16-Joinson-Chap16 11/20/06 11:06 AM Page 235
CHAPTER 16
Self-disclosure, privacy
and the Internet
Adam N. Joinson and Carina B. Paine
n this chapter, we examine the extant research
(Jourard and Lasakow 1958: 91). This shared
literature on self-disclosure and the Internet,
knowledge might exist between pairs of people,
Iin particular by focusing on disclosure in
within groups, or between an individual and an
computer-mediated communication (CMC)
organization. It has a variety of purposes, in part
and web-based forms  both to surveys and in
dependent on the context in which disclosure
e-commerce applications. We also consider the
occurs. For instance, within dyads, particularly
links between privacy and self-disclosure, and
romantic relationships, it serves to increase mutual
the unique challenges (and opportunities) that
understanding (Laurenceau et al. 1998), and
the Internet poses for the protection of privacy.
builds trust by making the discloser increasingly
Finally, we propose three critical issues that unite
vulnerable (emotionally or otherwise) to the other
the ways in which we can best understand the
person (Rubin 1975). Since self-disclosure is often
links between privacy, self-disclosure and new
reciprocated it frequently serves to strengthen the
technology: trust and vulnerability, costs and
ties that bind people in romantic or friendship-
benefits and control over personal information.
based relationships (Jourard 1971).
Central to the chapter is the notion that self-
Disclosure within groups can serve to enhance
disclosure is not simply the outcome of a commu-
the bonds of trust between group members, but
nication encounter: rather, it is both a product
it can also serve to legitimize group membership
and process of interaction, as well as a way for
and strengthen group identity. For instance,
regulating interaction dynamically. We propose
the admission of a negative identity (e.g.  I am
that by adopting a privacy approach to under-
an alcoholic ) within a shared identity group
standing disclosure online, it becomes possible
serves both to increase trust by revealing a
to consider not only media effects that encour-
stigmatized identity and act as a membership
age disclosure, but also the wider context and
card for a particular group (Galegher et al.
implications of such communicative behaviours.
1998). Personal growth may be an outcome of
honest self-disclosure (Jourard 1971). In a study
reported by Pennebaker et al. (1988), partici-
What is self-disclosure?
pants assigned to a trauma-writing condition
Self-disclosure is the telling of the previously (where they wrote about a traumatic and upset-
unknown so that it becomes shared knowledge, ting experience for four days) showed immune
the  process of making the self known to others system benefits, compared to a non-trauma
16-Joinson-Chap16 11/20/06 11:06 AM Page 236
236 � CHAPTER 16 Self-disclosure, privacy and the internet
writing group. Disclosure in this form has also In the case of content analysis, the issue of what
been associated with reduced visits to medical constitutes self-disclosure, and how it is scored,
centres and psychological benefits in the form is particularly important. One option would be
of improved affective states (Smyth 1998). For to count the number of instances within a con-
people using the Internet to talk about their versation in which a person discloses informa-
problems (or to publish weblogs), their activities tion about themselves. However, there are a
may well have unforeseen, positive, health and number of problems with this approach. First, it
psychological benefits. is not always clear what constitutes an act of self-
Finally, disclosure between an individual disclosure  for instance, to express an opinion
and an organization can serve authentication may well be classified in some contexts, but not
purposes  for instance, to establish identity, in others. Second, self-disclosure can only be prop-
allow authentication of a claim to identity and erly understood in terms of the ongoing interac-
to enable an organization to recognize you in tion. For instance, does one count answers to a
the future in order to personalize its offerings specific question   How old are you?  as self-
to you. Organizations might also ask for per- disclosure, or only spontaneous occurrences of
sonal information for marketing purposes  for disclosure (see Antaki et al. [2005] for a recent
instance, when registering to access a website or discussion of this issue). Moreover, given the
joining an online community. Of course, organ- dynamics of reciprocity, it may not even be
izations, in the form of researchers, might also possible to count occurrences of spontaneous
ask for personal information in the name of disclosure as independent of the conversational
academic research. dynamic. For these reasons, it is usual to treat
New technology, and in particular the Internet, discussions between people as a single unit of
might well change the demands upon people analysis (Kenny and Judd 1986).
to disclose personal information, as well as the Finally, not all self-disclosure is equal  disclosing
possible implications of such disclosure. For your season of birth is not the same as disclosing
instance, disclosing personal information to your age, which is not the same as disclosing your
another person online might not involve the sexual fantasies. One option is to use a three-layer
increased vulnerability that usually follows categorization scheme proposed by Altman and
self-disclosure of personal information offline Taylor (1973) to guide the content analysis of
(Ben-Ze2 ev 2003). Organizations might also depth. Altman and Taylor suggest that disclosure
demand more information in the name of authen- can be categorized into either peripheral, inter-
tication (although this need not always be per- mediate, and core layers. The peripheral layer is
sonal information). Furthermore, new technology concerned with biographic data (e.g. age), the
changes the scope of personal information that intermediate layer with attitudes, values and
can be disclosed or collected. For instance, the opinions and the core layer with personal beliefs,
development of ambient and ubiquitous devices, needs, fears, and values. Joinson (2001b) instead
such as smart mobile phones and RFID tags, used a 7-point Likert scale with which two scor-
makes it likely that information about location, ers allocated the degree to which an utterance
movements and social interactions are likely to  revealed vulnerability . However, Antaki et al.
be collected in the future in some form. How we (2005) argue that the act of disclosure needs
negotiate the disclosure of such information is a to take into account the interactional context
critical issue, equally as important as how sys- rather than simply being scored on a checklist.
tems are designed to minimize privacy viola- For instance, the phrase  I m the world s worst
tions while also providing adequate levels of cook could be disclosure, a plea for help or self-
functionality. deprecation. Without the context, they argue, it is
not possible to be certain.
Alternatively, lists of topics can be used to
Measuring self-disclosure
score intimacy  although again there are a num-
Within person-to-person and person to group ber of problems with their application in practise
interactions, self-disclosure has tended to be stud- to communication research (see Tidwell and
ied using either content analysis or self-report. Walther 2002, footnotes).
16-Joinson-Chap16 11/20/06 11:06 AM Page 237
Self-disclosure and the internet � 237
Self-report measures of disclosure have been personal questions. Indeed, Joinson et al. (in press)
used successfully, for instance to compare levels report that the provision of  I prefer not to answer
of disclosure in face-to-face (FtF) and online options in a salary question may improve data
relationships, or to link marital satisfaction with quality by reducing the number of non-responses
disclosure within the relationship. For instance, or default selections. In our own research (in
Parks and Floyd (1996) asked their participants preparation) we established that people are more
to report the level of self-disclosure in their likely to use an  I prefer not to say option when
Internet relationships using self-report (e.g. high faced with a sensitive rather than non-sensitive
scores on  I usually tell this person exactly how question, and that priming participants for online
I feel and low scores on  I would never tell this privacy (by asking them about their privacy con-
person anything intimate or personal about cerns and behaviours) significantly increases the
myself ). However, the same problems  a lack of use of  I prefer not say as an option to sensitive
context  arise for such self-report measures too. questions.
Measures of dispositional self-disclosure can Finally, self-disclosure can be measured using
also be used. For instance, within the International statistical techniques, for example the randomized
Personality Item Pool (IPIP) the RD3 subscale response technique (Musch et al. 2001). In the
of items  similar to the Temperament and randomized response technique, participants are
Character Inventory (TCI) has 10 items such as asked to answer a sensitive question either truth-
 Am open about myself to others (positive cod- fully or with a prespecified answer, depending on
ing) and,  Reveal little about myself  (negative the result of a random event such as a coin toss.
coding) to measure general self-disclosure. So, for instance, the question might be,  do you
However, it is not currently clear how such per- lie to your partner about anything important?
sonality type measures might interact with dif- Participants are asked to toss a coin, and if it is
ferent media, or indeed with people s behaviour heads, they tell the truth, if it is tails they say  yes
within a specific interaction. regardless of the truthful answer. Using statistical
Self-disclosure outside of person-to-person probabilities, a population estimate for a behav-
and group interactions can also be measured in a iour can be found, without knowing if any one
number of different ways. One system is to count individual told the truth or simply followed the
the number of words typed into text boxes in instructions for  tails .
response to a personal or sensitive question, and As noted earlier, self-disclosure in the age of
to rate those responses by their intimacy or ubiquitous computing poses novel challenges.
depth (e.g. Moon 2000; Joinson 2001b). Joinson For instance, it is likely that people will disclose
(2005) also describes the use of non-response as information without full awareness or control
a measure of self-disclosure in studies. There are (e.g. their location via a cell phone)  instead
two main ways in which non-response can be they may need to rely on privacy profiles or
operationalized in survey methodology and preferences to negotiate the disclosure on their
e-commerce. The first is non-response  either behalf. In these circumstances, discussion or
submitting a default selection, or where there is measurement of a single instance of disclosure
no default option, submitting no response. is meaningless without full consideration of the
A second is to add an option that allows partici- context in which disclosure occurred.
pants to select  I prefer not to answer (Buchanan
et al. 2002; Knapp and Kirk 2003). The use of
 I prefer not to answer as a response option to a
Self-disclosure and
sensitive question is methodologically similar
the Internet
to the provision of a  no opinion response in
attitudinal surveys. While it has been argued A rapidly increasing body of experimental and
that the provision of  no opinion choices may anecdotal evidence suggests that CMC and gen-
increase satisficing in attitude surveys (Holbrook eral Internet-based behaviour can be character-
et al. 2003), there is little reason to assume that ized as containing high levels of self-disclosure.
a similar process would operate in the use of For instance, Rheingold (1993) claims that new,
 I prefer not to answer responses to sensitive meaningful relationships can be formed in
16-Joinson-Chap16 11/20/06 11:06 AM Page 238
238 � CHAPTER 16 Self-disclosure, privacy and the internet
cyberspace because of, not despite, its limita- that heightened self-disclosure during CMC
tions. He further argues that  the medium will, may be due to people s motivation to reduce
by its nature . . . be a place where people often uncertainty. According to Uncertainty Reduction
end up revealing themselves far more intimately Theory (URT) (Berger and Calabrese 1975),
than they would be inclined to do without the people are motivated to reduce uncertainty in
intermediation of screens and pseudonyms . an interaction to increase predictability. In FtF
Similarly, Wallace (1999) argues that  The ten- interaction, uncertainty can be reduced through
dency to disclose more to a computer . . . is an both verbal and non-verbal communication and
important ingredient of what seems to be hap- cues. Tidwell and Walther hypothesize that dur-
pening on the Internet (1999: 151). Self-disclosure ing CMC, uncertainty reducing behaviours are
has been studied in a number of different text-based only, including increased levels of
settings using computers. For instance, Parks self-disclosure and question asking. To test this,
and Floyd (1996) studied the relationships Tidwell and Walther recruited 158 students to
formed by Internet users. They found that peo- discuss in opposite sex pairs with an unknown
ple report disclosing significantly more in their partner using a CMC system or FtF. The subse-
Internet relationships compared to their real life quent conversations were content-analysed for
relationships. Similarly, in their study of  coming disclosure using the breadth and depth indices
out on the Internet , McKenna and Bargh (1998) developed by Altman and Taylor (1973; see
argue that participation in online newsgroups above for a description).
gives people the benefit of  disclosing a long secret Tidwell and Walther found that those in the
part of one s self (1998: 682). Chesney (2005), in CMC condition displayed higher levels of both
a small-scale study of online diaries, reported question asking and self-disclosure compared
high levels of disclosure of sensitive information, the FtF condition. The questions asked by CMC
with half of his participants claiming to never discussants were also more probing and inti-
withhold information from their diaries. mate than those asked by those talking FtF,
In the series of studies reported by Joinson while both the questions and disclosure by
(2001a), the level of self-disclosure measured FtF interactants tended to be more peripheral
using content analysis of transcripts of FtF and than those in the CMC condition. Tidwell and
synchronous CMC discussions (Study one), and Walther conclude that the limitations of CMC
in conditions of visual anonymity and video encourage people to adapt their uncertainty-
links during CMC (Study two). In keeping with reducing behaviours  they skip the usual asking
the predicted effect, self-disclosure was signifi- of peripheral questions and minor disclosure,
cantly higher when participants discussed using and instead opt for more direct, intimate ques-
a CMC system as opposed to FtF. tioning and self-disclosure.
In the second study, incorporating a video Surveys and research administered via the
link while the participants discussed using the Internet, rather than using paper methodolo-
CMC program led to levels of self-disclosure gies, have also been associated with reductions
similar to the FtF levels, while the comparison in socially desirable responding (Joinson 1999;
condition (no video link) led to significantly Frick et al. 2001), higher levels of self-disclosure
higher levels of self-disclosure. (Weisband and Kiesler 1996) and an increased
These two studies together provide empirical willingness to answer sensitive questions (see
confirmation that visually anonymous CMC tends Tourangeau 2004).
to lead to higher levels of self-disclosure. The In a similar vein, survey methodology tech-
results of these studies also suggest that high levels niques that tend to reduce human involvement in
of self-disclosure can effectively be designed out question administration also increase responses
of an Internet interaction (e.g. through the use of to sensitive personal questions. For instance,
a video link or accountability cues (Joinson compared to other research methods, when data
2001a, Study 3), as well as encouraged. collection is conducted via computer-aided self-
Further empirical confirmation of increased interviews (where participants type their answers
self-disclosure during CMC comes from the work on to a laptop) people report more health-related
of Tidwell and Walther (2002). They proposed problems (Epstein et al. 2001), more HIV risk
16-Joinson-Chap16 11/20/06 11:06 AM Page 239
Models of self-disclosure online � 239
behaviours (Des Jarlais et al. 1999), more drug Internet: for commercial organizations, con-
use (Lessler et al. 2000), and men report less sex- sumers are often less than forthcoming, usually
ual partners, and women more (Tourangeau and because of a combination of privacy concerns,
Smith 1996). Medical patients tend to report lack of trust and concern about how personal
more symptoms and undesirable behaviours information will be used (Hoffman et al. 1999;
when interviewed by computer rather than FtF Metzger 2004). For instance, Olivero (2001)
(Greist et al. 1973). Clients at a STD clinic report studied the willingness to disclose information
more sexual partners, more previous visits and about the self to a commercial organization, and
more symptoms to a computer than to a doctor manipulated the level of trustworthiness of the
(Robinson and West 1992). Ferriter (1993) found organization, whether a financial reward was
that pre-clinical psychiatric interviews conducted offered for disclosure and the level of intrusive-
using CMC compared to FtF yielded more honest, ness of the questions. She found that the level
candid answers. Similarly, automated or com- of trust was associated with participants will-
puterized telephone interviews, compared to ingness to disclose to highly intrusive questions,
other forms of telephone interviewing, lead to but that an awareness of data mining/privacy
higher levels of reporting of sensitive informa- concerns moderated this effect of trust. Andrade
tion (see Lau et al. 2003; Tourangeau 2004). et al. (2002) conducted a similar study by examin-
Conversely, methods that increase the social ing three approaches to encourage self-disclosure
presence of the surveyor (e.g. by using photo- of personal information online  the complete-
graphs of the researcher) have been predicted to ness of a privacy policy, the reputation of a com-
lead to a reduced willingness to answer sensitive pany and the offer of a reward. They found that
questions (Tourangeau et al. 2003), although the the completeness of privacy policy and reputa-
findings of Tourangeau et al. were equivocal. tion of the company reduce the level of concern
However, Sproull et al. (1996) found that partic- over self-disclosure while the offer of a reward
ipants  present themselves in a more positive heightens concern.
light to the talking-face displays (1996: 116) However, there are a number of  counter sur-
than to text-only interfaces. Joinson et al. (in veys and empirical evidence suggesting that there
press) report that although personalizing the is a significant discrepancy between privacy
research experience leads to higher response principles and privacy practices. Very few indi-
rates to a self-administered survey, it also viduals actually take any action to protect their
reduces self-disclosure. personal information, even when doing so involves
Within the Human Computer Interaction limited costs (Berendt et al. 2005; Jenson et al.
(HCI) literature, the assumption seems to be 2005) i.e. there is a dichotomy between stated atti-
that people will avoid disclosing information to tudes and actual behaviours of people in terms of
commercial web services (Metzger 2004) due to their protection of personal information.
their privacy concerns (Jupiter Research 2002).
An online survey stated that the three biggest
consumer concerns in the area of online per-
Models of self-disclosure
sonal information security were: companies
online
trading personal data without permission, the
consequences of insecure transactions, and theft Explanations for high levels of self-disclosure in
of personal data (Harris Interactive 2002). For person-to-person CMC have tended to focus on
example, Hoffman et al. (1999) report that almost the psychological effects of anonymity:  This
95% of Internet users declined to provide per- anonymity allows the persecuted, the controver-
sonal information when requested to do so by a sial, and the simply embarrassed to seek infor-
website, and over 40% provided false demo- mation  and disseminate it  while maintaining
graphic information when requested. Quittner their privacy and reputations in both cyberspace
(1997) reports that 41% of survey respondents and the material world (Sobel 2000: 1522).
would rather exit a web page than reveal per- Theoretically, it has been argued that anonymity
sonal information. Clearly then, open self- in CMC works by replicating a  strangers on the
disclosure is not a universal experience on the train experience (Bargh et al. 2002), promoting
16-Joinson-Chap16 11/20/06 11:06 AM Page 240
240 � CHAPTER 16 Self-disclosure, privacy and the internet
private self-awareness and reducing accounta- is that it is the author to whom one is disclosing
bility concerns (Joinson 2001a), creating a need that is critical  if one trusts the recipient of the
for uncertainty reduction (Tidwell and Walther personal information, then one can act with
2002) or a combination of the media and the relative freedom in the pseudonymous world
process of interaction itself (Walther 1996). such disclosure purchases. Only by considering
Similarly, explanations for increased self- the wider context can such seemingly paradoxi-
disclosure to online surveys and web forms have cal impacts of new technology on personal
also tended to stress anonymity (Joinson 1999), disclosure be fully understood.
alongside the reduced social presence (and judge- This interpretation also strongly suggests
ment) of the researcher (Tourangeau 2004), that any explanation of self-disclosure online
reduced vulnerability (Moon 1998) and increased that relies solely on media effects (i.e. visual
privacy of the research environment (Tourangeau anonymity) is mistaken. Disclosure, while often
2004). Once privacy is reduced, or social presence  given away is also something that is carefully
increased, self-disclosure also tends to be reduced considered within the context of an ongoing
(Joinson et al. in press). interaction and wider context  regardless of
However, explanations for people s unwill- whether that interaction is interpersonal or
ingness to disclose personal information to human computer. We would suggest that a
e-commerce services invariably stress people s wider theoretical scope is needed  not only is it
privacy concerns (e.g. Hoffman et al. 1999), in important to consider the particular context of
particular, issues surrounding the level and type an interaction, but also how the person accessed
of information collected, and people s lack of that environment in the first place. For instance,
knowledge about how it may be used in the while the use of pseudonyms may enable expres-
future, or control over that use (Metzger 2004). sive freedom on a discussion board, we would
These differing approaches to understanding also ask how access was gained to the board,
disclosure and non-disclosure of personal infor- what registration process was in place, what
mation illustrate the paradox of self-disclosure records of postings are kept remotely and locally
on the Internet. On the one hand, the Internet and so on? Without this knowledge, one is forced
provides an environment in which people can to assume that people somehow dropped into an
express themselves with relative immunity online environment out of the sky, rather than as
via pseudonyms, but in order to access these a motivated act (see Joinson 2003).
services and sites they often need to disclose While concern about the privacy implications
high levels of personal information during the of new technology are nothing new (Home
registration process. Office 1972), the development and linking of
Within the privacy literature, this paradox is databases with biometrics, and the tension
relatively easy to solve  the provision of infor- between the need for identification, protection
mation about the self is treated quite separately of privacy and full participation in the e-society
from the use of privacy or pseudonymity to (Raab et al. 1996) makes an understanding of
express one s inner desires. However, it is rare the relations between privacy and the disclosure
for CMC self-disclosure research to explicitly and use of personal information critical. In the
consider privacy, in particular the multifactor next section of this chapter, we consider what
approaches to privacy discussed in the socio- privacy is, how the Internet and new technolo-
legal literature. gies threaten privacy, and the implications of
Within e-commerce, there are further para- privacy for understanding self-disclosure within
doxes which may be solved by looking at both an interaction.
the literature on interactional person-to-person
disclosure and the privacy literature concur-
What is privacy?
rently. For instance, there are occasions when
you need to disclose a lot of personal informa- There have been many attempts at definitions of
tion (e.g. purchasing online), but other factors privacy. In a legal context, privacy is largely syn-
(e.g. lack of social presence) make such privacy onymous with a  right to be let alone (Warren
concerns less pressing. The answer to this paradox and Brandeis 1890). However, others have argued
16-Joinson-Chap16 11/20/06 11:06 AM Page 241
What is privacy? � 241
that privacy is only the right to prevent the dis- different aspects of privacy, there are many
closure of personal information. Many researchers similarities between the theories. For example,
have referred to the difficulties involved in try- both theories are examples of  limited-access
ing to produce a definition (e.g. Burgoon et al. approaches to privacy in that they both empha-
1989) and despite various attempts to create a size controlling or regulating access to the
synthesis of existing literature (e.g. Parent 1983; self. Such commonalties between these theories
Schoeman 1984) a unified and simple account suggest that  their ideas provide a reasonable
of privacy has yet to emerge. Despite there being foundation for understanding the fundamentals
no unitary concept of privacy it is clear that of privacy as a psychological concept (Margulis
both individuals, and society, attach a level of 2003: 424). A large amount of work has supported
importance to privacy. For example, Ingham and extended both Westin s (e.g. Marshall 1974;
states that  man, we are repeatedly told is a social Pederson 1979) and Altman s work (e.g. Kupritz
animal, and yet he constantly seeks to achieve a 2000) and as such both of their theories have
state of privacy (1978: 45). stimulated much of the research and theory
Within psychological literature both Westin s development of privacy.
and Altman s theories figure prominently in the Since these earlier definitions, the highly
major reviews of privacy in the 1970s. Westin complex nature of privacy has resulted in an
(1967: 7) provides a link between secrecy and alternative way of defining it  through its vari-
privacy and defines privacy as  the claim of indi- ous dimensions. Burgoon et al. (1989) distin-
viduals, groups, or institutions to determine for guish four dimensions of privacy and define it
themselves when, how and to what extent infor- using these dimensions as  the ability to control
mation about them is communicated to others . and limit physical, interactional, psychological
At the psychological level, Westin states that pri- and informational access to the self or one s
vacy provides opportunities for self-assessment group (Burgoon et al. 1989: 132). Each of the
and experimentation and therefore the develop- dimensions they distinguish is briefly described
ment of individuality. Specifically, Westin (1967) below with some examples.
proposes four main functions of privacy:
1. The physical dimension Physical privacy is the
1. personal autonomy applies to the need for the
degree to which a person is physically accessi-
development of individuality and the avoid-
ble to others. This dimension is grounded
ance of manipulation by others;
within the human biological need for personal
2. emotional release refers to the need for oppor- space. Examples of violations to physical pri-
tunities to relax and escape from the tensions vacy include: surveillance, entry into personal
of everyday life in order to support healthy space and physical contact.
functioning;
2. The interactional dimension Interactional (or
3. self-evaluation is the application of individu- social/communicational) privacy is an indi-
ality onto events and the integration of expe- vidual s ability and effort to control social con-
rience into meaningful patterns, and
tacts (Altman 1975). Burgoon et al. (1989)
summarizes the elements of this dimension as
4. limited and protected communication refers to
control of the participants of, the frequency of,
both the sharing of personal information with
the length of and the content of an interaction.
trusted others and the setting of interpersonal
Non-verbal examples of violations to social
boundaries.
privacy include close conversational distance
Altman (1975) incorporates both social and
and public displays of affection. Verbal exam-
environmental psychology in understanding the
ples include violations of conversational
nature of privacy. He defines privacy as  the
norms (e.g. commenting on mood or appear-
selective control of access to the self (p. 24) and
ance) and initiating unwanted conversation.
believes privacy is achieved through the regula-
tion of social interaction, which can in turn pro- 3. The psychological dimension Psychological
vide us with feedback on our ability to deal with privacy concerns the ability of human beings
the world, and ultimately affect our definition of to control cognitive and affective inputs
self. Although Westin and Altman emphasize and outputs, to form values, and the right to
16-Joinson-Chap16 11/20/06 11:06 AM Page 242
242 � CHAPTER 16 Self-disclosure, privacy and the internet
determine with whom and under what cir- the ability to build interpersonal relation-
cumstances thoughts will be shared or intimate ships improves, while external social control
information revealed. As such, psychological over lifestyle choices and so on are restricted
privacy can either develop or limit human (Schoeman 1992).
growth. Examples of violations to psychological
Using these multidimensional approaches to
privacy include psychological assaults through
define privacy results in some overlap of the
name-calling and persuasion.
features between each dimension. For example,
4. The informational dimension Informational within Burgoon et al. s dimensions some features
privacy relates to an individual s right to of informational privacy overlap with psycho-
determine how, when, and to what extent logical privacy, and some features of social
information about the self will be released to privacy overlap with physical privacy. Within
another person (Westin 1967) or to an organ- DeCew s dimensions there is some overlap
ization. According to Burgoon et al. (1989), between accessibility and informational privacy,
this dimension is closely related to psycholog- and expressive privacy is conceptually linked
ical privacy: however, the control differs from with both of these dimensions. In addition, there
the individual self-disclosure associated with is also some overlap between Burgoon et al. s and
psychological privacy because it is partly gov- DeCew s dimensions. For example, the informa-
erned by law/custom and as it often extends tional dimension appears in both definitions and
beyond personal control. Examples of viola- Burgoon et al. s physical and social dimensions
tions to informational privacy include going appear to map onto DeCew s accessibility and
through another person s mail and sharing expressive dimensions respectively.
personal information with others. Of direct relevance to this chapter are the
dimensions of informational and expressive pri-
DeCew (1997) also reflects the multidimen-
vacy. Central to these dimensions are the desire
sional nature of privacy in her definition: how-
to keep personal information out of the hands
ever, she distinguishes only three dimensions:
of others, or in other words privacy concern
1. The informational dimension Informational
(Westin 1967), and the ability to connect with
privacy covers personal information such as
others without interference. In a systematic dis-
finances, medical details and so on that an
cussion of the different notions of privacy,
individual can decide who has access to and
Introna and Pouloudi (1999) developed a
for what purposes. If disclosed, this informa-
framework of principles that explored the
tion should be protected by any recipients of
interrelations of interests and values for various
it. By protecting informational privacy indi-
stakeholders where privacy concerns have risen.
viduals avoid invasions (or potential inva-
In this context, concern for privacy is a subjec-
sions) to their privacy.
tive measure  one that varies from individual to
2. The accessibility dimension Accessibility privacy
individual based on that person s own percep-
refers to physical or sensory access to a per-
tions and values. In other words, different
son. It  allows individuals to control decisions
people have different levels of concern about
about who has physical access to their persons
their own privacy.
through sense perception, observation, or bod-
One scheme for categorizing the different levels
ily contact (DeCew 1997: 76 7).
of privacy concerns is the Westin privacy segmen-
3. The expressive dimension Expressive privacy
tation (Harris and Associates Inc. and Westin
 protects a realm for expressing one s self-
1998). The Harris Poll is a privacy survey con-
identity or personhood through speech or
ducted by telephone across the United States
activity. It protects the ability to decide to
among approximately 1,000 people. This survey
continue or to modify ones behaviour
has been conducted regularly since 1995 and
when the activity in question helps define one- divides respondents into one of three categories
self as a person, shielded from interference,
depending on their answers to three statements.
pressure and coercion from government or The three categories of respondents are: Privacy
from other individuals (DeCew 1997: 77). As Fundamentalists who view privacy as an especially
such, internal control over self-expression and high value which they feel very strongly about.
16-Joinson-Chap16 11/20/06 11:06 AM Page 243
Privacy and the internet � 243
Currently about a quarter (35%) of all adults are However, this introduction of this technology
privacy fundamentalists (Computerworld 2005); raised concerns about privacy both among trav-
Privacy Pragmatists also have strong feelings ellers and aviation authorities (The Sunday
about privacy. They weigh the value to them and
Times 2004).
society of providing their personal information.
The concept of privacy has also been applied
Currently around approximately 55% of all
to the Internet (e.g. Cranor 1999). The increased
adults are privacy pragmatists (Computerworld
use of computers and of the Internet now fills
2005); Privacy Unconcerned who have no real
many parts of people s lives including online
concerns about privacy. Approximately 10% of all
shopping, the sharing of documents and vari-
adults are privacy unconcerned (Computerworld
ous forms of online communication. It is this
2005).
increased use of the Internet which raises
Although levels of concern may differ
concerns about privacy, in particular, those
between people, a failure to achieve any level
described above under informational privacy.
privacy will result in  costs . For example, by not
There are concerns that the Internet seems to
obtaining privacy a person will not benefit from
erode privacy (Rust et al. 2002) and that
the opportunities that the functions of privacy
offline privacy concerns are magnified online
provide  which could results in stress or
(Privacy Knowledge Base 2005). Indeed, the
negative feedback about the self. There are also
subject of online privacy has been appearing
costs of losing privacy either through privacy
in newspaper articles regularly over the last
invasion  when conditions for privacy are not
few years (e.g. The Times 2001; The Guardian
achieved, for example being overheard  or pri-
2004).
vacy violation when recipients of personal
Personal information is fast becoming one of
information, intentionally provided by the dis-
the most important ethical issues of our infor-
closer or gained through a privacy invasion, pass
mation age (Milberg et al. 1995): personal infor-
it on to others  for example, gossip). In the
mation has become a basic commodity and
early privacy research described, invasions and
users online actions are no longer simply
violations were not emphasized. Ingham (1978)
actions but rather data that can be owned and
states that  In everyday social life most individu-
used by others. Advances in technology and the
als are only rarely confronted with an invasion
increased use of the Internet have changed the
of their privacy, although the number of poten-
ways in which information is gathered and
tial threats is very large (1978: 40). However,
used. A wide variety of information data is now
more recently, technology has fuelled debate
collected with increasing frequency and in dif-
and controversy about potential invasions and
ferent contexts, making individuals become
violations to privacy (Dinev and Hart 2004), as
ever more transparent. The costs of obtaining
will be described below.
and analysing this are also decreasing with
the advances in technology. However, the value
of the users information which is collected is
Privacy and the Internet
increasing.
Since the concept of privacy has been applied to
At no time have privacy issues taken on greater
technology (e.g. Agre and Rotenberg 1997;
significance than in recent years, as technological
Austin 2003) there have been numerous cases
developments have led to the emergence of an
reported of the clash between privacy and
 information society capable of gathering, storing
new technology  how these technologies allow
and disseminating increasing amounts of data
intrusions into private, enclosed spaces, eroding
about individuals.
the distinction between public and private space
and therefore compromising the very idea of
(Schatz Byford 1996: 1)
private space. For example, at the end of last
year, a body scanning machine was introduced There are a number of specific threats to
in an airport in the UK. This x-ray machine pro- online privacy. For example, the impact of
duces  naked images of passengers enabling any  ubiquitous computing (Weiser 1988) means
hidden weapons or explosives to be discovered. that we leave data footprints in many areas of
16-Joinson-Chap16 11/20/06 11:06 AM Page 244
244 � CHAPTER 16 Self-disclosure, privacy and the internet
our lives that were previously considered integrated and multiplied electronically fur-
 offline . The extremely rapid development of ther increases this accessibility;
computing power, in terms of greater processing
Ć% Assembly  there are many effective tools for
speed, increased storage capacity, wider com-
searching for and assembling and reorgan-
munication connectivity and lower machine
izing information from many quite separate
size all impact on privacy (Sparck-Jones 2003).
sources;
These rapid advances mean that information
Ć% Remoteness  information collected is usually
can be efficiently and cheaply collected, stored
both physically and logically away from the
and exchanged  even data which may be
users to whom it refers. However, this infor-
deemed sensitive by the individuals concerned.
mation can be accessed and used by people
Information that is drawn from the physical
who the user does not know.
world is harboured in electronic databases,
which give these records permanence, malleabil- Each of the above features affects privacy and
ity and transportability that has become the their effect in combination is even greater.
trademark of technology. As such, massive data- Although massive data collection and storage is
bases and Internet records of information about possible in many environments, the online
individual financial and credit history, medical privacy problem is further exacerbated by the
records, purchases and so on exist. very structure of the Internet and its additional
Sparck-Jones (2003) labels a number of spe- feature of connectivity. The Internet allows for
cific properties of the information collected interactive two-way communication and is
which have consequences for privacy: woven into people s lives in a more intimate way
than some other media as it connects people
Ć% Permanence  once recorded, information rarely
with places and people with people. Accordingly
disappears. As such, fine-grained, searchable,
it poses unique information privacy threats that
persistent data exists on individuals and there
differ from issues previously addressed by
are sophisticated, cheap, data-mining devices
research (e.g. Smith et al. 1996) therefore, mak-
can also be used to analyse this information;
ing information collection, sharing and so on
Ć%  the ease with which information is
Volume
even easier.
now recorded using technology results in
There are also benefits to the technological
huge data sets. Furthermore, storage is cheap,
advances described, such as personalized services,
therefore large volumes of information sets
convenience and efficiency. In this way, the
can exist indefinitely;
collection of personal information can be con-
Ć% Invisibility  all information collected seems to
sidered a  double-edged sword (Malhotra et al.
exist within an opaque system and so any
2004). Users can trade off providing valuable
information collected may not be  visible to
information about themselves to take advantage
whom it relates. Even if information collected
of benefits  for example, providing personal
is available to a person they may not be able to
details and credit card information in order to
interpret it due to the use of incomprehensible
have the convenience of completing an online
coding;
transaction. Jupiter Research (2002) have found
Ć% Neutrality  the ease with which information
evidence that even privacy concerned individu-
can be collected means that any qualifying
als are willing to trade privacy for convenience or
information may be lost. So information may
to bargain the release of very personal informa-
be absorbed regardless of its metadata. i.e.
tion in exchange of relatively small rewards.
there are no distinctions between intimate,
However, consumer concern over disclosing per-
sensitive information and non-sensitive
sonal information is growing as they realize that
information;
data about their internet behaviours is being col-
Ć% Accessibility  there are a number of tools for lected without their knowledge and agreement.
accessing information meaning that any These privacy concerns can ultimately reduce
information collected can possibly be read by the personalization benefits that companies can
any number of people. The ease with which deliver to consumers. The question is whether
information can be copied, transferred, the benefits of the advances in technology and
16-Joinson-Chap16 11/20/06 11:06 AM Page 245
Cost and benefits � 245
the use of the Internet are diminished by endan- disclose personal information to a web service
gering privacy. that they do not trust (Hoffman et al. 1999).
However, many attempts to establish trust
between people and within groups rely on
Linking models of privacy
methods that increase the media richness of the
and CMC
interaction  for instance, by introducing video,
audio or photographs (see Olsen et al. 2002;
According to Berscheid (1977), privacy is the  hid-
Chapter 5 this volume). Quite apart from the
den variable in many social psychological studies.
substantial problems with media richness
In the years since her article was published, there
approaches to understanding online behaviour
has been relatively few attempts to expose this
(see Walther 1996), introducing cues that are
hidden variable to scrutiny in the psychological
supposed to improve trust may well serve to
literature. Privacy is particularly important for
reduce privacy in an interpersonal context.
understanding self-disclosure, since the relation-
However, in some instances trust will be criti-
ship between privacy and self-disclosure is some-
cal. For instance, if you register to a discussion
what paradoxical. Privacy is a prerequisite for
board, dating site or other web-based service,
disclosure, and yet, the process of disclosure serves
you will commonly be required to disclose to
to reduce privacy. The Internet may, in some
the owner of the site your real name, age, loca-
instances, serve to solve this paradox  disclosure
tion/ZIP or postal code, and email address. It is
and intimacy can be achieved without concur-
not uncommon to also be asked questions about
rent increases in vulnerability or losses of pri-
salary, occupation, marital status and other
vacy (see Ben-Ze2 ev 2003). But this introduces a
marketing-related queries. In the cases of dis-
further paradox  the Internet, and new media in
cussion boards and dating sites, this disclosure
general, have tended to erode privacy through,
of personal information purchases access to a
amongst others, the processes we outline above.
pseudonymous interactive environment in
Often the impression of privacy is a mirage  high
which participants can seek help, be intimate or
levels of personal information are held by a num-
just play, with little concern for the repercus-
ber of gatekeepers  whether it is through the
sions in their offline lives. In this situation,
process of registration, caches and logs kept on
expressive privacy has been obtained through
various servers or even locally based records.
the loss of informational privacy to a third
It therefore becomes critical to understand the role
party. Critically, we would argue that it is this
of these gatekeepers to understand fully disclosure
separation between the location of the expres-
of personal information online. We propose that
sive environment, and the third party, that is
as well as looking at the micro-level impacts of the
important. Obviously too, one would also
media environment on disclosure, one also needs
expect that for this bargain to work, the third
to look at the macro-level  the wider context in
party must be trustworthy.
which the micro-level behaviour is enacted.
For trust to be established, it is not always
necessary for privacy to be reduced. For instance,
Trust and disclosure
reputation systems (as used on eBay, the auction
Trust is a critical issue in both FtF and online site) allow trust to be established through the
disclosure of personal information. By disclos- use of peer-ratings of pseudonyms (Utz 2004).
ing information, we are making ourselves vul- However trust is established, it is clearly criti-
nerable  one reason it is often easier to disclose cal to understanding online behaviour, and is
to strangers than to close friends and family likely to become more important as we leave our
(Rubin 1975). This applies equally to disclosure personal data at the door of pseudonymous
to web-based forms  for instance, Moon (1998) environments.
found that people are more willing to disclose
personal information to geographically distant
Cost and benefits
servers  presumably because the vulnerability
of doing so is reduced. In e-commerce, the issue In the example above, access to an environment
of trust is also critical  people will generally not in which expressive privacy is enabled has been
16-Joinson-Chap16 11/20/06 11:06 AM Page 246
246 � CHAPTER 16 Self-disclosure, privacy and the internet
effectively purchased with personal informa- and to have them delivered at their home address
tion. This commodification of personal infor- in a plain envelope. In this second case, the level
mation is nothing new  witness the growth in of information privacy is low  they will need to
customer cashback or  loyalty cards provided by disclose their name, address and credit card
grocery shops  but what is interesting is that details, but expressive or social privacy is high.
one form of privacy is lost to gain another form. The method chosen will illustrate the relative
Andrade et al. (2002) adopt a social exchange costs and benefits our fictional teenager attaches
framework to study consumers willingness to to information and expressive/social privacy.
disclose personal information, although their A critical issue in applying such an economic
results suggest that considering people s decision- model to understanding privacy and disclosure
making within this framework alone does not is the value placed upon personal information
explain the results of the study. For instance, by the individual, and their interpretation of the
while manipulations that seemed to reduce the likely costs of disclosure. As such, people s
cost of personal disclosure (e.g. privacy policy) privacy concerns and the level of trust they have
did indeed have the desired effect, the offer of a in the recipient of the disclosure will determine
reward worked to reduce disclosure. Presumably the outcome of any cost benefit analysis.
this may be because offering financial rewards
opened questions of trust.
There are many other occasions when deci-
Control
sions about whether or not to disclose personal
A further context issue that we believe is impor-
information can be interpreted from a social
tant to understanding self-disclosure online is
exchange approach. For instance, in many cases
control  that is, control over what information
a loss of privacy provides benefits in terms of
is collected, and how and with whom informa-
convenience rather than financial gain. Within
tion is shared.
person-to-person interaction, self-disclosure can
Information is often collected online with or
also be understood in terms of costs and bene-
without the user s knowledge or consent. From a
fits. As Antaki et al. (2005) note, disclosure needs
technical standpoint, some types of information
to be  brought off  it does not occur without
are easier to obtain than others. Information can
repercussions for both interactants. By disclosing
be gathered unobtrusively, which requires little
personal information, the cost to a person is
cooperation on the part of the person supplying
increased vulnerability and a loss of privacy.
the information. For example, information may
However, in many cases, the benefits  a building
be collected by means of cookies and other
of trust, rapport, and reciprocation  will out-
software designed to track users movements
weigh the costs. However, this is not to say that
over the Internet. Other types of information
disclosure is not without risks. For instance, a
are less accessible, forcing companies to rely on
teenager agonizing about whether to confess to a
more intrusive means to obtain important data.
romantic crush is likely to be acutely aware that
This typically involves asking people to engage
disclosure to the object of their desire is a poten-
in some type of self-disclosure.
tially risky business that will lead to either a
Individual control over personal information
joyful reciprocation of feelings, or rejection.
is more difficult than ever before. Even when
In terms of e-commerce, there are also clear
personal information is voluntarily provided,
cost benefit issues regarding privacy and disclo-
privacy may still be compromised due to the
sure. For instance, imagine the same teenager
inability of an individual to control the use of
has successfully arranged their date, and they
the information. For example, privacy may be
now wish to purchase prophylactics. They have
comprised on two dimensions (Culnan and
two options: the first, to pay in their local town
Armstrong 1999):
with cash, is reasonably high in privacy  there is
no data trail, and unless the server behind the 1. Environmental control  if personal informa-
counter knows them, they have high informa- tion is accessed by unauthorized means (e.g.
tion privacy. The alternative is to use a credit through a security breach or an absence of
card to purchase the desired products online, appropriate internal controls);
16-Joinson-Chap16 11/20/06 11:06 AM Page 247
References � 247
2. Control over secondary use of information  if
Conclusions
information provided for one purpose is used
Self-disclosure is one of the few widely replicated
for unrelated purposes, without the individ-
and noted media effects of online interaction.
ual s knowledge or consent. (e.g. through the
However, despite the evidence that self-disclosure
duplication and sharing of computerized
occurs in a number of different contexts online,
information).
including CMC, weblogs and submission of
The secondary use of information and the
web forms, most approaches to understanding
fact that information may be logged and
the phenomenon confine themselves to consid-
preserved for future access mean that threats to
ering the impact of a single factor  anonymity.
privacy on the Internet can be immediate as well
We argue that by focusing solely on this micro-
as future threats.
level media effect, the wider context in which
Most people do not know what information is
disclosure is given, or required, is ignored  and
stored about them or who has access to it.
that ignoring this context limits how we can
However, there is now a growing awareness, as well
conceptualize online behaviour. By considering
as resentment, of the routine practice of collecting
the wider context, and in particular its implica-
and analysing personal information (Nissenbaum
tions for privacy, it is possible to develop a more
1998). This is partly due to reports in newspapers
nuanced picture of online behaviour across
and on online news sites. For example the  extent
situations.
of UK snooping revealed story reported that  offi-
cials in the UK are routinely demanding huge
quantities of information about what people do
Acknowledgements
online and who they call, say privacy experts (BBC
The writing of this chapter was supported by a
News 2003a). Also the  Top UK sites  fail privacy
grant from the Economic and Social Research
test  story reported  98% do not give enough
Council (RES-341 25 0011).
information about the text files which track user
movements, or provide a single-click opt-out
option (BBC News 2003b). A February 2002
References
Harris Interactive Survey (Harris Interactive 2002)
stated that the three biggest consumer concerns in Agre, P. E. and Rotenberg, M. (eds) (1997). Technology
and privacy: The new landscape. Cambridge, MA:
the area of online personal information security
MIT Press.
were: companies trading personal data without
Altman, I. (1975). The environment and social behavior.
permission, the consequences of insecure transac-
Monterey, CA: Brooks/Cole.
tions, and theft of personal data.
Altman, I. and Taylor, D. (1973). Social penetration:
Within the context of person-to-person The development of interpersonal relationships. New York:
Holt, Rinehart and Winston.
interaction, clearly control is also a critical issue.
Andrade, E. B., Kaltcheva, V. and Weitz, B. (2002).
Walther (1996) argues that hyperpersonal social
Self-disclosure on the web: the impact of privacy policy,
interaction online occurs, at least in part, because
reward, and company reputation. Advances in Consumer
of the increased control afforded by synchro-
Research 29, 350 353.
nous, visually anonymous CMC. For instance, we
Antaki, C., Barnes, R. and Leudar, I. (2005). Self-disclosure
can control what information we choose to as a situated interactional practice. British Journal of
Social Psychology 44, 181 200.
disclose, in what manner, and how we disclose it.
Austin, L. (2003). Privacy and the question of technology.
If privacy and self-disclosure are viewed as
Law and Philosophy 22, 119 166.
dynamic processes, then the removal of control
Bargh, J. A., McKenna, K. Y. A. and Fitzsimons, G. M.
affects the ability of people to effectively regulate
(2002). Can you see the real me? Activation and
their social interactions. We may wish to control
expression of the true self on the Internet. Journal of
Social Issues 58, 33 48.
the flow of personal information for a number of
BBC News (2003a). Extent of UK snooping revealed. 16
reasons  for instance, to maintain a desired level
May. Available at http://news.bbc.co.uk/1/hi/technology/
of intimacy and privacy, or to maintain social
3030851.stm. Accessed 20 June 2005.
harmony by not disclosing specific information 
BBC News (2003b). Top UK sites  fail privacy test .
but without control over what is disclosed and to
11 December. Available at: http://news.bbc.co.uk/1/hi/
whom, this is not possible. technology/3307705.stm. Accessed 20 June 2005.
16-Joinson-Chap16 11/20/06 11:06 AM Page 248
248 � CHAPTER 16 Self-disclosure, privacy and the internet
Ben-Ze2 ev, A. (2003). Privacy, emotional closeness, and Greist, J. H., Klein, M. H. and VanCura, L. J. (1973).
openness in Cyberspace. Computers in Human A computer interview by psychiatric patient target
Behavior 19, 451 467. symptoms. Archives of General Psychiatry 29, 247 253.
Berendt, B., Gunther, O. and Spiekerman, S. (2005). Privacy Guardian, The (2004). The privacy debate: this time it s
in E-commerce: stated behaviour versus actual personal. 26 April.
behaviour. Communications of the ACM 48, 101 106. Harris and Associates Inc. and Westin, A. (1998).
Berger, C. R. and Calabrese, R. J. (1975). Some explorations E-commerce and privacy: what net users want. Privacy
in initial interaction and beyond: toward a and American Business and Pricewaterhouse Coopers
developmental theory of interpersonal communication. LLP. Available at
Human Communication Theory 1, 99 112. http://www.pandan.org/ecommercesurvey.html.
Buchanan, T., Joinson, A. N. and Ali, T. (2002). Accessed 20 June 2005.
Development of a behavioural measure of self- Harris Interactive (2002). First major post-9/11 privacy
disclosure for use in online research. Paper presented at survey finds consumers demanding companies do more
German Online Research 2002, Hohenheim, Germany. to protect privacy; public wants company privacy
Burgoon, J. K., Parrott, R., LePoire, B. A., Kelley, D. L., policies to be independently verified. Available at
Walther, J. B. and Perry, D. (1989). Maintaining and http://www.harrisinteractive.com/news/allnewsbydate.asp?
restoring privacy through communication in different NewsID=429. Accessed 20 June 2005.
types of relationship. Journal of Social and Personal Hoffman, D. L., Novak, T. P. and Peralta, M. (1999).
Relationships 6, 131 158. Building consumer trust online. Communications of the
Chesney, T. (2005). Online self disclosure in diaries and its ACM 42, 80 85.
implications for knowledge managers. In UK Academy Holbrook, A. L., Green, M. C. and Krosnick, J. A. (2003).
for Information Systems Conference proceedings, 22 24 Telephone vs. face-to-face interviewing of national
March, Northumbria University, UK. probability samples with long questionnaires: comparisons
Computerworld (2005). Available at of respondent satisficing and social desirability response
http://www.computerworld.com/printthis/2005/ bias. Public Opinion Quarterly 67, 79 125.
0,4814,101766,00.html. Accessed 20 June 2005. Home Office (1972). Report of the Committee on Privacy.
Cranor, L. F. (1999). Internet privacy. Communications of Rt. Hon. Kenneth Younger, Chairman. London: HMSO.
the ACM 42, 29 31. Ingham, R. (1978). Privacy and psychology. In Young, J. B.
Culnan, M. J. and Armstrong, P. K. (1999). Information (ed.) Privacy (pp. 35 59). Chichester: Wiley.
privacy concerns, procedural fairness, and impersonal Introna, L. D. and Pouloudi, A. (1999). Privacy in the
trust: an empirical investigation. Organization Science information age: stakeholders, interests and values.
10, 104 115. Journal of Business Ethics 22, 27 38.
DeCew, J. (1997). In pursuit of privacy: Law, ethics, and the Jenson, C., Potts, C. and Jenson, C. (2005). Privacy
rise of technology. Ithaca, NY: Cornell University Press. practices of internet users: self-reports versus observed
Des Jarlais, D. C., Paone, D., Milliken, J., Turner, C. F., behaviour. International Journal of Human Computer
Miller, H., Gribble, J., Shi, Q., Hagan, H. and Friedman, Studies. Special issue on HCI Research in Privacy and
S. (1999). Audio-computer interviewing to measure risk Security.
behaviour for HIV among injecting drug users: a quasi- Joinson, A. N. (1999). Anonymity, disinhibition and social
randomised trial. The Lancet 353(9165), 1657 1661. desirability on the Internet. Behaviour Research Methods,
Dinev, T. and Hart, P. (2004). Internet privacy concerns Instruments and Computers 31, 433 438.
and their antecedents  measurement validity and a Joinson, A. N. (2001a). Self-disclosure in computer-
regression model. Behaviour and Information Technology mediated communication: the role of self-awareness and
23, 413 423. visual anonymity. European Journal of Social Psychology
Epstein, J. F., Barker, P. R. and Kroutil, L. A. (2001). Mode 31, 177 192.
effects in self-reported mental health data. Public Joinson, A. N. (2001b). Knowing me, knowing you:
Opinion Quarterly 65, 529 550. reciprocal self-disclosure on the internet.
Ferriter, M. (1993). Computer-aided interviewing and the Cyberpsychology and Behavior 4, 587 591.
psychiatric social history. Social Work and Social Sciences Joinson, A. N. (2003). Understanding the psychology of
Review 4, 255 263. internet behaviour: Virtual worlds, real lives. Basingstoke
Frick, A., B�chtiger, M. T. and Reips, U.-D. (2001). and New York: Palgrave Macmillan.
Financial incentives, personal information and drop-out Joinson, A. N. (2004). Self-esteem, interpersonal risk and
in online studies. In Reips, U.-D. and Bosnjak, M. (eds), preference for e-mail to face-to-face communication.
Dimensions of internet science (pp. 209 219). Lengerich: CyberPsychology and Behaviour 7, 472 478.
Pabst. Joinson, A. N. (2005). Internet behaviour and the design of
Galegher, J. Sproull, L. and Kiesler, S. (1998). Legitimacy, virtual methods. In C. Hine (ed.), Virtual methods: Issues in
authority and community in electronic support groups. social research on the internet (pp. 000 000). Oxford: Berg.
Written Communication 15, 493 530. Joinson, A. N., Woodley, A. and Reips, U-R. (in press).
Garfinkel, S. (2000). Database nation: The death of privacy Personalization, authentication and self-disclosure in
in the 21st century. Sebastopol, CA: O Reilly and self-administered Internet surveys. Computers in Human
Associates, Inc. Behavior.
16-Joinson-Chap16 11/20/06 11:06 AM Page 249
References � 249
Jourard, S. M. (1971). Self-disclosure: An experimental Musch, J., Broder, A. and Klauer, K. C. (2001). Improving
analysis of the transparent self. New York: Krieger. survey research on the World-Wide Web using the
Jourard, S. M. and Lasakow, P. (1958). Some factors in randomized response technique. In U. D. Reips and
self-disclosure. Journal of Abnormal and Social M. Bosnjak (eds), Dimensions of Internet science
Psychology 56(1), 91 98. (pp. 179 192). Lengerich, Germany: Pabst Science
Jupiter Research (2002). Security and privacy data. Publishers.
Presentation to the Federal Trade Commission Nissenbaum, H. (1998). Protecting privacy in an
Consumer Information Security Workshop. Available information age: the problem of privacy in public. Law
at http://www.ftc.gov/bcp/workshops/security/ and Philosophy 17, 559 596.
0205201leathern.pdf. Accessed 20 June 2005. O Neill, D. (2001). Analysis of Internet users level of online
Kenny, D. A. and Judd, C. M. (1986). Consequences of privacy. Social Science Computer Review 19, 17 31.
violating the independence assumption in the analysis Olivero, N. (2001). Self-disclosure in e-commerce
of variance. Psychological Bulletin 99, 422 431. exchanges: relationships among trust, reward and
Knapp, H. and Kirk, S. A. (2003). Using pencil and paper, awareness. Paper presented at the Psychology and the
Internet and touch-tone phones for self-administered Internet: A European Perspective conference, DERA,
surveys: does methodology matter? Computers in Farnborough, UK.
Human Behaviour 19, 117 134. Olson, J. S., Zheng, J., Bos, N., Olson, G. M. and Veinott, E.
Kupritz, V. W. (2000). The role of the physical environment (2002). Trust without touch: jumpstarting long-distance
in maximising opportunities for the aging workforce. trust with initial social activities. In CHI2002 Conference
Journal of the Industrial Teacher Education 37, 66 88. Proceedings (pp. 141 146). New York: ACM Press.
Lau, J. T. F., Tsui, H. Y. and Wang, Q. S. (2003). Effects of two Parent, W. (1983). Privacy, morality and the law. Philosophy
telephone survey methods on the level of reported risk and Public Affairs 12, 269 288.
behaviours. Sexually Transmitted Infections 79, 325 331. Parks, M. R. and Floyd, K. (1996). Making friends in
Laurenceau, J. P., Barrett, L. F. and Pietromonaco, P. R. cyberspace. Journal of Communication 46, 80 97.
(1998). Intimacy as an interpersonal process: the Pederson, D. M. (1979). Dimensions of privacy. Perceptual
importance of self-disclosure, partner disclosure, and and Motor Skills 48, 1291 1297.
perceived partner responsiveness in interpersonal Pennebaker, J. W., Kiecolt-Glaser, J. K. and Glaser, R.
exchanges. Journal of Personality and Social Psychology (1988). Disclosure of traumas and immune function:
74, 1238 1251. health implications for psychotherapy. Journal of
Lessler, J. T., Caspar, R. A., Penne, M. A. and Barker, P. R. Consulting and Clinical Psychology 56, 239 245.
(2000). Developing computer-assisted interviewing Privacy Knowledge Base (2005). Available at
(CAI) for the National Household Survey on Drug http://privacyknowledgebase.com. Accessed 20 June 2005.
Abuse. Journal of Drug Issues 30, 19 34. Quittner, J. (1997). Invasion of privacy. Time Magazine,
Malhotra, N. K., Kim, S. S. and Agarwal, J. (2004). Internet 25 August, pp. 30 31.
users information privacy concerns (IUIPC): the Raab, C. Bellamy, C., Staylor, J., Dutton, W. H. and Peltu, M
construct, the scale and a causal model. Information (1996). The information polity: electronic democracy,
Systems Research 15, 336 355. privacy and surveillance. In Dutton, W. H. (ed.)
Margulis, S. T. (2003). On the status and contribution of Information and communication technologies: Visions
Westin s and Altman s theories of privacy. Journal of and realities. Oxford: Oxford University Press.
Social Issues 59, 411 429. Rheingold, H. (1993). The virtual community, revised edn.
Marshall, N. J. (1974). Dimensions of privacy preferences. London: MIT Press.
Multivariate Behavior Research 9, 255 272. Robinson, R. and West, R. (1992). A comparison of
McKenna, K. Y. A and Bargh, J. (1998). Coming out in the computer and questionnaire methods of history-taking in
age of the Internet: identity demarginalization through a genito-urinary clinic. Psychology and Health 6, 77  84.
virtual group participation. Journal of Personality and Rubin, Z. (1975). Disclosing oneself to a stranger:
Social Psychology 75, 681 694. reciprocity and its limits. Journal of Experimental Social
Metzger, M. J. (2004). Privacy, trust and disclosure: Psychology 11, 233 260.
exploring barriers to electronic commerce. Available at Rust, R. T., Kannan, P. K. and Peng, N. (2002). The
http://www.jcmc.indiana.edu/vo19/issue4/metzger.html. customer economics of internet privacy. Journal of the
Accessed 20 June 2005. Academy of Marketing Science 30, 455 464.
Milberg, S. J., Burke, S. J. and Smith, H. J. (1995). Values, Schatz Byford, K. (1996). Privacy in cyberspace:
personal information privacy, and regulatory constructing a model of privacy for the electronic
approaches. Communications of the ACM 38, 65 74. communications environment. Rutgers Computer and
Moon, Y. (1998). Impression management in computer-based Technology Law Journal 24, 1 74.
interviews: the effects of input modality, output Schoeman, F. (1984). Privacy and intimate information.
modality, and distance. Public Opinion Quarterly In Schoeman, F. (ed.) Philosophical dimensions of privacy
62, 610 22. (pp. 403 417). Cambridge: Cambridge University Press.
Moon, Y. (2000). Intimate exchanges: using computers to Smith, H. J., Milberg, S. J. and Burke, S. J. (1996). Information
elicit self-disclosure from consumers. Journal of privacy: measuring individuals concerns about
Consumer Research 27, 323 339. organizational practices. MIS Quarterly 20, 167 196.
16-Joinson-Chap16 11/20/06 11:06 AM Page 250
250 � CHAPTER 16 Self-disclosure, privacy and the internet
Smyth, J. M. (1998). Written emotional expression: effect Tourangeau, R. and Smith, T. W. (1996). Asking sensitive
sizes, outcome, types, and moderating variables. Journal questions: the impact of data collection mode, question
of Consulting and Clinical Psychology 66, 174 184. format, and question context. Public Opinion Quarterly
Sobel, D. L. (2000). The process that  John Doe is due: 60, 275 304.
addressing the legal challenge to Internet anonymity. Tourangeau, R., Couper, M. P. and Steiger, D. M. (2003).
Virginia Journal of Law and Technology 5. Humanizing self adminstered surveys: experiments on
Sparck-Jones, K. (2003). Privacy: what s different now? social presence in web and IVR surveys. Computers in
Interdisciplinary Science Reviews 28, 287 292. Human Behaviour 19, 1 24.
Sproull, L., Subramani, M., Kiesler, S., Walker, J. H. and Utz, S. (2004). Trust at eBay  influenced by the reputation
Waters, K. (1996). When the interface is a face. of the seller or the description of the product? Paper
Human Computer Interaction 11, 97 124. presented at German Online Research 2004, University
Tidwell, L. C. and Walther, J. B. (2002). Computer-mediated of Duisburg-Essen, Germany.
communication effects on disclosure, impressions, and Wallace, P. (1999). The psychology of the internet.
interpersonal evaluations: getting to know one another a Cambridge: Cambridge University Press
bit at a time. Human Communication Research 28, Walther, J. B. (1996). Computer-mediated communication:
317 348. impersonal, interpersonal, and hyperpersonal
Times, The (2001). Governments seem determined to interaction. Communication Research 23, 3 43.
overthrow online privacy  even if they have to behave Warren, S. and Brandeis, L. D. (1890). The right to privacy.
like hackers to do so. Technobabble, 3 December. Harvard Law Review 4, 193 220.
Times, The (2001). Junk e-mails cost Ł6.4bn. 3 February. Weisband, S. and Kiesler, S. (1996). Self-disclosure on
Times, The (2004). Plane passengers shocked by their x-ray computer forms: meta-analysis and implications.
scans. 7 November. Proceedings of CHI96. Available at http://www.acm.org/
Tourangeau, R. (2004). Survey research and societal sigchi/chi96/proceedings/papers/Weisband/sw_txt.htm.
change. Annual Review of Psychology 55, 775 801. Accessed 20 June 2005.