Millisecond Forums

Hello,

I have a problem with missing data. We have been using Inquisit v3 (lab, not internet based version) over the past three years to collect data for a large trial. We recently found that data output was not saved for 31 out of 335 participant visits. We have been unable to find any pattern to the missing data and are hoping there is some way to find the missing data or understand what problem caused the data not to save.

Thank you for any guidance you can provide!

+x

 

sticelab - 8/31/2020

Hello,

I have a problem with missing data. We have been using Inquisit v3 (lab, not internet based version) over the past three years to collect data for a large trial. We recently found that data output was not saved for 31 out of 335 participant visits. We have been unable to find any pattern to the missing data and are hoping there is some way to find the missing data or understand what problem caused the data not to save.

Thank you for any guidance you can provide!

(1) Where were you saving the data to? Local harddiks, network location, etc?
(2) What were the <data> element's /file and /separatefile settings for the respective script(s)? If /seperatefiles was not set at all or set to false, then you'll have data from multiple participants in the same file(s).
(3) How did you conclude that data was not saved at all rather than, for example, the data files having been moved or accidentally deleted by someone?

+x

 

Dave - 8/31/2020

+x

 

sticelab - 8/31/2020

Hello,

I have a problem with missing data. We have been using Inquisit v3 (lab, not internet based version) over the past three years to collect data for a large trial. We recently found that data output was not saved for 31 out of 335 participant visits. We have been unable to find any pattern to the missing data and are hoping there is some way to find the missing data or understand what problem caused the data not to save.

Thank you for any guidance you can provide!

(1) Where were you saving the data to? Local harddiks, network location, etc?
(2) What were the <data> element's /file and /separatefile settings for the respective script(s)? If /seperatefiles was not set at all or set to false, then you'll have data from multiple participants in the same file(s).
(3) How did you conclude that data was not saved at all rather than, for example, the data files having been moved or accidentally deleted by someone?

(1) We were saving the data to local hard drives
(2) The data did not have /seperatefiles so all data one each laptop used was saved to one large file.
(3) Because data was saved in one large file, and the data surrounding each missing file is present, we felt we could reasonable assume the data wasn't deleted or moved. That being said, it's always possible but if there is a way to eliminate all other possibilities, we'd like to do that.

Thanks for the help!

+x

 

sticelab - 9/1/2020

+x

 

Dave - 8/31/2020

+x

 

sticelab - 8/31/2020

Hello,

I have a problem with missing data. We have been using Inquisit v3 (lab, not internet based version) over the past three years to collect data for a large trial. We recently found that data output was not saved for 31 out of 335 participant visits. We have been unable to find any pattern to the missing data and are hoping there is some way to find the missing data or understand what problem caused the data not to save.

Thank you for any guidance you can provide!

(1) Where were you saving the data to? Local harddiks, network location, etc?
(2) What were the <data> element's /file and /separatefile settings for the respective script(s)? If /seperatefiles was not set at all or set to false, then you'll have data from multiple participants in the same file(s).
(3) How did you conclude that data was not saved at all rather than, for example, the data files having been moved or accidentally deleted by someone?

(1) We were saving the data to local hard drives
(2) The data did not have /seperatefiles so all data one each laptop used was saved to one large file.
(3) Because data was saved in one large file, and the data surrounding each missing file is present, we felt we could reasonable assume the data wasn't deleted or moved. That being said, it's always possible but if there is a way to eliminate all other possibilities, we'd like to do that.

Thanks for the help!

Have you ruled out simple data entry errors? E.g. if you're counting unique subject IDs to establish the number of data sets present, you'll get a wrong count if somebody mistakenly assigned the same subject ID to multiple different subjects. Beyond that, there are only two scenarios I can think of in which no data would have been written to the overall file:
(1) There simply was no data to record, i.e. the experiment was terminated before any data had been gathered.
(2) The overall data file was locked by another process (e.g. openend in Excel) during the experiment, making it impossible for Inquisit to write to it.

+x

 

Dave - 9/1/2020

+x

 

sticelab - 9/1/2020

+x

 

Dave - 8/31/2020

+x

 

sticelab - 8/31/2020

Hello,

I have a problem with missing data. We have been using Inquisit v3 (lab, not internet based version) over the past three years to collect data for a large trial. We recently found that data output was not saved for 31 out of 335 participant visits. We have been unable to find any pattern to the missing data and are hoping there is some way to find the missing data or understand what problem caused the data not to save.

Thank you for any guidance you can provide!

(1) Where were you saving the data to? Local harddiks, network location, etc?
(2) What were the <data> element's /file and /separatefile settings for the respective script(s)? If /seperatefiles was not set at all or set to false, then you'll have data from multiple participants in the same file(s).
(3) How did you conclude that data was not saved at all rather than, for example, the data files having been moved or accidentally deleted by someone?

(1) We were saving the data to local hard drives
(2) The data did not have /seperatefiles so all data one each laptop used was saved to one large file.
(3) Because data was saved in one large file, and the data surrounding each missing file is present, we felt we could reasonable assume the data wasn't deleted or moved. That being said, it's always possible but if there is a way to eliminate all other possibilities, we'd like to do that.

Thanks for the help!

Have you ruled out simple data entry errors? E.g. if you're counting unique subject IDs to establish the number of data sets present, you'll get a wrong count if somebody mistakenly assigned the same subject ID to multiple different subjects. Beyond that, there are only two scenarios I can think of in which no data would have been written to the overall file:
(1) There simply was no data to record, i.e. the experiment was terminated before any data had been gathered.
(2) The overall data file was locked by another process (e.g. openend in Excel) during the experiment, making it impossible for Inquisit to write to it.

I have ruled out simple data entry errors. I've checked every ID, noted duplicates and compared to visit dates.

Thank you for those possibilities! I will test running the program with the data file open in another process and see if Inquisit can't save the file. I never noticed or had participant mention that the experiment ended early but will also try to see if I can replicate a situation in which that might happen with testing. I appreciate the insight!