ReportingSpatial Duirbin Model Results obtained using xsmle

[Martin Hulenyi]

Dear all,

I have conducted a SDM model using xsmle in Stata and obtained the following results:


SDM with spatial fixed-effects Number of obs = 4734

Group variable: regid36 Number of groups = 263
Time variable: year Panel length = 18

R-sq: within = 0.1923
between = 0.3338
overall = 0.0602

Mean of fixed-effects = 0.3286

Log-likelihood = 10715.9469
------------------------------------------------------------------------------
gdppcgr | Coef. Std. Err. z P>|z| [95% Conf. Interval]
-------------+----------------------------------------------------------------
Main |
laglgdppc | -.1611488 .0078934 -20.42 0.000 -.1766196 -.1456781
linvr | .0027371 .0030948 0.88 0.376 -.0033286 .0088028
lpopgr | -.033131 .0033712 -9.83 0.000 -.0397385 -.0265235
wgipca | .0116851 .0016333 7.15 0.000 .0084838 .0148864
lefpayr | -.0052668 .1060317 -0.05 0.960 -.2130852 .2025515
wgilefp | -.0188772 .0426143 -0.44 0.658 -.1023997 .0646454
-------------+----------------------------------------------------------------
Wx |
laglgdppc | .1314624 .0093483 14.06 0.000 .11314 .1497847
linvr | .0219235 .0055082 3.98 0.000 .0111276 .0327195
lpopgr | .0339427 .0070075 4.84 0.000 .0202083 .0476771
wgipca | -.0097747 .002402 -4.07 0.000 -.0144824 -.0050669
lefpayr | .2987033 .1790174 1.67 0.095 -.0521644 .6495709
wgilefp | .0781617 .0630748 1.24 0.215 -.0454626 .201786
-------------+----------------------------------------------------------------
Spatial |
rho | .7397675 .0133931 55.24 0.000 .7135175 .7660174
-------------+----------------------------------------------------------------
Variance |
sigma2_e | .0005853 .0000122 47.83 0.000 .0005613 .0006093
-------------+----------------------------------------------------------------
LR_Direct |
laglgdppc | -.1591031 .0079131 -20.11 0.000 -.1746125 -.1435938
linvr | .0059994 .0029658 2.02 0.043 .0001865 .0118123
lpopgr | -.0313583 .0033271 -9.43 0.000 -.0378792 -.0248374
wgipca | .0114947 .00153 7.51 0.000 .0084958 .0144935
lefpayr | .0380498 .0998539 0.38 0.703 -.1576602 .2337598
wgilefp | -.0077706 .0400321 -0.19 0.846 -.086232 .0706907
-------------+----------------------------------------------------------------
LR_Indirect |
laglgdppc | .0455653 .0202692 2.25 0.025 .0058383 .0852923
linvr | .0878984 .0159565 5.51 0.000 .0566242 .1191726
lpopgr | .0370957 .0236823 1.57 0.117 -.0093207 .0835121
wgipca | -.0039658 .0058497 -0.68 0.498 -.015431 .0074995
lefpayr | 1.064335 .5218013 2.04 0.041 .0416234 2.087047
wgilefp | .2333985 .1701132 1.37 0.170 -.1000172 .5668142
-------------+----------------------------------------------------------------
LR_Total |
laglgdppc | -.1135378 .020262 -5.60 0.000 -.1532506 -.0738251
linvr | .0938978 .0164368 5.71 0.000 .0616822 .1261133
lpopgr | .0057374 .0246923 0.23 0.816 -.0426586 .0541334
wgipca | .0075289 .0055652 1.35 0.176 -.0033786 .0184364
lefpayr | 1.102385 .5294337 2.08 0.037 .0647141 2.140056
wgilefp | .2256279 .1705046 1.32 0.186 -.108555 .5598107
------------------------------------------------------------------------------

My question, whether in the paper I should report the Main and WX part of the regression or the long-run direct, indirect and total results.


Best,

Martin Hulenyi

[Phil Bromiley]

You didn't get a quick answer. You'll increase your chances of a helpful answer by following the FAQ on asking questions - provide Stata code in code delimiters, readable Stata output (fixed spacing fonts help), and sample data using dataex.

What is reported varies by discipline. I'd look at recent papers in your discipline to see what they report. When in doubt, including additional results is often a good idea.