Unrepresentative Sample (original) (raw)

Disciplines > Argument > Fallacies > Unrepresentative Sample

Description | Discussion |Example | See also

Description

Sample X is taken from Population Y. Conclusion Z is drawn from sample X. It is assumed that Z is also true about Y.

Take a biased or otherwise statistically invalid sample. Analyze the data. Draw conclusions and declare the results significant.

Example

We surveyed homes during the day and found that 66% of the population enjoy soaps.

I asked four people in the street and three liked red. 75% of people like red.

Nine out of ten of cat owners we asked agreed that their cats like KitaKit.

Discussion

Most people believe they are pretty good at making statistical assessments. In fact we are generally pretty poor at it, and there are many traps into which we fall. Taking an unrepresentative sample is one of the most basic of these.

Where a sample is deliberately biased by leaving out data, this is the Fallacy of Exclusion.

Classification

Assumptive, Inductive, Statistical

Also known as

Biased Sample

See also

Theories about forecasting