SPSS Statistics

Expand all | Collapse all

Can copying data from MS ACCESS into SPSS cause issues?

  • 1.  Can copying data from MS ACCESS into SPSS cause issues?

    Posted Mon October 26, 2020 01:23 PM
    I'm working with a dataset for a project that has been runnning for a few years. We used MS ACCESS as the data collection tool. Each week, the data would be manually selected and copied to an SPSS file via the COPY/PASTE function. The file would be 'Saved As' with a new date, and the old version moved to an ARCHIVE folder. I'm cleaning data now, and notice that data values will randomly change, or cases will duplicate and overwrite the data of several other cases. i.e. Instead of having 250 different cases, I can suddenly have 30 of the same case, and 220 different cases.

    Does anybody have any suggestion for what may be causing this?

    ------------------------------
    Tegan Murray
    ------------------------------


  • 2.  RE: Can copying data from MS ACCESS into SPSS cause issues?

    Posted Tue October 27, 2020 06:45 AM

    Hi Tegan,

    In general, using copy and paste to get data into SPSS is an error prone operation, and what you describe seems like user error rather than the fault of the SPSS editor. That being said, there is a number of ways that copying can og wrong, for example due to how the clipboard interprets things. The most common I know is that the tables use different decimal point conventions, and troubles with definitions of  string versus numerical variables.

    It is probably much better to use the SPSS database connectivity to read the Access table directly With appropriate filtering for dates.

    All the best

    jon

     

     

     






  • 3.  RE: Can copying data from MS ACCESS into SPSS cause issues?

    Posted Tue October 27, 2020 10:05 AM
    Adding to what the other Jon said - and I agree with him - I wonder whether what you are seeing is a problem with the DE display or with the actual data.  Check with a procedure such as FREQUENCIES to see where the problem lies.  What you see in the DE display is just a view on the data, which actually are managed by the backend where the procedures run.

    --