Hi --
Raghunathan, T. E., Rosenthal, R., & Rubin, D. B. (1996) describe the Zpf
test (and an approximation) for testing whether two doubly dependent
correlation coefficients (Pearson r's) are significantly different. Does
anyone know of a parallel test for comparing the slopes (B's or Betas)
estimated by the regression?
By 'doubly dependent' I mean that for a single sample of participants, you
have
Rxy and Rqr
A concrete example might be, does the slope of the regression relating
heart rate and cortisol at time1 (R_hr.cort_t1) differ from that collected
at time2 ((R_hr.cort_t2) in the same sample.
Thanks much, Alex Shackman
=====================
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
