I take the following steps: Factor analysis on Likert scale data Seemingly unrelated regression on resulting factors (total of 5) get marginal effects
These are my Stata commands:
factor b1_r1 b1_r2 b1_r3 b1_r4 b1_r5 b1_r6 b1_r7 b1_r8 b1_r9 b1_r10 b1_r11 b2_r1 b2_r2 b2_r3 b2_r4 b2_r5 b2_r6 b2_r7 b2_r8, pcf
rotate, oblique quartimin
predict Factor1 Factor2 Factor3 Factor4 Factor5
order Factor1 Factor2 Factor3 Factor4 Factor5, after (b2_r8)
tabstat Factor1 Factor2 Factor3 Factor4 Factor5, stat (N, mean, sd, min, max)
cor Factor1 Factor2 Factor3 Factor4 Factor5
label variable Factor1 "financial"
label variable Factor2 "social"
label variable Factor3 "policy"
label variable Factor4 "land"
label variable Factor5 "labour/succession"
********************************
*SUR
********************************
global x2list q2_q1 q4_2 q5_age q5_age2 q6_sucage q7 q11_2 r0_r1 t0_r1
sureg (Factor1 $x2list) (Factor2 $x2list) (Factor3 $x2list) (Factor4 $x2list) (Factor5 $x2list), small cor
margins, dydx(q2_q1 q4_2 q5_age q5_age2 q6_sucage q7 q11_2 r0_r1 t0_r1) atmeans
The results from the sureg are exactly the same as the results from the margins. Am I doing something wrong?
I do not understand why the results of the sureg and margins are the same. I have tried to see if there are any online explanations and spoken to my colleagues but we can't seem to figure it out.