Web surveys have become very popular in recent years because they are
inexpensive relative to phone or mail surveys and because they are
very
fast. The main drawbacks of web surveys are the challenge to draw
representative samples of the general population, whether and how to
deal with respondents without Internet access and how to mitigate
nonresponse bias. I will discuss some proposed solutions including
adjustments for selection bias using propensity scoring, the use of
WebTV's for people without Internet access, and web survey panels.
Examples are drawn from work with the American Life Panel, the Health
and Retirement Study, the Harris Interactive Panel and the Knowledge
Network panel.
Meet the speaker in Room 212 Cockins Hall at 4:30
p.m. Refreshments will be served.