A question about list-wise deletion and multiple imputation using SPSS. The foll
ID: 2946447 • Letter: A
Question
A question about list-wise deletion and multiple imputation using SPSS.
The following output was run by SPSS. Address the issue of missing raven data (only missing raven data) using two methods: list-wise exclusion and multiple imputation. Then run multiple regression analyses using raven, tai_w, tai_e, nfcp, and nfcn as the predictors. How are the results similar and how are they different? Are the overall models the same? What do you conclude?
List-wise deletion:
Multiple imputation:
Model Summary Adjusted R Square Std. Error of the Estimate Model R Square 099* 010 .003 3.36925 a. Predictors: (Constant), nfcn, tai_e, nfcp, tai_WExplanation / Answer
If we se the 'R square adjusted' values (refer first table in bothb the cases) they either negative or very close to zero implying in both the cases models fit are not very good (infact very bad). If we observe the regression coefficients ( see the column of either unstandardized coeff or standardized coeff in the third table in both the cases) they are more or less same and more importantly they are all insignificant as the p-values (see the columns named 'sig.' in third table in both the cases) are large (usually if p-values are less than 0.05 then corresponding coefficients are considered significant provided the level of significance is given as 0.05 )