Response rate

The response rate or response rate (synonym: exhausting ) is the ratio of samples drawn units to units actually achieved.

Response rates play especially in survey research a role. Here, the utilization rate is the ratio of realized to attempted Interviews: So try an interviewer to interview 150 people, but it creates only 30 of them to carry out the complete survey (perhaps because most respondents refuse to cooperate ) is the return rate 20%. It is similar with written surveys: Receives the market researcher filled out of 1000 questionnaires sent 100 back, the return rate is 10%.

General, it should strive to achieve exhaustion, because in case of failure often is not clear whether it is systematic failures ( it refuse example, particularly many supporters of a particular party, the survey ) or to not distort no systematic failures (for example, in a household sample guessed premises ).

The different failure types are also taken into account in the calculation of response rates. For the calculation of net utilization of non-systematic failures are not considered.

In practice, there are different methods to calculate the response rate. The reason is that there are several ways to divide the number of satisfied interviews by the number of " trials ". Tells you about the number of satisfied interviews by the total target sample, we obtain a lower response rate than if you divided by the uncharged failures (address errors, the target person is not part of the population ) adjusted sample. The members of scientific associations of market and opinion research are required to adhere to agreements on the calculation of the different response rates to avoid manipulation and deception of customers and the public.

Generally lie but the return rates in written surveys significantly lower than oral surveys. This applies to written examinations - unless it is special surveys of a defined group of people - response rates of more than 15 percent already remarkably high.

As in the empirical high representativeness of the responses is sought, a moment of silence distortion ( non-response bias) to exclude.