Coder variance refers to nonsampling error that arises from inconsistencies in the ways established classification schemes are applied to the coding of research observations. In survey research, coder variance is associated with the process of translating the raw or verbatim data obtained from open-ended survey items into a quantitative format that can be analyzed by computers.
To appreciate how coder variance can occur, it is useful to review the process of preparing open-ended survey item data for analysis. Once all or a representative sample of the data have been collected, verbatim answers are examined for the purpose of denning a list of response categories (i.e. "code labels") that may be used for shorthand representations of the item data collected from each respondent. This list is known ...
Looks like you do not have access to this content.