Privacy

1
Privacy Self-disclosure online Implications
for Web surveys

• Carina Paine (1), Adam Joinson (1), Tom Buchanan
  (2)
• Ulf-Dietrich Reips (3)
• (1) The Open University, UK (2) Westminster
  University, UK
• (3) University of Zurich, Switzerland

2
Presentation Overview

• Introduction
• Self-Disclosure Privacy Online
• 2. Overview of 2 Studies
• Aim to explore the relationship between privacy
  self-disclosure online
• 3. Summary

3
Self-Disclosure Online
Introduction Study 1
Study 2
Summary

• People are increasingly required to disclose
  personal information online
• Self-disclosure online is a paradox
• Increased disclosure in CMC (Joinson, 2001)
• Due to anonymity? (e.g. Joinson, 1999 Reips
  Franek, 2004)
• Decreased disclosure for commercial websites
  (e.g. Jupiter, 2002)
• Due to privacy concerns? (e.g. Hoffman et
  al.,1999) lack of knowledge about how
  information will be used (Metzger, 2004)

4
Self-Disclosure Online Surveys
Introduction Study 1
Study 2
Summary

• Surveys administered online have been associated
  with higher levels of self disclosure (e.g.
  Weisband Kiesler, 1996)
• Due to anonymity likely audience? (Joinson)
• Due to increased privacy of research environment?
  (Tourangeau, 2004)
• Reduction in privacy results in reduction in
  self-disclosure (Joinson et al., in press)
• Does privacy play an important role in
  understanding peoples responses to web-based
  surveys?
• Privacy concerns rarely considered in
  self-disclosure research in the context of surveys

5
Privacy Online Surveys
Introduction Study 1
Study 2
Summary

• Privacy Online
• Offline privacy concerns magnified online?
  (Privacy Knowledge Base, 2005)
• Information disclosed is an increasingly valuable
  (but cheaply available) commodity
• Online Surveys Privacy Concerns
• Privacy is more sensitive issue online surveys
  can commit multiple violations of privacy more
  intense than those found in conventional surveys
  (Cho LaRose, 1999)
• Perceived Privacy?
• Anonymity confidentiality measures an issue
  of trust? (Singer et al., 1993, 1995)

6
Overview of Studies
Introduction Study 1
Study 2
Summary

• Overall Aim
• to explore the relationship between privacy
  self-disclosure, in particular, any mediating
  factors (such as trust perceived privacy)
• Initial 2 studies
• Interest in Trait State privacy
• Study 1 The development testing of privacy
  self-disclosure measures
• Study 2 2 part study detailed examination of
  relationship between privacy self-disclosure
• General Methodology
• Online participant panels of OU students
• Email invitation to Web based surveys

7
Study 1 methodology
Introduction Study 1
Study 2
Summary

• Online Survey Sample of 685 from panel 1
  (response rate75)
• Option posing items
• Privacy Attitude 48 items, 5-point scale e.g.
  how concerned are you about people online not
  being who they say they are?
• Privacy Behaviour 34 items, 5-point scale e.g.
  do you only register for web sites that have a
  privacy policy?
• Self-Disclosure 18 items, response options
  included please select as default prefer
  not to say option
• e.g. How many different sexual partners have
  you had?
• Order of item presentation manipulated

8
Study 1 results
Introduction Study 1
Study 2
Summary

• Psychometric procedures used to develop privacy
  self-disclosure scales
• Self-disclosure items pool of 16 items
• Privacy attitude 16 items
• Privacy behaviour 12 items equally split into
• General caution items
• Technical protection of privacy items
• Criterion validation study with Usenet groups
  OU online discussion groups confirmed scales
• Self-disclosure significantly lower when items
  presented after privacy items F(1,511)13.167,
  plt0.005, partial eta20.025

9
Study 2 methodology
Introduction Study 1
Study 2
Summary

• Trait privacy (from study 1 measures) State
  effects (perceived privacy)
• Study 2, part 1
• Online Survey Sample of 1931 from panel 2
  (response rate59)
• Option posing items
• Privacy Attitude
• Study 1 16 items, 5-point scale
• Westin (e.g. 1998) 3 items, 4-point scale
• IUIPC (Malhotra et al., 2004) 10 items,
  7-point scale
• Privacy Behaviour
• 6 general caution 6 technical protection
  items, 5-point scale

10
Study 2 methodology (cont)
Introduction Study 1
Study 2
Summary

• 1 month later.
• Study 2, part 2
• Online Survey Sample of 1931 OU students
  (response rate51) (67 retention rate)
• Option posing items
• Behavioural self-disclosure
• Study 1 16 items, 5-point scale
• Dispositional self-disclosure
• From IPIP, 10 items, 3-point scale
• Social desirability
• BIDR (IM SDE), 40 items, 7-point scale
• Trust, Anonymity, Confidentiality items, 5-point
  scale
• e.g. I am sure that my responses will remain
  confidential

11
Study 2 results
Introduction Study 1
Study 2
Summary

• Non-disclosure
• Active Mean 0.45, SD 1.05, range 0 - 10
• Passive Mean 0.09, SD 0.57, range 0 - 13
• Is there a link between privacy
  self-disclosure?
• Multiple regression analysis -gt Significant model
  F(3,748)5.478,plt0.005 Adjusted R square
  0.018 Privacy Concern Beta 0.086,
  plt0.05 Privacy Behaviour (general caution)
  Beta 0.102, plt0.05
• Is the link direct?
• Multiple regression analysis -gt Significant model
  F(17,572)5.198,plt0.0005 Adjusted R square
  0.108
• Trust Beta -0.251, plt0.0005
• BIDR IM score Beta 2.316, plt0.05
• Individual effects on self-disclosure?

12
STATE PROCESS
TRAIT PROCESS
13
Summary
Introduction Study 1
Study 2
Summary

• 1. Privacy is important
• i. State - as designed into a study (i.e. steps
  to ensure anonymity confidentiality, encourage
  trust)
• ii. Trait - general concern / behaviour of
  respondents
• 2. Different types of privacy seem to act
  independently
• they have unique effects on self-disclosure but
  a cumulative effect too
• Some evidence that link between perceived privacy
  self-disclosure may be via Trust ..
• 4. So, it is possible that steps to increase
  trust can mitigate impact of lowered perceived
  privacy
• 5. Peoples trait privacy will always remain,
  which isnt going to respond to survey design

14
Further Information

• Dr. Carina Paine / Dr. Adam Joinson,
• Institute of Educational Technology,
• The Open University,
• Milton Keynes,
• UK,MK7 6AA
• Web www.prisd.net (for slides references -
  next week)
• Email info_at_prisd.net

15
Model 1
16
Model 2