Demo

1
Demo
Robert Kraut Katie Bessiére Sara Kiesler Carnegie
Mellon University Human-Computer Interaction
Institute
2
Demo
Overview
Research Question How do varying levels of social
integration and different ways of using the
Internet interact to influence psychological
well-being? Assumption Social integrationleads
to better mental health, directlyand through
buffering fromstress.
3
Demo
Internet
Internet use and social integration The
Augmentation Hypothesis Online communication is
an added opportunity for interaction in addition
to more traditional modes of communication,
resulting in positive effects on psychological
well-being. The Substitution Hypothesis Online
communication replaces higher quality off-line
communication with thinner online communication,
resulting in negative effects on psychological
well-being. Evidence for both!
4
Demo
Internet

• Why the Disparate Results?
• Methodological Cross-sectional vs. Longitudinal
  data
• Substantive Effects depend upon other factors
• Details of online behavior
• Communication vs. other online activities
• Communication with friends family versus
  communication with strangers.
• Existing levels of social integration (e.g.,
  people with many versus few social contacts
  offline)

5
Demo
Hypothesis
Hypothesis Individuals with lower levels of
social resources will benefit more from using the
Internet for social purposes (i.e., communicating
with existing friends and family and
communicating with new people) than those with
high levels of social resources.
6
Demo
Data Collection

• Nationally representative longitudinal survey of
  US, self reports across two time periods (6-8
  months apart).
• Third wave collected, but not yet analyzed.
• 41 initial response rate, 82.8 completed the
  survey at time 2.
• N980

7
Measures of Internet use Frequency of use in
several domains
8
Demo
Measures
Measures of Social Resources

• Informal social activity
• Talk with friends
• Dinner out with friends
• Phone calls with friends family
• Card games
• I am part of a circle of friends

• Other indices of potential social resources
• Participation in formal groups
• Other adult in the household
• Has a job
• Perceived social support
• Perceived community belonging
• Extraversion
• Shyness (reversed)

9
Demo
Measures
Measure of Depression CES-D Scale (Cronbach
alpha .89) Measured on a 4-point scale (0 days
in past week 5-7 days)

• I felt that everything I did was an effort
• I felt depressed
• I felt fearful
• My sleep was restless
• I felt lonely
• I could not get "going"

• I had trouble keeping my mind on what I was doing
• I was bothered by things that don't usually
  bother me
• I felt that I could not shake off the blues even
  with help from my family or friends
• I felt hopeful about the future (Reversed)
• I was happy (Reversed)

10
Analysis

• Latent growth models/multi-level models to
  account for non-independence of observations
• Questionnaire administrations nested within
  respondent
• Predicting changes in depression from
• Existing social resources
• Ways of using the Internet
• Demographic controls
• Lag independent variables to increase confidence
  about causal direction
• Does Internet use and social resources at one
  time predict future changes in depression

11
Demo
Results
Predicting depression
Social resource Interpersonal Activity
12
Demo
Results
Cross-sectional Associations

• Less depression symptoms among
• Males
• Older respondents
• Richer respondents
• Those with more interpersonal activity

13
Demo
Results
Cross-sectional associations of depression with
Internet use

• More depression symptoms associated with \
• Using Internet for escape
• Using Internet to meet new people
• No reliable associations with
• Using Internet for information
• Using Internet for communicating with friends and
  family

14
Demo
Results
Longitudinal Results

• No overall change in depression with time
• Change depends upon how the Internet is used
• More Use Escape ? less depression
• More Use Meet New People ? more depression

15
Demo
Results
Effects of Internet use on changes in depression
depend upon available social resource

• .
• Using the Internet for escape decreases
  depression most for those who were most
  interpersonally active
• Using the Internet for information increases
  depression most for those who were most
  interpersonally active
• Using the Internet for meeting new people
  increases depression most for those who were most
  interpersonally active

16
Demo
Results
Use of the Internet for Meeting New People

• Increase in depression with Internet use is
  largest for those with most pre-existing social
  activities
• Reverses for those with few pre-existing social
  activities

17
Summary

• Cross sectional results differ from longitudinal
  ones
• Cross-sectional
• People who use the Internet socially
  (communication with friends and family or with
  new people) are less depressed than those who
  dont
• People who use the Internet for escape are more
  depressed than those who dont
• Longitudinal
• Using the Internet for meeting new people is
  associated with increases in depression
• Using the Internet for escape is associated with
  declines in depression
• Effects depend upon pre-existing social resources
• Association of using the Internet for meeting new
  people and increases in depression is greatest
  for those with most social resources
• Reverses for those with fewer social resources

18
Similar results for other measures of social
resources

• Interpersonal Activities(current analysis)
• Perceived Social Support
• Group Involvement

• Extraversion
• Shyness (reversed)
• Other adults are in the household

19
Demo
Results

• Possible Explanations
• Resource Effect addition of new social
  resources more beneficial for those who have
  fewer resources to begin with.
• Social Identity Theory addition of new,
  meaningful social identities is beneficial for
  those with few to begin with, but may saturate
  those who already have many.

20
Demo
More information
Contact Information Robert Kraut
robert.kraut_at_cmu.edu
21
extras
22
HomeNet Project

• Studying the way Americans are integrating the
  Internet into their lives since 1995
• Quantitative methods
• Longitudinal surveys
• Usage tracking
• Qualitative methods
• Family interviews
• Observations
• Today
• US longitudinal surveys about psychological
  well-being

23
Demo
Methods
Different people use the Internet differently

• Controlling for frequency of using the Internet
  for other purposes

24
Demo
Methods

• Measures of Internet Use
• Communicating With friends family
• keeping in touch, communicating with friends,
  communicating with family, communicating locally
• Meeting New People
• meeting new people, participating in an online
  group
• Information Seeking
• news, local events, work, hobby
• Escape and Entertainment
• playing games, killing time, releasing tension,
  downloading music,

25
Results
Interpersonal Activity
Change in depression levels due to the
interaction between use of the Internet for
Escape and Time, as well as use of the Internet
for Meeting New People and Time.
26
Demo
Conclusion

• Clear and consistent support for our hypothesis
  adding new online social contacts benefits those
  with low levels of offline social interaction,
  more than those with high levels of social
  interaction.
• Personal characteristics do affect the
  relationship between Internet use and depression.