The perception of privacy : a multidimensional scaling analysis


Material Information

The perception of privacy : a multidimensional scaling analysis
Physical Description:
ix, 184 leaves : 28 cm.
Wilmoth, Gregory Hicks, 1947-
Publication Date:


Subjects / Keywords:
Privacy   ( lcsh )
Social psychology   ( lcsh )
Psychology thesis Ph. D   ( lcsh )
Dissertations, Academic -- Psychology -- UF   ( lcsh )
bibliography   ( marcgt )
non-fiction   ( marcgt )


Thesis--University of Florida.
Bibliography: leaves 175-182.
General Note:
General Note:
Statement of Responsibility:
by Gregory H. Wilmoth.

Record Information

Source Institution:
University of Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
aleph - 023403868
oclc - 06952010
System ID:

Table of Contents
    Title Page
        Page i
        Page ii
        Page iii
        Page iv
    Table of Contents
        Page v
        Page vi
        Page vii
        Page viii
        Page ix
    Chapter 1. Introduction
        Page 1
        Page 2
        Page 3
        Page 4
        Page 5
        Page 6
        Page 7
        Page 8
        Page 9
        Page 10
        Page 11
        Page 12
        Page 13
        Page 14
        Page 15
        Page 16
        Page 17
        Page 18
        Page 19
        Page 20
        Page 21
        Page 22
        Page 23
        Page 24
        Page 25
        Page 26
        Page 27
        Page 28
        Page 29
        Page 30
        Page 31
        Page 32
        Page 33
        Page 34
        Page 35
        Page 36
        Page 37
        Page 38
        Page 39
        Page 40
        Page 41
        Page 42
        Page 43
        Page 44
        Page 45
        Page 46
        Page 47
        Page 48
        Page 49
        Page 50
        Page 51
        Page 52
        Page 53
        Page 54
        Page 55
        Page 56
    Chapter 2. Method
        Page 57
        Page 58
        Page 59
        Page 60
    Chapter 3. Results
        Page 61
        Page 62
        Page 63
        Page 64
        Page 65
        Page 66
        Page 67
        Page 68
        Page 69
        Page 70
        Page 71
        Page 72
        Page 73
        Page 74
        Page 75
        Page 76
        Page 77
        Page 78
        Page 79
        Page 80
        Page 81
        Page 82
        Page 83
        Page 84
        Page 85
        Page 86
        Page 87
        Page 88
        Page 89
        Page 90
        Page 91
        Page 92
        Page 93
        Page 94
        Page 95
        Page 96
        Page 97
        Page 98
        Page 99
        Page 100
    Chapter 4. Discussion
        Page 101
        Page 102
        Page 103
        Page 104
        Page 105
        Page 106
        Page 107
        Page 108
        Page 109
        Page 110
        Page 111
        Page 112
        Page 113
        Page 114
        Page 115
    Appendix A. Conceptual explanation of multidimensional scaling
        Page 116
        Page 117
        Page 118
        Page 119
        Page 120
        Page 121
        Page 122
    Appendix B. Collection and selection of experimental stimuli
        Page 123
        Page 124
        Page 125
        Page 126
        Page 127
        Page 128
        Page 129
        Page 130
        Page 131
        Page 132
        Page 133
        Page 134
        Page 135
        Page 136
        Page 137
    Appendix C. Selection of the MDS target stimuli
        Page 138
        Page 139
        Page 140
        Page 141
        Page 142
        Page 143
        Page 144
        Page 145
        Page 146
    Appendix D. Selection of defining attributes and defining attribute procedure
        Page 147
        Page 148
        Page 149
        Page 150
        Page 151
        Page 152
        Page 153
        Page 154
        Page 155
        Page 156
        Page 157
        Page 158
        Page 159
        Page 160
        Page 161
        Page 162
        Page 163
        Page 164
        Page 165
        Page 166
    Appendix E. Experimental stimuli
        Page 167
        Page 168
        Page 169
        Page 170
        Page 171
        Page 172
        Page 173
        Page 174
        Page 175
        Page 176
        Page 177
        Page 178
        Page 179
        Page 180
        Page 181
        Page 182
    Biographical sketch
        Page 183
        Page 184
        Page 185
        Page 186
Full Text








Just as few individuals have the requisite knowledge

and skills for single-handedly constructing a house, few

can complete a dissertation without assistance. In my case

many individuals provided knowledge, skills, labor, and mor-

al support.

First, I wish to acknowledge the eager, patient, and

professional guidance and support rendered by my committee

members. Dr. Lawrence J. Severy always offered encourage-

ment, faith and especially much needed latitude. He gen-

uinely facilitated the entire project, by his inspiration

and by his timely provision of materials and resources

needed to carry-out the research. Conceptual and methodo-

logical assistance was provided by Dr. Marvin E. Shaw. Dr.

Hernan Vera repeatedly inspired my thinking with his orig-

inal insights and questions. Dr. William Froming encour-

aged me to design the research to test wider issues than

privacy. Dr. Scott Miller helped clarify the presentation

of the methodology through his practical, specific questions.

Above all else the committee members showed patience and

faith in my less than orthodox execution of the dissertation.

Second, special thanks are directed to Michael Conlin

and the staff of CIRCA, the University of Florida. Without

Michael Conlin's humor, patience and programming expertise,

this dissertation would not now be complete.

Third, many deserve recognition for their valuable la-

bor. Starr Silver, Jeff Elliott, Charles Green, Sonja

Peterson, Darlene Puckett, and Jill Schepler all served as

coders for one or more phases of the research. Mercedes

de los Santos, Yvette Garcia, and Greta Garrett ably con-

ducted the experimental sessions. Karen Long typed the

rough draft. Lois Rudloff performed a miracle by typing

my disorganized, sporadic writings into a completed disser-

tation within minimal time limits. Her patience, profes-

sionalism, and friendliness greatly reduced my anxiety.

Fourth, my sincere thanks are extended to all those

students who endured the long and tedious experimental tasks.

Their conscientiousness underlies the quality of the data.

Fifth, my warmest thanks to especially to Karen Long

and Starr Silver. Without their concern, magnamity, and

emotional support, the months of labor on this dissertation

would have been desolate and hollow. Only I and they can

know their full contribution and my deep appreciation. Dr.

Lawrence Severy again deserves acknowledgement for his emo-

tional support. I could not have been more fortunate in my

selection of Dr. Severy as chairman of my dissertation.

Finally but not least, I wish to acknowledge the dur-

able understanding and material support of my parents, Leslie

and Eileen Wilmoth. Without them, I would never have reached

this goal.

My sincerest apologies to anyone whom I have forgotten

to acknowledge.



ACKNOWLEDGEMENTS ....... ..........

ABSTRACT . . . . . . . . . . .



INTRODUCTION ...........
Definitions of Privacy ......
Theories of Privacy .. .........
Biological Theories ......
Developmental-Learning Theories
Interpersonal Theories .......
Social-Normative Theories . .
Ecological Theory of Privacy .
Functional Theories of Privacy .


Research on Privacy .... ..........
Measures of Privacy ........
Research Related to Biological
Theories ..... ..............
Developmental Research .. ........
Research on Interpersonal Theories o
Privacy . . . . . . .
Research Related to Social-Normative
Theories ...................
Research on Functional Theories of
Privacy . . . . . . .

II METHOD . . . . . . . .


* 37


. 47

Phase 1 ..... .............
Phase 2 ..... .............
Phase 3 ..... .............

RESULTS ..... .............
Multidimensional Privacy Results
Multidimensional Intimacy Results
Multidimensional Reserve Results
Multidimensional Solitude Results




IV DISCUSSION ....... ...............
Comparison of Dimensions Across Types of
Privacy . . . . . . . . .
Dimensions Predictive of Perceived
Privacy . . . . . . . . .
Testing Multidimensional Models ....
Comparison of Results to Privacy
Theories ...... .................
Toward a Multidimensional Model of
Privacy . . . . . . . . .


SCALING . . . . . . . .

STIMULI . . . . . . . . .




REFERENCES . . . . . . . . . .

BIOGRAPHICAL SKETCH ....... ................













Abstract of Dissertation Presented to the Graduate
Council of the University of Florida in Partial
Fulfillment of the Requirements for the Degree
of Doctor of Philosophy



Gregory H. Wilmoth

August 1980

Chairman: Lawrence J. Severy

Major Department: Psychology

The multidimensional nature of perceived privacy was

explored through a multidimensional scaling analysis (MDS)

of forty privacy situations. Three dimensions labeled "So-

cial Density," "Normativeness," and "Interpersonal Involve-

ment" were judged to best explain privacy. These dimen-

sions in combination significantly predicted the privacy

scale scores of the privacy situations. The data were par-

titioned into three sets according to three types of pri-

vacy situations (intimacy, reserve, and solitude) and an-

alyzed. Two (Personal-Impersonal and Normativeness) of

four dimensions found for intimacy were significantly re-

lated to the privacy scale scores. An Empathy and Self-

Disclosure dimension, along with an Undefined dimension of

a five-dimensional solution for reserve, accounted for sev-

enty percent of the privacy scale scores and was marginally


significant. No reliable dimensional solution was obtained

for the solitude situations.

The Personal-Impersonal, Self-Disclosure, and Empathy

dimensions derived from the partitioned analyses shared de-

fining attributes with the Interpersonal Involvement dimen-

sion. This was interpreted as indicating that the Inter-

personal Involvement dimension was a higher-order dimension

with three components. Intimacy and reserve were distinct

based on their respective combinations of these component,

lower-order dimensions. Intimacy was further specified by

the Normativeness dimension.

The results were evaluated against current privacy the-

ories. The data were supportive of Altman's interpersonal

boundary-regulation theory and Kelvin's social-normative

theory of privacy. There was little evidence supporting a

theory of privacy as control and choice freedom.

A U-shaped three dimensional model consisting of Social

Density, Normativeness, and Interpersonal Involvement was

proposed to explain the perception of privacy. The model

specifies that both high Interpersonal Involvement (inti-

macy) and low Interpersonal Involvement (solitude) are as-

sociated with high perceived privacy. Solitude and anonym-

ity are primarily distinguished by being low and high re-

spectively on the Social Density dimension. Perceived pri-

vacy increases as Social Density decreases. The Normative-

ness dimension is speculated to interact additively with the


other two dimensions. As Normativeness of the behavior de-

scribed decreases the perception of privacy increases.


Privacy has been proposed as the core, integrating

construct in environmental psychology by Altman (1975).

The class of proxemic behaviors including personal space,

interpersonal distance, immediacy, territoriality, social

density, and crowding are all explained using this construct.

Beyond proxemic behaviors, Kelvin (1973) contends that

privacy is a basic state relevant to an understanding of so-

cial behavior in general and crucial to social psychological

theory overall. The sociologists Georg Simmel (1957),

Erving Goffman (1959, 1963) and Barry Schwartz (1968), all

conceptualize privacy as the inherent, counterpattern to so-

cial interaction. The 'right to privacy' which is consid-

ered a fundamental value to democracy is an increasingly

addressed topic in political science and is a focal concern

among those studying the socio-political impact of informa-

tion technologies (Packard, 1964; Miller, 1971; Rule,

McAdam, Stearns & Uglow, 1980; and Westin, 1967).

In spite of privacy's central position in Altman's

(1974, 1975, 1976) conceptualization of the environment and

social behavior, and its importance within sociology and po-

litical science, there is remarkably little agreement on the

definition of privacy and little research on its


determinants and consequences. Given these circumstances

surrounding privacy, it seems appropriate to undertake an

empirical investigation of the dimensions underlying the

experiencing of privacy as a first step toward a more sys-

tematic understanding of an increasingly important concept.

Such an endeavor would also provide a set of parameters to

be manipulated in future research as well as possibly sug-

gesting a taxonomy of privacy situations.

The complexity of the concept has been noted by

Proshansky, Ittelson and Rivlin (1970) and others (see

Laufer, Proshansky & Wolfe, 1973; and Wolfe, 1978). An es-

sential, preliminary issue concerns this complexity: Is

privacy as defined by sociologists, anthropologists, polit-

ical scientists, and social and environmental psychologists

one concept or many?

The myriad definitions of privacy can be categorized

into three types. One category of definition centers around

the concept of information flow. Control or choice freedom

is the focal element of a second definitional type. The

third type defines privacy within the context of normative


Definitions of Privacy

Westin (1967, p. 30) defined privacy as "the right of

the individual to decide what information about himself

should be communicated to others and under what conditions."

This seminal definition has been used by Pastalan (1970a,

1970b), Roberts and Gregor (1971) and Derlega and Chaikin

(1977). Three key elements of this definition, i.e. deci-

sion autonomy, information about self, and communication,

are central to related definitions. Bates (1964) stated

that privacy is a person's feeling that others should be

excluded from a structured portion of this total phenomeno-

logical field. Similarly for Jourard (1966) privacy is an

outcome of a person's wish to withhold from others certain

knowledge about his past and present experience and action,

and his intentions for his future. Finally, Bennett (1967)

defined privacy as the selective control of communication.

Despite the core elements shared by each of these def-

initions, a unified concept is negated by differences in

the ontological locus of privacy. In Westin's perspective,

privacy is a legal-political prerogative as opposed to the

emotional state proposed by Bates. Alternatively, privacy

is an informational outcome state between social actors re-

sulting from selective communication (Jourard, 1966).

Privacy definitions which focus exclusively on the

autonomy (choice freedom) dimension tend to claim the onto-

logical status of "process" rather than of "state." At the

most inclusive level, privacy is a claim to immunity from

intrusion or control by others (Proshansky & Ittelson, 1970;

and Weinstein, 1971). Margulis (1974 and used by Berscheid

(1977)) defined privacy as a process of controlling per-

sonal transactions through control over boundaries between

self and others, and Marshall (1974) defined it as the abil-

ity to control the degree to which people and institutions

encroach upon one's life. In Altman's (1975) privacy the-

ory, privacy is defined as "selective control of access to

the self or to one's group." (p. 18) These definitions

tend to share the characteristic that privacy is a bidirec-

tional process--both input from others to self and output

from the self to others are controlled. Definitions which

emphasize the selective control of information communicated

about self (Jourard, 1966; Westin, 1967; and others) do not

contain this bidirectionality dimension.

One further definitive dimension of privacy, its nor-

nativeness, is emphasized by sociologists. Warren and Las-

lett (1977) distinguish between secrecy and privacy by the

normative acceptance and protection of the behavior in ques-

tion. Privacy protects those behaviors which are either

morally neutral or valued by society through consensual,

social legitimation and further seen as nonthreatening to

others. Schwartz (1968) and Kelvin (1973) also primarily

discussed this normative element of privacy. None of the

previously cited definitions of privacy include this norma-

tive distinction.

Laufer, Proshansky, and Wolfe (1973) contend that many

behaviors, events, and affective experiences are encompassed

in the word "privacy." Rather than attempting to define

privacy, they have worked at identifying the types of phe-

nomenon subsumed under this multi-dimensional concept.

Westin (1967) preceded them in this multi-dimensional per-

spective. Westin's four categories of privacy are solitude,

anonymity, reserve, and intimacy. Solitude exists when a

person is alone and free from observation. Anonymity oc-

curs when a person is in a public place with others present

but none of these others know his identity. Reserve is both

the selective control over information about self communi-

cated to others and the selective attention to others' com-

munications. Intimacy is a state in which two or more per-

sons are in a situation of solitude from the observation of

others. Intimacy may be thought of as group solitude.

Proshansky, Ittelson and Rivlin (1970) have noted that

Westin's types of privacy are not defined by the same di-


Nine analytic dimensions of privacy were defined by

Laufer, Proshansky, and Wolfe (1973), only three of which

qualify as categories of privacy. The remaining six dimen-

sions represent functions and processes of privacy. These

are the interaction, task, and ritual privacy dimensions.

Interaction privacy is the boundary-control process of com-

ing together and withdrawing from others. Task privacy is

restricted to only those situations in which task comple-

tion is enhanced by separation from others. Ritual privacy

covers those behaviors for which there is a normative pre-

scription that these be performed in nonpublic places free

of observation. Such behaviors include elimination and sex-

ual acts.

Through a factor analysis of questionnaire items, Mar-

shall (1974) identified two additional types of privacy to

those proposed by Westin (1967). These two were seclusion

and 'not neighboring.' An interpretation of the seclusion

factor suggests that it is the state of having one's ter-

ritory free from the visual and auditory observation of

others. It is distinguished from solitude by its territori-

al locus; solitude is privacy of the person while seclusion

is privacy of one's territory. Seclusion includes solitude

but the inverse relationship does not hold. Not neighbor-

ing concerns attitudes toward selective involvement with

neighbors and norms governing visiting.

This profusion of types of privacy and the multi-dimen-

sional nature of the existent definitions emphasizes the is-

sue of whether privacy is one concept or many. The choice

of a single definition of privacy or selection of one of

the concepts of privacy has implications for the variables

and relationships hypothesized as essential in explaining

that phenomenon. The result could be the appearance of sev-

eral seemingly conflicting theories of privacy. Any attempt

to encompass all of the different types and meanings of pri-

vacy into a unitary concept will require a much more

extensive range of variables and relationships, and more

complex theories to successfully integrate the diverse phe-

nomena involved.

Theories of Privacy

Given the variety of conceptualizations and types of

privacy, the existence of a wide and diverse range of pri-

vacy theories is to be expected. One classification of

these theories results in some six distinct orientations

toward privacy: biological, developmental-learning, inter-

personal, social-normative, ecological, and functional.

Many of the theories to be reviewed are multi-dimensional

in conceptualization and thus overlap these somewhat arbi-

trary classifications. Those theories which, however, as-

sume either a biological causation (genetic) or a mediating

biological process (e.g. arousal) are classified as biolog-

ical regardless of the presence of learning or social fac-

tors in the theories. Theories which are constructed on

learning principles or life-cycle maturation changes are

identified as developmental. Only those theories which

emphasize the regulatory nature of privacy as a process

involving adjustments of self-boundaries to permit various

levels of contact with others are included within the inter-

personal category. Where the institutional nature of pri-

vacy is central to a theory, it is distinguished as

social-normative. The ecological orientation is reserved

solely for explanations of privacy based on Barker's (1968)

analytic concepts. Those theories which conceive of pri-

vacy as a mediating process toward the attainment of a cer-

tain class of choice freedom goals are demarcated as func-


Biological Theories

Biological theories of privacy can be further subdi-

vided into three classes: genetic, arousal mediating, and

adaptation theory (Helson, 1964) derived. Genetic theories

postulate an innate propensity toward either gregariousness

(Halmos, 1953), territoriality (Pastalan, 1970a), or survival

(Klopfer & Rubenstein, 1977). Mehrabian and Russell's (1974)

arousal mediation theory proposes that the underlying, fun-

damental quality of all situations which elicit a desire

for privacy is the emotive response of the individual to

that situation which in turn is a function of the situa-

tion's arousal eliciting level. Both Hill (1969) and Mar-

shall (1972) explain desired privacy as a function of adap-

tation level processes.

Halmos (1953) postulated that the human species through

biological evolution is genetically programmed to be gre-

garious. This bio-social need has been frustrated as a re-

sult of an increasing division of labor concomitant with

cultural evolution. Primitive societies are characterized

by primary relationships (Gemeinshaft) and intimacy while

post-primitive societies are defined by secondary relation-

ships (Gesellshaft) and reserve. The social fragmentation

resulting from a Gesellshaft society creates social isola-

tion which engenders individual anxiety. It is this anxi-

ety which causes the individual's desire for privacy. Ano-

nymity is impossible (except possibly in selected rituals)

in a Gemeinshaft society. The desire for solitude is nat-

ural to both types of society and functions to "regenerate

social life for its more harmonious living" (Halmos, 1953,

p. 168).

Unlike Halmos, Pastalan (1970a) does not directly claim

territoriality to be an inherited drive. He does, however,

treat it as an universal in human behavior. Human territo-

riality has, furthermore, unique behavioral states (pri-

vacy) which have no conceptual counterparts in animal ter-

ritoriality. Westin's (1967) four states of privacy and

their related functions (personal autonomy, emotional re-

lease, self-evaluation, and protected communication) have

no parallel in animal behavior. Human territoriality, there-

fore, has social determinants such as the situational con-

text, one's position in a role system, and one's emotional

goals. These will be explicated later in reviewing Pasta-

lan's (1970b) other work.

Without proposing discrete instincts, Klopfer and Ru-

benstein (1977) argue that some degree of privacy is

essential to many animals' survival. Although animals like

the social, colonial insects (bees and ants) apparently

display no privacy, most animal species exhibit some pat-

terned social withdrawal (privacy) and/or concealment of

information about their motivational state. Privacy aids

survival by reducing competition for food and reproductive

resources, by providing protection from predators, and by

concealing information about one's motivational state and

intentions from a competitor. An economic model is applied

to predict under what environmental contingencies a species

will evolve privacy behaviors. The costs of attaining and

maintaining a particular form and level of privacy must be

counterbalanced by the benefits resulting from this privacy

in order to enhance survival. Klopfer and Rubenstein (1977)

discuss several classes of contingencies in an animal's phy-

sical and animate environment which govern these cost-bene-

fit ratios. To the extent that privacy in humans is moti-

vated by a desire for 'happiness' rather than survival, pri-

vacy becomes a peculiarly human concept.

Although Mehrabian and Russell (1974) do not present a

systematic theory of privacy, their discussion of privacy

suggests certain systemic relations. They formulated a gen-

eral approach to understanding the relationships of behav-

ior to the environment. Between the environment and any be-

havior is an emotional mediating response to the environment.

These emotional responses are in turn grounded on the

physiological responses of arousal, pleasure, and muscular

tension. Activity, evaluation, and dominance (potency) are

the emotional-connotative (Osgood, Suci & Tannenbaum, 1957)

responses corresponding to the respective physiological re-

sponses. Arousal is a direct function of the information

load of an environment. An environment's information load

is measured by its novelty and complexity.

Privacy is equated with low environmental information

load (Mehrabian & Russell, 1974, p. 299). A person is in

a state of privacy to the extent that he is receiving a

low level of output (information load) from others. The

desire for privacy should be an inverse function of the in-

formation load to which one is exposed. Reciprocally, the

rate, breadth, and depth of information load from the self

to others determines the degree of privacy. A low informa-

tion load from self (another's environmental stimulus) to

another constitutes high privacy. To the extent that en-

vironments enhance positive (pleasure) feelings, self-dis-

closure (information load to others) is also enhanced.

Since either an excessively high or low information load

is aversively arousing and thus unpleasant, the maximum

preference for environments will be slightly discrepant

from the adaptation level. People will be engaged, there-

fore, in an ever changing flow between privacy and social


Whereas the discrepancy hypothesis is central to

Mehrabian and Russell's (1974) approach, adaptation level

theory is focal for Hill (1969) and Marshall (1972). Adap-

tation level theory has been applied differently by Hill and

Marshall. Hill applies the theory to two dimensions: the

relation of perceived to actual privacy and the optimization

of visual privacy requirements with visual-access-outward

requirements. In neither instance is a formal, systematic

explanation proposed in terms of adaptation level theory.

Only focal stimuli and background stimuli, but not residual

stimuli (Helson, 1964), are implied by Hill. On the issue

of the relation of perceived to actual privacy, Hill hypoth-

esizes that "as people become aware of their surrounding

environment (both physical and social) so their judgements

of perceived privacy are modified and approach a level which

is roughly equivalent to the real amount of privacy present

in a given situation" (Hill, 1969, p. 39). The fit between

past privacy experience (residual stimuli) and present pri-

vacy opportunities and/or constraints is not discussed.

The problem of optimization is a function of the in-

herent boundary permeability involved in privacy. While

privacy is desired, isolation is not; where freedom from

observation is sought, the complete lack of access to out-

ward observation is not. The process of optimization be-

tween these extremes of visibility is an adaptation (balance)

between outward and inward visibility (privacy). There is,

however, no single such optimization; rather the achieved

optimization is influenced by such focal and background

stimuli as room function and outside view--activity. The

compromise between the opposing requirements for outward

and inward vision would be optimized more toward privacy

for a bedroom fronting on the street as opposed to a kitch-

en facing a secluded backyard.

Marshall (1972) explicitly treats the residual stim-

uli of childhood home density and childhood neighborhood

density. She deduced that: (1) those reared in a high

density home (large number of siblings and without a room

of one's own) would prefer low levels of privacy as a re-

sult of adapting to the attendant lack of privacy in the

childhood home, and (2) those spending their childhood in

cities, as opposed to small towns, would prefer more ano-

nymity as a result of adapting to the increased availabil-

ity of anonymity in the city. Both actual density and per-

ceived crowdedness of childhood environments were consid-


Background stimuli such as present environmental den-

sity in the home and neighborhood were also hypothesized to

be related to preferred privacy (Marshall, 1972). An in-

dividual's preference for privacy represents an adaptation

to the amount of privacy available within one's environment.

Preference for privacy should be inversely related to den-

sity in the present home and neighborhood environment.

Predictions about the type of privacy effected by home

and neighborhood density are different and thus the de-

sired type of privacy also changes respective to these two

forms of density. Note that whereas Marshall addresses

the adaptation relation between preferred (desired) pri-

vacy and environmentally available privacy, Hill (1969)

focuses on the relation between perceived privacy and

achieved privacy.

Developmental-Learning Theories

Marshall's (1972) hypotheses about the effect of

childhood density experiences on adult privacy preferences

assumes that experiences perceived as either private, inva-

sions of privacy, or crowding are the same in both child-

hood and adulthood. Developmental psychology challenges

such assumptions, partially through the impact of learning.

Ittelson, Proshansky, Rivlin, and Winkel (1974) believe

that our conceptions of what constitutes privacy, the areas

in which privacy is appropriate, the settings which support

it, and the techniques available to enforce privacy are

learned through the observation of adults, primarily through

experiencing exclusion from adults' lives.

The earliest developmental conceptualization of privacy

(Smith et al., 1969) focused on the family developmental cycle

as opposed to the individual developmental cycle. Different

stages of the family cycle represent distinguishable activity

patterns. The priorities awarded to privacy and interac-

tion within the home are influenced by these activity pat-

terns. The desires for specific types of privacy, there-

fore, are a function of family-cycle stages. Family life-

cycle stage interacts with dwelling size and arrangement,

family size, and occupational status of the parents.

Six dimensions of development relevant to privacy were

identified by Laufer, Proshansky, and Wolfe (1973). The

first of these, ego development, concerns itself with de-

velopment of independence and autonomy. Both behaviors in-

volve the individual's volition of freedom to choose being

alone versus being a member of a group. Coping with forced

aloneness (isolation), chosen aloneness (solitude), being a

stranger (anonymity), forced intimacy (invasion of privacy),

and chosen intimacy are all developmental tasks in the ma-

turation of independence, autonomy, and interpersonal inter-


Aging from infancy to adulthood inherently transfers

the individual through different roles imposing different

demands and requirements. New concepts and patterns of pri-

vacy are defined by the changing role requirements. New

modes of privacy which accompany increased status must be

added to the individual's behavioral repertoire.

In addition to the socially structured behaviors of

roles, other activities and environments change through the

life cycle. New physical settings such as school present

new problems for privacy requiring new adaptations in pri-

vacy behaviors.

A less obvious developmental task which takes on par-

ticular importance in industrial-secular societies is adap-

tation to socio-cultural change. The patterns and expecta-

tions of privacy learned in pre-adulthood may be incompat-

ible with rapidly emerging life styles, mores, and technol-


Since control is a critical element in all definitions

of privacy, development of the awareness of being able to

control becomes important. The first successfully concealed

behavior begins awareness that one's own thoughts and behav-

iors are not known unless they are volitionally disclosed.

The learning of what thoughts and behaviors to control

(neither to perform in public or to make publicly announced)

is an important process of socialization.

Of the six dimensions (Laufer, Proshansky & Wolfe,

1973) the last is conceptual complexity. With maturation

the concept of privacy should become both more complex and

integrated with related conceptual systems like moral reason-

ing. The complexity of the behavioral repertoire used to

create and maintain privacy and the functions for which pri-

vacy is engaged should also increase with age. Laufer,

Proshansky and Wolfe (1973) rather than presenting a devel-

opmental theory of privacy have identified six analytic

developmental dimensions which are hypothesized to be re-

lated to privacy.

The only specifically learning theory of privacy to

appear is that of Smith and Swanson (1979). They apply

Rotter's (1954) social learning theory to privacy. Through

the learning history of the individual, both preferences

for types of reinforcements and expectancies of reinforce-

ment resulting from various actions are formed. "[PIrivacy

as a process of interactional control is a function of the

value of such control and the expectancy of achieving such

control" (Smith & Swanson, 1979, p. 5). The immediate

and enduring psychological situation is the third factor

influencing the behavior potential toward privacy. The

psychological situation refers to those aspects of the in-

ternal and external environment as perceived by the indi-

vidual. The primary hypothesis states that "individual no-

tions of privacy vary as a function of their environment"

(Smith & Swanson, 1979, p. 17). On the psychological lev-

el, the environment influences behavior potential not

through changing values but through impacting expectancies

that potential behaviors will be reinforced. Expectancies

should be object- and situation-specific whereas values are

intra-individual entities transcending situational specifics.

Failure to achieve privacy will be more aversive as the ex-

pectancy for attainment and value of attainment increase.

Social learning theory suggests a model of reactions to in-

vasion of privacy which Smith and Swanson (1979) do not de-


Interpersonal Theories

Some of the developmental dimensions presented by

Laufer, Proshansky, and Wolfe (1973) confront the interper-

sonal, social nature of privacy. Jourard (1966, 1971),

most known for his theory and research on interpersonal

self-disclosure, has clearly conceived of privacy as one

extreme of the self-disclosure process. He frames self-

disclosure and concealment (privacy) as bi-polar opposites.

Jourard's (1966, 1971) definitions of self-disclosure and

privacy are conceptually interdependent: privacy is the

achieved concealment of knowledge about self while self-

disclosure is the revealment of knowledge about self. Self-

disclosure is a process that changes the state of privacy

existing between two people. Intimacy is both achieved by

and a state of high self-disclosure. In Westin's (1967)

frequently cited types of privacy, intimacy is the polar

opposite of solitude.

The psychological functions attributed to privacy

(Jourard, 1966; and Altman, 1975) are identical to those

claimed for self-disclosure (Jourard, 1971). That which

constitutes the content of privacy is conceptually equiva-

lent to that of self-disclosure: centrality to the person's

social identity. The interpersonal, environmental, and

psychological conditions conducive to self-disclosure are

the same as those for the privacy state of intimacy (Der-

lega & Chaikin, 1977). Solitude, anonymity, and reserve

all function to conceal knowledge about the self. Jourard

(1966) notes that solitude and anonymity conceal behavioral

knowledge about the self while reserve (the opposite of

self-disclosure?) conceals psychological (attitudes, in-

tentions, emotions, etc.) knowledge from others. Self-

disclosure is championed (Jourard, 1971) as the cure to

isolation (forced privacy) and alienation whereas privacy

is championed (Jourard, 1966) as the cure to constant pub-

lic scrutiny and alienation.

Altman's (1975) theory of privacy owes a heavy intel-

lectual debt to Jourard's self-disclosure work. Altman and

Taylor (1973) expanded the self-disclosure process into a
'social penetration' process. Privacy emerges as the su-

preme self-other interpersonal boundary-regulation process

(Altman, 1975). An important aspect of Altman's theory is

his contention that privacy is a dialectic process, as op-

posed to a personal disposition or situational contingency.

As a boundary-regulation process, privacy is a constantly

shifting balance between being accessible to others (so-

cial penetration) and being closed to others (privacy).

This shifting balance manifests the inherent nature of pri-

vacy toward optimization. The individual desires a specific

level of privacy-interaction, either a deficit or an ex-

cess activates behaviors to restate the optimization bal-

ance. Both inputs from others and outputs to others are

governed by this boundary-regulation process.

Altman's theory has conceptual appeal due to the fact

that the conceptualization of crowding and isolation can be

derived from the privacy theory. Crowding exists when de-

sired privacy is high and inputs from others are simultane-

ously high. When the desire for social contact is high

(low desired privacy) and actual contact (output to others)

is low then isolation is experienced. The motivational

element (desired privacy) of the theory introduces the cen-

tral concept of control. Whether intimacy or invasion of

privacy occurs is determined by whether the contact is cho-

sen. Choice is the volitional aspect of control. To the

extent that privacy is optimized, control has been success-

fully achieved. An achieved privacy which is discrepant

from the desired privacy motivates the re-exertion of con-


Whereas Jourard (1966, 1971) restricts privacy to the

control exercised over information about self, Altman (1975)

encompasses the control over all self-other contacts within

privacy. The implications of this distinction are impor-

tant: in Jourard's conceptualization there is a single,

essential motivation for privacy, the concealment of knowl-

edge about self; in Altman's theory there can be multiple

motivations for privacy other than the concealment of self

such as task completion, restoration of sleep, etc. The

principles governing and dimensions of Altman's privacy

should be more numerous, diverse, and complex than those

applying to Jourard's privacy.

Social-Normative Theories

An emphasis on motivational states in explaining pri-

vacy suffers from the danger of not seeing the wider so-

cial matrix in which both motivation and privacy exist.

In spite of the proof of individual differences, most behav-

ior is predictable from knowing the social group to which a

person belongs and the social situation in which that per-

son is found. Privacy, as a form of social behavior, is

theorized to be structured and institutionalized like other

social behaviors. This normative theoretical perspective

is exemplified by Schwartz (1968), Kelvin (1973), and Pas-

talan (1970.

Social life is a constant flow into contact with others

and ebbs from such contact. This ebb and flow is all of a

single whole. Just as social interaction is governed by in-

stitutionalized expectations (norms), so social withdrawal

is also guided by norms (Schwartz, 1968). Norms for privacy

function identically as do norms for social interaction:

to stabilize the social order. These norms stabilize the

social order in two ways: (1) one set of norms prescribe

what behaviors must be performed in private thereby

maintaining intact those rules which would be subverted

by public performance, and (2) a second set of norms which

guarantee privacy through describing who may observe and/

or reveal information about whom and under what highly cir-

cumscribed circumstances. Authority typically has the pre-

rogative and duty to intrude another's privacy when the pub-

lic safety or morality is perceived to be threatened. Sur-

veillance i legitimated, institutionalized privacy intru-

sion (Schwartz, 1968). Simultaneously, authority is gener-

ally immune legally from intrusion.

Privacy, as an institutionalized mode of withdrawal,

will reflect the social structure of which the norms are a

part. The set of privacy norms form subsets of norms asso-

ciated with respective statuses. Both the privilege and

power to require information (intrude upon) of another are

normative aspects of the status hierarchy. Other norms in

turn govern the use of this power in order to guarantee

some level of confidentiality of the subordinate's infor-

mation. To the extent that the social structure is a func-

tion of the economic system, privacy will also be a luxury

restricted to those who can purchase it in the forms of se-

cluded estates, butlers, answering services, etc.

Group solidarity is enhanced by privacy in two ways.

First, privacy protects self-defaming information from be-

coming group knowledge which could result in expulsion or,

potentially worse, create the formation of coalitions

splintering the group. Second, privacy allows for the so-

cial withdrawal from the group necessary to prevent the hos-

tility which is likely to arise from too much group inti-

macy (Schwartz, 1968).

Role performance is public performance with its con-

comitant surveillance thereby enforcing conformity. This

constant conformity restricts the expression of individual

differences and in turn suppresses the individual's iden-

tity. Privacy provides an institutionalized outlet for en-

actment of non-role behavior and helps prevent "the ego

from identifying itself too closely with or loosing itself

in (public) roles" (Schwartz, 1968, p. 752).

Finally, privacy is institutionalized in architecture.

Spaces must be included for prescribed private behavior

(bathrooms) and the privileged privacy connected with stat-

us must be provided and constructed to symbolize status.

Pastalan (1970b), unlike Schwartz, specifically fo-

cused on role theory model for desired privacy. As members

of a society, individuals occupy a number of roles in the

social order. Characteristics of one's role complex act

as contingencies conducive to eliciting desires for privacy.

The cumulative demands of multiple roles may be stress pro-

ducing, privacy in the form of solitariness may be desired

to escape from this stress and to recuperate mental, emo-

tional, and physical energies. Role behavior is both pre-

scribed behavior and public behavior; therefore, it is

conformity constrained behavior and constantly open to

sanction. Under these conditions, privacy is desired as

both an escape for personal expression and an escape from

the sanctions inherent with surveillance. In addition to

cumulative demands of multiple roles, the requirements of

some roles may conflict with the successful performance of

requirements for other roles. The stress engendered by

this role conflict can be reduced by keeping the incompat-

ible roles separate which may be viewed as the form of se-

lective, task privacy. Privacy will be also desired in

order to maintain the enactment of incompatible roles which

would not be able to be both enacted if one status system

knew of this dual, incompatible role enactment. In states

where a person may not be a member of the Communist party

and a teacher in the public schools, privacy functions to

conceal information about performing these incompatible

roles from those in the public education status system.

Kelvin (1973) presented a social psychological theory

of privacy as opposed to the strictly sociological theories

of Schwartz and Pastalan. Privacy is a subjective state,

it is perceived privacy (Kelvin, 1973). The perception of

privacy is a function of the potential vulnerability ex-

pected in an immediate situation from the presence of

others. As this feeling of vulnerability increases, the

probability that one's actions are changed by the presence

of another increases and perceived privacy decreases.

Privacy is an aspect of a relationship; it is the negation

of potential power relations between self and others.

The major source of another's power is a result of

norms, the presence of norms legitimizes another's negative

sanctioning of our behavior through criticism, rebuke, rid-

icule, etc. Specific norms prescribing privacy function to

protect an individual's independence in situations in which

the person would be vulnerable by another's surveillance.

These privacy norms are higher-order norms which counter-

mand other norms. It is these other norms which create the

vulnerability. Since norms constrain behavior choices

through their prescriptions and proscriptions, privacy norms

provide freedom of choice in these countermanded, protected

behavioral content areas. Privacy norms constrain the lim-

iting power of social norms. Ambiguity arises from this

normative conflict which in turn creates intrapsychic anx-

iety which is a source of the desire for privacy.

Kelvin (1973) postulates that privacy norms arise in

those areas of behavior which are in transition from rigid

normative control to public permissiveness. In situations

of permissiveness our independence is not vulnerable to

others' socially sanctioned power. Changes in technology,

social organization, standard of living, etc. effect atti-

tudes toward norms and permissiveness.

Ecological Theory of Privacy

Ecological psychology (Barker, 1968) integrates the

constructs of norms and roles into the built environment.

Behavior settings are defined by K-21 (Wicker, 1979) scores

which measure various aspects (penetration and autonomy are

two) of the normative role patterns in specifiable spacio-

temporal sites. Behavior programs are the particular role

action patterns of a behavior setting.

Bechtel (1977) has applied ecological theory to the

discussion of privacy. Behavior settings require boun-

daries to control access of others and outside stimulation

which would disrupt the setting's behavior program. The

determinants of privacy needs are the characteristics of

the specific behavior program. Requirements for privacy

vary with each different behavior program. Anything that

interrupts or influences the enactment of a behavior pro-

gram is an invasion of privacy. This is usually some ex-

traneous source of stimulation which interferes with task


Vulnerability and a lack of privacy exist when the pen-

etration and autonomy levels of a behavior setting are too

low. Penetration is the amount of population interdepen-

dence across settings. A low penetration score indicates

that many people other than those enacting the behavior pro-

gram enter that behavior setting. The autonomy score mea-

sures the amount of decision making possessed by those

enacting the behavior program. Low scores on both penetra-

tion and autonomy indicate that the behavioral program is

receiving high levels of external stimulation and external

interference (decisional influence). Vulnerability results

from this inability to control external stimulation and in-


The conceptual unit in ecological psychology for mea-

suring privacy is not the person but rather a behavior set-

ting. In the absence of physical boundaries which maintain

privacy for a behavior setting, however, psychological bar-

riers must be erected by the setting's occupants to regain

privacy. Since psychological boundaries require the exer-

tion of energy beyond that required for the behavior pro-

gram, it is considered stressing. Boundary problems are

created by either work settings differing in behavior pro-

grams of work settings unrelated in organization being too

close physically.

Functional Theories of Privacy

Proshansky, Ittelson, and Rivlin (1970) have taken a

functional, analytic approach to privacy. Their conceptual

comparison of Westin's (1967) four types of privacy resulted

in their proposition that they all share the common property

of maximizing the individual's freedom of choice by removing

social constraints. The exercise of privacy is the exercise

of power over who shall or shall not control our actions

(Ittelson, Proshansky, Rivlin, & Winkel, 1974). The psy-

chological experience of privacy occurs when the person

perceives and believes that he has the freedom in a given

situation to determine when and how to behave. The desire

or need for privacy arises when the individual is faced

with social constraints or limitations which prevent him

from behaving in ways appropriate to his particular pur-

poses. Crowding occurs to the extent that the sheer num-

ber of people present restricts freedom of choice in pur-

suing one's goals. Territorial behavior is an expression

of a desire for privacy in so far as it decreases crowding

and social constraints.

A conceptual refinement of this functional approach

was provided by Johnson (1974). Freedom of choice is

equated with personal control which was distinguished into

primary and secondary control. Secondary control consists

of those behaviors which create conditions which are pre-

requisites or facilitative to primary controlling behaviors.

Privacy comprises those behaviors intended to establish sec-

ondary control over outcomes. As such privacy does not

function to satisfy any inherent or fundamental need but

rather functions to facilitate the attainment of other

goals. Formally, Johnson defines privacy as "those behav-

iors which enhance and maintain one's control over outcomes

indirectly by controlling interactions with others" (1974,

p. 90). Once freedom of choice has been exercised in the

selection of a goal, freedom of choice is performed in the

selection of behaviors expected to lead to that goal. Pri-

vacy is this behavior selection control designed to attain

pre-selected goals. Privacy, therefore, is not an outcome

but a behavior conducive to an outcome.

Whether privacy is satisfying is a function of the ex-

tent to which it achieves the desired outcome. The selec-

tion of privacy behavior automatically leads to multiple

consequences in the form of eliminated options in addition

to the intended consequence. Ambivalence then is fre-

quently experienced with privacy. Since secondary control

is distal from the desired outcome, the relationship be-

tween privacy and outcomes may be highly uncertain. This

uncertainty in turn is a source of stress and anxiety asso-

ciated with privacy.

Research on Privacy

As this review of privacy theories has indicated, pri-

vacy is a complex construct. The research on privacy mir-

rors this complexity. At the present embryonic stage of

theory and its testing, a detailed critique of privacy re-

search design and methodology seems premature. Instead

only the findings reported as significant will be presented.

Thus the word 'significant' is not repeated for each find-

ing. The methodological soundness of the reported research

and the appropriateness of the statistical analyses are but

briefly reviewed. Conflicting findings may be due, there-

fore, to a variety of operational, methodological, and an-

alytic differences and inadequacies.

Measures of Privacy

A large proportion of the research in the area of pri-

vacy has addressed such concerns as personality dispositions

toward privacy and attitudes about privacy. Before review-

ing the literature reporting on privacy research it is nec-

essary to briefly describe these personality and attitudi-

nal scales for privacy and report on their reliability and


Marshall (1972, 1974) has developed the most compre-

hensive dispositional measure of privacy preferences. The

fifty-six item Privacy Preference Scale (PPS) consists of

six statistically determined factors: Intimacy, Not Neigh-

boring, Seclusion, Solitude, Anonymity, and Reserve. While

no reliability level was reported for the scale as a whole,

tests of subscale (factors) item homogeneity using coeffi-

cient alpha yielded reliability coefficients ranging from

.80 for Not Neighboring to .56 for Intimacy. The subscale

totals when correlated with the PPS total score ranged from

.47 (Intimacy) to .65 (Not Neighboring). The reported va-

lidity for the PPS is based on its correlation with seven-

teen predicted relations with behavior. Of these seventeen

hypotheses, nine of them were confirmed. Only one relation-

ship of thirty-seven non-predicted hypotheses was signif-


The Need-for-Privacy Subscale (NPS) of McKechnie's

(1977) Environmental Response Inventory (ERI) is a gener-

alized measure with no distinctive factors. The major

themes of the subscale are the need for physical isolation

from stimulation, need for freedom from distraction, and

dislike of neighboring. The only published scale reliabil-

ity and validity information appears in the scale's manual

which is available only through purchase.

The Environmental Privacy Scale (EPS; Windley, 1973)

was developed to assess preferences to control access to

others by using features of the physical environment. The

reliability (item homogeneity) for this scale was intoler-

ably low at .40 indicating high measurement error. Because

of this low reliability no validity data were reported for

the EPS.

Research Related to Biological Theories

There is no directly relevant research on privacy ap-

plicable to Halmos' (1953) bio-social need theory, Mehrabian

and Russell's (1974) semantic arousal mediation theory, or

Bechtel's (1977) ecological theory. Research is available

on territorial theories (Klopfer & Rubenstein, 1977; and

Pastalan, 1970b) and on adaptation theories (Hill, 1969; and

Marshall, 1972, 1974) of privacy.

Sommer (1966) has distinguished two strategies of ter-

ritorial defense to protect privacy or attain privacy:

avoidance and offensive display. In task situations free-

dom from distraction can be attained either by being phys-

ically removed from others (avoidance) or by laying effec-

tive claim to a large work area (offensive display). At

library study tables, the avoidance strategy for privacy

resulted in a significantly greater choice of end chairs

at tables. Chairs at the middle of tables were chosen more

for the offensive display strategy. The avoidance strategy

was used more than offensive displays. Sommer hypothesized,

but did not test, that the effectiveness of the offensive

display strategy decreases as social density and size of

the area claimed increases.

Barriers to interaction are defensive territorial mark-

ers; fencing is an example of this territorial strategy.

Marshall (1974) found that persons scoring high on the PPS

Reserve subscale had a greater amount of fencing around

their homes.

Within the home, Rosenblatt and Budd (1975) obtained

differences in the use of territory for privacy between mar-

ried and unmarried couples. Unmarried couples were signif-

icantly more likely than married couples to have an area

within their residence for being alone. Married couples,

however, had more separate storage areas than unmarried


While territory is one strategy to attain privacy,

territoriality and privacy have different psychological

effects (Edney & Buda, 1976). Students preferred a situ-

ation which was both in their territory and afforded pri-

vacy for a wide range of activities: (1) personal activ-

ities: taking care of personal business, personal hobby,

and writing a book about self; (2) work activities: study-

ing, reading and creative writing; (3) basic biological

functions: sex and toilet-grooming; and (4) thinking about

anything and singing out loud. Territory without privacy

was more preferred for relaxation activities: watching TV,

'just relaxing-being yourself,' casual conversation with

one friend, and eating but not for relaxing after a hard

day. For drinking and 'doing something very different and

unconventional' subjects preferred no territory and no pri-

vacy. Across all seventeen activities, a majority of peo-

ple preferred territory without privacy indicating that peo-

ple distinguish between privacy and territory.

Edney and Buda (1976) and Bower (1979) also found that

territory and privacy differentially affect attribution,

creativity, perception of the environment, and self-percep-

tion. Subjects without privacy and those without territory

attributed their behavior more to the influence of others.

Bower (1979) was unable to replicate this finding for the

territory condition but did replicate the privacy effect on

attribution. Subjects in privacy but not territory scored

higher on a creativity test than those without privacy.

Rooms which were private were perceived as more stimulating

than non-private ones. Both privacy alone and the inter-

action of privacy and territory increased the felt comfor-

tableness of the room (Bower, 1979). Territoriality did

not affect room stimulation perception. Territory did, how-

ever, increase reports of feelings of possession whereas

the privacy conditions did not. Bower (1978) replicated

this finding with the semantic differential but also found

an interaction between privacy and territoriality on seman-

tic differential ratings of possession.

Hill's (1969) use of adaptation level theory relates

present privacy opportunities to perceived privacy. As ac-

tual privacy opportunities decrease the preference for pri-

vacy should increase with all other things being equal.

The preference for privacy increased as the number of po-

tential observers passing by windows outside increased

(Hill, 1969). Marshall (1972) provided additional support

for this hypothesis. Preference for privacy scores on her

PPS scale for the total PPS, Seclusion, Solitude, and Re-

serve subscales all decreased as perceived privacy within

the present home increased. Inversely, as perceived crowd-

ing within the present home increased, the preference for

Reserve increased. People who perceived their present

house as too crowded (insufficient privacy) reported ob-

jective conditions which indicated a lack of actual privacy

opportunities: (1) more persons per room, (2) less ade-

quate insulation between noisy and quiet activities in the

home, (3) more able to overhear neighbors, (4) and more

houses visible from the living room windows. Those who per-

ceived their homes as sufficiently private had more actual

insulation between noisy and quiet activities, had fewer

houses visible from their living room windows, and lived

further away from neighbors. People increased the use of

drapes drawn closed during the day as their perceived pri-

vacy from neighbors decreased (Marshall, 1972). Contrary

to the adaptation hypothesis of perceived privacy, per-

ceived privacy from neighbors increased as reported noise

from traffic and from neighbors increased (Marshall, 1972).

Pastalan (1974) reported that PPS scores decreased

for nursing home residents who moved to a new nursing home

which provided more actual opportunities for privacy (pri-

vate rooms). Those patients reporting having a successful

strategy for attaining privacy had lower preferences for

solitude. The desire for Intimacy (PPS subscale) was most

clearly related to changes in actual levels of privacy.

The preference for reserve (PPS subscale), however, was not

related to changes in actual reserve.

Results reported by Lawton and Bader (1970) attenuate

Hill's adaptation hypothesis. Regardless of actual privacy

opportunities, the preference for privacy increases with

age and socio-economic status.

Hill's (1969) other adaptation hypothesis dealt with

the optimization of privacy and boundary permeability.

Privacy boundaries function to control interaction and ac-

cess, not to eliminate these. There is a functional bal-

ance between privacy (access to self) and visibility (ac-

cess to others). Hill found this optimization process, as

the desire for privacy increased the visual access by the

resident to others outside the home decreased.

Rather than interpreting adaptation level theory in

terms of present privacy conditions, Marshall (1972, 1974)

focuses on the impact of childhood privacy conditions on

adult privacy preferences. Those people experiencing lit-

tle privacy in childhood will develop a low preference for

privacy as adults. Their adult preferences are a result of

childhood adaptations. Marshall (1972, 1974), however,

found minimal relations between the density of childhood

environments and adult privacy preferences. There were no

significant relationships between the childhood density

measures and preference for seclusion or solitude (PPS

subscale scores). Privacy preference (total PPS score) was

not significantly related to childhood perceived privacy,

number of siblings, childhood perceived crowding, having

own room, amicability of siblings, amount of open space in

childhood neighborhoods, childhood town size, and childhood

spatial mobility.

Some relationships between certain childhood density

variables and isolated PPS subscales were found. Prefer-

ence for anonymity increased as the size of the town one

had grown-up in increased. An increasing size of childhood

town increased the preference for adult intimacy. The de-

sire for intimacy was higher amongst those adults who had

lived in single family residences as children. While shar-

ing a room as a child, having many siblings, and having

open space around the childhood home were not related to

adult privacy preferences, these variables were related

to perceived crowding of the childhood home. Those adults

who in turn remembered their childhood homes as crowded

preferred both anonymity, intimacy, and reserve.

Developmental Research

Developmental research on privacy can be divided into

four categories: (1) research on changes in privacy atti-

tudes over the developmental cycle, (2) research on changes

in the meaning of privacy over the developmental cycle,

(3) research on changes in household privacy behaviors and

their socialization correlates over the developmental cycle,

and (4) research on changes in household privacy behaviors

over the stages of the family cycle. These will be reviewed

in turn.

Viewing preference as an attitude, preference for a

private room increases with age from years ten through forty

(Lawton & Bader, 1970). This privacy preference stabilized

after age forty. Analyses of Marshall's (1970 and 1974)

Privacy Preferency Scale (PPS) scores resulted in several

age related differences. Adults (the parents of the stu-

dent respondents) had higher preferences for reserve and

not neighboring while the students had higher preferences

for solitude and intimacy. Attitudes about what content

should be private indicated that older persons are less

concerned than younger persons about others possessing

knowledge of one's interests, values, and social adjust-

ment (Rosenbaum, 1973).

Wolfe and Laufer's (1974) research substantiated the

theoretical predictions by Laufer, Proshansky, and Wolfe

(1973) that with maturation the concept of privacy should

become more complex and the salience of the defining attrib-

utes of privacy should change. The ability to give a def-

inition of privacy is a function of age, 53 percent of five-

year-olds could not give a definition even though they

could give examples of private times and places. The ma-

jor shift point in this definitional ability occurred be-

tween the ages of five and seven. only 7 percent of seven-

year-olds failed to provide a definition. The complexity

of these definitions as indexed by nonredundant responses

increased with age. The greatest increase in conceptual

complexity occurred between ages eleven and thirteen.

The salient meanings of privacy changed frequently

between years five and seventeen. Five meanings were sa-

lient at all these ages: alone, doing what you want with-

out being bothered, quiet, alone with somebody, and con-

trolling access to information. These correspond to sol-

itude, control, task privacy, intimacy, and reserve. The

most salient dimension at age five is quiet. Controlling

access to spaces, being away from people, doing and think-

ing alone, and no one bothering me were not given by five-

year-olds but were given by older children. The most sa-

lient meaning at age seven was 'controlling information.'

The salience of controlling interaction (being alone when

you want to) as a privacy meaning occurred for all ages

nine and above. Controlling individual activity, i.e.

autonomy (being able to do what you want) is a salient mean-

ing for children eleven and older.

The age points at which the salience for various mean-

ings change vary by specific meaning. The increase for

the salience of 'no one bothering me' occurs between ages

seven and nine, (2) for 'alone' between seven and eleven,

and (3) for 'controlling access to information' between

thirteen and seventeen. 'Controlling access to information'

decreased from age seven to thirteen and 'controlling access

to spaces' decreased between eleven and thirteen.

Comparing the salience of alternative meanings across

age groups revealed a complex pattern of results.

"Controlling information" was emitted more than "alone"

for only seven-year-olds, than "no one bothering me" for

all ages except thirteen-year-olds, and than "controlling

access to spaces" by all ages. The meaning "alone" was

given by a greater percentage of children than "no one

bothering me" at all ages except seven, and more than "con-

trolling spaces" at all ages except seven and nine. Only

children at age thirteen gave "no one bothering me" more

than "controlling spaces." More children at all ages gave
"alone" and "controlling information" than "no one bother-

ing me" and "control over spaces." The percent of chil-

dren using "when you want to" as part of the meaning for

privacy increases with age and is used more often after

age fifteen.

Averaged across all ages, the most frequently pro-

vided meanings for privacy were: (1) controlling access

to information (39%), (2) alone (33%), (3) no one bother-

ing me, and (4) controlling access to spaces. These four

categories of mean were given by four largely independent

groups of children controlling for age. This result indi-

cates that control per se cannot be viewed experientially

as unidimensionally.

Knowledge about privacy behavioral changes and the so-

cialization influences on these changes is provided Parke

and Sawin's (1979) research. They found that the use of

physical privacy markers and privacy rules both increased

with the child's age. Both of these developmental trends

were focused around two household areas: the bathroom and

the child's bedroom. The primary privacy marker was the

closed door and the most prominent privacy rule was knock-

ing on closed doors. A closed door signalled restricted

access to the room. As the age of the child increased,

both parents and siblings were more restricted in their

access to the bathroom occupied by the child. The largest

shift (increase) in the child restricting access to others

occurred during early adolescence. Adolescents of the same

age who were more physically mature as measured by an ob-

jective index were more likely to keep the bathroom door

closed and to restrict access to other family members even

after they knocked.

Viewing the family as a social system, the exercise

of privacy by one member is largely contingent on the co-

operative behavior of the remaining family members. As a

result not only are parents and siblings increasingly re-

stricted in access to the bathroom as the child ages but

parents knock more on both the bathroom and bedroom doors

of older children. These mutually contingent behaviors

are strongly attenuated by the gender composition of inter-

actants and by the activity being engaged in by the pri-

vacy-seeking member. The child's restriction of access to

others is less for same sex versus opposite sex parents and

siblings. Inversely parents knock on the bathroom door

more when it is occupied by a child of the opposite gender.

While fathers knock on daughter's bedroom doors more than

their son's door, mothers' knocking show no difference be-

tween sons and daughters. The greatest increase in the per-

centage of girls restricting access to the bathroom was be-

tween ten and thirteen but for fathers only. The changes

for boys occurred later (fourteen to seventeen) and was di-

rected toward both parents.

Not surprisingly the type of activity influences pri-

vacy. The more intimate the bathroom behavior (toilet use,

bathing, dressing, and grooming), the more others are re-

stricted access as the child gets older. This restriction

is applied more to parents of the opposite rather than like


Parke and Sawin (1979) compared the influence of matu-

ration with house size and family size. Neither overall

household square footage nor family size were related to

the privacy measures. Family density (space divided by

family size), however, was related to the privacy measures.

As the number of bedrooms in a house increased, more chil-

dren kept their bedroom doors closed. It can be hypoth-

esized that parents with greater privacy preferences choose

houses with more bedrooms and encourage greater privacy be-

haviors in their children. In support of this hypothesis,

the more bathrooms in a house the more likely that children

between two and nine, but not for those between ten and

seventeen (a ceiling effect possibly for this latter group?)

closed the bathroom door. Family density was curvilinearly

related to both sibling access restriction to bathrooms and

parental knocking on bathroom doors. Both these behaviors

were high in low and high density homes.

A number of maternal socialization practices impacted

on children's privacy and children's privacy behaviors.

Restrictive and coercive mothers exercise more control over

their children's privacy habits. Children's bedroom doors

were kept open more during recreational periods and while

entertaining friends in homes where the mothers were more

protective, imposed more restrictive regulations, clar-

ified and enforced household rules more, and made more co-

ercive suggestions. Children of such mothers sought more

privacy as indicated by keeping the bathroom door closed


Affectionate and approving mothers permit more private

behaviors by their children and have children who seek less

privacy. Children of mothers who were more affectionate

and more approving in their critical appraisals kept the

bathroom open more while grooming and dressing (Parke &

Sawin, 1979).

Developmental changes in privacy continue beyond child-

hood and adolescence. Relationships and families also move

through a developmental cycle. Unmarried co-residing cou-

ples exercise different privacy behaviors than similar

married couples (Rosenblatt & Budd, 1975). While married

couples are more likely to have a place in their residence

to be alone than married couples, married couples have more

separate storage areas. Neither type of couple showed a

difference in access restrictiveness toward their partner

when occupying the bathroom. Couples who co-resided before

marriage were less territorial (separate storage areas)

than couples married without prior co-residence (Rosenblatt

& Budd, 1975).

Once a married couple has children further modifica-

tions in privacy behavior occur. Mothers with children

have less time alone in the house and less time alone with

guests (Smith et al., 1969). There was no difference, how-

ever, in the amount of time mothers and nonmothers spent in

the home. Those with pre-school children have less privacy

than those mothers with only school age children. More pro-

tective mothers (those who always accompany their children

outside and require the child to play in the same room in

which the mother is working) have less privacy. Alterna-

tively those mothers who participate more in their chil-

dren's play and encourage their children to help more in

household chores have less privacy. Household rules also

affect the mother's privacy. When one room is the chil-

dren's play area and other rooms are "off limits" and when

all the pre-school children are required to nap at the same

time, the mother has more privacy.

Fathers have less privacy (time alone in the home)

than husbands without children. Men with children achieved

privacy by spending less time in the home than did those

without children who acquired their privacy by having a

separate room in the home for themselves. Although fathers

spent less time in the home than husbands without children,

the amount of time they spent in interaction with other mem-

bers of the family did not vary across family stages (Smith

et al., 1969).

Family stage also affects the amount of children's

privacy. The youngest pre-school child has more privacy

than other age children due to nap time which may not al-

ways be spent in sleep. Children with private bedrooms

were able to choose more privacy and had more privacy than

those who shared bedrooms (Smith et al., 1969). On a cog-

nitive level, however, Wolfe and Laufer (1974) found no re-

liable differences in the salience of privacy meanings be-

tween children sharing and not sharing bedrooms.

Research on Interpersonal Theories of Privacy

When privacy is conceptualized as control over when and

to whom to transmit information, then research on self-dis-

closure becomes relevant to privacy. Jourard (1966, 1971)

has explicitly related these two research areas as points

along the same continuum of self-communication. In both the

privacy (Westin, 1967) and self-disclosure literature,

intimacy is a key concept. The meaning of the concept is

not the same, however, in both contexts. As a state of

privacy, intimacy is the chosen seclusion of a dyad or

group (Westin, 1967). Intimacy in self-disclosure refers

to the self-centrality of the information communicated

(Derlega & Chaikin, 1977). Although the privacy state of

intimacy may imply a depth of disclosure between the par-

ties, this disclosure depth is not an explicit definitional

component. The term intimacy will be used subsequently in

this paper to refer only to the depth of disclosure and not

to Westin's (1967) state of privacy for which 'group seclu-

sion' will be substituted.

Reserve, which is one of Westin's privacy states

rather than Westin's intimacy, is conceptually more sim-

ilar to self-disclosure intimacy. Defined as "a psycho-

logical barrier against unwanted intrusion" (Westin, 1967,

p. 32), Westin unfortunately includes three discrete types

of behavior under reserve: (1) withholding disclosure,

(2) discretion by the recipient of disclosure, and (3) se-

lective attention to others (tuning others out). Reserve

as withholding disclosure is clearly related to the self-

disclosure research. This research (Derlega & Chaikin,

1977) has shown that one's level of self-disclosure is par-

tially determined by the discloser's trust in the recipi-

ent's discretion.

The implicit notion that 'group seclusion' (Westin's

intimacy) is a facilitating condition for self-disclosure

intimacy has been supported by Holahan and Sleikin (1977).

When physical-environmental barriers were not available for

privacy, the psychological barrier of reserve became oper-

ative as demonstrated by the reduced level of self-disclo-


Altman (1975) theorizes that privacy is a dialectic

process; there is a shifting balance between social con-

tact (penetration) and withdrawal (privacy). Self-disclo-

sure operates as a dialectic process; "the best predictor

of the level of intimacy B will use in disclosing to A is

the level of intimacy with which A initiated the communica-

tion" (Derlega & Chaikin, 1977, p. 106). The intimacy of

communication between members of a dyad tends to stabilize

at the level at which one member stops reciprocating the

level of intimacy (Altman & Taylor, 1973).

Research Related to Social-Normative Theories

Chaikin and Derlega (1974) and Derlega and Chaikin

(1977) have identified several norms current in American

society that regulate the amount and level of self-disclo-

sure. These norms serve as one form of the institutional-

ization of privacy. Such institutionalized privacy modes

are hypothesized to reflect and express the social structure

(power hierarchies). Not only are women expected to

disclose more than men, they in fact do disclose more and

the content of this increased disclosure exposes their weak-

nesses (Derlega & Chaikin, 1977). The content of men's

self-disclosure consists more of information attesting to

their strengths. Social dominance by men is manifest both

by the amount (less) and content of their self-disclosure.

When one person or group knows more about another person or

group, that person or group is in a position to influence

or exploit the other.

Control over when, where, how and by whom one's body

is touched is another form of privacy. Research shows that

a touched person is more likely to be of a lower occupa-

tional status than the toucher (Henley, 1973). Another

instance of male dominance and privacy is the finding that

men initiate touch with females more than vice versa (Hen-

ley, 1973a). If increasing age confers increasing status,

it is consistent that a touched person is younger than the

toucher (Henley, 1973). Henley (1974) concluded that

personal information flows opposite to the chain of author-

ity. The prerogative of eliciting personal information,

however, is in line with the authority hierarchy (Kelvin,


Kelvin (1973) theorized that perceived privacy de-

creases as one's vulnerability increases. The perception

of vulnerability increases with the power differential be-

tween the self and the other. The specific information

that places one in a position of vulnerability is deter-

mined by norms and the reference group in which one is

functioning. Anxiety accompanies the perception of vulner-


Self-disclosure is much less in those content areas

that provoke anxiety (Derlega & Chaikin, 1977). Altman and

Taylor (1973) found evidence that personal characteristics

negatively valued by the self and those socially forbidden

by one's reference groups were scaled as more intimate.

These attributes were also perceived as having higher costs

(vulnerability). One of five factors found for questions

judged to be sensitive to privacy invasion was 'social ad-

justment' (Rosenbaum, 1973). Items of this factor dealt

with negatively valued personal characteristics such as

drug and alcohol use, and psychiatric history. Those groups

lower in the social structure (women, younger respondents,

the less educated, and those with lower income) were more

sensitive to questions about social adjustment.

Questions inquiring about finances and family back-

ground (including religion and race) were rated as the

most vulnerable to privacy invasion (Rosenbaum, 1973). Re-

ligion was also an objectional content area in MIPI ques-

tions (Gynther, 1972). Dominant groups (males and those

with more income) were more sensitive to financial questions.

An exception to this result was that lower educated persons

rated these items as more private. Only the less educated

and those from urban areas were more concerned about ques-

tions delving into family background. In a rural setting

one's family background is already public knowledge.

Privacy norms function to protect those behaviors pro-

scribed to private performance (Kelvin, 1973). The two con-

tent areas of the MMPI with the most items judged as intrud-

ing on privacy were those dealing with sex and elimination

processes (Gynther, 1972).

No research was found bearing on Pastalan's (1970b)

theory of role complexity, role incompatibility, and pri-

vacy. Neither has research been conducted to test Bechtel's

(1977) privacy theory which is derived from ecological psy-


Research on Functional Theories of Privacy

While the social-normative theories focus on the ef-

fects of power on privacy, the functional theories empha-

size the control nature of privacy. Privacy functions to

facilitate freedom of choice and to reduce social con-

straints on behavior. The only reported research relevant

to these issues investigated cognitive and emotional dimen-

sions of control.

The meanings 'controlling access to information' and

'doing what you want without being bothered' were salient

for all age groups from five to seventeen (Wolfe & Laufer,

1974), while 'controlling spaces' was salient for all those

above age six. Control over information and over spaces

were salient to distinct groups of children. Wolfe and

Laufer (1974) interpreted this finding as indicating that

while the notion of control may be central to meanings of

privacy, the behaviors over which control is incorporated

into the meaning of privacy vary between groups.

A study (Harmon & Betak, 1974) of privacy constructs

related to housing showed that the value of privacy was in

controlling intrusion from neighbors. An analysis of these

privacy constructs led to the conclusion that freedom of

choice was a superordinate construct unifying the privacy

constructs. A major concern elicited from elderly res-

dents about neighborhood designs was the privacy concern

for autonomy or freedom of choice over involvement with

children in particular (Steinfeld, 1973). Finally, sub-

jects in a privacy versus non-privacy condition reported

feeling more freedom (Edney & Buda, 1976; and Bower, 1978)

and control (Dower, 1978).

The conceptualization of privacy remains a crucial is-

sue. The choice of one definition rather than another re-

sults in different theories, programs of research, and em-

pirical results. As a first attempt at clarifying this is-

sue, two types of definitions are considered: formal and

subjective. Formal definition is one which consists of a

set of properties external to the organism. The occurrence

of this set of properties constitutes an instance of the

referent. The definition of reinforcement in behaviorism

is such a formal definition. Jourard's (1966) conceptual-

ization of privacy as an informational outcome state exist-

ing between interactants is another formal definition. Any

condition which changes the informational flow between two

individuals would be a variable impacting privacy. Poten-

tially independent of this informational exchange is the

person's emotional experience of 'privacy.' A subjective

definition, therefore, is one whose referent is some inter-

nal state of the organism, in this case an emotional state

associated with the linguistic label--'privacy.' It is sub-

jective precisely in the sense that the meaning of the lex-

eme associated with the emotional state changes with age

(Wolfe & Laufer, 1974) and across culture (Roberts & Gregor,

1971) and history (Flaherty, 1972).

The types of theories required to deal with subjective

privacy must address both meaning and learning. A theory

of meaning should provide an explanation of the semantic

components of the referent 'privacy.' Such a theory should

locate the concept within a semantic domain at a specified

level of hierarchical inclusion. Privacy is not an iso-

lated concept but rather one within an interlocking network

of concepts. Using the lexeme 'woman' as an example, woman,

which is a unit in the domain 'humans,' is an adult and a

female of the species. Evidence already exists demonstrat-

ing that the meaning of privacy changes with age (Wolfe &

Laufer, 1974). This suggests that some aspect (either

components or structure) of the domain in which privacy is

an element is changing. An explanation of subjective pri-

vacy, therefore, will require some theory of learning to

account for these semantic changes. Wolfe and Laufer (1974)

conclude that the complexity of the meaning of privacy in-

creases with age thus indicating that the environmental

stimuli eliciting the emotional state associated with the

cognitive-linguistic category 'privacy' become more diverse.

What are the dimensions underlying these diverse, complex

stimuli which elicit the feeling of privacy? What explains

variations in these underlying dimensions across age and


A program of research on subjective privacy should con-

sist of explorations into meaning such as that by Wolfe and

Laufer (1974) and multi-dimensional analysis of situations

to identify the dimensions underlying privacy across a wide

range of environmental (both physical, social, and conative)

contingencies. Research should also be directed toward the

role of socialization (learning) on the meaning of privacy.

Parke and Sawin (1979) have examined privacy behaviors in

light of socialization practices but have not yet addressed

the differences in meaning that privacy takes on under these

varying types of socialization. Both the referents of

linguistic categories and the experiencing of emotions are

largely consensual within a social group or culture. One

manifestation of this consensus is the presence of norms.

Norms are a given existing before an infant's birth and

constitute a part of the infant's socialization. These

larger contexts of privacy meaning must be investigated

for an adequate theory of subjective privacy.

As an initial approximation to understanding privacy,

the proposed research investigates subjective or perceived

privacy. In this instance subjects will themselves decide

what is "private" by responding to an elicitation frame in-

structing them to describe a situation in which they felt

private. Within this conceptual context and following the

reasoning of Schopler et al. (1978), this research asks,

"What are the cognitive and stimulus dimensions that deter-

mine peoples' perceptions of being private?" The preced-

ing review of the privacy literature indicates a wide di-

vergence about what constitutes the necessary and sufficient

conditions for eliciting the state of feeling private. Nine

major conditions can be identified in the literature and

these are the dimensions hypothesized to be the determi-

nants of the experience of feeling private:

1. a physical situation characterized by the absence

of others (solitude), Westin (1967);

2. an interpersonal situation involving limited so-

cial stimulation (the absence of distraction),

Bechtel, (1977) and Mehrabian and Russell, (1974)

(and the absence of stimulus overload), Altman


3. the selective control to which others encroach

upon one's life (personal control), Proshansky,

Ittelson & Rivlin (1970);

4. the selective control of information communicated

about self (reserve), Jourard (1966), Westin (1967),

and Chaikin and Derlega (1977);

5. a situation in which the person is an outsider or

unknown to others (anonymous), Westin (1967);

6. a situation of desired physical and/or psycholog-

ical immediacy between two or more people (inti-

macy), Westin (1967);

7. a situation in which one is free from observation,

Shaw (1980);

8. a situation characterized by perceived adequacy

of space, Bechtel (1977), Sommer (1966), and

Stokols (1972);

9. the amount of information one person possesses

about another.

Other conditions have been posited as leading to the desire

for privacy. An example is the enactment of anti-normative

behaviors (Kelvin, 1973). The study of the variables con-

ducive to eliciting a desire for privacy is, however, an-

other issue from the investigation of the perception of pri-

vacy. But, although theoretically there should be a syste-

matic relationship between the two sets of variables, but

the investigation of motivational determinants is beyond the

scope of this research proposal. The same caveat applies

for the consequences of privacy.

Given the formative stage of theory and research on

privacy, the selection of a research strategy which samples

a large variety of situations in which people may experi-

ence privacy and which is independent of any one theory

and/or conceptualization of privacy is desirable. Such a

promising exploratory methodology is multidimensional scal-

ing (Shepard, Romney & Nerlove, 1972). This methodology

has been gainfully applied to the concept of crowding by

Schopler et al. (1978) and Stockdale et al. (1978) and

was adopted for this study (see Appendix A for a conceptual

explanation of multidimensional scaling).


The study consisted of three phases. The first phase

involved collecting paragraphs describing situations in

which students felt private. These served as the experi-

mental stimuli. In Phase 2 subjects produced similarity

rankings between the privacy stimuli. These data enabled

the determination of the dimensions used by respondents in

comparing and discriminating among privacy situations. In

order to identify and label these major dimensions, Phase 3

respondents rated each privacy situation on 40 relevant

qualities. An overview of the procedures employed will be

described separately for each of these phases. A complete,

detailed description of each phase is provided in respec-

tive appendices noted.

Phase 1

Paragraphs describing privacy situations were collected

from 36 undergraduates from the University of Florida. Sub-

jects received class credits for responding five times to

the following instructions:

Please think of a time in your life when you felt very


Describe the incident you have in mind, and your

feelings at the time, in as much detail as pos-

sible. (Refer to Appendix B for complete instruc-


Forty privacy situations were selected from the 187 col-

lected to serve as the experimental stimuli. These 40

situations represented: (1) all four types of privacy

(solitude, anonymity, reserve, and intimacy) and (2) the

most and least private situations for each of the four

types (refer to Appendix B for a description of this se-

lection procedure).

Phase 2

Twenty males and 20 females participated in Phase

2 in partial fulfillment of the requirements for an intro-

ductory psychology class. Groups composed of 5 same-sex

persons working at two, two-hour sessions one week apart

completed the experimental tasks.

The experimenter explained that he was interested in

discovering what caused people to feel private. Subjects

were told that the task they were about to perform would

provide this information. Respondents were given a deck

of 4x6 cards on which were typed the experimental stimuli

(the privacy situations collected in Phase 1) and were first

instructed to familiarize themselves with the descriptions.

Subsequently the subjects were presented response booklets.

At the top of each page of the booklet was a "target situ-

ation" and subjects were instructed to rank order all of

the privacy situations in terms of their similarity in the

experience of being private to each target. The nine sep-

arate target situations were members of the 40 experi-

mental stimuli and were selected to represent each of the

nine hypothesized conditions causing the subjective expe-

rience of privacy (see Appendix C for a description of this

selection procedure). Order of targets was randomized over

subjects. One target was repeated to provide a reliability

estimate (refer to Appendix C).

Phase 3

The dimensions constituting the experience of subjec-

tive privacy were determined by analysing the data obtained

in Phase 2 via SAS ALSCAL (Young & Lewyckyj, 1979). In

order to identify and label these resultant dimensions,

45 introductory psychology students participated in

Phase 3 for course credit. One female and two males failed

to appropriately complete the experimental task and their

data were excluded from the analysis. Subjects were tested

in same-sex groups of 10.

The experimenter explained that the purpose of the ex-

periment was to identify the qualities that best character-

ized situations in which students felt private. Each

subject received a booklet containing 41 pages, each of

which contained a privacy situation and listed 40 rele-

vant qualities on which to be rated. (See Appendix D for

the selection of the forty qualities.) These qualities

reflected hypotheses about the dimensions of subjective

privacy. Subjects were instructed to rate the situation

and the actors on a ten-point scale and to record their

responses on computer-scored answer sheets. Both the or-

der of qualities and situations were randomly determined.

The presentation order of the privacy situations was pro-

gressively staggered over the ten booklets such that fa-

tigue effects were distributed equally over all 40 exper-

imental stimuli (see Appendix D). One experimental stim-

ulus was repeated at the end of the booklet to provide a

response reliability estimate (see Appendix D).


The ranking data collected in Phase 2 were analyzed

with SAS-ALSCAL (Young & Lewyckyj, 1979). The following

parameters guided the analysis: (1) the measurement lev-

el of the datawere ordinal since a ranking task was used;

(2) the data were dissimilarity as opposed to similarity

in type; (3) the data we row conditional; and (4) the

data were asymmetric in the sense that not all possible

comparisons of stimuli were performed. Row conditional

data obtain when the experimental stimuli are ranked

against target stimuli. The meaning of rank values var-

ies for each standard target stimulus against which an

item is compared. The characteristics of the observations

are conditional on some aspect of the experimental proce-

dure (the targets) in such a way that some observations

cannot be meaningfully compared with other observations.

Since multidimensional scaling is primarily an ex-

ploratory analysis, a triangulation approach was used for

analyzing and interpreting the data. Although ALSCAL pro-

vides for each dimensional solution a measure of goodness

of fit (stress) and the amount of variance accounted for,

there is no statistical criterion which signals which di-

mensional solution is the best or appropriate one. Two

sources of information are resorted to in determining the

most appropriate and illuminating dimensional solution.

First, correlations between the dimensional coordinates of

each solution and the defining attribute ratings are exam-

ined and reported. Second, correlations between the dimen-

sional coordinates of each solution and the privacy unidi-

mensional scale values obtained in Phase 1 are examined and

reported. This triangulation of data sometimes converges

to clarify the interpretation of the results and at other

times suggests noncongruent, alternative interpretations.

Multidimensional Privacy Results

The stress values and r-square values for the one

through six dimensional solutions are presented in Table 1.

The reduction in stress values due to the generation of the

third and subsequent dimensions is small. This suggests

that a two-dimensional solution is adequate. Subsequent ex-

ploratory analyses of the three- through five-dimensional

solutions indicate that the four-dimensional solution is

noteworthy. Thus both the two-dimensional and four-dimen-

sional analyses are presented.

The first dimension of the two-dimensional solution is

defined by three attributes. The attribute, "too much going

on around the actor," is the only one achieving an accepted

level of significance (r=-.36, p=.02). The other two

Table 1.

Privacy MDS Results

Dimension Stress RSQ

1 .516 .196

2 .394 .244

3 .320 .308

4 .264 .368

5 .223 .425

6 .192 .480

attributes (too many people for space available, r=-.31,

p=.053; and crowded, r=-.30, p=.06) are marginally signif-

icant. Together these attributes are interpreted as "So-

cial Density" and this dimension is thus labeled.

The second dimension of the two-dimensional solution

has one attribute which is correlated at a level of margin-

al significance. "The actor's behavior is against the

rules" (r=.29, p=.068) is cautiously interpreted as a "Nor-

mativeness" dimension. This dimension is unique to this so-

lution, not reappearing in any other dimensional solution.

This argues for yet increased caution.

The four-dimensional solution further elaborates the

first dimension previously discussed. The correlations of

the attributes "too much going on around the actor" (r=-.48,

p=.03) and "the actor feels claustrophobic" (r=-.31, p=.05)

are also now significantly related to this dimension. The

one marginally significant attribute ("the actor wants to

escape" (r=-.29, p=.07)) also seems relevant to the inter-

pretation of this dimension. This dimension represents a

continuum from crowded to not crowded. Four consequences

are apparently associated with this crowding continuum. The

amount of stimulation impinging on the actor, the degree of

privacy available, the motivation to escape the situation,

and the feeling of claustrophobia all vary along the same

dimension as crowding.

No attribute is significantly or marginally correlated

with the second dimension of the four-dimensional solution.

The attribute defining the second dimension of the two-di-

mensional solution is only correlated r=.05 with this dimen-

sion. This dimension is uninterpretable and thus unlabeled.

The nine attributes defining the third dimension indi-

cate that it is a complex one. In order of correlational

magnitude, the nine are: (1) no one knows the actor, anon-

ymous (#12, r+.50, p=.001); (2) the actor is visible (#16,

r=-.45, p=.004); (3) the actor is self-revealing (#22, r=

-.44, p=.005); (4) the actor is solitary (#30, 4=-.45, p=

.004); (5) another knows the actor's thoughts and feelings

(#6, r=-.41, p=.008); (6) the actor's feelings are caused

by the presence of others (#29, r=-.41, p=.03); (7) the

actor trusts others present (#20, r=-.34, p=.03); (8)

others understand how the actor feels (#15, r=-.33, p=.04);

and (9) the actor is free to be self (#10, 4=.32, p=.04).

The attribute "the actor is careful about what they say"

(#13, r=.27, p=.09) is marginally significant. A single

rubric which encompasses all of these attributes is not

immediately obvious.

"Privacy" suggests itself as one candidate. One end

of this dimension is anchored by two types of privacy,

i.e., anonymity and solitude. The other end, however, is

marked by two attributes (self-revealing and being care-

ful about what is said) which are frequently believed to cre-

ate two other types of privacy, i.e., intimacy and reserve

respectively. Partially for this reason, this dimension

is not interpreted as privacy.

What is desired is a construct which encompasses all

of these attributes without contradiction or exception. A

careful, detailed examination of the attributes varying

along this dimension suggests such a construct. As one ap-

proaches the end of the dimension anchored by anonymity,

solitude, and freedom to be oneself, the following changes

occur: other's knowledge of self decreases, one's visibil-

ity decreases, one's self-revealingness decreases, other's

understanding of one's feelings decreases, one's trust of

others decreases, one's guardedness over what is said de-

creases, and the extent to which one's feelings are caused

by others decreases. Moving away from this anchor point,

anonymity or ignorance of one's identity (both name and self-

hood) decreases, solitude or withdrawal from others de-

creases, and freedom to be oneself (interpreted here as

spontaneity) decreases. These processes suggest a movement

toward or away from engaging others and being involved with

others. This dimension is, therefore, interpreted and la-

beled as "Interpersonal Involvement."

The fourth dimension shares three attributes (#4, #24,

and #40) with the first dimension. The polar opposite of

these attributes is, however, different in the two dimen-

sions. Only "the actor wants to escape" (r=.32, p=.04) is

significantly correlated to this dimension. The attributes,


Table 2.

Privacy MDS Dimensions Correlated with
Defining Attributes

2 4
Dimension Item # r r

37 too much going on around
19 too many people for space
38 crowded
28 too little privacy
24 actor feels intruded upon
40 actor feels claustrophobic
4 actor wants to escape

8 actor's behavior is against
the rules

-.36 -.48

-.313 -.40



no one knows actor, anonymous
another knows actor's thoughts
and feelings
actor is visible
actor is solitary
actor's feelings caused by
presence of others
actor is self-revealing
actor is free to be self
actor is careful about they say
others understand how actor
actor trusts others present

actor wants to escape
situation is largely physical
situation is sexual
actor feels claustrophobic
actor feels intruded upon
actor's feelings due to







+ 28





Table 2--continued.

iDimension #2 of the two-dimensional solution is distinct
from dimension #2 of the three-dimensional solution which
is undefined.
2Correlation coefficients for the 2-dimensional solution.

3Probability is greater than .05 and less than .10.

4Correlation coefficients for the four-dimensional solution.

"the situation is largely physical" (r=-.29, p=.07), "the

actor feels claustrophobic" (r=.28, p=.09), "the situation

is sexual" (r=-.27, p=.09), and "the actor feels intruded

upon" (r=.26, p=.10) are marginally significant. This con-

tinuum ranges from a situation which is sexual and physical

to nonsexual and nonphysical. Simultaneously the continuum

covers both the desire to escape (avoidance), and the feel-

ing of intrusion, and claustrophobia on one end to the de-

sire to not escape (approach), and the absence of feeling

intruded upon and claustrophobic. In the first dimension

the attributes common to both dimensions were associated

with crowdedness but in this dimension they are related to

sexuality. A cautious interpretation of the tenuous (mar-

ginal) relationship between approach-avoidance (escape) and

sexuality is "Physical Involvement."

The five-dimensional solution is similar to the four-

dimensional solution. This solution is not insightful since

no attribute is related to the dimension and it adds nothing

to the variance accounted for in a stepwise regression of

dimensional coordinates onto the privacy scale scores.

To what extent are the dimensions used by subjects in

discriminating privacy situations related to the perceived

degree of privacy in those situations? To answer this ques-

tion correlations between the dimension coordinates for each

privacy situation and the privacy scale values for each pri-

vacy situation were computed. Table 3 presents the results.

Table 3.

Privacy MDS Dimensions Correlated with
Privacy Scale Values

Dimensional Dimension
Solution 1 2* 3 4 5

2 -.3211 -.3012

3 -.3261 .1393 -.3102

4 -.3801 .0963 -.3701 .0583

5 -.4001 .1323 -.4431 -.1023 -.055

*Dimension 2 is defined by distinct attributes in solution 2
versus solutions 3, 4 and 5.
2 P<10


Dimension #1 (Social Density) is significantly correlated

with the degree of perceived privacy. The "Interpersonal

Involvement" dimension (#3) is also significantly related

to perceived privacy. Finally, the second dimension (NOR-

MATIVENESS) of the two-dimensional solution is correlated

(r=-.301, p=.06) at a marginally significant level. "Phys-

ical Involvement" shows no relationship to perceived privacy.

The stepwise regression of the two-dimensional solution

indicates that both dimensions together significantly pre-

dice (F(2,30)=4.7, p=.02) the privacy scale scores. Each

dimension accounts for one-half of the explained variance


The four-dimensional solution also significantly pre-

dicts the privacy scale scores. When dimension #1 is en-

tered first into the stepwise regression followed by the

inclusion of dimension #3 29 percent of the variance in

privacy scale scores is predicted (F(2,30)=7.68, p=.002).

Although the F-value (F2,30)=3.97, p=.009) for all four

dimensions entered into the stepwise regression is signif-

icant, less than 2 percent of explained variance is added.

Note that dimension #2 in the four-dimensional solution is

distinctly different from the "Normativeness" dimension of

the two-dimensional solution.

While the above analyses add support to the view that

privacy is a single, multidimensional construct, a more

atomistic analysis might provide further evidence

Table 4.

Regression of Privacy MDS Dimensions on
Privacy Scale Values


Dimension RSQ








.1601 .0173







.0033 .4061

*Dimension 2 is defined by distinct attributes in solution
2 versus solutions 3, 4 and 5.

3* >10

demonstrating the specific dimensional combinations and

properties of the four types of privacy. The multidimen-

sional data set was, therefore, subsequently partitioned

into subsets containing data for the privacy situations

previously coded into the four types. Three usable subsets

resulted. Too few stimuli existed for privacy situations

which were anonymous to be further analyzed. The results

for the remaining three privacy types are reported in al-

phabetical order.

Multidimensional Intimacy Results

The MDS stress values and r-square values for one

through five dimensional solutions appear in Table 5. The

insubstantial increase in the stress and r-square values be-

tween the two and three, and four and five dimensional solu-

tions suggests that either a two or four dimensional solu-

tion is optimal. Both solutions are, therefore, explored

via the defining attribute data. The five dimensional solu-

tion is also discussed due to its potentially interesting

relationship to the privacy scores.

The first dimension of the two dimensional solution is

defined by six attributes (see Table 6). In order of their

correlational magnitude they are (1) the situation is im-

personal (r=-.87, p=.002), (2) the situation is unpredict-

able (r=.84, p=.004); the actor is self-revealing to others


Table 5.

Intimacy MDS Results

Dimension Stress RSQ

1 .441 .573

2 .265 .702

3 .211 .763

4 .149 .858

5 .106 .917

(r=-.73, p=.025); (4) the situation is largely psycholog-

ical (r=.72, p=.03); (5) the actor feels comfortable (r=

-.68, p=.05); and (6) others understand how the actor feels

(r=-.67, p=.05). The most salient attribute continuum along

which subjects discriminate intimacy situations is the de-

gree to which the interpersonal interaction is personal or

impersonal. An interpretation of this personal-impersonal

dimension is suggested by three attributes. The more the

actor is self-revealing and the more others understand the

actor's feelings the more personal is the situation. Dis-

closing feelings and having others understand one's feel-

ings makes the situation more psychological rather than

mundane. Interaction is mundane to the extent that it fo-

cuses on task or external content. Apparently a prominent

affect accompanying this personal-impersonal dimension is

emotional comfort. Felt comfort decreases along this di-

mension as the situation becomes more personal. The extent

to whicn the situation (the other's behavior and one's own?)

is predictable increases as the interaction becomes more

personal. This interpretation of the attribute loadings

and combination suggests that an appropriate labeling of

this dimension is "Personal-Impersonal."

Only two attributes correlate significantly with the

second dimension of the two-dimensional solution. These

are "the situation is familiar" (r=.98, p=.0001) and "the

actors' feelings are caused by the presence of others"

(r=.81, p=.008). The dimension is clearly a "familiarity"

one but it is unclear what constitutes "the situation."

Two alternative explanations are that (1) the situation

is the physical and behavioral environment and (2) the sit-

uation refers to the other persons present. One hypothesis,

therefore, is that as the physical and behavioral environ-

ment becomes more familiar, the actor's emotions are a func-

tion of the others present, the more novel and unfamiliar

the situation the more the actor's emotions are associated

with the physical and behavioral environment and less with

the others present. An alternative hypothesis results when

the situation refers to the other people present. In this

case the more familiar are the other people present the more

the actor's emotions are associated with these people and

the less the emotions are related to either the actor's in-

ternal motivations or to the external, physical, non-human

environment. Regardless of which of these alternatives or

others is correct, this dimension is labeled as "familiar-


The first two dimensions of the four-dimensional solu-

tion are essentially the same in their defining attributes

as those of the two-dimensional solution. Seven attributes

are significantly correlated to the first dimension of the

four-dimensional solution. These are (1) the situation is

impersonal (r=-.88, p=.002), (2) the situation is unpre-

dictable (r=-.79, p=.01); (3) the actor is self-revealing

Table 6.

Intimacy MDS Dimensions Correlated with Defining Attributes
(Two, Four and Five Dimensional Solution)

Dimension Item # DIMENSION LABEL / Item 2-dimens 4-dimens 5-dimens
r r r

26 situation is impersonal .871 -.882 -.921
2 situation is unpredictable .841 -.792 -.703
32 actor is self-revealing to others -733 .772 .841
23 situation is public .663 .762 -.733
34 situation is largely psychological -.723 .713 .693
15 others understand how actor feels .673 .773
12 no one knows actor .72

18 situation is familiar .981 .862 841
29 actor's feelings caused by presence :812 .683 723
of others
23 situation is public .474 .384
16 actor is visible .57

5 actor's behavior is shameful .782 39 4
8 actor's behavior is against the rules .772 :024

10 actor is free to be self .812
5 actor's behavior is shameful .683
13 actor is careful about what they say .67
2P .005
3p<. 01
4 .05

to others (r=.77, p=.02); (4) the situation is public

(r=-.76, p=.02); (5) the situation is largely psychological

(r=.71, p=.03); (6) the actor feels comfortable (r=.70,

p=.04); and (7) others understand how the actor feels

(r=.67, p=.05). The only change in the first dimension

between the two and four-dimensional solutions is the ad-

dition of the attribute "the situation is public. This

attribute is marginally (p=.052) significant for the two-

dimensional solution. The more impersonal the situation,

the more public it is. The interpretation of this dimen-

sion remains the same, "Personal-Impersonal."

The second dimension of the four-dimensional solution

is identical to that for the two-dimensional solution ex-

cept that the magnitudes of the significant correlations is

reduced. The "familiarity" attribute remains the most prom-

inent, defining one.

No attribute is significantly correlated with the

third of the four dimensions. The attribute (the situation

is public r=.47, p=.2) with the highest correlation on

this dimension is significantly correlated with the first

dimension. No interpretation is possible of this dimension.

This finding potentially argues for a two-dimensional solu-

tion as most parsimonious to the data.

The fourth dimension is, however, defined by two at-

tributes which are significantly correlated with it. Both

attributes concern the actor's behavior; one evaluates it

while the other compares it against norms. These attri-

butes are "the actor's behavior is shameful" (r=.78, p=

.01); and "the actor's behavior is against the rules" (r=

.77, p=.01). One interpretation is that the actor's behav-

ior in intimate situations is discriminated along a con-

tinuum of being shameful apparently in association with

the degree to which the behavior conforms to the rules.

This dimension is therefore interpreted as "Normativeness."

Given that people can discriminate situations which

are intimate along a number of dimensions, which of these

dimensions or combination of dimensions is most related to

the perception of privacy? The correlations (see Table 7)

between the dimensional coordinates of each intimacy situ-

ation and their privacy scale score indicate that only the

first dimension is significantly correlated (r=.725, .747,

and .770 respectively, p=.01) with the privacy scores. The

direction of this relationship is such that as the situation

becomes more personal the perception of the amount of pri-

vacy present increases. Regression analyses (see Table 8)

further show that the Personal-Impersonal dimension pre-

dicts over 50 percent of the variance in privacy scale

scores. The addition of the second dimension results in

neither a significant multivariate F nor a substantial in-

crease in the variance accounted for. In spite of the

strength and clarity of the second dimension, it plays lit-

tle role in the perception of privacy. The same is true of

Table 7.

Intimacy MDS Dimensions Correlated with
Privacy Scale Value

Dimensional Dimension
Solution 1 2 3 4 5

2 -.747* .230

3 .725* .177 -.254

4 .770* .088 -.234 .464
5 .770* .085 -.197 .157 .350


Table 8.

Regression of Intimacy MDS Dimensions on
Privacy Scale Values

Dimensional Dimensions Stepwise
Solutions 1 2 3 4 5 MRSQ

2 .558* .053 .580

3 (l)** (2)

3 .526* .031 .064 .601

(1) (2) (3)
4 .593* .007 .054 .215 .865*

(1) (4) (3) (2)
5 .593* .007 .039 .025 .122 .799

(1) (2) (3) (4) (5)


**Designates the order in which the dimension was optimally
entered into the stepwise regression.

third dimension. Entering the fourth dimension into a

setpwise regression following the first dimension results,

however, in both a significant prediction (F (2,8)=13.08,

p=.007) and an increase of 22 percent of variance

accounted for over that for the first dimension alone. The

more normative and less shameful the actor's behavior the

less private is the intimacy situation perceived. These

two dimensions of Personal-Impersonal and Normativeness

combined account for eighty-one percent of the privacy scale

score variance.

Multidimensional Reserve Results

Table 9 shows the stress values and r-square values

for one through five-dimensional solutions. The addition

of dimension two substantially reduces the stress value and

increases the r-square. The stress reduction due to the in-

clusion of the third and fourth dimensions is more modest.

While two dimensions is the minimum number of dimensions,

the maximum number is uncertain. Although the five dimen-

sional solution results in a small decrease in the stress

value, this solution offers the best prediction of the pri-

vacy scale scores. For this reason, both the two- and five-

dimensional solutions are interpreted based on the defining

attribute data.

Table 9.

Reserve MDS Results

Dimension Stress RSQ

1 .454 .439

2 .292 .617

3 .210 .742

4 .151 .841

5 .111 .905

The first dimension of the two-dimensional solution

is an anomaly. Although this dimension has the least dis-

persion of actual scores from the computed dimension, no

attribute is significantly correlated with it. Two attri-

butes are marginally significant in their correlation with

this dimension. The attributes are "the actor is self-re-

vealing to other" (r=-.66, p=.054) and "the actor's feel-

ings are due to self" (r=.59, p=.092). A cautious inter-

pretation of this dimension is "Self-Disclosure." Subjects

discriminate situations which are reserved to the extent

that self-disclosure is present. Such an interpretation is

conceptually relevant to the construct "reserved." Appar-

ently the level of self-disclosure occurring and the extent

to which the actor's feelings are due to internal motiva-

tions are associated. The more inhibited the actor feels

about self-disclosing, the more the actor's feelings are due

to his/her own motivations and concerns. This interpreta-

tion, albeit based on questionable statistical significance,

is conceptually reasonable.

Five attributes define the second dimension of the two

dimensional solution. In descending order of correlational

magnitude, they are (1) too many people for space avail-

able (r=-.84, p=.005); (2) the actor is visible (r=-.81,

p=.008); (3) the actor feels claustrophobic (r=-.79, r=.01);

(4) others are too close to actor (r=-.71, p=.03); and (5)

too little privacy (r=-.70, p=.04). Four other attributes

are marginally significant; they are (1) actor is soli-

tary (r=.66, p=.051); (2) others interfere with or block

the actor (r=-.66, p=.053); (3) the actor feels anxious

(r=-.66, p=.053); and (4) crowded (r=.61, p=.08). Overall,

four attributes (#19, 40, 7, and 38) suggest that this di-

mension primarily represents a continuum of social density.

Another attribute (#33) has been hypothesized to be a cause

of feeling crowded. Visibility is closely associated with

this dimension and thus with crowding. Further related to

this dimension is inadequate privacy. Based on the prepon-

derance of attributes denoting and implying crowding, this

dimension is labeled 'Social Density."

For the five-dimensional solution, only the "actor is

self-revealing" attribute is marginally correlated (r=-.60,

p=.09) with the first dimension. The cautious interpreta-

tion of this dimension as "Self-Disclosure" is maintained.

The second dimension of the five-dimensional solution

is defined by nine attributes (#19, 28, 40, 22, 33, 29, 16,

38, and 7). Items #19, 28, 40, 22, 33, 16, 38, and 7(see

Table 10) are identical to those for the two dimensional

solution. The new attribute which is added to the defini-

tion of this dimension is "the actor's feelings are caused

by the presence of others" (r=.70, p=.04). This does not

change the original interpretation and labeling of this di-

mension as "Social Density." It does suggest that as crowd-

ing increases the actor's emotions are more related to the

presence of others.

TaDie 1U.
Reserve MDS Dimensions Correlated with Defining Attributes
(Two and Five Dimensional Solutions)

Dimension Item # DIMENSIONAL LABEL / Item 2-Dim 5-Dim
r r
32 actor is self-revealing .662 -.60 3
36 feelings due to self .59 46
19 too many people for space -.841 .83 1
16 actor is visible -.81 1 69 1

40 actor feels claustrophobic -.79 .782
7 others are too close -.71 1671
28 too little privacy -.70 .823
39 actor is solitary .66 2 -.50 1
33 others interfere with actor -.662 741
22 actor feels anxious -.662 .761
38 crowded .61 691
29 feelings caused by presence of others 701
10 actor is free to be self :642
24 actor feels intruded upon .62
5 actor's behavior is shameful -.771
23 situation is public -721
26 situation is impersonal -652
25 actor's feelings due to situation -. 90 1
34 situation is largely psychological 751
8 actor's behavior is against the rules .68

15 others understand how actor feels -.602

Dimension 3 has two attributes with significant corre-

lations. They are "the actor's behavior is shameful" (r=-.77,

p=.02), and "the situation is public" (r=-.72, p=.03). An-

other attribute, "the situation is impersonal" (r=-.65, p=

.059) is marginally significant. These attributes are con-

sistent with Schneider's (1977) conception of shame, thus

this dimension is labeled as "Shame."

The fourth dimension of the five-dimensional solution

is defined by three attributes: (1) the actor's feelings

are due to the situation (r=.90, p=.001); (2) the situation

is largely psychological (r=-.75), p=.02); and (3) the ac-

tor's behavior is against the rules (r=-.68, p=.05). This

dimension seems to best represent the level of situationally

motivated feelings. At one end of the continuum character-

istics of the situation largely influence the actor's emo-

tions while at the other extreme of the continuum the ac-

tor's emotions are mostly unrelated to characteristics of

the situation. Along this same dimension but in the oppo-

site direction is the degree to which the situation is psy-

chological. The more the actor's feelings are due to sit-

uational contingencies the less psychological is the situa-

tion. Similarly the more the actor's feelings are due to

the situation the less the actor's behavior is against the

rules. The conjunction of these attributes suggests the

actor is making an attribution of external causation for

his/her emotions. This dimension, therefore, is interpreted

as "Situational Attribution of Emotions."

Only one attribute potentially defines the fifth di-

mension. The attribute, "others understand how the actor

feels" (r=-.60, p=.09), is marginally significant when cor-

related with dimension #5. This dimension is cautiously

interpreted as "Empathy."

Of the five dimensions, only the last one is margin-

ally correlated (r=.60, p=.088) with the privacy scale

scores for the reserve situations. The perception of pri-

vacy for situations in which people felt reserved increases

as the emotional empathy of the others present decreases.

A stepwise regression of all five dimensions onto the

privacy scale scores again reveals marginally significant

findings. Dimension #5 accounts for 36 percent of

the variance in privacy scores (F(2,8)=3.93, p-.088). The

inclusion of dimension #3 next increases the r-square to

.56 (F(2,8)=4.01, p=.078). The addition of dimension #1 is

the last inclusion to a stepwise regression which results

in a substantial increase in r-square (r =.70) and is still

marginally significant (F(2,8)=3.81, p=.09). Although

based on marginally significant findings, this result cau-

tiously suggests that reserved situations are perceived as

more private as they are characterized by increasing emo-

tional empathy from others, by the increasingly personal na-

ture of the content of interaction, by the increasing ex-

tent to which the content is not public knowledge, and by

Table 11.

Reserve Correlations with Privacy Scale Values

Dimensional Dimensions
Solutions 1 2 3 4 5

2 .111 -.130

3 .171 -.234 .251

4 .191 -.125 .198 .565

5 .369 .158 .464 -.051 .600*

*p=. 088

Table 12.

Regression of Reserve MDS Dimensions on Privacy
Scale Values

Dimensional Dimensions RSQ Stepwise
Solutions 1 2 3 4 5 MRSQ

2 .012 .017 .029
(2)** (1)

3 .029 .055 .063 .152
(3) (2) (1)

4 .036 .016 .039 .320 .404
(3) (4) (2) (1)

5 .136 .025 .215 .003 .360* .719
(3) (4) (2) (5) (1)


**Designates the order in which the dimension was optimally
entered into the stepwise regression.

decreasing self-disclosure. While this is suggestive at

best, these findings are consistent with various theories

of privacy.

Multidimensional Solitude Results

The one through six dimensional stress and r-square

values are presented in Table 13. The addition of the

fourth dimension results in a small decrease in stress thus

the three dimensional solution is considered the most accu-

rate and parsimonious. This decision is partially sup-

ported by the defining attribute analysis. An examination

of the attributes highly correlating with each dimension

for the three, four, and five dimensional solutions reveals

that only those for the first dimension in each analysis

are similar. Given the unstable nature of these relation-

ships, it seems that three dimensions is the maximum reli-


A total of sixteen attributes define the first dimen-

sion of the three-dimensional solution. Four of the six

attributes with the highest correlations with this dimen-

sion concern the actor's affect. These attributes are

(1) the situation is comfortable (r=-.76, p-.0001); (2) the

actor feels distressed (r=.69, p=.0008); (3) the actor feels

helpless (r=.68, p=.001); and (4) the actor feels anxious

(r=.62, p=.003). One aspect of this dimension along which