Missing Data/Imputation Discussion > SPSS Imputation Errors
Haha..
Okay, so I just did what it said (i.e., changed it to a scale variable) and it worked, but... some of the data points are still missing in the imputed datasets. Is this supposed to occur?
Again, thank you and I apologize for the erroneous message above!
L
Hi L!
Thanks for the questions and it is good to see that you are figuring your own questions out! There is not supposed to be missing data remaining after imputation, unless there are exclusionary options selected in SPSS.
For example, in the "constraints" tab of the "Multiple Imputation" dialogue box, there is a box that will exclude variables with large amounts of missing data (if checked).
See an example of this box on page 29 of the SPSS Missing Data Manual:
Jeremy Taylor
Hi Jeremy,
My name is Filipa, I am a PhD student and when I was working with my data, I also had the error when conducting imputation of missing values The imputation model contains more than 100 parameters. No missing values will be imputed. Reducing the number of effects in the imputation model, by merging sparse categories of categorical variables, changing the measurement level of ordinal variables to scale, removing two-way interactions, or specifying constraints on the roles of some variables, may resolve the problem. Alternatively increase the maximum number of parameters allowed on the MAXMODELPARAM keyword of the IMPUTE subcommand."
However, for solving these problems, I don't know what is the best thing I should do: in fact, my variables are nominal and ordinal, so I think that the solution to change for scale will not work in my case. Therefore, I would like to know with you what will be the best solution to solve this problem, do you think it will be to specify constraints? Or do you think it's better to increase the number of parameters through subcommand MAXMODELPARAM? In fact, I have read in the spss manual that this later is not the best option. Therefore, can you please advice me?
many thanks in advance for all your help,
Filipa
Filipa Calado
Hi Jeremy!
Thanks for your other replies! They were incredibly helpful!! So I have another question for you. I keep getting an error message with two of my larger variable sets (each variable has ~ 60 items) and I was wondering if you might be able to help me troubleshoot. Below is the actual message I am getting (the ERNS variable is the larger one that I just tried to impute):
"The imputation model for ERNS_21r contains more than 100 parameters. No missing values will be imputed. Reducing the number of effects in the imputation model, by merging sparse categories of categorical variables, changing the measurement level of ordinal variables to scale, removing two-way interactions, or specifying constraints on the roles of some variables, may resolve the problem. Alternatively increase the maximum number of parameters allowed on the MAXMODELPARAM keyword of the IMPUTE subcommand."
I actually cannot find the MAXMODELPARAM in the command and do not know to what it refers, so I am concerned about how modifying it may influence my results. Any assistance would be most appreciated!!