My end users generally use SPSS to analyse our research data from FileMaker. I have been exporting to Excel and then to SPSS but was wondering if there was a simpler way to eliminate the middle step.
Can anyone help?
IBM SPSS can use ODBC to connect directly to an FM database.
If that is the SPSS package your talking about.
I'm sure that if you research IBM SPSS documentation and tech support forums you can find out how to make the product behave.
Finding out about ODBC only took me 5 mins and i've never used the product.
the .sav "format" contains binary data. so no, there is no direct method to get FMP data export into SPSS. However comma-separted and tab-delimited text can also be imported into SPSS, so it's an alternative to Excel (.xlsx)
Perhaps FMS can be used as an ODBC source, so that SPSS can read it directly. I have no experience with that.
Can you send us an example of the SPSS language you're using for defining the data? Is the data usually just a bunch of fixed length records?
In my copy of SPSS, I'm most familiar with something like:
Data List File = Myfile.txt Records=1
/Var1 to Var10 1-10 Var11 11-12 Var12 to Var 20 13-21.
Something like that. Is that similar to what you have?
You could make database tables for the value labels, but how about, instead, converting you numeric variable values to their labels and exporting the strings? Then you could use autorecode to restore the numeric values when you read the data back into SPSS. Autorecode can use a lookup table for the recode values, so you can set it up to guarantee that the recoding is consistent.
There is a valuelabel function that you can you to "reverse autorecode" the data. This could all be automated with Python programmability.
May be this link hels you solve the issue: FIleMaker Pro: Exporting Data - YouTube
Retrieving data ...