Coefplot for multiple regression

[Khati Zolfaghari]

Hello and good day, I estimate through a cross-sectional model and I did one by one estimation . I want to do two different plots.
1) coefplot based on the following regressions:

. reg fertility lngdpper Unemployment WPUI StringencyIndex

      Source |       SS           df       MS      Number of obs   =       130
-------------+----------------------------------   F(4, 125)       =     54.44
       Model |  43.5590529         4  10.8897632   Prob > F        =    0.0000
    Residual |  25.0034792       125  .200027834   R-squared       =    0.6353
-------------+----------------------------------   Adj R-squared   =    0.6236
       Total |  68.5625321       129  .531492497   Root MSE        =    .44724

---------------------------------------------------------------------------------
      fertility | Coefficient  Std. err.      t    P>|t|     [95% conf. interval]
----------------+----------------------------------------------------------------
       lngdpper |  -.3097261   .0322072    -9.62   0.000    -.3734681   -.2459841
   Unemployment |  -.0069286   .0077619    -0.89   0.374    -.0222904    .0084332
           WPUI |  -.0007319   .0003823    -1.91   0.058    -.0014886    .0000247
StringencyIndex |  -.0151191   .0058869    -2.57   0.011    -.0267701   -.0034681
          _cons |   5.534372   .2454153    22.55   0.000     5.048665    6.020079
---------------------------------------------------------------------------------

estimates store D

. reg fertility lngdpper Unemployment WPUI schoolclosing

      Source |       SS           df       MS      Number of obs   =       130
-------------+----------------------------------   F(4, 125)       =     53.58
       Model |  43.3045478         4   10.826137   Prob > F        =    0.0000
    Residual |  25.2579843       125  .202063874   R-squared       =    0.6316
-------------+----------------------------------   Adj R-squared   =    0.6198
       Total |  68.5625321       129  .531492497   Root MSE        =    .44952

-------------------------------------------------------------------------------
    fertility | Coefficient  Std. err.      t    P>|t|     [95% conf. interval]
--------------+----------------------------------------------------------------
     lngdpper |  -.3234715   .0316028   -10.24   0.000    -.3860173   -.2609257
 Unemployment |  -.0068205   .0078007    -0.87   0.384    -.0222591    .0086181
         WPUI |  -.0008157   .0003773    -2.16   0.033    -.0015625    -.000069
schoolclosing |  -.2478858   .1079816    -2.30   0.023    -.4615948   -.0341769
        _cons |   5.546419    .247489    22.41   0.000     5.056608    6.036231
-------------------------------------------------------------------------------

estimates store E

. reg fertility lngdpper Unemployment WPUI Cancelpubliceven

      Source |       SS           df       MS      Number of obs   =       130
-------------+----------------------------------   F(4, 125)       =     52.25
       Model |   42.902468         4   10.725617   Prob > F        =    0.0000
    Residual |   25.660064       125  .205280512   R-squared       =    0.6257
-------------+----------------------------------   Adj R-squared   =    0.6138
       Total |  68.5625321       129  .531492497   Root MSE        =    .45308

----------------------------------------------------------------------------------
       fertility | Coefficient  Std. err.      t    P>|t|     [95% conf. interval]
-----------------+----------------------------------------------------------------
        lngdpper |  -.3184778   .0322615    -9.87   0.000    -.3823274   -.2546282
    Unemployment |  -.0070328   .0078681    -0.89   0.373    -.0226049    .0085392
            WPUI |  -.0008469   .0003845    -2.20   0.029    -.0016079   -.0000858
Cancelpubliceven |  -.3190479   .1775599    -1.80   0.075    -.6704609    .0323652
           _cons |   5.499037   .2478894    22.18   0.000     5.008433    5.989641
----------------------------------------------------------------------------------

estimates store F

coefplot D E coefplot D F E, vertical drop(_cons) yline(0)

but when I did the coefplot , it is not so clear and I think there is a mistake.



2) just scatter plot among the fertility and my variables without doing a regression.
I did this command which the results are not clear:

twoway (scatter fertility gdpgrowth,mcolor(blue) msymbol(t))(scatter fertility StringencyIndex ,mcolor(red) msymbol(D)) (scatter fertility schoolclosing , msymbol(S))

I hope I receive your advice so soon. thank you so much in advance.

[Andrew Musau]

coefplot is from SSC, as you are asked to explain (FAQ Advice #12). There are quite a number of examples in http://repec.sowi.unibe.ch/stata/coe...g-started.html. Otherwise, you need to be specific on how you want the result to look like.

[Khati Zolfaghari]

@Andrew Musau thank you so much for your reply. Actually, I saw http://repec.sowi.unibe.ch/stata/coe...g-started.html and I did it. But I want to do a coefplot , and I want to add one observation at each point.

For example the first estimation :
reg fertility lngdpper Unemployment WPUI StringencyIndex the second :
reg fertility lngdpper Unemployment WPUI schoolclosing and so on.. I want that the coefplot also represent this things like "Different plot styles per subgraph" but in the yline all explanatory variables be included and show different regression. Also for another scatterplot I want to show the relationship between fertility and other explanatory variables without runing the regression at the on graph. (I hope I could explain good ) I try to attach my graph but I want to have another plot. I could not. thanks alot.

[Andrew Musau]

I cannot completely follow #3. It appears that you are running a number of regressions with some fixed variables and adding a different variable at each iteration. If you want to plot the coefficients on the added variables, here is one way.

sysuse auto, clear
eststo m1: regress mpg weight turn 
eststo m2: regress mpg weight displacement
eststo m3: regress mpg weight gear_ratio
coefplot (m1, keep(turn)) (m2, keep(displacement)) (m3, keep(gear_ratio)), vert nokey

To change the marker and CI colors, look at the link in #2.

[Khati Zolfaghari]

@Andrew Musau thank you so much for your reply.