Hi All,
I would like to perform a stage-sequential mediation analysis
involving an IV, a Putative Mediator, and an Outcome:
Dichotomous Variables:
IV (0 = low, 1 = high)
Mediator (0 = low, 1 = high)
Outcome (0 = low, 1 = high)
Generally: there is an event that has a “high” value on the IV,
which then increases the probability of a “high” value on the
Mediator, which in turn increases the probability of a “high”
value on the Outcome.
Specific test conditions: (a) the probability of a person undergoing
the sequence “no Mediator to Mediator to Outcome” is greater in
the “high” value IV group than it is in the “low” value IV
group; (b) being in the “high” value IV group increases the
probability of the “high” value Mediator group; and (c) at both
levels of the IV, the “high” value Mediator group increases the
probability of the “high” value Outcome group.
Below are some rough thoughts on syntax. I would greatly appreciate
any help with syntax to accomplish my above goals. I am running SPSS
version 10.0 for Windows. Thank you.
Best,
Frank
---condition (a and c)
LOGISTIC REGRESSION VAR=outcome
/METHOD=ENTER iv /METHOD=ENTER mediator
/CONTRAST (iv)=Indicator
/CLASSPLOT
/PRINT=GOODFIT
/CRIERIA PIN(.05) POUT(.10) ITERATE(20) CUT(.5).
---condition (b)
LOGISTIC REGRESSION VAR=mediator
/METHOD=ENTER iv
/CLASSPLOT
/PRINT=GOODFIT
/CRIERIA PIN(.05) POUT(.10) ITERATE(20) CUT(.5).
====================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
Locked
