A survey of public opinion from a particular sample.
A few thousand is typical.
- Literary Digest in 1936
- 1992 British General Election
- There are no advantages. (Experienced pollsters do not aspire to conduct unscientific polls. Bias is typically introduced in poll designs because of carelessness or expedience.)
- Less accurate than polls conducted with random sampling (scientific polls).
- Represents raw, not refined, public opinion.