No matter how well a market research project is planned and executed, the reality is that there are usually some responses which are partially complete. Dealing with these holes in your data is a delicate matter. Leaving them in could distort the distributions. In addition, some systems cannot handle the empty values. Throwing away the responses is a waste and going back to the field would not be economically feasible and cannot maintain the statistical integrity of the project.

The best solution is to ascribe the data using statistical methods. The objective of the ascription process is that the ascribed data will have the same distribution as the original distribution of the non-empty data.

Ascription is one of the statistical services offered by SM Research. Two technologies can be applied during the ascription process. The first involves accumulated distribution filling, which fills holes with no special conditions. The second is called key demographic matching, which fills the holes by the answers from matched demographic respondents.

Both methods will create a non-distorted, no black-hole dataset that can be used in further studies. Attached is a real example. The survey contains 500 respondent records. Originally in Question Q7_Rec, there were 252 empty cells. In Question D7 there were 11 cells filled with not valid answer “99”. After the ascription, Q7_Rec were filled with values 1~3 and D7 were filled with 1~6. Both distributions were very close to the original distributions based on the valid answers.