syntax question

[joost]

Hi,

i have the following syntax:
UNIANOVA
  FS_L_G_SUPFR_CT  BY DIAGNOSIS  WITH AGE_MONTHS
  /METHOD = SSTYPE(3)
  /INTERCEPT = INCLUDE
  /EMMEANS = TABLES(DIAGNOSIS) WITH(AGE_MONTHS=MEAN) COMPARE ADJ(LSD)
  /PRINT = DESCRIPTIVE ETASQ OPOWER
  /CRITERIA = ALPHA(.05)
  /DESIGN = AGE_MONTHS DIAGNOSIS .

i want to repeat this analysis and for every repetition i want to replace
the variable "FS_L_G_SUPFR_CT" for another variable. i have many variables
and do not want repeat the code for every analysis.so, can a loop be
scripted that runs UNIANOVA for a list of variables?

best,
-joost

=====================
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

[Bruce Weaver]

You can stick it in a macro.

define !MyModel (Ylist = !cmdend)

!do !Y !in (!Ylist)
UNIANOVA
  !Y  BY DIAGNOSIS  WITH AGE_MONTHS
  /METHOD = SSTYPE(3)
  /INTERCEPT = INCLUDE
  /EMMEANS = TABLES(DIAGNOSIS) WITH(AGE_MONTHS=MEAN) COMPARE ADJ(LSD)
  /PRINT = DESCRIPTIVE ETASQ OPOWER
  /CRITERIA = ALPHA(.05)
  /DESIGN = AGE_MONTHS DIAGNOSIS .
!doend

!enddefine.

* Now call the macro, handing it a list of dependent variables .

!MyModel Ylist = Y1 Y2 Y3 .

joost wrote

Hi,

i have the following syntax:
UNIANOVA
  FS_L_G_SUPFR_CT  BY DIAGNOSIS  WITH AGE_MONTHS
  /METHOD = SSTYPE(3)
  /INTERCEPT = INCLUDE
  /EMMEANS = TABLES(DIAGNOSIS) WITH(AGE_MONTHS=MEAN) COMPARE ADJ(LSD)
  /PRINT = DESCRIPTIVE ETASQ OPOWER
  /CRITERIA = ALPHA(.05)
  /DESIGN = AGE_MONTHS DIAGNOSIS .

i want to repeat this analysis and for every repetition i want to replace
the variable "FS_L_G_SUPFR_CT" for another variable. i have many variables
and do not want repeat the code for every analysis.so, can a loop be
scripted that runs UNIANOVA for a list of variables?

best,
-joost

=====================
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

[Edwin Meijdam-2]

Hi Joost,

Perhaps this will works (untested):

define uniList (!POS !CHAREND('/'))

!DO !i !IN (!1)

UNIANOVA

 !i  BY DIAGNOSIS  WITH AGE_MONTHS

 /METHOD = SSTYPE(3)

 /INTERCEPT = INCLUDE

 /EMMEANS = TABLES(DIAGNOSIS) WITH(AGE_MONTHS=MEAN) COMPARE ADJ(LSD)

 /PRINT = DESCRIPTIVE ETASQ OPOWER

 /CRITERIA = ALPHA(.05)

 /DESIGN = AGE_MONTHS DIAGNOSIS .

!DOEND.

!ENDDEFINE.

* Run command.

uniList VAR1 VAR2 VAR3  /.

On Mon, Dec 12, 2011 at 1:27 PM, joost <[hidden email]> wrote:

Hi,

i have the following syntax:
UNIANOVA
 FS_L_G_SUPFR_CT  BY DIAGNOSIS  WITH AGE_MONTHS
 /METHOD = SSTYPE(3)
 /INTERCEPT = INCLUDE
 /EMMEANS = TABLES(DIAGNOSIS) WITH(AGE_MONTHS=MEAN) COMPARE ADJ(LSD)
 /PRINT = DESCRIPTIVE ETASQ OPOWER
 /CRITERIA = ALPHA(.05)
 /DESIGN = AGE_MONTHS DIAGNOSIS .

i want to repeat this analysis and for every repetition i want to replace
the variable "FS_L_G_SUPFR_CT" for another variable. i have many variables
and do not want repeat the code for every analysis.so, can a loop be
scripted that runs UNIANOVA for a list of variables?

best,
-joost

=====================
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

--
Edwin Meijdam
[phone] 

DASC B.V. 

Postbus 88 4130 EB Vianen
085 - 8774406
[hidden email]
http://www.linkedin.com/in/emeijdam

[David Marso]

Alternatively you could slaughter the data using VarsToCases followed by SORT and SPLIT FILE.
Using SPLIT FILE LAYERED keeps all related material from the collection of vars together.
OTOH: Isn't a multivariate approach generally preferable (statistically) to a slew of univariate analyses?
--
Example using one of the canned data files:
GET
  FILE='C:\Program Files\SPSS\1991 U.S. General Social Survey.sav'.
VARSTOCASES /ID = id
 /MAKE trans1 FROM sex race region life sibs childs age paeduc maeduc speduc
  prestg80 occcat80 tax usintl obey popular thnkself workhard helpoth hlth1
  hlth2 hlth3 hlth4 hlth5 hlth6 hlth7 hlth8 hlth9 work1 work2 work3 work4
  work5 work6 work7 work8 work9
 /INDEX = Index1(trans1)
 /KEEP =  educ happy
 /NULL = KEEP.
SORT CASES BY Index1.
SPLIT FILE LAYERED BY index1.
UNIANOVA
  TRANS1 BY educ  WITH happy
  /METHOD = SSTYPE(3)
  /INTERCEPT = INCLUDE
  /EMMEANS = TABLES(educ) WITH(happy=MEAN) COMPARE ADJ(LSD)
  /PRINT = DESCRIPTIVE ETASQ OPOWER
  /CRITERIA = ALPHA(.05)
  /DESIGN = educ happy .

Bruce Weaver wrote

You can stick it in a macro.

define !MyModel (Ylist = !cmdend)

!do !Y !in (!Ylist)
UNIANOVA
  !Y  BY DIAGNOSIS  WITH AGE_MONTHS
  /METHOD = SSTYPE(3)
  /INTERCEPT = INCLUDE
  /EMMEANS = TABLES(DIAGNOSIS) WITH(AGE_MONTHS=MEAN) COMPARE ADJ(LSD)
  /PRINT = DESCRIPTIVE ETASQ OPOWER
  /CRITERIA = ALPHA(.05)
  /DESIGN = AGE_MONTHS DIAGNOSIS .
!doend

!enddefine.

* Now call the macro, handing it a list of dependent variables .

!MyModel Ylist = Y1 Y2 Y3 .

joost wrote

Hi,

i have the following syntax:
UNIANOVA
  FS_L_G_SUPFR_CT  BY DIAGNOSIS  WITH AGE_MONTHS
  /METHOD = SSTYPE(3)
  /INTERCEPT = INCLUDE
  /EMMEANS = TABLES(DIAGNOSIS) WITH(AGE_MONTHS=MEAN) COMPARE ADJ(LSD)
  /PRINT = DESCRIPTIVE ETASQ OPOWER
  /CRITERIA = ALPHA(.05)
  /DESIGN = AGE_MONTHS DIAGNOSIS .

i want to repeat this analysis and for every repetition i want to replace
the variable "FS_L_G_SUPFR_CT" for another variable. i have many variables
and do not want repeat the code for every analysis.so, can a loop be
scripted that runs UNIANOVA for a list of variables?

best,
-joost

=====================
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

[Leah Girshkin]

Thanks!  That worked perfectly for me.  Can't believe it was so hard to find this thread, such a useful thing to be able to do.


>>>
You can stick it in a macro.

define !MyModel (Ylist = !cmdend)

!do !Y !in (!Ylist)
UNIANOVA
  !Y  BY DIAGNOSIS  WITH AGE_MONTHS
  /METHOD = SSTYPE(3)
  /INTERCEPT = INCLUDE
  /EMMEANS = TABLES(DIAGNOSIS) WITH(AGE_MONTHS=MEAN) COMPARE ADJ(LSD)
  /PRINT = DESCRIPTIVE ETASQ OPOWER
  /CRITERIA = ALPHA(.05)
  /DESIGN = AGE_MONTHS DIAGNOSIS .
!doend

!enddefine.

* Now call the macro, handing it a list of dependent variables .

!MyModel Ylist = Y1 Y2 Y3 .