Dear all,
I'd like to get some help from you about panel data analysis. There
are two-wave panel data with three variables (x1, y1, z1 for wave 1;
x2, y2, z2 for wave 2). I'd like to test whether the influence of
independent variables (x, z) on the dependent variable (y) will become
stronger (or weaker) from time 1 to time 2. To examine this question,
I can run regressions (y=a + b1*x + b2*z) separately against wave 1
and 2 data and compare regression coefficients ( e.g., b1 for x1 and
b1 for x2) using z-statistics to test the difference statistically.
However, my question is whehter and how I can set up one regression
model where the dependent variable (y) is regressed on independent
variables (x, z) with a dummy variable for time (t: 0=wave 1, 1=wave
2) and interaction terms between independent variables and the time
dummy variable (x*t and z*t). The equation will be like this: y=a +
b1*x + b2*z + b3*t + b4*(x*t) + b5*(z*t). If this is possible, the
interaction terms, I think, can test my research question. That is, a
positive interaction (b4) between the time dummy (t) and an
independent variable ( e.g., x) would suggest that the influence of
the independent variable (x) on the dependent variable (y) is stronger
in wave 2 (t=1) as compared to that in wave 1. I did some research
about this possibility and found that research in medical science
often uses regression models similar to this, which are called
"repeated measures regression." Any suggestion will be welcome and
appreciated.
Jason
*
* For searches and help try:
* http://www.stata.com/support/faqs/res/findit.html
* http://www.stata.com/support/statalist/faq
* http://www.ats.ucla.edu/stat/stata/
