Journal of Environmental Psychology (1996) 16, 221–233 0272-4944/96/030221+13$18.

00/0
 1996 Academic Press Ltd

ENVIRONMENTAL
Journalof

PSYCHOLOGY
RESTORATIVE QUALITIES OF FAVORITE PLACES

KALEVI KORPELA* AND TERRY HARTIG†

*Department of Psychology, University of Tampere, Finland †School of Public Health,
University of California, Berkeley, CA, U.S.A.

Abstract

Theory and research dealing with place identity and restorative environments have for the most part
proceeded independently. Assuming that emotional- and self-regulation are processes underlying the develop-
ment of place identity, and that a person’s favorite place is an exemplar of environments used in such regu-
lation processes, the present study goes beyond preliminary observations about restorative aspects of favorite
places to consider how individuals evaluate their favorite places using terms set out in restorative environ-
ments theory. Finnish university students (n=78) evaluated the central square of their city (Tampere) and
favorite and unpleasant places of their own designation using the Perceived Restorativeness Scale (PRS), an
instrument based on attention restoration theory. Consistent with notions of self-regulation, PRS subscale
scores for Being Away, Fascination, Coherence, and Compatibility were all high in the favorite place evalu-
ations, but Coherence and Compatibility were reliably higher than Being Away, which was in turn reliably
higher than Fascination. Also, PRS subscale scores for the favorite places were reliably higher than those for
the central square, which were in turn higher than those for the unpleasant places. Furthermore, differences
were also found in self-reported emotional states associated with each place. The discussion suggests ways to
develop further mutually reinforcing relations between restorative environments research and research on
place identity.  1996 Academic Press Limited

Introduction freedom of expression and escape from social

Questions about place identity and restorative
environments are receiving more attention from
researchers in the environment–behavior-design
field (e.g. Stokols, 1995). Work in these areas has
for the most part proceeded independently. How-
ever, Korpela (1991) has suggested that restorative
experiences figure in emotion- and self-regulation
processes through which individuals develop
place identity. Taking a person’s favorite place as
one to which the person has become attached,
studies by Korpela (1989, 1991, 1992) have provided
self-report data which suggest that adolescents
use their favorite places to anchor emotion- and
self-regulation (see also Silbereisen et al., 1986;
Noack & Silbereisen, 1988; Owens, 1988;
Silbereisen & Noack, 1988; Sommer, 1990). Korpela
found that his subjects often went to their favorite
places to relax, to calm down, and to clear their
minds after threatening or emotionally negative
events. In addition, experiences of beauty, control,
221
pressures were reported. Favorite places, it
appears, enable experiences like those of interest in
research on restorative environments (e.g. Hartig et
al., 1991; Kaplan & Kaplan, 1989; Ulrich et al.,
1991). The present study goes beyond these prelimi-
nary observations to consider how individuals
evaluate their favorite places using terms set out in
restorative environments theory. It is thus a more
deliberate effort to bring research on place identity
into a mutually reinforcing relationship with res-
torative environments research. Some foundations
for such a relationship are identified in the follow-
ing overview of theory in each of the areas to be
joined.
Place identity, emotion-regulation, and self-
regulation
Proshansky et al. (1983) conceived of place identity
as ‘a substructure of self-identity which consists of
222 K. Korpela and T. Hartig
. . . cognitions about the physical world in which the
individual lives’ (p. 59). They added:
These cognitions are memories, ideas, feelings, atti-
tudes, values, preferences, meanings, and concep-
tions of behavior and experience which relate to the
variety and complexity of physical settings that
define the day-to-day existence of every human
being. At the core of such physical environment-
related cognitions is the ‘environmental past’ of the
person; a past consisting of places, spaces, and their
properties which have served instrumentally in the
satisfaction of the person’s biological, psychological,
social, and cultural needs. (p. 59).
In elaborating on this conceptualization, Korpela
(1989) addressed the criticism Proshansky et al.
received from Sarbin (1983) for not applying an
organizing principle that would explain how con-
duct might be influenced by place identity at any
given time. Building on work by Epstein (e.g., 1983,
1994) and Vuorinen (1990), Korpela proposed that
place identity is constituted in on-going processes of
emotion- and self-regulation; that is, physical
environments can be used to regulate pleasure/pain
and self-experience, and place identity is partly for-
med by the experiences and cognitions produced in
those regulation processes. 1
Ideas about emotion- and self-regulation provide
means to join the place identity concept with res-
torative environments theory (Korpela, 1989, 1992,
1995). Emotion-regulation is not only an inner
homeostatic process, but one that also involves
interaction with the environment (Vuorinen, 1990;
Dodge & Garber, 1991; Fuhrer et al., 1993; Kaiser,
1993). Emotion-regulation has been defined with
reference to the ‘intra- and extraorganismic factors
by which emotional arousal is redirected, controlled,
modulated, and modified to enable an individual to
function adaptively in emotionally arousing situ-
ations’ (Cicchetti et al., 1991, p. 15). Of the types of
emotion-regulation system discussed by Izard and
Kobak (1991), it is the intentional and self-initiated
type that may best encompass the use of physical
environments. Other forms of intentional and self-
initiated emotion-regulation include self-soothing
and gratification techniques; techniques for mod-
ifying expressive behavior; skeletal-muscular tech-
niques such as play and exercise; and perceptual-
cognitive techniques such as attention focusing and
shifting, self-instructions, or self-monitoring.
With respect to self-regulation, Epstein (1983,
1985, 1994) assumes that human beings construct a
personal theory of themselves and the world. This
personal theory of reality exists at a preconscious
level of awareness and directs behavior. It is not
developed for its own sake, but for the purpose of
fulfilling four fundamental functions. In other
words, four principles governing motivation, cog-
nition and action are set forth. These principles are
consistent with psychoanalytical, learning and
phenomenological theories and integrate them
(Epstein, 1983, 1985). One principle is the need to
maximize the pleasure/pain balance. It corresponds
to the pleasure principle in Freudian theory and to
the principle of reinforcement in learning theory. A
second principle—the unity principle—is the need
to assimilate the data of reality into a coherent con-
ceptual system and to maintain this system
(Epstein, 1983). This need is consistent with a need
in all higher animals to establish an adequate rep-
resentation of their environment so that they can
cope with demands more flexibly and efficiently
than if their responses were instinctive reactions to
discrete stimuli (Epstein, 1981). A third principle is
the need to maintain a favorable level of self-
esteem; it is derived from the pleasure principle,
and becomes important once a self-theory has been
formulated. The fourth principle is the need to
maintain relatedness, or favorable relationships
with significant others. The principles are assumed
to overlap and interact. For example, pleasure is
usually experienced when there are elevations in
self-esteem and when previously unassimilable
material is assimilated into a person’s self-experi-
ence. Displeasure is experienced when self-esteem
is diminished or when there is a failure in assimi-
lation (Epstein, 1990). The four principles are all of
basic importance, and behavior is viewed as a
compromise among them (Epstein, 1985). For
example, individuals are generally motivated to set
a level of self-esteem only as high as they can
realistically maintain without exposing themselves
to frequent disappointments (Epstein, 1985).
Emotion- and self-regulation imply the likelihood
of periodic changes from negative to more positive
emotions and self-cognitions. Such changes are also
characteristic of restoration. In describing processes
through which person–environment interactions aid
change from negative antecedents to more positive
states, theories about restorative experience and
restorative environments thus also reveal processes
that may promote the development of place identity
(cf. Fuhrer et al., 1993). Furthermore, one perspec-
tive on restoration, attention restoration theory
(Kaplan & Kaplan, 1989; Kaplan, 1995), goes
beyond more momentary changes in psychological
and physiological states to encompass reflection on
oneself and one’s place in the world as an aspect of
restorative experience. This is presumably import-
ant for self-regulation; reflection may aid in the
 Favorite Places
assimilation of new experiences into one’s self-
theory and in the maintenance of one’s conceptual
systems.
Restoration and restorative environments
Attention restoration theory (e.g. Kaplan & Kaplan,
1989; Kaplan, 1995) and Ulrich’s psychophysiolog-
ical stress reduction framework (Ulrich, 1983;
Ulrich et al., 1991) dominate the restoration per-
spectives in environmental psychology. These theor-
ies differ in the emphasis placed on emotional,
physiological, and attentional factors in their speci-
fications of both antecedent conditions and the res-
toration process. Briefly, the stress reduction frame-
work (Ulrich, 1983; Ulrich et al., 1991) assumes an
antecedent condition of psychophysiological stress,
defined as a process of responding emotionally,
physiologically, and behaviorally to a situation in
which well-being is challenged or threatened.
Consequences of stress include negative emotions
and heightened autonomic arousal. Restoration is
brought about in a visual encounter with a scene
having particular properties, such as natural con-
tents, moderate complexity, and the presence of a
focal point. Perception of these properties prompts a
shift toward more positively-toned emotional states,
drives down activity in different physiological sys-
tems, evokes sustained attention, and blocks nega-
tive emotions and thoughts.
In contrast, attention restoration theory (e.g.
Kaplan & Kaplan, 1989; Kaplan, 1995) posits an
antecedent condition of directed attention fatigue,
which is incurred in any prolonged mental effort
which requires steady exercise of the inhibitory
mechanism assumed to make directed attention
possible. Among the negative corollaries of directed
attention fatigue are irritability, an inability to
plan, a reduced sensitivity to interpersonal cues,
and increased likelihood of errors in performance.
Restoration can proceed when four factors charac-
terize the person–environment interaction. One,
being away, involves getting psychological and poss-
ibly geographical distance from one’s usual context,
including the work one ordinarily does and the pur-
suit of particular goals and purposes. A second fac-
tor is fascination, or effortless attention. When func-
tioning is supported by fascination, efforts at
inhibiting distractions can be relaxed and directed
attention capacity can be restored. Fascination can
be engaged by environmental contents such as
water or by processes of exploring and making sense
of an environment. A third factor, extent, refers to
the possibility for immersion in a coherent physical
223
or conceptual environment that is of sufficient scope
to sustain exploration and interpretation. Finally,
compatibility refers to a match between personal
inclinations and purposes, environmental supports
for intended activities, and environmental con-
straints on action (see also Kaplan, 1983).
Other differences between the stress reduction
and attention restoration frameworks have to do
with the duration of the restorative process and the
immediacy, depth, and persistence of effects. As
presently formulated, the psychophysiological
stress reduction approach emphasizes the first
moments of an encounter with an esthetically pleas-
ing scene. Possible changes in affect and
arousal then initiated are modeled in some detail
(Ulrich, 1983). Longer-term effects are possible
(Parsons, 1991), but as yet have been little dis-
cussed. In contrast, attention restoration theory
allows for a potentially long duration in a restora-
tive experience, such that a person may pass
through successive levels (Kaplan & Kaplan, 1989).
The first level involves ‘clearing the head’. The
second is recovery of directed attention capacity.
The third entails facing accumulated matters on
one’s mind. The fourth is reflection on one’s priorit-
ies, prospects, actions and goals in life. Penetration
into successive levels requires increasing time and
intensity of the factors thought to work in restora-
tive experience. Thus, this theory describes a broad
temporal range within which restoration and its
effects might be seen.
In describing possible restoration processes and
outcomes, both the attention restoration and stress
reduction frameworks add to an understanding of
emotion- and self-regulation phenomena. Both
frameworks also point to qualities of person–envi-
ronment interactions that might contribute to the
development of place identity. However, by allowing
for greater temporal extension of the restoration
process and a greater range of effects, including
changes in perspectives on oneself and one’s place
in the world (see e.g. Kaplan & Talbot, 1983), atten-
tion restoration theory allows greater latitude for
relating restorative experiences to place identity as
an outcome of environmental self-regulation.
Appropriately, attention restoration theory pro-
vides means for relating restoration to place ident-
ity. In its specification of operative factors, atten-
tion restoration theory acknowledges the interests,
purposes and environmental constraints that can
structure people’s lives and activities. An open
question has to do with the varying significance of
the four factors for environmental self-regulation.
Compatibility might seem to be most closely aligned
224 K. Korpela and T. Hartig
with the idea of self, in that it explicitly draws in
the individual’s own purposes and inclinations. Fur-
thermore, high degrees of compatibility are
assumed to be necessary for the kind of reflection
thought to contribute to the deep restoration which
may most involve one’s sense of self. Yet it is
thought that high compatibility cannot be found in
situations lacking high degrees of being away, fasci-
nation, and extent. Moreover, having a sense of
being away in a particular environment, being fasci-
nated by what is found there, and finding it to be of
some coherence and extent are all behaviors pre-
sumably susceptible to processes that also work in
the development of self. Thus, when thinking of res-
torative experiences available in places to which
people are attached, such as favorite places, the
relative importance of the different factors is a
question in need of investigation (cf. Hartig et al.,
1991).
Study objectives
Favorite places appear to afford restorative experi-
ences that aid emotion- and self-regulation pro-
cesses which are basic to the development of place
identity. Attention restoration theory (Kaplan &
Kaplan, 1989; Kaplan, 1995) provides a set of con-
structs that are useful for understanding restora-
tive person–environment interactions. One aim of
the present study was to examine the absolute
levels and relative importance of these factors in
people’s experiences of their favorite places, as
reflected in evaluations obtained with the Perceived
Restorativeness Scale (PRS) (Hartig et al., 1996).
The study also compared the evaluations of favorite
places with evaluations of a place in the everyday
public life of all subjects and an unpleasant place of
each subject’s designation. The two additional
places were likely to have different, lower restora-
tive potentials; the intent was to examine how the
restorative factors varied across environments the
subjects might move among and to which they
would have different degrees of attachment. Fur-
thermore, the subjects used an emotional state scale
to report how they would feel in each place. Aside
from portraying the emotions felt in each place, the
added scale made it possible to examine the
relations between those emotions and perceptions of
restorativeness.
Another concern here was with the kind of places
that become favored. Building on assumptions
about psychological and psychophysiological mech-
anisms evolved in natural environments, restora-
tive environments theory posits that natural
environments in general will be more restorative
than many of the urban environments in which
people have concentrated themselves. Some pre-
liminary experimental evidence supports this
assertion (Ulrich, 1979; Hartig et al., 1991; Ulrich et
al., 1991). Although the extant theory is open to the
idea that other places serve restoration, few of the
studies in this area have considered other than
natural and urban environments (for an exception,
see Kaplan et al., 1993). In contrast, previous
research on favorite places (e.g. Korpela, 1992)
found that subjects did frequently identify natural
environments as their favorite places, but that they
also named a variety of other settings, such as local
cafés or their room at home. It was of interest
whether the favorite places identified in the present
study would be similarly variable.
The present study also served a methodological
function. The reliability and validity of the PRS has
been assessed in other studies by having subjects
evaluate a set of sites expected to differ in theoreti-
cally meaningful terms, with the sites presented in
different ways (on-site, video, slides) to subjects
from different populations (Hartig et al., 1996). In
the present study, subjects evaluated the favorite,
everyday, and unpleasant places on the basis of
their memory or imagination. Thus, two of the
environments were the same across subjects only in
that they were identified as either a favorite or an
unpleasant place. Having each subject refer to his
or her own favorite and unpleasant place allowed
for variability in the characteristics of the environ-
ments while holding the personal significance of the
place roughly constant. Also, not providing any sen-
sory information about the environment in question
is an appropriate next step beyond simulations;
arguably, the commonly used (visual) simulation
methods already require that a subject draw on
memory or imagination to flesh out the visual and
nonvisual aspects of the setting not provided in the
simulation. It was of interest whether reliability
and validity information obtained with the use of
these methods would be consistent with that
obtained previously (cf. Craik & Feimer, 1987).
Method
Design
All subjects evaluated seven settings. The first was
Tampere’s central square, a setting chosen because
of its presumed familiarity. The next two settings
were a favorite and an unpleasant place in the
 Favorite Places
given subject’s life. Some subjects (n=48) evaluated
a favorite place first, the others (n=30) an
unpleasant place. Subjects were instructed to
imagine the setting while completing their evalu-
ations. The remaining settings were a rock garden,
an outdoor shopping mall, a study room with plants,
and the bottom floor of a parking garage, all on or
near the campus of the University of California,
Irvine (UCI). Each of these was represented with a
selection of color slides, and the simulations were
randomly ordered for each group of subjects being
led through the procedure. The rationale for the use
of these four sites is discussed elsewhere (Hartig et
al., 1996), as are the results based on their evalu-
ation by the present sample.
Subjects
Undergraduate students (49 female, 29 male) from
the University of Tampere volunteered for partici-
pation. Neither course credit nor monetary compen-
sation was provided. Subjects’ ages ranged from 19
to 46 years (M=25·46, S.D.=6·19).
Measures
Hartig and colleagues (Hartig et al., 1996) have
initiated development of a Perceived Restora-
tiveness Scale (PRS) to represent factors set out in
attention restoration theory (Kaplan & Talbot,
1983; Kaplan & Kaplan, 1989; Kaplan, 1995). Six-
teen items are intended to tap the factors Being
Away (e.g. Spending time here gives me a good
break from my day-to-day routine); Fascination (e.g.
My attention is drawn to many interesting things);
Coherence (e.g. There is a great deal of distraction);
and Compatibility (e.g. I have a sense that I belong
here). Note that the Coherence items are intended
to tap an aspect of Extent, and that they all define
coherence in negative terms (e.g. as with regard to
pervasive distraction).
In the present study the 16 PRS items were
included in a form comprised of 44 items altogether.
Some of the additional items were to represent
other constructs (e.g. preference, familiarity), and
some were included to serve scale development pur-
poses. For each item, respondents use a 7-point
scale (0=Not at all; 6=Completely) to indicate the
extent to which the statement describes their
experience in the given setting. Four random item
orders were balanced across subjects; each subject
used the same version of the form for all evalu-
ations. For the central square and the favorite and
unpleasant places, the subjects were asked to
225
imagine themselves in the setting and to then indi-
cate the extent to which each item applied to their
experience there.
Validation studies in the United States and
Sweden (Hartig et al., 1996) have separately ana-
lysed the evaluations of each of the four UCI sites
mentioned above. With relatively few exceptions,
adequate internal consistencies were found in each
site in each of three studies for the a priori Being
Away, Fascination, Coherence, and Compatibility
subscales. Principal factor analyses of the same
data indicated that the 4-factor solution, obliquely
(oblimin) rotated to allow for potentially correlated
factors, was less stable across sites than was a
solution in which one factor was loaded on primarily
by Being Away, Fascination, and Compatibility
items and a second was loaded on by Coherence
items. For each of the four sites in each of the three
studies, composite scores based on the 2-factor
solution showed reasonable patterns of positive,
negative, and null correlations with other measures
of environmental characteristics and with self-
reported emotions. The composite scores also were
useful in differentiating between the sites in pre-
dicted and theoretically meaningful ways. Further-
more, for each of the sites similar evaluations were
obtained on-site and with simulations. A fourth
study reported by Hartig et al. (1996) involved the
present subjects’ evaluations of the four UCI sites
and produced results consistent with those obtained
previously.
The Zuckerman Inventory of Personal Reactions
and Feelings (ZIPERS) was used to assess
emotional states (Zuckerman, 1977). Twelve items
represent the factors Fear Arousal, Sadness,
Anger/Aggression, Positive Affect, and Attentive-
ness. The instructions were modified to obtain sub-
jects’ views on how they would feel if they were in
the given setting, imagining that they were actually
within the setting. Responses were given on 5-point
scales (1=Not at all, 5=Very Much). The ZIPERS
has been a sensitive measure in previous exper-
imental research on the relative restorative poten-
tials of natural and urban environments (Ulrich,
1979; Hartig et al., 1991; Ulrich et al., 1991). It was
expected that the PRS subscales would correlate
positively with Positive Affect and Attentiveness
and negatively with Sadness, Fear Arousal, and
Anger/Aggression.
A 16-item form was used to obtain descriptions of
the favorite and unpleasant places. Subjects used a
7-point scale (0=Not at all; 6=Completely) to indi-
cate the extent to which certain features were pre-
sent or the extent to which the place could be
226 K. Korpela and T. Hartig

characterized in a particular way. The items
referred to the presence of shops, other people, per-
sonal belongings, traffic, water, plants/greenery,
peace and quiet, beautiful views, sunlight, and fac-
tories, and the extent to which it could be described
as urban or natural, inside or outside. An open-
ended item enabled mention of other features or
qualities that a subject considered salient. Finally,
subjects were asked how often they visited their
favorite place.
Procedure
Groups of subjects (n=3–11) were scheduled for ses-
sions of approximately 65 minutes duration. At the
beginning of each session subjects were given a
packet which included an instruction sheet and
seven sets of forms, with the ZIPERS stapled in
front of the PRS. The form for describing favorite
and unpleasant places was stapled in front of the
ZIPERS in the second and third sets. Subjects were
asked to first complete the evaluation of the central
square and to then proceed to the evaluations of
favorite and unpleasant places. In each of these
they were to draw on their memory of place and to
imagine being in it while completing their evalu-
ation. When all subjects were finished with the first
three sites, a brief pause was taken. The subjects
were then shown the slides of the first UCI site in
the given random sequence. After the slides for the
site had been shown, the subjects were directed to
fill out the questionnaires while bearing in mind the
place they had just seen. This process was repeated
without pause until the subjects had evaluated all
four of the simulated sites.
Results
Favorite and unpleasant places and their
characteristics
For both favorite and unpleasant places, the sub-
jects’ selections varied considerably in their physi-

TABLE 1
Favorite and unpleasant places identified by Finnish university students (n=78)
Favorite places Unpleasant places
Home n Traffic or Crowds n
home or own room 14 a busy street or tunnel 15
home yard 5 a large store 12
childhood home 1 a rush-hour crowd 7
the bus-railway station 3
Water a parking lot 3
a lake shore 13 a train bridge 2
at sea in a boat 2 a traffic jam 1
an island 2 a bus stop 1
on lake ice 1 being with people 1

Greenery Urban Locations

a wooded area 8 a suburb 8
a recreation area 4 an industrial area 3
a city park 3 a street 2
a square 1
Urban Locations a yard outside a high-rise 1
a restaurant or cafe 4 a disco 1
a building 4 the inland revenue 1
a suburb 3 a library 1
an ice hockey hall 1
the old town 1 Various Functional Settings
a bus stop 1 a university hall 3
a cinema 1 a lunch room 2
a library 1 a dentist’s waiting room 2
a workplace 2
Summer Cottage 6 a women’s bathroom 1
a lecture hall 1
Scenic View 3 a hospital ward 1
in an exam 1
a student dormitory 1
 Favorite Places 227

cal and social characteristics (see Table 1). Many of
the favorite places were natural settings.
Unpleasant places often had traffic or crowds.
For both favorite and unpleasant places, categor-
ies in Table 1 overlap to some degree. For example,
the Finnish summer cottage is typically located in
the countryside or near the coast, and may be
favored in large because of its natural surroundings
and scenic views. This overlap is reflected in the
ratings of the various characteristics of the places
each subject identified. So, for example, with
respect to favorite places, 83·3 per cent of the sub-
jects noted the presence of a beautiful view (vs 2·6%
for unpleasant places); 83·1 per cent (vs 16·7%)
noted that sunlight was at least rather much pre-
sent; 73·1 per cent (vs 17·9%) indicated that water
was at least rather much present, and 56·4 per cent
(vs 2·6%) noted that personal belongings were
rather much or present. Conversely, factories were
rather much or more present in the unpleasant
places of 48·1 per cent of the subjects (vs 2·6% for
the favorite places), and unpleasant places were
described as rather much or more urban by 84·6 per
cent of subjects (vs 26% for favorite places).
Finally, subjects varied in the frequency with
which they visited their favorite places, with a
range from zero to 21 visits in a typical week. One
half reported that they visited only once weekly,
while 24 per cent indicated that they visited daily.
As for the unpleasant places, 33 per cent of the sub-
jects visited at least seven times each week, while
31 per cent visited from two to six time weekly
(median=4).
Psychometric properties of the Perceived
Restorativeness Scale
Validation studies have yielded evidence of suf-
ficient reliability and validity for the Perceived Res-
torativeness Scale (PRS) (Hartig et al., 1996). How-
ever, the previous studies analysed evaluations
obtained on-site or with simulations. Of interest
here was whether the psychometric properties of
the PRS as used to evaluate imagined places would
be consistent with those described previously.
The results do in fact largely agree with those
from the earlier validation studies. Analyses of the
internal consistency of each a priori subscale in the
evaluations of each place show 11 of the 12 Cron-
bach’s α values falling in the range 0·74 to 0·92

TABLE 2
For the favorite place evaluations, Perceived Restorativeness Scale (PRS) item communalities, factor structure matrices,
and factor correlation matrices from the 4- and 2-factor solutions
PRS items arranged 4-Factor Solution 2-Factor Solution
by a priori subscale
h2 I II III IV h2 I II
Being Away 1 0·46 0·27 −0·05 0·24 0·67 0·17 0·41 −0·12
2 0·88 0·47 −0·24 0·29 0·91 0·39 0·61 −0·25

Fascination 3 0·56 0·72 −0·07 0·26 0·44 0·54 0·73 −0·06

4 0·67 0·78 0·16 0·24 0·43 0·68 0·76 0·17
5 0·56 0·73 −0·06 0·30 0·42 0·56 0·74 −0·06
6 0·79 0·83 0·16 0·33 0·51 0·81 0·85 0·13
7 0·69 0·79 −0·27 0·36 0·43 0·65 0·80 −0·26

Coherence 8 0·79 −0·11 0·88 −0·26 −0·04 0·64 −0·17 0·80

9 0·20 0·13 0·39 −0·16 0·09 0·21 0·08 0·43
10 0·41 −0·16 0·62 −0·27 −0·15 0·40 −0·23 0·62
11 0·40 −0·11 0·61 −0·06 −0·12 0·26 −0·14 0·51

Compatibility 12 0·42 0·61 −0·13 0·37 0·14 0·35 0·59 −0·16

13 0·82 0·40 −0·27 0·90 0·17 0·43 0·55 −0·46
14 0·72 0·29 −0·21 0·85 0·26 0·33 0·47 −0·42
15 0·48 0·33 −0·13 0·68 0·25 0·28 0·47 −0·32
16 0·69 0·79 −0·17 0·27 0·11 0·45 0·67 −0·12

I II III I
Factor Correlations II −0·06 −0·19
III 0·34 −0·24
IV 0·39 −0·04 0·21
An item’s highest factor-variable correlation is given in bold type.
228 K. Korpela and T. Hartig

(median=0·765). 2 The exception is with the Coher-
ence subscale in the favorite place, for which α=
0·68. To examine the correspondence between the a
priori subscales and possible empirical factors, prin-
cipal factor analyses were carried out as in the pre-
vious studies, with squared multiple correlations as
the starting values for the communality estimate
procedure, four factors specified for extraction, and,
allowing for correlations among the factors, oblique
(oblimin) rotation. This solution accounted for 63·9,
59·5, and 50·0 per cent of the total variance in the
evaluations of the central square, favorite place,
and unpleasant place, respectively. The items defin-
ing the different factors in the 4-factor solution
obtained for each place were for the most part
grouped according to a priori subscale. Results of
the factor analyses for the favorite and unpleasant
places are given in Tables 2 and 3 (regarding the
results for the central square see note 2).
As in the previous studies, the 4-factor solution
included cross-loadings >0·40 and correspondingly
substantial correlations between some factors. This
was particularly the case for the central square.
Consequently, following with the previous studies, a
second factor analysis was completed for each site,
as before but with the extraction of two factors
specified. The result was much the same as was
seen in the previous studies; Being Away, Fasci-
nation, and Compatibility items load primarily on
one factor and Coherence items load on the second
(see Tables 2 and 3, Note 2). The 2-factor solution
accounted for 56·9, 44·7, 35·5 per cent of the total
variance in the evaluations of the central square,
favorite place, and unpleasant place, respectively;
thus, there were reductions in variance accounted
for of 7, 14·8, and 14·5 per cent, respectively, over
the 4-factor solution for each site. Note that if a
standard extraction criterion (eigenvalues >1·00) is
applied, the number of factors extracted varies
across places as follows: central square=two;
favorite places=4; unpleasant places=5 (an
additional 5·1% of variance accounted for as com-
pared to the 4-factor solution).
The results of the factor analyses are somewhat
equivocal with respect to the calculation of com-
posite scores for use in subsequent analyses; the
scores for different sites could be based on either the
a priori subscales or on the 2-factor solution. The
latter strategy was more clearly indicated in the
earlier validation studies by Hartig et al. (1996).

TABLE 3
For the unpleasant place evaluations, Perceived Restorativeness Scale (PRS) item communalities, factor structure
matrices, and factor correlation matrices from the 4- and 2-factor solutions
PRS items arranged 4-Factor Solution 2-Factor Solution
by a priori subscale
h2 I II III IV h2 I II
Being Away 1 0·63 0·55 −0·07 0·70 0·25 0·37 0·61 0·00
2 0·78 0·07 0·05 0·86 0·11 0·08 0·28 0·08

Fascination 3 0·40 0·50 −0·17 0·18 0·53 0·38 0·61 −0·11

4 0·56 0·39 −0·02 0·25 0·73 0·39 0·62 0·04
5 0·41 0·28 −0·07 0·40 0·57 0·28 0·53 0·00
6 0·29 0·07 −0·07 −0·03 0·51 0·07 0·26 −0·04
7 0·57 0·50 −0·03 0·34 0·69 0·51 0·71 0·04

Coherence 8 0·45 −0·10 0·66 0·05 −0·13 0·44 −0·13 0·65

9 0·39 0·25 0·56 0·00 0·00 0·33 0·14 0·56
10 0·61 −0·23 0·74 −0·08 −0·27 0·63 −0·31 0·73
11 0·62 −0·09 0·71 −0·04 0·27 0·50 0·14 0·69

Compatibility 12 0·17 0·31 −0·13 0·27 0·29 0·17 0·41 −0·08

13 0·73 0·85 −0·13 0·27 0·36 0·54 0·74 −0·06
14 0·34 0·53 0·09 0·07 0·35 0·27 0·50 0·13
15 0·76 0·86 −0·01 0·25 0·20 0·41 0·64 0·04
16 0·29 0·43 −0·19 0·34 0·33 0·29 0·52 −0·14

I II III I
Factor Correlations II −0·04 −0·02
III 0·27 −0·06
IV 0·37 −0·09 0·20
An item’s highest factor-variable correlation loading is given in bold type.
 Favorite Places 229

However, in the present study the remaining
results are based on composite scores for the four a
priori subscales. The 4-factor solution more
adequately represents the variability in evaluations
of the favorite and unpleasant places, and it also
allows a qualitatively richer description of the
experiences available in the different places.
Experiential qualities of favorite and unpleasant
places
On a scale from 1–7 (recoded from the original 0–6)
scale, the favorite place ratings are near the high
end for all of the PRS subscales (see Table 4). Com-
patibility and Coherence appear to be most charac-
teristic of the subjects’ experiences in their favorite
places, and in about equal degree; a paired-samples
t-test suggests that the mean scores for those subs-
cales do not reliably differ from one another. Com-
patibility (and so by implication Coherence) is, how-
ever, significantly higher than Being Away (and so
Fascination) (t (76)=1·98, p=0·05). Finally, Being
Away seems to be more characteristic of the sub-
jects’ favorite place experiences than is Fascination
(t (74)=2·31, p<0·05).
Parallel analyses were conducted for the central
square and the unpleasant places to determine
whether the relative standing of the PRS subscale
scores was the same as in the favorite place evalu-
ations. As seen in Table 4, the mean Fascination
and Coherence scores for the central square were
both at moderate levels and not reliably different
from one another. Coherence (and so Fascination)
was, however, reliably higher than Compatibility
(t (72)=5·44, p<0·001), which was in turn reliably
higher than Being Away (t (75)=2·79, p<0·01). In
contrast, in the unpleasant places mean Coherence
was at a moderate level but was reliably greater
than Fascination (t (73)=6·12, p<0·001). Fascination
was in turn only marginally higher than Being
Away (t (74)=1·97, p=0·05), which was in turn
higher than Compatibility (t (76)=2·62, p=0·01).
These results not only point to the varying signifi-
cance of the different factors across experiences in
different places, but also indicate that the PRS is
sensitive to differences at the subscale level; the
subscales do not merely move up or down in tandem
across environments.
With respect to differences between the environ-
ments in terms of the PRS subscale scores, paired-
samples t-tests confirm that the favorite places
were given higher mean ratings than the central
square in terms of Being Away (t (76)=13·80), Fasci-
nation (t (73)=8·31), Coherence (t (72)=14·23), and
Compatibility (t (75)=18·66) (for all, p <0·0001). The
tests likewise confirm that the central square
received higher evaluations than the unpleasant
places in terms of Being Away (t (76)=3·57, p <
0·001), Fascination (t (73)=7·90, p<0·0001), and
Compatibility (t (74)=8·10, p<0·0001). However, the
difference between the central square and the
unpleasant places in terms of Coherence scores only

TABLE 4
Descriptive statistics for the Perceived Restorativeness Scale (PRS) and Zuckerman Inventory of Personal Reactions
(ZIPERS) subscales for each site
Subscale Site
Favorite Place Central Square Unpleasant Place
M (S.D.) M (S.D.) M (S.D.)
a b
PRS Being Away 5·47 (1·26) 2·42 (1·28) 1·88 (0·98)
Fascination 5·16 (1·13)a 3·50 (1·28)b 2·13 (0·76)
Coherence 5·89 (0·89)a 3·46 (1·09) 3·10 (1·19)
Compatibility 5·77 (0·81)a 2·71 (1·22)b 1·59 (0·54)

ZIPERS Positive Affect 4·04 (0·66)c 2·26 (0·80)d 1·54 (0·45)

Sadness 1·94 (0·95) 1·83 (0·83)d 2·74 (1·07)
Attentiveness 3·06 (1·20) 2·97 (1·04)e 2·64 (1·07)
Anger/Aggression 1·25 (0·40)c 1·96 (0·78)d 3·17 (0·87)
Fear Arousal 1·69 (0·73)c 1·97 (0·73)d 2·60 (0·95)
PRS values fall on a scale from 1–7, where lower values indicate, for example, lower fascination. ZIPERS values fall on
a scale from 1–5, where lower values indicate, for example, lower sadness. The n’s range from 74–78. aReliably greater
than corresponding score for the central square, p<0·0001; bReliably greater than corresponding score for the
unpleasant place, p<0·001; cReliably different from corresponding score for the central square, p<0·01; dReliably
different from corresponding score for the unpleasant place, p<0·0001; eReliably different from corresponding score for
the unpleasant place, p<0·05.
230 K. Korpela and T. Hartig

verges on statistical significance (p=0·06). Note that
the order in which the favorite and unpleasant
places were evaluated did not affect the evalu-
ations; analysis of variance in the evaluations of
each site produced no statistically significant F-
ratio for any of the subscales.
Differences between the places are also reflected
in subjects’ reports of the emotions they would feel
in each, as assessed with the Zuckerman Inventory
of Personal Reactions (ZIPERS). Subjects associated
higher Positive Affect (t (73)=16·00, p<0·0001) and
lower Anger/Aggression (t (77)=7·55, p<0·0001) and
Fear Arousal (t (75)=2·70, p=0·009) with their favor-
ite place in comparison to the central square. They
associated higher Positive Affect (t (74)=8·02, p<
0·0001) and Attentiveness (t (77)=2·14, p=0·035),
and lower Sadness (t (77)=6·60, p<0·0001),
Anger/Aggression (t (76)=10·23, p<0·0001), and
Fear Arousal (t (75)=6·08, p<0·0001) with the cen-
tral square as compared to the unpleasant places.
Finally, as shown in Table 5, the various PRS
subscale scores were reliably correlated with diff-
erent subscale scores for the ZIPERS. Two sets of
relationships were found in all three sites. One con-
sisted of positive correlations between Being Away,
Fascination, and Compatibility scores on the one
hand and the Positive Affect score on the other.
The other involved negative correlations between
Coherence and ZIPERS Anger/Aggression and Fear
Arousal scores. These results are largely consistent
with expectations and with results from Hartig et
al. (1996), which were discussed with respect to the
criterion validity of the PRS.
Discussion
The environmental self-regulation hypothesis
(Korpela, 1989, 1995) provides a bridge between
restorative environments research and research on
place identity. Taking a person’s favorite place as
an exemplar of environments used in emotion- and
self-regulation, the present study was an initial
empirical examination of how experiences in favor-
ite places might be evaluated using terms set out in
attention restoration theory (Kaplan and Kaplan,
1989; Kaplan, 1995). That favorite place experi-
ences are characterized by high levels of being
away, fascination, coherence, and compatibility
affirms that it is useful to think of favorite places as
sources of restorative experience. Of the restorative
factors, the results suggest that, on average, coher-
ence and compatibility are most characteristic of
favorite place experiences, followed by being away
and then fascination. This ranking conforms to the
idea that maintenance of a coherent conceptual sys-
tem is a fundamental aspect of environmental self-
regulation, one which may rely on opportunities for

TABLE 5
Correlations between Perceived Restorativeness Scale (PRS) subscales and Zuckerman Inventory of Personal Reactions
(ZIPERS) subscales, calculated for each site
Site and PRS Subscale ZIPERS Subscale
PA S A A/A FA
Central Square
Being Away 0·57*** 0·10 0·16 −0·22 −0·06
Fascination 0·65*** 0·00 0·27* −0·28* 0·12
Coherence 0·36*** −0·03 −0·08 −0·49*** −0·34**
Compatibility 0·58*** 0·10 0·32** −0·25 0·05

Favorite Place
Being Away 0·38*** −0·05 0·23 −0·24 0·16
Fascination 0·49*** −0·11 0·11 −0·11 0·22
Coherence 0·14 0·03 −0·20 −0·56*** −0·46***
Compatibility 0·37*** −0·05 0·06 −0·11 −0·01

Unpleasant Place
Being Away 0·31** 0·09 −0·11 −0·16 0·02
Fascination 0·37*** 0·29* 0·17 −0·04 0·10
Coherence 0·13 −0·05 0·01 −0·28* −0·31**
Compatibility 0·50*** 0·04 0·06 −0·24 −0·04
The n’s range from 69 to 78. PA=Positive Affect; S=Sadness; A=Attentiveness; A/A=Anger/Aggression; FA=Fear
Arousal.
*p<0·01; **p<0·005; ***p<0·001.
 Favorite Places
reflection in nondistracting circumstances to which
one may withdraw. Conceivably, some environ-
ments with less personal significance might serve
restoration, yet have different profiles of the four
operative factors; they might thus afford outcomes
differing in temporal extension, depth and other
qualities. For example, a video game might enable
high levels of being away and fascination and some-
what lower levels of compatibility and coherence;
time spent playing one might change mood without
having profound effects on the person’s perspectives
on the world and their place in it.
Arguments about the restorativeness of favorite
place experiences are strengthened by three other
components of the results. First, the favorite places
identified by subjects were, in keeping with the
literature on restorative environments, most often
places with greenery, water, and scenic quality.
Second, the reliable differences between evaluations
of the favorite place, Tampere’s central square, and
the unpleasant place were as expected, with levels
of being away, fascination, coherence, and compati-
bility highest in the favorite places, intermediate in
the central square, and, with the exception of coher-
ence, lowest in the unpleasant places. Thus, the
four restorative factors provided a meaningful basis
for differentiating favorite place experiences from
experiences of the central square and unpleasant
places. Third, the self-reported emotional states
associated with each place were about as expected,
in that the pattern of differences between places in
terms of positive affect, anger/aggression, and fear
arousal paralleled that seen with the PRS scores.
The seeming match between self-reported emotions
and evaluations in terms of the restorative factors is
substantiated to an extent by the various corre-
lations between PRS and ZIPERS scores.
Characterization of the favorite places as restora-
tive environments is here based on scores from an
instrument still in an early stage of development.
Previous studies (Hartig et al., 1996) and the results
presented here show that the present version of the
PRS has generally reliable subscales, is sensitive to
differences between environments, and has
adequate criterion validity (as reflected in corre-
lations between PRS and ZIPERS scores). Ques-
tions remain, however, about the factor structure in
responses obtained with the PRS. As in the previous
studies, the results presented here suggest that a 2-
factor solution is more stable across sites, although
here it accounts for a considerably smaller portion
of the variance in evaluations of the favorite and
unpleasant places than does the 4-factor solution.
As with many instruments used in psychological
231
research, additional studies using alternative item
combinations and analytic strategies (e.g. confirma-
tory factor analysis) will be needed to refine the
PRS further.
The present study suggests several ways to inte-
grate further research on place identity and restora-
tive environments. Views on restorative person–
environment interactions might be modified in par-
allel with the further development of ideas about
emotion- and self-regulation. Such development can
help elaborate the relations between place identity
and constructs such as place attachment. It can
simultaneously consider the possibility that a per-
son’s experience of psychological bonds with a
specific locale may generalize to the experience of
psychological bonds with that type of locale
(Feldman, 1990). It should also consider Lalli’s
(1992) argument that each spatial level on which
place attachment (and place identity) may be mani-
fest needs its own theoretical account and specific
empirical operationalizations (see also Altman &
Low, 1992). This possibility bears on a larger set of
questions concerning the ways in which evolution-
ary, cultural, social, and individual level processes
combine to influence the restorativeness of environ-
ments and so their suitability for service in
emotion- and self-regulation (cf. Hartig, 1993;
Hartig & Evans, 1993).
Acknowledgements
This study was supported by grants to the first
author from the Academy of Finland and the Scien-
tific Foundation of the City of Tampere. Preparation
of this article was partially supported by Grant T32
HL07365-14 from the U.S. National Heart, Lung,
and Blood Institute to the second author. The
authors are grateful to Tommy Gärling for com-
ments on the study design and to Florian Kaiser,
Daniel Stokols, and the anonymous reviewers for
their constructive criticisms.
Notes
Reprint requests and correspondence should be addressed
to Kalevi Korpela, Institute of Migration, Piispankatu 3,
SF-205 00 Turku, Finland.
(1) The common features of Epstein’s and Vuorinen’s the-
ories of self-regulation are described in detail elsewhere
(Korpela, 1989, 1995).
(2) To conserve space, some results bearing on the
psychometric properties of the PRS are not reported here
in detail. For the full set of results from this study and
232 K. Korpela and T. Hartig
the other validation studies completed to date, (see Har-
tig et al., 1996). Validation of a measure of perceived
environmental restorativeness (Göteborg Psychological
Reports, 26:7). Göteborg: Göteborg University, Depart-
ment of Psychology.
References
Altman, I. & Low, S. (1992). Place Attachment. New York:
Plenum Press.
Cicchetti, D., Ganiban, J. & Barnett, D. (1991). Contri-
butions from the study of high-risk populations to
understanding the development of emotion-regu-
lation. In J. Garber & K. A. Dodge, Eds, The Develop-
ment of Emotion-regulation and Dysregulation.
Cambridge: Cambridge University Press, pp. 15–48.
Craik, K. & Feimer, N. (1987). Environmental assess-
ment. In D. Stokols & I. Altman, Eds, Handbook of
Environmental Psychology Vol 2. New York: Wiley,
pp. 891–918.
Dodge, K. A. & Garber, J. (1991). Domains of emotion-
regulation. In J. Garber & K. A. Dodge, Eds, The
Development of Emotion-regulation and Dysregul-
ation. Cambridge: Cambridge University Press, pp.
3–11.
Epstein, S. (1981). The unity principle versus the reality
and pleasure principles, or the tale of the scorpion and
the frog. In M. D. Lynch, A. A. Norem-Hebeisen, & K.
J. Gergen, Eds, Self-concept. Cambridge, MA:
Ballinger, pp. 27–37.
Epstein, S. (1983). The unconscious, the preconscious,
and the self-concept. In J. Suls & A. G. Greenwald,
Eds, Psychological Perspectives on the Self Vol. 2.
Hillsdale, NJ: Lawrence Erlbaum, pp. 219–247.
Epstein, S. (1985). The implications of cognitive-experien-
tial self-theory for research in social psychology and
personality. Journal for the Theory of Social Behavior,
15, 283–310.
Epstein, S. (1990). The self-concept, the traumatic neur-
osis, and the structure of personality. In D. Ozer, J.
M. Healy, Jr & A. J. Stewart, Eds, Perspectives on
Personality, Vol. 3. Greenwich, CT: JAI Press, pp.
63–98.
Epstein, S. (1994). Integration of the cognitive and the
psychodynamic unconscious. American Psychologist,
49, 709–724.
Feldman, R. M. (1990). Settlement-identity: psychological
bonds with home places in a mobile society. Environ-
ment & Behavior, 22, 183–229.
Fuhrer, U., Kaiser, F. & Hartig, T. (1993). Place attach-
ment and mobility during leisure time. Journal of
Environment Psychology, 12, 187–198.
Hartig, T. (1993). Nature experience in transactional
perspective. Landscape and Urban Planning, 25,
17–36.
Hartig, T. & Evans, G. W. (1993). Psychological foun-
dations of nature experience. In T. Gärling & R. G.
Golledge, Eds, Behavior and Environment: Psycho-
logical and geographical approaches. Amsterdam:
North-Holland, pp. 427–457.
Hartig, T., Mang, M. & Evans, G. W. (1991). Restorative
effects of natural environment experiences. Environ-
ment & Behavior, 23, 3–26.
Hartig, T., Korpela, K. M., Evans, G. W. & Gärling, T.
(1996). Validation of a measure of perceived environ-
mental restorativeness. (Göteborg Psychological
Reports, 26:7). Göteborg: Göteborg University,
Department of Psychology.
Hartig, T., Böök, A., Garvill, J., Olsson, T. & Gärling, T.
(in press). Environmental influences on psychological
restoration. Scandinavian Journal of Psychology.
Izard, C. E. & Kobak, R. R. (1991). Emotions systems
functioning and emotion-regulation. In J. Garber & K.
A. Dodge, Eds, The Development of Emotion-regu-
lation and Dysregulation. Cambridge: Cambridge
University Press, pp. 303–321.
Kaiser, F. (1993). Mobilität als Wohnproblem: Ortsbin-
dung im Licht der emotionalen Regulation (Mobility
as a dwelling problem: Place attachment as emotional
regulation) Bern: Lang.
Kaplan, R. & Kaplan, S. (1989). The Experience of Nature:
A psychological perspective. Cambridge: Cambridge
University Press.
Kaplan, S. (1983). A model of person-environment com-
patibility. Environment & Behavior, 15, 311–332.
Kaplan, S. (1995). The restorative benefits of nature:
toward an integrative framework. Journal of Environ-
mental Psychology, 15, 169–182.
Kaplan, S. & Talbot, J. F. (1983). Psychological benefits of
a wilderness experience. In I. Altman & J. F. Wohlwill
Eds, Human Behavior and Environment: Advances in
theory and research, Vol. 6: Behavior and the Natural
Environment. New York: Plenum Press, pp. 163–203.
Kaplan, S., Bardwell, L. V. & Slakter, D. B. (1993). The
museum as a restorative environment. Environ-
ment & Behavior, 25, 725–742.
Korpela, K. M. (1989). Place-identity as a product of
environment self-regulation. Journal of Environmen-
tal Psychology, 9, 241–256.
Korpela, K. M. (1991). Are favorite places restorative
environments? In J. Urbina-Soria, P. Ortega-
Andeane & R. Bechtel, Eds, Healthy Environments.
Oklahoma City, OK: Environmental Design Research
Association, pp. 371–377.
Korpela, K. M. (1992). Adolescents’ favourite places and
environmental self-regulation. Journal of Environ-
mental Psychology, 12, 249–258.
Korpela, K. M. (1995). Developing the environmental self-
regulation hypothesis (Doctoral dissertation). (Acta
Universitatis Tamperensis, Series A, Vol. 446).
Vammala: University of Tampere.
Lalli, M. (1992). Urban-related identity: theory, measure-
ment, and empirical findings. Journal of Environmen-
tal Psychology, 12, 285–303.
Noack, P. & Silbereisen, R. K. (1988). Adolescent develop-
ment and choice of leisure settings. Children’s
Environments Quarterly, 5, 25–33.
Owens, P. E. (1988). Natural landscapes, gathering
places, and prospect refuges: characteristics of out-
door places valued by teens. Children’s Environments
Quarterly, 5, 17–24.
Parsons, R. (1991). The potential influences of environ-
mental perception on human health. Journal of
Environmental Psychology, 11, 1–23.
Proshansky, H. M., Fabian, A. K. & Kaminoff, R. (1983).
Place-identity: physical world socialization of the self.
Journal of Environmental Psychology, 3, 57–83.
Sarbin, T. R. (1983). Place identity as a component of self:
 Favorite Places
an addendum. Journal of Environmental Psychology,
3, 337–342.
Silbereisen, R., Noack, P. & Eyferth, K. (1986). Place for
development: Adolescents, leisure settings, and devel-
opmental tasks. In R. K. Silbereisen, K. Eyferth & G.
Rudinger, Eds, Development as Action in Context:
Problem behavior and normal youth development.
Berlin: Springer-Verlag, pp. 87–107.
Silbereisen, R. & Noack, P. (1988). Adolescence and
environment. In D. Canter, M. Krampen & D. Stea,
Eds, Environmental Policy, Assessment and Com-
munication. Aldershot: Avebury, pp. 19–34.
Sommer, B. (1990). Favorite places of Estonian ado-
lescents. Children’s Environments Quarterly, 7,
32–36.
Stokols, D. (1995). The paradox of environmental psy-
chology. American Psychologist, 50, 821–837.
233
Ulrich, R. S. (1979). Visual landscapes and psychological
well-being. Landscape Research, 4, 17–23.
Ulrich, R. S. (1983). Aesthetic and affective response to
natural environment. In I. Altman & J. F. Wohlwill,
Eds, Human Behavior and Environment: Advances in
theory and research, Vol. 6: Behavior and the natural
environment. New York: Plenum Press, pp. 85–125.
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.
Vuorinen, R. (1990). Persoonallisuus ja minuus
(Personality and self). Juva: WSOY.
Zuckerman, M. (1977). Development of a situation-
specific trait-state test for the prediction and
measurement of affective responses. Journal of Con-
sulting and Clinical Psychology, 45, 513–523.