Really helpful video!! I watched three videos of multiple imputation but still not solve my problems. After using the syntax you taught I finally get my imputed dataset!
This was really helpful, great video. I would be very interested to learn more about how to do multiple imputation with repeated measures. Have you any videos on this or resources that might help me? Thanks
Great video! It looks like the imputed file was sent to the folder called /Multiple Imputation, not /Multiple Imputation Example. Just add that folder into last path in your syntax. It looks like you named the file name to Multiple Imputation Example not the folder.
This is very useful, thanks so much for sharing. Do you have any papers you recommend that could validate this method alongside any citable methods for doing in such a way?
Fantastic video!! One small question: what to do when you have categorical variables? I have a few categorical variables that I want to , but with the syntax from your video, SPSS doesn't recognize them as categorical variables.
Super helpful!!! Have you ran repeated measures ANOVA on MI data? It seems SPSS is yet to implement this, and R appears to be the only package capable of pooling MR ANOVA from MI'd datasets. Any thoughts?
great video and instruction. You do a good job explaining the process. I have a question. Do you need the ROLE=IND if you do not have a variable to use it with?
Hi. Thanks for the video! The same sintaxe would work if we had not the full data for Attention? I ask because in real setting if we are missing data for memory (Age we would if it was a case of pre-post), we would be also probably be missing for Attention? PS: Our context are probably different - I am not from psyc :P
Hi, if the data were missing for attention then you could impute attention as well. I would impute whatever variable has missing data. In any case, the model still uses the available data from each variable to predict what values to impute in the missing data.
Can I just impute one variable or two ? I don't know if this method is useful for me. I want to run parallel mediation analysis but one mediator has to exclude one person but for the other one not. How could I do this?
Can this be used without the Missing Values add-on for SPSS? My university doesn't have the add-on available, but I'd like to to MI in SPSS if at all possible (am using SPSS for other related analyses and want to be consistent). Thanks for the great content!
Unfortunately SPSS gives me an error message, because the number of parameters is too small. SPSS says it must be increased. How do I do this in the syntax? I would be very grateful for help!!!
Hi Annika, apologies for such a late reply. If increasing the number of parameters using /IMPUTE maxmodelparam = 500 doesn't work, try changing the model in the syntax from SCALEMODEL=LINEAR to SCALEMODEL=PMM (see video at 6 min and 15 sec). I hope this helps!
Also i have variables that were showen to only for the experimental group, so the control group got a -777 there = reasond missig, item was not shown. And in other variables, I have -999 for when people just did not answer or missed the entire measurement. Of course, I only want to impute the -999 how do I specify that? Do you know if there is any way I could use FIML in SPSS?
Really helpful video!! I watched three videos of multiple imputation but still not solve my problems. After using the syntax you taught I finally get my imputed dataset!
SO HELPFUL. You are very good at explaining things and making it clear.
Hi. Thank you for a very clear video. One question though, how do you decide on which version of 5 imputed time series you choose?
This was really helpful, great video. I would be very interested to learn more about how to do multiple imputation with repeated measures. Have you any videos on this or resources that might help me? Thanks
Great video! It looks like the imputed file was sent to the folder called /Multiple Imputation, not /Multiple Imputation Example. Just add that folder into last path in your syntax. It looks like you named the file name to Multiple Imputation Example not the folder.
This is very useful, thanks so much for sharing. Do you have any papers you recommend that could validate this method alongside any citable methods for doing in such a way?
Genius! Super clear!
you are a genius! thank you so much for this video
Fantastic video!! One small question: what to do when you have categorical variables? I have a few categorical variables that I want to , but with the syntax from your video, SPSS doesn't recognize them as categorical variables.
Super helpful!!! Have you ran repeated measures ANOVA on MI data? It seems SPSS is yet to implement this, and R appears to be the only package capable of pooling MR ANOVA from MI'd datasets. Any thoughts?
great video and instruction. You do a good job explaining the process. I have a question. Do you need the ROLE=IND if you do not have a variable to use it with?
I keep getting this error: "Subcommand IMPUTE: MAXPARAMDRAWS must specify a single integer value."
This was really helpful, thank you!!
Can you post a video on what you do with the imputed dataset to create a clean final file with one row per case?
Does using a multiple variables as the ROLE=IND give you a more valid imputation data? Or is one variable as a ROLE=IND is sufficient?
Hi. Thanks for the video! The same sintaxe would work if we had not the full data for Attention? I ask because in real setting if we are missing data for memory (Age we would if it was a case of pre-post), we would be also probably be missing for Attention? PS: Our context are probably different - I am not from psyc :P
Hi, if the data were missing for attention then you could impute attention as well. I would impute whatever variable has missing data. In any case, the model still uses the available data from each variable to predict what values to impute in the missing data.
Can I just impute one variable or two ? I don't know if this method is useful for me. I want to run parallel mediation analysis but one mediator has to exclude one person but for the other one not. How could I do this?
Can this be used without the Missing Values add-on for SPSS? My university doesn't have the add-on available, but I'd like to to MI in SPSS if at all possible (am using SPSS for other related analyses and want to be consistent). Thanks for the great content!
Hey! I think it should be possible. Perhaps try to see if it works and let me know if it doesn't work.
@@thepsychologyseries6998 @themtb11 Did it work without having the add-on available?
How can we implement Predictive mean matching for missing values?
Thank you so much for the video!
I keep getting this error: "Subcommand IMPUTE: MAXPARAMDRAWS must specify a single integer value."
Unfortunately SPSS gives me an error message, because the number of parameters is too small. SPSS says it must be increased. How do I do this in the syntax? I would be very grateful for help!!!
/IMPUTE maxmodelparam = 500 (or higher)
Hi Annika, apologies for such a late reply. If increasing the number of parameters using /IMPUTE maxmodelparam = 500 doesn't work, try changing the model in the syntax from SCALEMODEL=LINEAR to SCALEMODEL=PMM (see video at 6 min and 15 sec). I hope this helps!
Will I be able to do a mediation analysis?
Also i have variables that were showen to only for the experimental group, so the control group got a -777 there = reasond missig, item was not shown. And in other variables, I have -999 for when people just did not answer or missed the entire measurement. Of course, I only want to impute the -999 how do I specify that? Do you know if there is any way I could use FIML in SPSS?