A quota sample is a sample picked to match the population with respect to some summary characteristics. It is not a random sample. For example, in an opinion poll, one might select a sample so that the proportions of various ethnicities in the sample match the proportions of ethnicities in the overall population from which the sample is drawn. Matching on summary statistics does not guarantee that the sample comes close to matching the population with respect to the quantity of interest. As a result, quota samples are typically biased, and the size of the bias is generally impossible to determine unless the result can be compared with a known result for the whole population or for a random sample. Moreover, with a quota sample, it is impossible to quantify how representative of the population a quota sample is likely to be—quota sampling does not allow one to quantify the likely size of sampling error. Quota samples are to be avoided, and results based on quota samples are to be viewed with suspicion. See also convenience sample.