Paper for conference, Plants for People
International Exhibition Floriade 2002
Health Benefits of Gardens in Hospitals
Roger S. Ulrich, Ph.D.
Center for Health Systems and Design
Colleges of Architecture and Medicine
Texas A & M University
College State, TX 77843
This paper selectively reviews scientific research on the influences of gardens and
plants in hospitals and other healthcare settings. The discussion concentrates mainly on
health-related benefits that patients realize by simply looking at gardens and plants, or in
other ways passively experiencing healthcare surroundings where plants are prominent.
The review also briefly addresses other advantages of gardens and plants in hospitals,
such as lowering the costs of delivering healthcare and improving staff satisfaction.
It might be asked at the outset: why is worthwhile to focus exclusively on gardens
located in hospitals and other healthcare facilities? One important reason is linked to the
fact that extraordinary amounts of money are spent internationally for construction of
healthcare environments. This funding for hospitals potentially represents a major source
of resources for gardens, plants, and related features such as atriums. Consider the
example of only one large medical complex in the United States, the Texas Medical
Center in Houston, which plans to spend about $1.8 billion on new construction during
the next two years. In the State of California alone, new spending for hospital buildings
will be upwards of $14 billion by 2010. Even individual buildings can be extremely
costly -- Northwestern University’s recently opened main hospital in Chicago cost $687
million. Spending in the United States for new hospitals has averaged about $15 billion
annually during the last decade. The United Kingdom plans to spend at least $4 billion on
new hospital construction within the next three years or so. When substantial additional
spending is considered for the many other types of healthcare environments -- for
example, nursing homes, primary care clinics, rehabilitation facilities -- it becomes even
clearer that healthcare design and construction directly accounts for vast amounts of
money. This reality implies great opportunities for funding and creating new gardens to
enrich and improve the lives of patients and the environments of hundreds, if not
thousands, of existing medical facilities.
Background: Gardens and Hospital Design
The belief that plants and gardens are beneficial for patients in healthcare
environments is more than one thousand years old, and appears prominently in Asian and
Western cultures (Ulrich and Parsons, 1992). During the Middle Ages in Europe, for
example, monasteries created elaborate gardens to bring pleasant, soothing distraction to
the ill (Gierlach-Spriggs et al., 1998). European and American hospitals in the 1800s
commonly contained gardens and plants as prominent features (Nightingale, 1860).
Gardens became less prevalent in hospitals during the early decades of the 1900s,
however, as major advances in medical science caused hospital administrators and
architects to concentrate on creating healthcare buildings that would reduce infection risk
and serve as functionally efficient settings for new medical technology. The strong
emphasis on infection reduction, together with the priority given to functional efficiency,
shaped the design of hundreds of major hospitals internationally -- that are now
considered starkly institutional, unacceptably stressful, and unsuited to the emotional
needs of patients, their families, and even healthcare staff (Ulrich, 1991; Horsburgh,
1995). Despite the intense stress often caused by illness, pain, and traumatic hospital
experiences, little attention was given to creating environments that would calm patients
or otherwise address emotional needs (Ulrich, 2001).
A growing awareness has developed in recent years in the healthcare community
of the need to create functionally efficient and hygienic environments that also have
pleasant, stress reducing characteristics. An important impetus for this awareness has
been the major progress achieved in mind-body medical science. A substantial body of
research has now demonstrated that stress and psychosocial factors can significantly
affect patient health outcomes. This knowledge strongly implies that the psychological or
emotional needs of patients be given high priority along with traditional concerns,
including infection risk exposure and functional efficiency, in governing the design of
hospitals (Ulrich, 2001). It also follows that conditions or experiences shown by medical
researchers to be stress reducing and healthful, such as pleasant soothing distractions and
social support, must become important considerations in creating new healthcare
facilities. The fact that there is limited but growing scientific evidence that viewing
gardens can measurably reduce patient stress and improve health outcomes has been a
key factor in the major resurgence in interest internationally in providing gardens in
hospitals and other healthcare facilities.
Importance of Health Outcomes Evidence
Healthcare administrators everywhere are under strong pressures to control or reduce
costs yet increase care quality. Faced with imperative demands such as paying for costly new
medical technology, administrators may often consider gardens as desirable but nonessential.
Convincing the medical community to assign priority and resources usually requires
providing credible evidence that gardens or plants produce benefits yet are cost-effective
compared to alternatives, including not providing gardens/plants.
It should be emphasized here that most healthcare administrators and especially
physicians consider evidence from health outcomes research to provide the most sound and
persuasive basis for assessing whether a particular medical treatment or service (here
providing a garden or plants) is medically beneficial and financially sensible. (Ulrich, 1999,
Health outcomes are numerous and varied, but most refer to measures of a patient’s
medical condition or to indicators of healthcare quality. These measures include (1)
observable clinical signs or medical measures, (2) subjective measures such as reported
satisfaction, and (2) economic measures (Ulrich, 2002).
• Clinical indicators that are observable signs and symptoms relating to patients’
conditions. (Examples: length of stay, blood pressure, intake of pain drugs)
• Patient/staff reported outcomes. (Examples: patient reports of satisfaction with
healthcare services, staff reported satisfaction with working conditions)
• Economic outcomes. (Examples: cost of patient care, recruitment or hiring costs
due to staff turnover)
Clinical and economic outcomes data traditionally have carried the greatest
weight in decisions, but in recent years evidence regarding effects of treatments or
services on patient satisfaction has gained much importance as healthcare providers in the
United States and Europe have faced mounting pressures to become more patient or
STRESS REDUCING EFFECTS OF VIEWING PLANTS AND NATURE
Several studies of nonpatient groups (such as university students) as well as
patients have consistently shown that simply looking at environments dominated by
greenery, flowers, or water -- as compared to built scenes lacking nature (rooms,
buildings, towns) -- is significantly more effective in promoting recovery or restoration
from stress. (See Ulrich, 1999, for a survey of studies.) A limited amount of research
suggests that viewing settings with plants or other nature for a few minutes can promote
measurable restoration even in hospital patients who are acutely stressed.
There is considerable evidence that restorative effects of nature scenes are
manifested within only three to five minutes as a combination of psychological/emotional
and physiological changes. Concerning the first, psychological/emotional, many views of
vegetation or garden-like features elevate levels of positive feelings (pleasantness, calm),
and reduce negatively toned emotions such as fear, anger, and sadness. Certain nature
scenes effectively sustain interest and attention, and accordingly can serve as pleasant
distractions that may diminish stressful thoughts. Regarding physiological manifestations
of stress recovery, laboratory and clinical investigations have found that viewing nature
settings can produce significant restoration within less than five minutes as indicated by
positive changes, for instance, in blood pressure, heart activity, muscle tension, and brain
electrical activity (Ulrich, 1981; Ulrich et al., 1991).
One controlled experiment, for example, measured a battery of physiological
responses in 120 stressed persons (non-patients) who were randomly assigned to a
recovery period consisting of one of six different videotapes of either nature settings
(vegetation or vegetation with water) or built settings lacking nature (Ulrich et al., 1991).
Findings from four continuously recorded physiological measures (blood pressure, heart
rate, skin conductance, muscle tension) were consistent in indicating that recuperation
from stress was faster and much more complete when individuals were exposed to the
nature settings rather than any of the built environments. The quickness of nature-induced
restoration was manifested as significant changes in all physiological measures within
about three minutes. The pattern of physiological data further supported the interpretation
that nature, compared to the built settings, more effectively lowered activity in the
sympathetic nervous system. (Heightened sympathetic nervous system activity involves
energy consuming mobilization or arousal and is central in stress responding.) Moreover,
data from self-reports of feelings indicated that the nature environments likewise
produced substantially more recuperation in the psychological component of stress.
Persons exposed to the settings with plants and other nature, in contrast to the built
environments, had lower levels of fear and anger, and reported far higher levels of
positive feelings (Ulrich et al., 1991).
Hartig (1991) also used both physiological and psychological measures to study
restoration in non-patient subjects who were stressed because they either had driven an
automobile through urban traffic or completed a series of difficult tests. His findings were
broadly similar to those described above -- more specifically, blood pressure data and
emotional self-reports converged to indicate that recovery was appreciably greater if
persons looked at a nature setting dominated by vegetation rather than a built
environment without nature (Hartig, 1991).
Nakamura and Fujii have carried out two studies in Japan (1990, 1992) that
measured brain wave activity as unstressed persons (non-patients) looked either at plants
or human-made objects. In an intriguing first experiment, the researchers analyzed alpha
rhythm activity as subjects viewed: two types of potted plants, each with and without
flowers (Pelargonium and Begonia); the same pots without plants; or a cylinder similar to
the pots (Nakamura and Fujii, 1990). Results suggested that persons were most wakefully
relaxed when they observed plants with flowers, and least relaxed when they looked at
pots without plants. In the second study they recorded the electroencephalogram (EEG)
while persons were seated in a real outdoor setting and viewed a hedge of greenery, a
concrete fence with dimensions similar to the hedge, or a mixed condition consisting of
part greenery and part concrete (Nakamura and Fujii, 1992). The EEG data supported the
conclusion that the greenery elicited relaxation whereas the concrete had stressful
Benefits of Nature and Gardens in Healthcare Settings
The research examples described above, all based on non-patient groups, indicate
that visual exposure to plants and other nature lasting only a few minutes can foster
considerable restoration or recovery from stress.
It is important to emphasize that broadly parallel findings have been obtained
when stressed patients in healthcare settings have been visually exposed to nature. A
study by Heerwagen and Orians, for instance, found that anxious patients in a dental fears
clinic were less stressed on days when a large nature mural was hung on a wall of the
waiting room in contrast to days when the wall was blank (Heerwagen, 1990). The
restorative benefits of the nature scene were evident both in heart rate data and self-
reports of emotional states.
In the case of hospitals and other healthcare facilities, there is mounting evidence
that gardens function are especially effective and beneficial settings with respect to
fostering restoration for stressed patients, family members, and staff (Ulrich, 1999).
Cooper-Marcus and Barnes (1995) used a combination of behavioral observation and
interview methods to evaluate four hospital gardens in California. They found that
restoration from stress, including improved mood, was by far the most important category
of benefits derived by nearly all users of the gardens -- patients, family, and employees.
Likewise, a recent study of a garden in a children’s hospital identified mood
improvement and restoration from stress as primary benefits for users (Whitehouse et al.,
2001). This conclusion was supported by convergent results from behavioral
observations, interviews, and surveys. The fact that stress is a pervasive, well-
documented, and very important health-related problem in hospitals implies major
significance for the finding that restoration is the key benefit motivating persons to use
gardens in healthcare facilities (Ulrich, 1999).
Well-designed hospital gardens not only provide calming and pleasant nature
views, but can also reduce stress and improve clinical outcomes through other
mechanisms, for instance, fostering access to social support and privacy, and providing
opportunities for escape from stressful clinical settings (Ulrich, 1999; Cooper-Marcus
and Barnes, 1995). Concerning the last of these, escape, Cooper-Marcus and Barnes
(1995) concluded that many healthcare employees used gardens as an effective means for
achieving a restorative pleasant escape from work stress and aversive conditions in the
hospital. They also included in their report statements by several patients which
suggested that the gardens fostered restoration in part by providing positive escape (and
sense of control) with respect to stress. For example, a patient interviewed in a hospital
garden commented: “It’s a good escape from what they put me through. I come out here
between appointments. . I feel much calmer, less stressed” (Cooper-Marcus and Barnes,
1995, p. 27).
In addition to ameliorating stress and improving mood, gardens and nature in
hospitals can significantly heighten satisfaction with the healthcare provider and the
overall quality of care. Evidence from studies of a number of different hospitals and
diverse categories of patients (adults, children, and elderly patients; ambulatory or
outpatient settings, inpatient acute care wards) strongly suggests that the presence of
nature -- indoor and outdoor gardens, plants, window views of nature -- increases both
patient and family satisfaction (Cooper-Marcus and Barnes, 1995; Whitehouse et al.,
2001; Picker Institute and Center for Health Design, 1999).
The capacity for gardens and plants to heighten satisfaction, as well as reduce
stress, is attracting considerable attention from hospital administrators who, as noted
earlier, are facing strong pressures to become more patient/consumer oriented and
improve the consumer’s healthcare experience. A nationally prominent hospital
administrator in the United States recently evaluated the role of gardens in the highly
competitive marketplace of managed care, and endorsed their effectiveness for increasing
care quality and patient/consumer satisfaction (Sadler, 2001). Further, the administrator
advocated creating gardens as an effective means for helping hospitals and providers to
achieve more positive market identities and thereby improve economic or financial
outcomes (Sadler, 2001).
Benefits of Healthcare Gardens for Staff
Healthcare staffing problems are a critical issue in most European countries and
North America. It has been known for decades that healthcare occupations such as
nursing are stressful because they often involve overload from work demands, lack of
control or authority over decisions, and stress from rotating shifts (Ulrich, 1991).
Workloads and pressures have mounted further, however, as healthcare providers
everywhere have been forced to control or cut costs (Ulrich, 2002). These conditions
have in many locations lowered lower job satisfaction, increased absenteeism and
turnover, contributed to shortages of qualified personnel, increased providers’ operating
costs, and eroded the quality of care that patients receive (Ulrich, 2002).
These serious staff related problems imply major importance for the
aforementioned finding that healthcare staff heavily use gardens for positive escape from
workplace pressures and to recuperate from stress. Additionally, it should be emphasized
that evidence has begun to appear showing that hospital gardens increase staff
satisfaction with the workplace, and may help hospital administrators in hiring and
retaining qualified personnel (Whitehouse et al., 2001; Sadler, 2001; Cooper-Marcus and
Barnes, 1995, 1999).
Effects of Nature on Clinical Outcomes
Findings from a few studies focusing on hospitals and other healthcare facilities
suggest that views of nature can have important benefits in terms of improving patient
clinical outcomes. At Uppsala University Hospital in Sweden, Outi Lundén, John Eltinge,
and I (1993) investigated whether exposing heart surgery patients to simulated nature
views would improve recovery outcomes. We assigned each 160 patients in intensive
care to one of six visual stimulation conditions: two nature pictures (either a view of trees
and water, or an enclosed forest scene); two abstract pictures; and two control conditions
(either a white panel, or no picture or panel). Results suggested that patients who viewed
the trees/water scene were significantly less anxious during the postoperative period than
patients assigned to the other pictures and control conditions. Moreover, patients exposed
to the trees/water view suffered less severe pain, as evidenced by the fact they shifted
faster than other groups from strong narcotic pain drugs to moderate strength analgesics.
By contrast, a rather surprising finding was that an abstract picture dominated by
rectilinear forms produced higher patient anxiety than control conditions of no picture
Another medical outcomes study compared the recovery records of gall bladder
surgery patients who had a bedside window view of either trees or a brick building wall
with no nature (Ulrich, 1984). To keep other factors constant that could affect outcomes,
the methods ensured that the tree and wall view groups were equivalent, for example, in
age, weight, tobacco use, and general medical history. The outcomes data showed that
those with the nature view, compared to those who looked out at the wall, had shorter
hospital stays and suffered fewer minor post-surgical complications (such as persistent
headache or nausea) (Ulrich, 1984). Further, patients with the view of trees more
frequently received positive written comments from staff about their conditions in their
medical records (“patient is in good spirits”). Those in the wall view group, however, had
far more negative evaluative comments (“patient is upset,” “needs much
encouragement”). Another major difference was that persons with the view of trees,
compared to the wall view patients, needed far fewer doses of strong narcotic pain drugs.
The above findings not only indicated that views of nature in hospitals could
enhance clinical or medical outcomes; as well, the results suggested that nature could
improve economic outcomes by reducing the costs of care. The findings clearly implied
that by providing nature it would be possible to achieve cost savings, for instance,
because length of hospital stays might be shortened, and some patients would have
reduced need for costly injections of strong pain drugs.
QUALITIES OF EFFECTIVE RESTORATIVE GARDENS
Few studies have examined rigorously how different design approaches and
specific environmental characteristics affect hospital garden performance with respect to
fostering restoration from stress or improving medical outcomes. No well controlled
experiment has investigated, for instance, whether designing flowers beds with
curvilinear in contrast to rectilinear forms or edges influences a garden’s effectiveness in
producing stress recovery. Nonetheless, the studies described in earlier sections have
yielded a few broad conclusions and general guidelines regarding design directions for
creating successful healthcare gardens.
The limited evidence to date suggests that gardens will likely calm or ameliorate
stress effectively if they contain verdant foliage, flowers, water (not tumultuous),
congruent or harmonious nature sounds (birds, breezes, water), and visible wildlife
(birds) (Ulrich, 1999, pp. 74-75). Additionally, nature settings with savanna-like or park-
like qualities (grassy spaces with scattered trees) are known to foster restoration. In their
study of users of four hospital gardens, Cooper-Marcus and Barnes (1995, p. 55) found
that the most frequently mentioned positive garden qualities were visual nature elements,
especially trees, greenery, flowers, and water. Respondents strongly associated these
nature features with restorative influences on their moods.
By contrast, a characteristic that usually worsens garden effectiveness in reducing
stress is predominance of hardscape (concrete, for example) or other starkly built content
(Ulrich, 1999). Whitehouse and her associates (2001) found that users of a children’s
hospital garden disliked and avoided areas having a high percentage of concrete ground
surface and/or starkly built features. Persons interviewed in this study consistently
recommended that the garden should have “more greenery and flowers” and less concrete
(Whitehouse et al., 2001). Based on this evidence the administration of the hospital
directed that the garden be reconstructed to include many more plants and less hardscape,
in order to become more effective in promoting restoration.
In addition to predominance of hardscape rather than vegetation, other garden
qualities that can hamper recovery or even aggravate stress include: cigarette smoke;
intrusive, incongruent urban or machine sounds (traffic, for example); crowding;
perceived insecurity or risk; prominent litter; and abstract, ambiguous sculpture or other
built features that can be interpreted in multiple ways (Ulrich, 1999). Regarding
abstraction and ambiguity, there is mounting evidence that designers of hospital gardens
should exercise considerable caution before including abstract art works or ambiguous
design features. It appears that acutely stressed patients may be vulnerable to having
stressful rather than positive reactions to ambiguous art or design (Ulrich, 1991). Current
evidence suggests that the safest, most consistently effective general strategy for
designers of hospital gardens is simply to feature the restorative, unambiguously positive
qualities of greenery, flowers, and most other nature content (Ulrich, 1999).
A documented example of adverse patient reactions to ambiguous features
occurred when a major university hospital installed a large-scale series of sculptures and
other artworks to form a “bird garden” in a rooftop space overlooked on all sides by
rooms for cancer patients (Ulrich, 1999). Although called a “garden,” the space actually
contained no greenery, flowers, or other nature. Soon after this sculpture garden was
installed, administrators and physicians began to receive many anecdotal reports of strong
negative reactions by patients. Accordingly, a questionnaire study was conducted of
patient reactions to the artwork (Hefferman et al., 1995). The study showed that more
than 20% of the cancer patients reported having a negative emotional or psychological
reaction to the “garden.” Several patients had strongly negative responses, interpreting
some rectilinear metal bird sculptures, for instance, as frightening predatory animals
(Ulrich, 1999). The administration and medical staff decided that the rate and intensity of
negative effects on patient outcomes was too high, so the art installation was removed for
medical reasons (Ulrich, 1999).
Findings from several studies have converged in indicating that simply viewing
certain types of nature and garden scenes significantly ameliorates stress within only five
minutes or less. Further, a limited amount of research has found that viewing nature for
longer periods not only helps to calm patients, but can also foster improvement in clinical
outcomes -- such as reducing pain medication intake and shortening hospital stays.
Well-designed hospital gardens not only provide restorative and pleasant nature
views, but also can reduce stress and improve clinical outcomes through other
mechanisms such as increasing access to social support, and providing opportunities for
positive escape from stressful clinical settings.
As well, evidence from studies of a number of hospitals strongly suggests that
gardens and other nature helps to heighten patient and family satisfaction with the
healthcare provider and the overall quality of care. Research has begun to appear
suggesting that hospital gardens also increase staff satisfaction with the workplace, and
can be advantageous in hiring and retaining qualified personnel. The potential for
hospital gardens to improve medical outcomes, satisfaction, and economic outcomes is
notably increasing the attention and priority accorded to gardens, as administrators and
providers everywhere face strong pressures to increase quality, become more
consumer/patient oriented, control costs, and in some locations establish a positive
market identify in the face of strong competition from other providers.
Cooper-Marcus, C. and M. Barnes (1995). Gardens in Healthcare Facilities: Uses,
Therapeutic Benefits, and Design Recommendations. Martinez, CA: The Center for
Cooper-Marcus, C. and M. Barnes (1999). Healing Gardens: Therapeutic Benefits and
Design Recommendations. New York: John Wiley.
Gierlach-Spriggs, N. Kaufman, R. E., and S. B. Warner, Jr. (1998). Restorative Garden:
The Healing Landscape. New Haven: Yale University Press.
Hefferman, M. L., Morstatt, M., Saltzman, K., and L. Strunc (1995). A Room with a
View Art Survey: The Bird Garden at Duke University Hospital. Unpublished
research report, Cultural Services Program and Management Fellows Program, Duke
University Medical Center, Durham, NC.
Hartig, T. (1991). Testing restorative environments theory. Unpublished doctoral
dissertation, Program in Social Ecology, University of California, Irvine.
Heerwagen, J. (1990). The psychological aspects of windows and window design. In K.
H. Anthony, J. Choi, and B. Orland (Eds.), Proceedings of 21st annual conference of
the Environmental Design Research Association. Oklahoma City: EDRA, 269-280.
Horsburgh, C. R. (1995). Healing by design. New England Journal of Medicine, 333:
Nakamura, R. and E. Fujii (1990). Studies of the characteristics of the
electroencephalogram when observing potted plants: Pelargonium hortorum
“Sprinter Red” and Begonia evansiana. Technical Bulletin of the Faculty of
Horticulture of Chiba University, 43: 177-183. (In Japanese with English summary)
Nakamura, R. and E. Fujii (1992). A comparative study of the characteristics of the
electroencephalogram when observing a hedge and a concrete block fence. Journal of
the Japanese Institute of Landscape Architects, 55: 139-144. (In Japanese with
Nightingale, F. (1860) (1996). Notes on Nursing (Revised with Additions). London:
Picker Institute and Center for Health Design (1999). Assessing the Built Environment
from the Patient and Family Perspective: Health Care Design Action Kit. Walnut
Creek, CA: The Center for Health Design (www.healthdesign.org)
Sadler, B. (2001). Design to Compete in Managed Healthcare. Facilities Design &
Ulrich, R. S. (1981). Natural versus urban scenes: Some psychophysiological effects.
Environment and Behavior, 13: 523-556.
Ulrich, R. S. (1984). View through a window may influence recovery from surgery.
Science, 224: 42-421.
Ulrich, R. S. (1991). Effects of health facility interior design on wellness: Theory and
recent scientific research. Journal of Health Care Design, 3: 97-109. [Reprinted in:
Marberry, S.O. (Ed.) 1995. Innovations in Healthcare Design. New York: Van
Nostrand Reinhold, pp. 88-104.]
Ulrich, R. S. (1999). Effects of gardens on health outcomes: Theory and research. In C.
Cooper-Marcus & M. Barnes (Eds.), Healing Gardens: Therapeutic Benefits and
Design Recommendations. New York: John Wiley, pp. 27-86.
Ulrich, R. S. (2001). Effects of healthcare environmental design on medical outcomes. In
A Dilani (Ed.) Design and Health: Proceedings of the Second International
Conference on Health and Design. Stockholm, Sweden: Svensk Byggtjanst, 49-59.
Ulrich, R. S. (2002). Communicating with the healthcare community about plant benefits.
In C. Shoemaker (Ed.) Proceedings of the Sixth International People Plant
Symposium. Chicago: Chicago Botanic Garden.
Ulrich, R. S., Lundén, O., and J. L. Eltinge (1993). “Effects of exposure to nature and
abstract pictures on patients recovering from heart surgery.” Paper presented at the
Thirty-Third Meeting of the Society for Psychophysiological Research, Rottach-
Egern, Germany. Abstract in Psychophysiology, 30 (Supplement 1, 1993): 7.
Ulrich, R. S. and R. Parsons (1992). Influences of passive experiences with plants on
individual well-being and health. In D. Relf (Ed.), The role of horticulture in human
well-being and social development. Portland, OR: Timber Press, pp. 93-105.
Ulrich, R. S., Simons, R. F., Losito, B. D., Fiorito, E., Miles, M. A., & Zelson, M. (1991).
Stress recovery during exposure to natural and urban environments. Journal of
Environmental Psychology, 11: 201-230.
Whitehouse, S., Varni, J. W., Seid, M., Cooper-Marcus, C., Ensberg, M. J., Jacobs, J. J.
and R. S. Mehlenbeck (2001). Evaluating a children’s hospital garden environment:
Utilization and consumer satisfaction. Journal of Environmental Psychology, 21: 301-