SPSS Statistics

 View Only
Expand all | Collapse all

Handling missing valued in SPSS (multiple imputation?)

  • 1.  Handling missing valued in SPSS (multiple imputation?)

    Posted Thu May 28, 2020 07:09 PM
    Hi there,

    I need to calculate the total of numerical values from a variable. The values actually represent the number of persons. But since the data is based on documents from the past there is few missing values.

    Still, I have 88% of the value whereas the rest is unavailable data. 

    Is there any ways that SPSS provides to estimate the value of the missing cases and get a total (estimated of course) that could be close to the reality if I had 100% of the values.

    Is multiple imputation one of the strategies I can use.

    At this level of my work I only need to get the sum.

    Thank you for helping me with this.

    Cheers.

    --
    Guillaume

    #SPSSStatistics


  • 2.  RE: Handling missing valued in SPSS (multiple imputation?)

    IBM Champion
    Posted Fri May 29, 2020 09:41 AM
    There are several approaches you could use depending on what you intend to do with the sum.

    Simplest would be to simply scale up the sum according to the proportion of missing values.

    If there are ten variables, V1 to V10, you could do this (computes could be combined).
    compute  nmiss=nmiss(V1 to V10).
    compute mean = mean.1(V1 to V10).
    compute imputedsum = (mean * (10 - nmiss) + (nmiss * mean)).

    If you have reason to believe that the missing values are related to other variables, you  might use the imputation procedure to calculate estimated means.

    --