When this box is checked, any record that is part of the input data source that does not meet the filter criteria will be written to the task output data source, along with the records that do meet the filter criteria. If you uncheck this box, only the input data source records that meet the filter criteria will be part of the task's output data source.
Typically, when doing lead cleaning, you’ll want to leave this box checked. However, if you are reducing the input data set using a filter and want to process only those records, leave the box unchecked. An example is to create a filter to select all non-surviving records after a Deduplication and merge task, add a “reject” reason to an attribute field and then write the records to an export data file. In this case, you only want to continue to process the filtered-only records since the goal is to create an output file containing only non-surviving records.