Data cleansing involves the detection and removal of errors and inconsistencies in a data set or database due to the corruption or inaccurate entry of the data. Data cleansing is a subset of data preparation, which also includes scoring tests, matching data files, selecting cases, and other tasks that are required to prepare data for analysis.
We all know that respondents don’t always complete surveys in the manner we expect. Sometimes the data we are working with reminds us that, despite our very best efforts, survey research can be a less-than-perfect process. We often need to take steps to balance the integrity of the data, while being careful not to alter responses in a way, which may introduce unintended biases.