(Feedback from INSEE made on 21 September 2017)
From the statisticians viewpoint, data editing precedes the 5.3 "review and validate" phase such as the automatic coding. They would tend to have data editing in the 5.1 "Integrate phase".
(Feedback from INSEE made on 21 September 2017)
From the statisticians viewpoint, data editing precedes the 5.3 "review and validate" phase such as the automatic coding. They would tend to have data editing in the 5.1 "Integrate phase".
3 Comments
InKyung Choi
Meeting 3 April: similar issue was discussed in Design and Collect phases and automatic correction was indeed mentioned as one of editing examples happening before Process phase. Nilgun proposed some changes in this regard.
Action: Nilgün Dorsan to check text in 5.3
Nilgün Dorsan
After combining data at 5.1 (integrate data); classification of input data, control of the codes and automatic coding, are done at 5.2 (classify and code). Unless doing this work, one cannot run the edit rules.
At 5.3 (review and validate), errors, outliers, etc. are detected with the help of edit rules.
At 5.4 (edit and impute), after the errors are detected, correction/imputation procedure of the data has been started.
In line with those definitions, the issue raised at #37 about the "data editing" covers both 5.3 and 5.4.
In addition, 5.4 (edit and impute) especially defines the data correction activities. At this sub-process, "Edit" term is used as a correction which might be confused the users of GSBPM.
The title of this sub-process can be changed as "Correct/Impute".
user-8e470
Meeting 15/5: Agree with the point, but do not agree to change the name. Too many other things reference this sub process so the impact of the change is great.