|
I have inherited a set of SPSS files for several ongoing clinical trials, some of which involve numerous follow-up assessments. The files for each of the assessments are in a horizontal file structure with separate sets of variables for each of the follow-ups (i.e. dep01_f1 for follow-up 1, dep01_f2 for follow-up #2, etc.). This has resulted in data sets that are fairly shallow (<100 records) but quite wide (>3000 variables). Eventually, the researcher will be wanting to do multilevel modeling, in addition to time-series analysis.
I come from a database management background with only limited statistical analysis experience. I'm uncomfortable with the horizontal file structure...it goes against the grain of my RDBMS training. Furthermore, I've read that SAS (and SPSS?) needs to have the data in vertical structure for mixed-method analysis. Before I propose that the data be converted to vertical file structure, I would like to know if there are other considerations that make the horizontal structure better? Thanks. Karen R. Harker, MLS UT Southwestern Medical Library 5323 Harry Hines Blvd. Dallas, TX 75390-9049 214-648-8946 http://www.utsouthwestern.edu/library/ ===================== To manage your subscription to SPSSX-L, send a message to [hidden email] (not to SPSSX-L), with no body text except the command. To leave the list, send the command SIGNOFF SPSSX-L For a list of commands to manage subscriptions, send the command INFO REFCARD |
|
Hi Karen,
To do multilevel/mixed effects modeling then yes, the files will need to be in the "vertical" structure. However, I often find that I end up needing both formats of the files - eg, vertical for the analysis, but horizontal for some of the charts. SPSS has the CASETOVARS command (Data> Restructure Wizard) for transforming from one structure to the other. Cheers, Kylie. -----Original Message----- From: SPSSX(r) Discussion [mailto:[hidden email]] On Behalf Of Karen Harker Sent: Monday, 16 February 2009 10:15 AM To: [hidden email] Subject: Horizontal or vertical file structure for repeated measures I have inherited a set of SPSS files for several ongoing clinical trials, some of which involve numerous follow-up assessments. The files for each of the assessments are in a horizontal file structure with separate sets of variables for each of the follow-ups (i.e. dep01_f1 for follow-up 1, dep01_f2 for follow-up #2, etc.). This has resulted in data sets that are fairly shallow (<100 records) but quite wide (>3000 variables). Eventually, the researcher will be wanting to do multilevel modeling, in addition to time-series analysis. I come from a database management background with only limited statistical analysis experience. I'm uncomfortable with the horizontal file structure...it goes against the grain of my RDBMS training. Furthermore, I've read that SAS (and SPSS?) needs to have the data in vertical structure for mixed-method analysis. Before I propose that the data be converted to vertical file structure, I would like to know if there are other considerations that make the horizontal structure better? Thanks. Karen R. Harker, MLS UT Southwestern Medical Library 5323 Harry Hines Blvd. Dallas, TX 75390-9049 214-648-8946 http://www.utsouthwestern.edu/library/ ===================== To manage your subscription to SPSSX-L, send a message to [hidden email] (not to SPSSX-L), with no body text except the command. To leave the list, send the command SIGNOFF SPSSX-L For a list of commands to manage subscriptions, send the command INFO REFCARD ===================== To manage your subscription to SPSSX-L, send a message to [hidden email] (not to SPSSX-L), with no body text except the command. To leave the list, send the command SIGNOFF SPSSX-L For a list of commands to manage subscriptions, send the command INFO REFCARD |
| Free forum by Nabble | Edit this page |
