Eta value (partial n) with planned contrasts in one-way

[J.D. Haltigan]

Is there a way to manually generate the eta (partial effect size) statistic
when one uses planned contrasts in a ONE-WAY scenario. I can only seem to
get this if each contrast is specified as its own ONE-WAY but not
separately.

As below:

ONEWAY V1 V2 BY
    FACTOR1 (3-level)
  /CONTRAST=1 0 -1
  /CONTRAST=0 1 -1
  /STATISTICS DESCRIPTIVES
  /PLOT MEANS
  /MISSING ANALYSIS
  /POSTHOC=LSD BONFERRONI ALPHA(0.05).




--
Sent from: http://spssx-discussion.1045642.n5.nabble.com/

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

[J.D. Haltigan]

As a further illustration, the below syntax only gives partial eta for the
overall or omnibus ANOVA. I want the partial eta for each contrast (probably
can do in head, but for the time being let us proceed).

  UNIANOVA Y1 BY FACTOR1
  /CONTRAST(FACTOR1)=SPECIAL(1 0 -1)
  /METHOD=SSTYPE(3)
  /INTERCEPT=INCLUDE
  /POSTHOC=FACTOR1(LSD BONFERRONI T2)
  /PLOT=PROFILE(FACTOR1) TYPE=BAR ERRORBAR=CI MEANREFERENCE=YES
  /PRINT ETASQ DESCRIPTIVE HOMOGENEITY OPOWER
  /CRITERIA=ALPHA(.05)
  /DESIGN=FACTOR1.



--
Sent from: http://spssx-discussion.1045642.n5.nabble.com/

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