Tuesday, January 11, 2011

Off to a good start!

Eleven days in, 2011 has been off to a rocking start for me:


  1. An article I submitted came back a promising revise and resubmit
  2. I convinced the powers that be at my school to let me teach Course Y instead of Course X in the Fall, which will be killer for my job apps
  3. I got an awesome job for this coming summer in a place where I have lots of friends
  4. I have had an excellent week of working and folk dancing--three days of dancing in a week and my achilles tendon seems to be mostly under control!
  5. My advisor officially proved Phase 4 of my research, which is basically just filling in some gaps in my data.  In other words, even if it fails, I've got enough data for a dissertation!
  6. I got a foot pedal for Christmas.  While it doesn't make transcription bearable, it does make it faster.  
  7. I figured out how to conduct Kendall's partial tau in SPSS after an update changed the syntax rules and ruined my previous method.  This is an essential part of (1), and I'm going to paste the directions here, in addition to saving them on my computer, so I can find them in the future.  
How to conduct Kendall's Partial Tau in SPSS with Variable1, Variable2, and Variable3:

Paste the following syntax into the syntax editor and run it:

NONPAR CORR 
  /VARIABLES=Variable1 Variable2 Variable3 
  /PRINT=KENDALL TWOTAIL NOSIG 
  /MISSING=PAIRWISE
  /MATRIX=OUT(*).

In addition to the Kendall's tau correlations, you will get a matrix in a new data editor window.  Under rowtype, change all instances of "TAUB" to "CORR".  

Then paste the following into the syntax editor and run it:
PARTIAL CORR 
  /VARIABLES=Variable1 Variable2 BY Variable3 
  /SIGNIFICANCE=TWOTAIL 
  /FORMAT=MATRIX
  /MISSING=LISTWISE
  /MATRIX=IN(*).

This should also work for Spearman's partial rho.  In the first part, replace "KENDALL" with "SPEARMAN" and then change "RHO" to "CORR" in the matrix.  

1 comment:

  1. I have absolutely no idea what on earth the latter part of this post is about, but I'm glad it pleases you. I'm also very glad to hear about all the awesome professional credentials you're burnishing! Lend me some karma if you have any left over, would you, please?

    ReplyDelete