Home / Summaries / Designing Surveys: A Guide to Decisions and Procedures / rate-unit-members

Data collection II: Controlling error in data collection - Measures of survey quality

5 important questions on Data collection II: Controlling error in data collection - Measures of survey quality

How can you get a sense of nonresponse error?

In the absence of data, we rely on indirect indicators. In a general population survey, we know from census data the expected distribution of some demographic characteristics. If we collect some of this information in the survey, we can compare resnpondent's demographic characteristics to the census. If we find that a) our respondents underrepresent some demographic groups, and b) some of the substantive survey questions tend to be answered somewhat differently by members of the underrepresented groups compared with outher respondents. This would raise concers that our estimates may be biased against the underrepresented group.

What is the unit response rate?

The percentage of eligible sample units for whom interviews are obtained.

What is the cooperation rate?

The percentage of sample members who are interviewed, divided by interviews plus refusals and partial interviews.

What is the difference between the unit response rate and the cooperation rate?

The unit response rate includes sample members who were not contacted, were not interviewed because of language barriers, and so forth. While the cooperation rate only includes those who were contacted and either participated or refused.