Missing values

General help and assistance with jamovi. Bug reports can be made at our issues page: https://github.com/jamovi/jamovi/issues . (If you're unsure feel free to discuss it here)
Sjscher
Posts: 1
Joined: Thu May 12, 2022 12:50 pm

Re: Missing values

Post by Sjscher »

Current research suggests that methods such as mean imputation for missing values produces biased estimates.

Are there modules in jamovi for implementing modern missing value techniques (multiple imputation or maximum likelihood)?
User avatar
jonathon
Posts: 2613
Joined: Fri Jan 27, 2017 10:04 am

Re: Missing values

Post by jonathon »

not to my knowledge. it would be great if someone would write one.

https://dev.jamovi.org

jonathon
User avatar
leehw
Posts: 20
Joined: Fri Nov 22, 2019 8:48 am

Re: Missing values

Post by leehw »

I've only just seen these responses, having come up against this issue again. HUGE thanks all round!
malinek
Posts: 2
Joined: Tue Nov 08, 2022 8:42 am

Re: Missing values

Post by malinek »

hi,

as one of the previous users asked, i also want to figure out how to calculate the sum (score) including some missing data, ie including subjects with 1 missing item, but exclude those with more than 1 missing (but not do a mean imputation).
is there a function for this in jamovi?

alternatively, is there a way of making a count of the number of missing items for each sum score so i can write a "if, then" to exclude those with more than 1 item missing?
thanks! :)
User avatar
jonathon
Posts: 2613
Joined: Fri Jan 27, 2017 10:04 am

Re: Missing values

Post by jonathon »

if you take a look, you'll see that MEAN() can take a min_valid argument ...

so if you take the MEAN(A, B, C, D, min_valid=3) ... it will only provide a mean if at least three of those values are not missing.

jonathon
malinek
Posts: 2
Joined: Tue Nov 08, 2022 8:42 am

Re: Missing values

Post by malinek »

thanks for your quick reply!
yes, i saw in this thread that is is possible with the mean, but it is possible with other functions such as SUM and if so, how? (as we are not interested in the mean but the sum score of the items)
User avatar
jonathon
Posts: 2613
Joined: Fri Jan 27, 2017 10:04 am

Re: Missing values

Post by jonathon »

SUM() supports this too
Screenshot 2022-11-15 at 08.33.08.png
Screenshot 2022-11-15 at 08.33.08.png (86.66 KiB) Viewed 1814 times
Bobafett
Posts: 76
Joined: Thu Jul 18, 2019 11:33 am

Re: Missing values

Post by Bobafett »

[quote="jonathon"]SUM() supports this too
Screenshot 2022-11-15 at 08.33.08.png
[/quote]

I started using this a while back, but you need to remember that the sum scores created here may not be comparable to others as they will be based on different numbers of trials/items included in the score. What you could do instead is calculate a pro-rated sum score by using the mean score (with the minimum valid qualifier) and multiplying this by the full number of items/trials that exist on the task.

Not a great workaround, but one that works...
adearif132
Posts: 2
Joined: Wed Jun 28, 2023 6:53 am

Re: Missing values

Post by adearif132 »

Hi,

When I run Descriptives all my variables show missing value, but no cells were empty. Is something wrong with my dataset, How to solve this problem, and does it affect all the analysis?

Thanks
Attachments
missing value.png
missing value.png (188.99 KiB) Viewed 946 times
Bobafett
Posts: 76
Joined: Thu Jul 18, 2019 11:33 am

Re: Missing values

Post by Bobafett »

adearif132 wrote: Wed Jun 28, 2023 7:05 am Hi,

When I run Descriptives all my variables show missing value, but no cells were empty. Is something wrong with my dataset, How to solve this problem, and does it affect all the analysis?

Thanks
At a guess, at one point you probably had 70 rows of data (maybe as a result of copy/paste?) and then cut this out, but jamovi still sees 70 rows in the data as the excess were not removed. Quick check - are rows 36-70 shaded differently?
Post Reply