designer laying out open offices in these two cultures needs
to be aware of these differences if he wants to control the behavioral side effects
Simultaneous connections and separations of his physical design decisions.
It is particularly important to record cultural contexts for behavior when
When observers see and can describe relationships like these, they try to you carry out observational studies in another country, in ethnic neighborhoods,
find out what the relationships mean to participants. Although they must use or in parts of your own country witl, strong regional differences. Otherwise,
other research methods as well to determine meaning, behavior observation designers using your data will be making decisions irrelevant to users. As in Le
provides clues to meaning. The clues are the ways people react when other Corbusier's Chandigarh, people may end up cooking on stoves on the floor in
people talk to them, touch them, and so on. efficiency kitchens and establishing illegal street markets in the plazas in front of
132 CHAPTERB OBSERVING ENVIRONMENTAL BEHAVIOR 133
modern government buildings (Brolin, 1972). To see behavior from a cultural quality of separation. For example, walls with no soundproofing between bed
perspective other than one's own requires general observation and study of rooms provide neighbors with aural opportunities (and inhibitions) that denser
another culture, awareness of one's own cultural biases, and at times requesting walls do not.
members of or experts on another culture to help interpret behavioral data once Screens-glass panels, a garden hedge, a shower curtain, doors, counters,
they are collected. As the basis for this interpretation, it is necessary to describe windows-separate and connect people more selectively than complete walls.
as fully as possible people's reactions to relationships they find themselves in. Glass can enable visual connection but tactile separation; a shower curtain, the
opposite. Materials can be combined to provide different degrees of connection
Setting and separation along any mix of dimensions. Screens can also be designed to give
selective control over the screen to users. For example, the lock and bell on a
The meaning of behavior in a particular setting depends on the potential of glass-paneled house door provide a range of permeability options for family
the setting for use-the options it provides (Gans, 1968). If people in an airport members, friends, and thieves (Hoogdalem, 1977).
waiting lounge are sitting on the floor surrounded by empty seats, their behavior
may have a different meaning than if no seats are available. Understanding
participants' choices and possibilities to act helps you interpret what they finally
choose to do.
i��·::
Relational design decisions. Barriers clearly determine potentials for re
lationships between people in settings. Barriers include walls of various materials
and consistencies, screens in different sizes and materials, objects used to mark
the edges of places, and symbols from color changes to verbal signs. Design
decisions defining.fields in space influence behavior relationships less obviously.
Field definitions include such characteristics of places as shape, orientation, size,
and environmental conditions-sound, light, air.
To define the ways these physical characteristics affect relationships be ��-- 1.
tween people, we can use the simple relational scheme developed earlier: seeing,
hearing, touching, smelling, and perceiving.
Barriers. Barriers are physical elements that can keep people apart or join
them together on one or more of the five dimensions-seeing, hearing, and so
An object, here a column in a shared interior porch, can help people divide
on. As one progresses from walls to symbols, barriers become more permeable. space perceptually. (Congregate House for Older People. Design-research
Walls separate people in places. The absence of walls allows people to be team: Barry Korobkin, John Zeise!, and Eric Jahan. Donham & Sweeney,
connected. The thickness, consistency, and materials of walls influence the associated architects.)
OBSERVING ENVIRONMENTAL BEHAVIOR 135
134 CHAPTERS
Finally, symbols can be barriers. Color changes in the rug around a public
telephone and cha�n ceiling height in a room signal that someone considers
this space to be two separate places, perceptually.
Depending on how people interpret spatial symbols, they may change their Functionally Functionally Functionally
behavior: not walking too close to the phone caller because of the floor color, distant nearer close
calling one part of a room by another name because of the shift in ceiling height.
Symbols can also be overt signs: "Do Not Walk on the Grass" potentially Degrees of functional distance
keeping people off; "Open for Business" potentially bringing people in. Sitting
on the grass near a "Keep Off " sign conveys another impression to observers than stairway forces residents of apartment 6 to pass apartment 1 whenever they come
if there is no such separator. or go. But people living in apartments 2 and 7 can leave home and return without
ever running into one another. As the hypothesis of Festinger et al. leads them to
Fields. Field characteristics of an entire place can alter people's ability to expect, residents in the functionally closer pair, 1 and 6, selected one another
be together or apart. Field characteristics do this not by standing between people, more often as friends than did residents in apartments 2 and 7.
like barriers, but by altering the physical context within which visual, aural,
tactile, olfactory, and perceptual relationships take place. Field characteristics of
places include their shape, orientation, size, and environmental condition.
The shape of a setting affects primarily visual and perceptual relationships.
If people want to, they can use the cues that shapes provide to consider areas
within one space as separate places. Comers in a square area, for example, can be
more easily seen as separate from one another than parts of a round place can. In
a study of children playing in different rooms, groups of children quickly claimed Schematic diagram of a building in Festinger' s dormitory study. (Reprinted
from Social Pressures in Informal Groups, by Leon Festinger, Stanley
as distinct territories the places in the leaves of clover-shaped rooms (Hutt,
Schachter, and Kurt Back, with the permission of the publishers, Stanford
1969). University Press. Copyright 1950, renewed 1978 by the Board of Trustees of
the Leland Stanford Junior University.)
Effects of shape
from the rest of the group. In the main hall of New York's Grand Central Station,
the same people could easily be dispersed.
Chapter 9
Loudness, light intensity, and air flow are environmental conditions that
directly affect possibilities for behavior relationships by limiting and augmenting
people's ability to hear, see, and smell other people-1ifil:t-activities. For example,
light turned low in a restaurant effectively separates people at different tables as if
there were a physical screen between them. A single worker in an open-plan
office listening to a radio at high volume acoustically invades the space of other FOCUSED INTERVIEWS
workers and separates himself from them aurally. Machines that emit high
pitched sound and mask background noise without participants' awareness pro
tect acoustical privacy as a closed door might. An exhaust hood and fan over a
kitchen stove keep kitchen smells out of adjacent rooms-olfactorily separating
people cooking in the kitchen from others. Asking questions in research means posing questions systematically to find
out what people think, feel, do, know, believe, and expect. Normally when we
think of an interview or a questionnaire, we think of the yes/no or multiple-choice
OVERVIEW questions of most public opinion polls. But such questions are fringe forms of a
research tool of potentially much more penetrating power. You can use a focused
To design environments suited to what people do in them, we must under interview with individuals or groups to find out in depth how people define a
stand environmental behavior: Who does what with whom? In what relationship, concrete situation, what they consider important about it, what effects they
sociocultural context, and physical setting? This chapter proposes that by looking intended their actions to have in the situation, and how they feel about it.
at how environments affect people's ability to see, hear, touch, smell, and Originally formulated to tap reactions to films of military instruction and prop
perceive each other, we can begin to understand how environments impinge on aganda, radio broadcasts, and other mass communication devices, focused inter
social behavior. views are particularly suited to the needs of environment-behavior researchers
Environmental elements that affect relationships include barriers, such as interested in reactions to particular environments. Many of the concepts this
walls, screens, objects, and symbols; and fields, such as shape, orientation, size, chapter explains and the way it explains them are based on Merton, Fiske, and
and environmental conditions. Design decisions about these elements have iden Kendall's insightful and inventive book The Focused Interview (1956).
tifiable side effects for social behavior.
Environmental-behavior descriptions that can enable designers to improve
control over behavioral side effects of their decisions include six elements: actor, PREINTERVIEW ANALYSIS AND INTERVIEW GUIDE
act, significant others, relationships, context, and setting.
The next three chapters discuss how to find out about people's feelings, To understand thoroughly how someone reacts to a situation, one must first
attitudes, perceptions, and knowledge-namely, by asking questions. analyze the structure of that situation, using theory and observational research
methods. This analysis can then be used as the basis for discussing the situation
in detail with the respondent. Such a situational analysis guides the discussion;
the interviewee's responses are used to test, refine, and modify the analysis. A
skilled focused interviewer negotiates with a respondent to find correspondence
between his own analytic structure and the respondent's mental picture of the
situation. By structuring the information themselves, focused-interview respon
dents become participants in the research.
The interview guide is a loose conceptual map, such as a family might
draw up before taking a cross-country camping trip. It lays out major sights to
see, places to stay, and so on. After the trip begins, the family members find
some of the sights closed, others uninteresting, others so arresting that they stay
longer than expected. They also find that they do not drive as many miles as
planned each day and that the children like to stop to eat more often. Every day
they adjust their plans, and they end up having a fine trip that mixes the plans
1 8 HAPT R
they made on the basis of advance analysis with reactions to events as encoun• Str ngth of R spond nt ' F Ii ngs
tered. Skilled focused interviewers similarly modify their original plans to corres
Throughout any design project, decisions about priorities are made. Is it
pond to the conceptual map reflected in the respondent's answers. That concep more important to plan direct access to cars from apartments or to keep cars
tual map is the respondent's definition of the situation for which the interviewer
parked far from the front door? Is it more important for patients in a cancer
is searching.
treatment center to wait with relatives, or is modesty more important for them,
In the focused-interview guide, the map is a set of topics, elements, pat•
maintained by waiting alone? Designers forced to make such tradeoffs can better
��s, and relationships that the interviewer tentatively intends to cover. Adjust
control the side effects of their decisions if they know the strength of respon
ments to the guide during the interview are carried out by skillful use of the major
dents' feelings about convenient access, a view free of automobiles, relatives'
focused-interview tool, .the probe: the interviewer's prompting for further elabor
support, and modesty.
ati�n _of an answer. An interviewer probes to find out how a respondent•s·definr.:·
tion of the situation differs from the hypothesized one; this information allows the
interviewer to adjust and refine the guide. The researcher's goal is to determine
which of the many hypothesized elements are important to the respondent and Intentions
then to understand as thoroughly as possible what these elements mean in the.
respondent's definition of the situation. Observing behavior and physical traces tells investigators about unintended
To avoid misunderstandings, one should know that for surveys in which consequences of activities. In Boston's West End "urban village," men spent
questions are posed with prescribed rigidity, a "good interviewer" is one who much time on the street washing and polishing their cars. They said they did this
adheres to the text and never develops initiative of his own. In a focused inter to keep the cars clean, which the cars are. Observations showed that the men
view, the opposite is true. polished their cars next to one another and talked to passers-by as well, creating a
close-knit network of neighborhood friends. This social contact is another con
OBJECTIVES OF FOCUSED INTERVIEWS sequence of car washing in this neighborhood. Both consequences could be
observed in the situation. Only by asking the actors what their intentions are can
Definition of the Situation -�esearchers distinguish conscious intent from unintentional side effects.
I-
FOCUSED INTERVIEWS 141
140 CHAPTER 9
Probes are primarily questions that interviewers interpose to get a respon Probe Purpose
dent to clarify a point, to explain further what she meant, to continue talking, or
to shift the topic. The probe is the systematic development of an everyday device Addition Flow
used in conversation when one person is interested in precisely what another has Encouragement
to say. Body movement
Attentive silence
• Addition probes encourage respondents to_ keep talking-to keep the flow of the
interview moving. Reflecting Nondirection
• Reflecting probes determine in a nondirected way which of the analyzed topics in
the interview guide are significant to the respondent and which new ones to add Echo
because they were overlooked. Question-to-question
• Transitional probes make sure that the respondent discusses a broad range of Attentive listening
salient topics.
• Situational probes stimulate the respondent to specify what parts of a situation Transition Range
prompted the responses. Cued
• Emotion probes encourage discussion in depth of how the respondent feels about Reversion
each specified part of the situation. Mutation
• Personal probes get respondents to describe how the context of their lives influ
enced their reactions. Situation Specificity
Re-presentation
This chapter goes on to discuss each type of probe, showing with examples
Environmental walk-through
how it can be used to enrich an interview. Reconstruction
Example Comment Graphic re-presentations can be used together with cued transition probes or even
with mutation probes if interviewers want to find out respondents' reactions to a
Resp: I feel that if I don't always stay General response describing reaction to broad range of environments. Re-presentations then take on the added directive
aware of where I am, I'll get lost. environment. nature of these other probes.
Int: What is it about your office that Probe requesting specification of A special case of re-presentation, the environmental walk-through probe,
makes you feel that way? environmental stimulus. can be used if the focused interview takes place in the environment that is the
Resp: The windows. I can't see any Focused specification of environmental topic of the interview. During a walk-through the interviewer asks the respondent
windows from where I work, so I never stimulus. to point out and describe places and objects that are important to him. For
know what time it is or which direction particularly salient items the interviewer and respondent stop to specify more
I'm facing. precisely what it is about the item that is relevant.
A walk-through is not just a guided tour. To get the most out of an
environmental walk-through, interviewers first question the respondent in one
When interviewers repeatedly request specification of an environmental place, asking him to describe the environment they will walk through, together
stimulus, respondents may revert to mere description of the environment. Inter with his general reactions to it. As places and objects are mentioned, the inter
preting a request for specification as a request for information, they may proceed viewer discusses them up to the point of requesting detailed specification, noting
to try to remember as many details about the environment as they can-even these items for later reference during the walk-through. In this way, the inter
irrelevant ones. To avoid this pitfall and to elicit sufficiently specifying re viewer uses first the respondent's personal definition of the situation to define
sponses, interviewers can use probes aimed at helping respondents remember important elements and then the walk-through to elicit further specification.
clearly the settings they are asked to specify. Reconstruction probes may be used when respondents have trouble remem
Using re-presentation probes, interviewers present respondents with a bering the setting they are asked about or when they remember it only in general
photograph or drawi!!$��<�J!� of the setting being discussed-a doorway, terms.Reconstruction probes ask respondents to think back to particular events in
an area, a piece of hardware. This active probe is least directive when the picture a place to recall their reactions to it at the time the event took place.
is presented only after the respondent has verbally identified an element or place
as relevant to him. Int: When you first entered the hospital
three weeks ago, which entrance did you
come in?
Example Comment
or
Int: What in the school causes the most Request for general information about Int: What do you remember about the last
maintenance problems? problems. time you sat at your old desk, before
Resp: Well, we have the most trouble Mention of an object. moving to this office?
keeping the thermostats in order.
Int: (Presenting photograph of thermostat Re-presentation probe, combining When the respondent refers to a complex set of phenomena that she re
to respondent) Here is a photograph we photograph of object with request for members only as a whole or when she replies "don't know" or "can't remember"
took recently of the thermostat you use specification. after being asked to specify her answers, reconstruction probes often help switch
throughout the building. What is there attention to specifics.
about it that gives you the most trouble?
Example Comment
Resp: lf you look closely, you can see Focused response specifying aspect of
how flimsy the adjustment switch is. object that causes reaction. How do you feel about the park?
/111: General request for information on feelings.
When kids fool with the switch or even
Resp: I think it is a particularly good General report of feelings.
when faculty members try to adjust the
temperature, the switch often breaks off. place to come with my children.
This means we have to replace the whole Int: What makes it a good place for Specifying probe.
unit at$ ... hildrcn?
150 CHAPTER 9 FOCUSED INTERVIEWS 151
Example Comment Resp: No.In fact, they are actually Specifying response.
well-behaved if you talk to them. They
Resp: I don't know, it's just the way it's "I don't know" meaning "I can't just walk across the lawn where there is
planned. verbalize it." no path and sometimes throw rocks at the
Int: Well, do you remember the last time Reconstruction probe. lights.
you went to the park with your kids? Int: What do you feel about this? Feeling probe.
Resp: Yes, we played hide-and-seek on Response beginning to reconstruct Resp: I am very angry that they do not Depth response.
the curved pathways, and ... specific situations. obey the rules, but I am glad that the kids
respect our being old and that they stay
out of our front yards.
Reconstruction probes help respondents to look retrospectively at the situation
they are commenting on-to put themselves backward in time and reexperience A series of feeling probes can bring to the surface strongly felt sentiments
the setting. Specifying probes in general do more than isolate for analysis specific that appear at first to be peripheral, and it can show seemingly deeply felt
parts of a whole situation. By linking specific parts to specific respondent reac sentiments to be no more than offhand remarks. As a rule, no briefly expressed
tions, they set up the interview so that each reaction can be explored in depth. sentiment ought to be taken at f ace value until it has been probed in depth.
Another probe for depth of emotion is the projection probe, in which
interviewers ask respondents to project feelings about a situation onto another,
Emotion Probes to Increase Depth
hypothetical person. This is useful when discussing sensitive emotions that the
Depth in a focused interview is the degree to which the respondent's respondent himself might not admit having but would be at ease admitting that
feelings about a situation are explored. Reports that a respondent '•likes" or "others" or "someone else" might have.
"dislikes" a place, that it is "very satisfying," or that it is a "frightening" place
can signify a variety of things. Someone, for example, can dislike her workplace Example Commellt
but choose to work there because it is better than any other place she has found.
Int: How do you feel about playing at the General request for feeling.
Or a street can be frightening to someone, but the fear can be such a peripheral central basketball court with the older
concern that it does not hinder his walking there. kids?
Interviewers use emotion probes to determine how strongly a person feels
Resp: I don't mind.I'll play anywhere. I Neutral feeling response.
about a response he has given. The probes encourage respondents to explore and
play there sometimes, and sometimes I
explain in depth the meaning and richness of general expressions of feelings.
play on the smaller court down the block.
Emotion probes keep respondents from merely describing a setting by directing
them to explain their feelings about it as well. Int: Why do you use the smaller court? Feeling probe.
Feeling probes continually use the term feel or feeling in questions or Resp: Because I just don't feel like Response indicating avoidance reaction.
repeatedly ask respondents to explain what they mean by a given generally hassling with the older kids.
expressed feeling. Int: Does anybody avoid the central Projection probe.
courts because he's afraid?
Example Comment
Resp: Sure, some kids are really afraid of Projective response describing feelings of
General reaction. getting picked on by the older kids. Some "some kids" in depth.
Resp: I am frightened by the teenagers
even avoid walking down the block if
who walk through the project.
they know that someone playing on the
Int: What do you mean, "I am Feeling probe. central court is after them.
frightened?"
Resp: The teenagers are rough and could Descriptive response. When respondents seem to avoid answering a feeling question, this is a
hurt us.We are old. clue to interviewers to try a projection probe-particularly when respondents
Int: Are you actually afraid they will Feeling probe. deny that they personally have a certain feeling.At the same time, interviewers
hann you? must be careful not to think that very response referring to "a friend" or "some-
FOCUSED INTERVIEWS 153
152 CHAPTER 9
Personal Probes to Tie In Context
· one else" actually describes respondents' unexpressed feelings. The ability to
make such distinctions increases with interviewing experience. Reactions to environments have, as a rule, a dual chain of causes, the
A final emotion probe is the attentive-listening probe, in which interview environment and characteristics of the reacting person. One such characteristic is
ers listen for the meaning implied in the respondent's answer and then make this his or her po ition in the environment: nurse, doctor, patient, visitor; teacher,
_ �
meaning explicit in a follow-up probe: student, pn�c'.pal, parent; tenant, janitor, landlor4, delivery person. More gener
al charactenstlcs, such as age, sex, and family status, can also heavily influence a
respondent's reactions. The most important factors, however, may derive from
Example Comment the respondent' �iography, his or her history, different personal backgrounds,
_ �
General request for feelings about a experiences, or 1d1osyncrasies that influence his or her feelings about things. For
Int: How do you feel about the rules that
specific subject. example, one high-rise urban tenant may have lived his whole life on a farm
the school principal makes about what
you can and cannot do on school whi�e another "'.a� brought up in high-rise city buildings. One doctor may .,;
grounds? particularly sens1t1ve to problems of cancer patients because she had a parent who
Response obliquely describing reactions suffered from the disease. If a researcher wants to understand a respondent's
Resp: He has a right to make any rules he
wants.But they should apply equally to and feelings. answers throughout a focused interview and to generalize data to any larger group
all grades, not just to us seventh- and of pe�ple, the r�searcher must know the personal context within which a respon
eighth-graders. dent 1s answenng questions: position in the system, personal characteristics,
Attentive-listening probe. background, personal idiosyncrasies. Biography is a dimension that can provide
Int: You mean you feel you are being
treated unfairly?
useful insights, but it requires an interviewer's particular tact.
Self-description probes directly request respondents to describe'themselves
Resp: Yes, it really makes me angry that In-depth feeling response.
and why they react to situations the way they do. This achieves results when
they can ...
respondents are self-analytic and conscious of underlying reasons for their
actions.
Sometimes stating implied feelings in terms of limits to action-in terms of
extremes-allows respondents first to reject the extreme statement and then to Example Comment
clarify what they were trying to say.
Resp: I hate people talking in the library. General statement on depth of feeling.
Int: ls there anything particular about you Self-description probe.
Example Comment that makes you feel so strongly?
Resp: Yes, my mother was a librarian, Context response explaining personal
Resp: I can't think of any place I'd rather Seemingly extreme statement of feeling.
and ... background.
live.
Int: Does that mean you like it here so Extreme attentive-listening probe.
When self-descriptive answers result from nonpersonal probes, they are also
much you wouldn't move for anything? significant.
Resp: Not exactly.If my best friend Rejection of extreme restatement and
bought a house where it is wanner, I'd clarification of attachment to residence. Example Comment
consider moving.
Resp: I am afraid to live in that area. Response stating general feeling.
/111: Why are you afraid? General probe.
Interviewers should be careful not to put words in a respondent's mouth by
restating implied feelings approximately and then using a feeling probe too He.rp: My age means that my legs are not Contextual response explaining reactions
o strong, so I am afraid of falling down in terms of personal characteristics.
forcefully: "You did mean this, didn't you?"
whoo I walk.And in that area the kids
Attentive-listening probes, like many others, have side effects as signifi
play ball and ride bicycles on the
cant as their direct effect. Respondents who see interviewers interested and • d walk.
listening tend to relax, be more talkative, and feel greater rapport.
154 CHAPTER 9 FOCUSED INTERVIEWS 155
Parallel probes help respondents talk about themselves in one setting by Appeals for Equal Time
requesting them t o find parallel situations in their own lives. This often has the
_ When one person takes over an inte rview,
effect of getting respondents to explicate the parallel by talkmg about personal that person and othe rs usually
contexts. �now it. Sometimes people even do so as a subtle challenge to t
he interviewer. It
1s your task to appeal t o the pe rson 's sen se
o f fair play in order to give
others a
chance to talk:
Example Comment
Resp: l find this office extremely General re sponse. Int: Good point. Perhaps we should hear some other opinions now.
inefficient and wasteful.
Int:To �et a broad enough picture, it might be good to see what other people think
Int: ln what way? General probe. about this as well.
Resp: l don't know, just "inefficient." Difficulty expressing se lf.
Int: ls there any setting you can think of Parallel probe. Attention to Body language
which is inefficient like this office or
which explains what you mean by Reticent responden ts in a group often remai
n quie t, leaving the fl oor to th
"inefficient"? self-�hosen leader. This does not, however, e
mean that quieter interviewees have
Resp: A submarine is efficient. When I Response explaining personal context. �o�l11ng to �ay; they just do not create their own openings in the conversa
was a sailor, we learned that . . . it 1s y�ur J ob to create openings for them tion . So
wh en y ou notice they want
somethmg. Cues tha t they have an opinion t o expr to say
ess include these:
1 7
STANDARDIZED QUESTIONNAIRES 159
158 CHAPTER 10
Table 10-2. Post-relocation "grief" by pre-relocation "feeling like home" Mutual exclusiveness
Exhaustiveness
Single abstraction level
Percent Severe GriefReactions
ments with people who do not understand the value of �ualitative knowing in
Intrusiveness
scientific research. This is an important characteristic l,f the method when re
Control in administering questionnaires raises the issue that respondents search results are to be used in a court of law, in a politkal setting, in applied
can change and distort answers. Respondents can be directed by the questions design-in any competitive decision-making situation. Such situations are
themselves, for example, to treat some issues in greater depth than others, define increasing as E-B research issues are brought into the publk eye by citizens'
things in certain ways, respond in provided categories. This is not a problem if groups and environmental legislation.
questions and response categories correspond to respondents' definitions of the Naive researchers sometimes are themselves convinced by the numbers '
situation. However, when they do not correspond, respondents sometimes feel they can get using questionnaires. They think they can kam something signifi
that the implicit contract to answer questions means they ought not to correct cant by asking a lot of questions and running answcn; through a computer.
obvious mistakes: the researcher surely knows what she is after. Something can, of course, be learned in this way. But it has a low probability of
Researchers using standardized questionnaires must decide before going solving the researcher's, the client's, or the respondents' problems.
into the field what level of refinement they want answers to achieve to solve their
problem. There is little room for adjustment once data gathering begins. One of Quantitative questionnaire data not augmented by researchers' qualitative
insight or by qualitative data from other methods can provide a hollow and
the most frustrating things that can happen to researchers i_n structured interview unscientific understanding of important problems (Camphcll, 1975).
research is to find that they have spent a great deal of time and energy finding out
everything except one item essential to explaining relations between crucial
ORGANIZATION
variables.
To avoid some of the side effects of control in any method and in any type of
If you are not careful, the way your questionnaire is structured can antago
interview (not only questionnaires), researchers carry out particularly thorough
nize, bore, confuse, and tire respondents. If it does, you might as well not ask
preliminary diagnostic research. Focused interviews may be used to determine
any questions.
how people similar to intended questionnaire respondents define a situation: what
is important; the names they use for places or things; the types of answers they Rapport
give. Observation methods can also be used during diagnostic studies. Using
diagnostic data, investigators structure standardized questionnaires. But that is Questionnaire respondents participate in a research project as informants
not all there is to it. about themselves. Research results are as valid as the relationship between inter
After the questionnaire is written, investigato!§_prete�t it with more people viewer and respondent is open and nondefensive. Rapport can be established by
like the expected respondents. Pretesting a questionnaire means administering it to introducing oneself and the purpose of the interview clearly, honestly, realistical
self-conscious respondents while asking them to comment on points such as these: ly, and without threatening the respondent. Environment-behavior research proj
what do they understand each question to mean; is it clear or confusing; what do ects may be introduced to respondents as attempts to ask their advice--how to
they think is its intent; do response categories give them ample opportunity to make future similar environments better, what could have been improved in that
express themselves? Pretests are invaluable aids not only in questionnaire setting, or just what people like and think. Respondents like to see themselves as
construction but in designing any research instrument to fit the needs of a particu advice givers rather than guinea pigs.
lar situation, group of respondents or elements to observe, and research problem. Questions requesting positive responses ("What do you like best about
_ J:'retest_!>� carried out skillfully also alert investigators to unforeseen prob working in this building?") can start an interview on a friendly note. Later,
lems in other dimensions of the research approach: problem definition, research requests can be made for suggestions on improvement,;. Initial questions can
design, methodological mix, observer training, interviewers' skills, "even of the request general impressions; ask for simple demographic information, such as
first steps of analysis." A pretest is a small, self-conscious pilot study, a micro previous residence; or, especially, be on interesting topics to elicit respondents'
cosm of the actual project carried out to identify, if possible, unintended side attention. For every situation and problem each investigator must work out the
effects (Galtung, 1967: 138). most appropriate way to begin.
Quantitative analysis of questionnaire data not only contributes precision to Early que tions can influence the way respondent, answer later ones. For
knowledge; it also can make research data convincing to others. The apparent example, if early questions give respondents the feeling that interviewers really
exactness and rigorousness of statistical analysis is a useful device to win argu- want to find out what is wrong with a place, they may criticize it more than they·
162 CHAPTER 10 STANDARDIZED QUESTIONNAIRES 163
with a neighborhood, with an event, with a set of activities in one place. For . # 130
129. What problems have you ha d? (Reco rd exact answer)
clarity each group can be introduced with a unifying sentence: "And now I would
· like to ask you some questions about . . . "
Interviewers can also group questions having similar types of response
categories, such as those discussed in the next section: a series o(_preference 130. Do you think the rules and regulations YES NO
questions,_ then semantic differential questions, then attitude questions. Both abo ut living here are fair? ......... ········ ·l 2
type·s of grouping� can lead to "response sets" among ans�ers-namely, respon GO TO
�
# 132
dents' natural tendency to answer questions in a way that seems logically consis 131. Why not? (Record exact answer)
tent. For example, respondents may tend not to admit to criticism of one part of a
setting while praising another. It is therefore sometimes necessary to mix up . I
questions about different dimensions of the same topic and to limit the length of
any one set of questions with identical response categories. 132. When o � first moved here did you go through
�
Another way to use wisely the time respondents give you is not to ask them a �raining program to learn how t o look after YES NO
this h ouse?................................... 1 2
questions that do not apply to them. Filtering questions can help you avoid
133. Would you recommend a training program like YES NO
inapplicable questions by, for example� finding out who drives to work before that f or other people moving in here?......... 1 2
asking how long it takes by car, how many people are in the car, what parking
134. Would you like to have a booklet explaining YES NO
conditions are like. When follow-up quesiions are used for explanation how to look after this house?................. 1 2
("Why?"), specification ("What precisely?"), or clarification of intensity ("How
135. Do you think it would be helpful t o have a
much?"), it saves time to target them only to. respondents to whom they apply. handbook to explain the rules and regulations YES NO
All this can be achieved with clear layout and written interviewer instruc about living here?.. ... • ......... ............ . l 2
tions to keep the interview flowing and to avoid confusing respondents with
irrelevant questions. Saile et al. (1972) faced many of the problems discussed
► 136. Do you think yo u'll stay in this house or do
you expect to move sometime? .......STAY MOVE SOMETIME
here. Their questionnaire helped resolve many of them (facing page).
I l�I 2
0 0
rise from 65° F to 67° F. The same problem holds for age differences. A year has this factory
a different quantity and quality at ages 5, 30, and 80. Therefore, interval catego are unfair [D [I] IT]
ries are not presented here as distinct from ordinal ones. Management is
0 0 0 DJ
very helpful in
Information. Ordinal precoding can be used for questio� s gath<:ring infor job training �
.
mation that reasonably are seen as "how much" or "how many questions-age, The work areas
income, size of household, number of clubs a respondent belongs to. we have could
0 0 0�-
easily be much
Age: Under 11 D, 11-200, 21-30 D, 31-40 D, 41-50 D, 51-600, 61 or better [D IT]
over □.
(Number scores in the boxes are not presented on the questionnaire.)
Club membership: None D , l or 2 D , 3 to 6 D , 7 or more D
Attitudes. Ordinal coding may also be useful for response categories fol When Likert-scaled questions are used, they can be grouped together in a
lowing questions that ask respondents to judge the intensity of an attitude about questionnaire so that once respondents understand how to use this system of
something, such as a situation, person, object, or setting. recording responses, they can use it for several questions. When this is done,
however, the list must be short enough and must mix up positive and negative
statements to avoid respondents' going down a long list checking only one
Would you say the rules in this factory are
column and not thinking.
very fair□, fair D , unfair D , very unfair D , or do you have no
opinion D
If researchers feel that using Likert-scale items on a questionnaire can help
about this? them solve their problems, they should carefully study the assumptions under
Would you say the work areas you have are lying this type of attitude quantification. If they decide they can make these
very supportive □, supportive D , unsupportive O , very unsupportive D
, or are assumption , th y arry ut careful empirical procedures to choose and score
you uncertain about thi D ?
168 CHAPTER 10 STANDARDIZED QUESTIONNAIRES 169
groups of statements that actually do relate to one another (as explained, for Re�po�deilts are in structed that marking the line abov e a response of l means the
example, in Shaw & Wright, 1967). O�Ject 1s extremely wi de, �ontemporary, or whatever; 2, quite wide; 3, slig htly
The same caveat ho lds for the use of every empirically developed measure- wide ; 4, n�utral, �ually wide and narrow, or wideness is unrelated to the obje ct.
ment scale, including the semantic differential scale, discussed next . Choice and interpretation of scal e items are difficult. Osgood et a l. carried
out s everal studies and m uch computer ana lysis to de velop a sca le of 50 pa ired
Meaning. When you look at the Eiffel Tower, Mount Fuji, or the chair items particul'."'ly �elev ant to �ene ral concepts. They found that the descriptive
you are sitting in, you react to it in part on the basis of what it "means" t� �ou. terms fell ma ml_ y m to ca te gones of eva luation , potency, and acti vi ty. A grea t
Y ou may, for example, feel unc omfortable and tense in the chair because it 1s an deal more work 1s necessary to ada pt thi s sca li ng techniq ue t o E-8 studies.
antique that you see as dainty, we ak, and silly, althou gh tasteful. M ost people Wh en sem
. _ antic differential scales are used , it is necessary to choose cate
find it difficult to express verbally the range of meaning things have to them. In a gones appropnate to the particular research situation and respondents' defi nition
taste test of different ice creams, few tasters could spon taneously manage to say of the situation. For e xample, in thei r study of housing for older persons, Howell
anything but "creamy" and "tasty" in attempts to differentiate b rands. B t ':he� and Epp ( 1976) pretested the followi ng sema nti c diff erenti al quest ion :
�
presented with lists of descriptive terms to choose from, the� could easily md1-
cate what the different tastes meant to them (Osgood, Suc1, & Tannenbau m, How would you describe the way the building looks from the outside?
1957).
The principle that people express the meaning things hold �or them m�re like a private __ / __ / _ _ / __ / __ / __ / __ / like a public
completely when presented with a set of appropriate alternatives u�derhes home (I) (2) (3) (4) (3) (2) (I) building
another analytic coding t echnique-the semantic differential scale. L1k� �he hard __ / __ / _ _ / __ / __ / __ / _ _/ soft
Likert and other scales, this one must entail carefu l procedures for determmmg simple __ / __ / __ / __ / __ / __ / __ / complex
what alternatives are "appropriate" for particular respondents and si tua�ions : It
also entails important and often questionable assumptions about quanuficat1on The older res pondents were able to choose between the attributes in the first pair,
and ensuing ana lysis of data (Osgood et a l., 1957).
but they could not understand what the oth er two pairs had to do with the wa y
If the scaling technique is critically examined, it can be selectively used to their building looked.
identify the quality an d intensity of meaning that E-8 topics such as environ- Question construction has d eveloped into a complex s kill-perhaps too
ments, persons, places, and-situations hol d for people. c?mplex. The semantic differential scal e exhibits a problem many such tech
The format for semantic differentiation presents respondents first with the mques face : They may cause you more damage than they are worth. It is unclear,
name or picture of an object (place, concept , and so on) or with the �bj� t its elf,
f?r example, that seven-point polar-opposite judgment tests yield more informa
followed by a series of polar opposite terms: good/bad, happy/sad, b1g/httle. For ti on tha_ n � th ree- or five-point agree/dis agree rating scale. If respondents feel that
each pair of terms, responde nt s are requested to indicate how the terms _ap�l y
to
th j c me s them. T y t is m
the a_ dj ectives they are asked to rate are nonsensical (gay an d dreary applied to a
the object on the basis of what e ob e t an to he do h the
cha ir, for example), the loss of rapport wi th the i nterviewe r may invalidate othe r
following format : parts of t he interview. Careful pretesti ng is one wa y to avoid such mistak es.
�nother is. to include on the team constructing questions some people who are
hke potential respondents. This is good ad vice no matter what type of question
Your Chair you are constructing. In sum, rating scales of any sort must be used on ly after
Wide __ I __ ! __ / __ / __ ! __ I __ / Narrow carefully examining their wording and the operational assumptions they embody.
(1) (2) (3) (4) (3) (2) (1)
Rank-ordering of items. It ma y be useful to precode responses to questions
.
Contemporary __ / -- I -- I --I -- I -- I __ I Traditional re�pondents to rank a group of items relative to one another on a single
ask '. ng
Functional __I __ I __I __ / __I __I __I Nonfunctional attnb ute: importance, beauty, useful ness, worthwhileness . For example :
Tasteful __ / __ / __ / __ / __ / __ / __ / Tasteless
Gay _I _I _I __ / __I _I __ ! Dreary Whi�h of the spaces on the following list do you feel it is most important to
Orderly __ / __ / __ / __ / __ / __ / __ / Cha�tic include in a house? (Please circle "1" for the most important, "2" for the second
Private __ / __ / __ I __ I __ I __ I __ I Public most important, "3" for the next, and so on until you have ranked all places in
Sparkling _t _I _! _I __ I _I _I Dingy terms of their importance to you.)
STANDARDIZED QUESTIONNAIRES 171
170 CHAPTER 10
It is interesting enough that we can use the idea of implicit mental maps to
bathroom 2 43 5 6 7 8 9 10 11 12 help design places more comprehensible to people. Still more interesting is that
kitchen 2 7 8 9 10 11 12
4·
3 5 6 people's cognitive maps only partly correspond to the measurable attributes of
laundry room 2 4 5 6 7 8 9 10 11 12
environme nts that might b e represented by a street map drawn to scale or an
3
living room 2 4,
3 5 6 7 8 9 10 11 12
9 10 11 12 aerial photograph. People's cognitive maps are influenced and distorted by their
bedrooms 2 43 5 6 7 8
9 10 II 12 backgroun d, their experience, their purposes, and so on. For exainple, in a
den or rec. room 2 43 5 6 7 8
2 9 10 II 12 hospital, workers estimated a path outside the building to be twice as long as a
study 43 5 6 7 8
storage attic 2 43 5 6 7 8 9 10 11 12 path inside the building, although the two were m easurably the same distance
vestibule 2 43 5 6 7 8 9 10 11 12 (Stea, 1974).
dining room 2 43 5 6 7 8 9 10 11 12 If designers know how people who use their environments see them, they
2 6 7 8 9 10 11 12 can better con trol the sid e e ffects of design decisions. In the hospital mentioned
Other: specify 43 5
2 3 4 6 7 8 9 10 11 12 above, for example, if the designer knows that outside paths are seen as longer
Other: specify 5
11 12 than insid e ones, he might make differen t decisions about enclosing them in order
2 3 4 5 6 7 8 9 10
Other: specify to provide alternative ways to get around that are actually seen as alternatives .
l�ernative li:e-styles, _for There is no one way to study the complex set of perceptions and attitudes
When items in the group are unduly comple x-a a that make up a person's cogn itive map. Lynch ( l 960) carried out a 90-min ute
example-it is easi er to presen
t them in pairs for sequential companson than in focused in terview with responden ts, one part of which requested them to draw
simultaneous list. freehand maps of the city. Some of Lynch's respondents were shown a series of
photographs of downtown areas and asked to choose those they felt were most
Each technique for pr ecoding responses cr eat� s opportun ities for rese�r ch-
typical of the way they saw the city. The voluntee rs in this group were also
ers, but each also limits what resear chers can do w1�h the data. �nly expene�ce
interviewed with a walk-through probe (see Chapter 9) of the downtown area.
with using classification and scaling methods-aski ng t�e q�est1on s, recording
During the trip, interviewers asked respondents why they took a particular path,
answers, tabu lating responses, and analyzing data-will give researchers the
what t hey saw, and when the y felt confident or lost (Lynch, 1960: 140-142).
knowledge and se lf-con fidence needed to choose a form for precoded responses .
The visual-response tec hniques discussed in this section are essen tial
used togethe r with verbal responses and observational methods-to study peo
VISUAL RESPONSES ple's attitudes , perceptions, and knowledge concerning physical environmen ts.
information about respon Broadly defined, people's "cognitive maps" comprise all these mental process
Some cognitive expressiv e, and perceptual es-requirin g fo r their study the s ame array of methods and techniques. In this
, expr essed visually than verball�
dents physical surro�ndings may be better : section, I will discuss freehand maps, the use of base maps, drawings, respon
han d area maps, base-map add1
through nonprecoded techniques, such as free dents' photographs, and games.
e nts, an d games.
tions drawings, photographs taken by respond .
'
This is especially true for peopl
e ' s cogn itive maps, the mental pictures of
nd act
e way the y look at, re act t o a
their surroundings that they use to structur e th '. Freehand Area Maps
Jonge, 1962; Ladd, 19?0, L�nch,
in their environment (Downs & Stea, 1973; d e
ing picture as a two-�1mens1onal
1960). One can envision this continually chang Lynch's i nstr uctions t o respon d ents were to draw "a quick map of
h�logr�m, o� � file of pictures �ept
map or drawing, a three-dimensional model, a . __ . . . . Make it just as if you were making a rapid description of the city to a
a map seen as a flat p1�ce
in one's mind. "A cognitive map is not necessanly stranger, covering all the main feat ures. We don't expect an accurate drawing
an ongoing "�rocess .. • • by which
of paper (Downs & Stea, 1973: 11). It is more t th
-just a rough sketch" (Lynch, 1960: 141).
s, and decodes in fo�at10� abou
an individual acquir es, codes, stores, recall nv o m ent
: Lynch analyzed the resulting maps for such things as omissions, precision,
a d a t ibu of .. . his e v e ryda y spati al e ir n
distortions, and differential knowledge of areas. He also established a coding
relative lo c ations n t r tes
sche me for map responses, which he called "city image elements": paths, edges,
(Downs & Stea, 1973: 9). . .
n e_n t.
to you og i iv map wh enev er you de al with an e_nv1ro � districts, n odes, landmarks, and element interrelations.
You refer r c n t e
dining
if you fi d your s lf_i t h e
Your so-called map tells you, for example, that
n e n
Lynch's study began a tradition of freehand-area-map drawing in inter
you have neve r bee m, ?n� �f the
room of a modem middle-class Western home
n
views, both t d v 1 p th m thod (de Jonge, 1962) and to look at cross-cultural
Y u will be urprl s d if I n t
doors around y u prob bly Ieod • t th kit hen.
1t
nd up dif nm p .
n th third fl -
-If, i x mpl-, you find lh t th kit I n i
STANDARDIZED QUESTIONNAIRES 173
172 CHAPTER 10
designers, this means that if they want to control the behavioral side effects of
places they design, they must understand the ways different groups see them.
Area maps can be used to find out where respondents feel at hon:ie, are
afraid, spend time. Limits on topics are imposed only by investigators' imagina
tion. However, there are unresolved difficulties of interpretation. Some people
draw particularly well; others refuse to draw. Some people can draw landmarks
easily only when starting with a base map with major streets already indicated;
others do so· just as easily on a blank piece of paper. As with all methods, the
more area maps are used and by more people, the more we learn about how to
tum the data they provide into useful infonnation.
(-�\; I I �� --�
.. ' �
flCtoe;,- --,
'I
I
,_,,
Interpreting area maps has shown that people with limited m�vement Composite path-map of respondents' trips from home to local store. (From
often poorer people-have detailed knowledge of their immediate neighborhood Charlesview Housing: A Diagnostic Evaluation, by J. Zeise! and M. Griffin.
but only an ill-fonned image of the city they live in as a whole (Orleans, 1973) Cambrid , Mo .: Harvard Oroduat School of Design, Architecture Re-
and that children ·c thin , diff r ntly from ndull nnd wom n from m n. For
s h m , 197 .
174 HAPT R 10 AN AR IZ 175
and asked them to draw or point out on it the paths they took on the way to their
cars, to the shops, and to the bus stop. These pathway maps were requested
instead of verbal descriptions because they seemed more reliable, more accurate,
and more expressive of the process of taking a trip.-
Maps like these may be quantitatively coded. They also lend themselves
particularly well to comparison and visual analysis on composite data maps.
Drawings
Sometimes people's mental pictures about the future can actually be ex
pressed in a picture. Sanoff and Barbour (1974) worked with an architect com
missioned to design a grade school. They were interested in finding out what a
"dream school" was like for students. One approach they used was to ask stu
dents involved in the programming research to draw typical African, Japanese,
and American schools and their dream school. They found particular contrast
between the factorylike drawings of typical American schools and the multilevel,
almost "treelike" dream schools. Although this type of response is still more
abstract and difficult to interpret quantitatively than maps, it can provide investi
gators with important qualitative insights.
Photographs
Games
Another way E-B researchers have recorded respondents' ideas has been to
develop games through which respondents express themselves by making a series
of linked choices (Robinson et al., 1975). One of the oldest such games is
Wilson's neighborhood game (1962). Alternative degrees of attributes such as
neighborhood physical quality and sanitation services each have a price tag
attached. Respondents are given a set of chips representing the total amount of
,....
money they can spend to "buy" the amenities on the game board. With the Drawings by children participating in the design charrette for the Wallace
amount of play money they have, they are forced to choose among attractive <?'Neill Alternative School, Pinehurst, North Carolina. (From "An Alterna
alternatives, not all of which they can afford. Their final judgments express not a uve Strategy for Planning an Alternative School," by H. Sanoff and G.
linear series of individual choices but a balanced set of simultaneous ones. �arbour. In G. T. Coates (Ed.), Alternative Leaming Environments. Copy
nght 1974 by Dowden, Hutchinson & Ross, Inc. Used by pennission.)
STANDARDIZED QUESTIONNAIRES
176 CHAPTER 10
presentation by respondents of freehand maps, additions to base maps, draw
Zeise!, in his design programming and evaluation studies of a low-income
ings, photographs, and games. The visual character of such data makes them
housing project in South Carolina (1971), and Zeisel and Griffin, in their housing
available both for quantitative analysis and for qualitative visual presentation on
evaluation (1975), developed a Dwelling Unit Aoor Plan Game to present re
composite maps or charts.
spondents with a series of simple design decisions: in which rooms oug�t en
Used together with observation methods and focused interviewing, stan
trances to be; how would you like the kitchen to relate to where you eat; kitchen
dardized questionnaires are particularly useful to gather information about such
to living room; living room to eating; and balcony location (Zeise!, 1971). Each
topics as people's perceptions, their attitudes, their values, and the meaning the
decision, a choice of three alternatives for separation and connection, is pre
environment holds for them.
sented in the context of earlier choices. All together they result in an entire floor
The next chapter presents E-B topics particularly suited to question�
plan. In addition to composite results, interviewers use the opportunity the game
naire and interview investigation and discusses some rules of thumb for asking
provides to ask respondents why they made each choice, probing to find out what
questions.
behavioral or cognitive side effects respondents were trying to achieve by the
choice.
OVERVIEW
Standardized questionnaires are useful if you know what you want to find
out from people, if you want to discover regularities among groups of people
with particular characteristics, and if you want to be able to quantify your data.
After discussing how to organize questions in a questionnaire so that it
establishes a nQndefensive, open interview situation, this chapter presents ways
to record responses to standardized questions. Open-ended responses can be
coded for analysis into mutually exclusive, exhaustive categories at a single level
of abstraction. The same criteria for coding categories may be used to precode
response categories if the investigator has developed the categories empirically to
be sure they fit respondents• definitions of the situation and enable respondents to
express themselves adequately on the topic. Otherwise, the control exerted by
using an intrusive method, such as a questionnaire, distorts data and makes them
worthless.
Some data, particularly visual data useful in assessing respondents' "cogni
tive maps," cannot be precoded. Response categories for such data include visual