Sampling Methods for Web and Email Surveys

1
Sampling Methods for Web and E-mail Surveys

• Associate Professor Ron Fricker
• Naval Postgraduate School
• Monterey, California

2
Today Anyone Can Survey for Little Cost
3
Internet-based Surveys and Sampling

• Focus is on survey data collection to support
  statistical inference
• Internet-based surveys often conducted via
  multiple media and mixed (survey) mode
• Contact mode respondent contact medium
• Survey or response mode survey medium
• Follow-up mode non-response follow-up medium

4
Bigger Samples Are Not Always Better

• When conducting Internet-based surveys,
  temptation is to (attempt to) survey everyone.
  Why not? Its cheap and easy
• Example Survey2000
• Goal Quantify how often people have moved, what
  role they play in their communities, and how
  geography has shaped their tastes in food, music,
  and literature
• More than 80,000 surveys initiated and 50,000
  completed

5
Types of Internet-based Surveys
6
Internet Surveys Pose Significant Challengesfor
General Populations

• Internet users do not reflect the general
  population
• Frame and coverage bias
• Very hard to impossible to generate a sampling
  frame
• There is no master list of e-mail addresses
• Recruiting respondents via the web can introduce
  bias
• Unethical/illegal to solicit survey participation
  via e-mail (spam)

7
Unsolicited E-mail is Spam
8
Problems Mitigated for Some Populations

• Internet-based surveys within individual
  organizations often easier
• List of e-mail addresses or common e-mail address
  syntax often available
• No problem with e-mail as spam
• But issues can remain
• How complete is the e-mail address list?
• Do all respondents have access to the Internet/
  e-mail?
• Non-response issues and bias

9
Summary