If the data set is non-linear with high variability, which approach to replacing missing data is preferred? In SPSS, I selected the mean of nearby points (and specified two points), but the imputed values looked unusual. For instance, SPSS imputed a value of 3.75, even though the surrounding two values horizontally were 4 and 5, while the surrounding two values vertically were 4 and 1. How did SPSS compute this value? We tried analyzing data with missing values using SPSS, but that approach was not working. We are now relying on imputation by interpolation methods to replace missing values.

More Subina Saini's questions See All
Similar questions and discussions