Missing F Statistic and Interpreting Multiple Dummy Variables

[Shruti Gorsia]

Hi all,

I am investigating the significance of environmental concern on sustainable clothing consumption using an OLS linear regression, I have collected my data via online survey and received 168 responses. I faced an issue with heteroskedasticity so I am using robust standard errors for my regression, however this seems to be causing my F-statistic to go missing. I am unsure why, however I believe it is something to do with the fact that I have multiple dummy variable in my regression. Is there a way I can get my F-statistic back?

I also am having trouble with interpreting multiple dummy variables, which are control variables like gender, employment and education. I have been looking online and can only find information of interpreting just one dummy variable by looking at the residual for the baseline dummy variable but as I have many, this is much more complicated. This may be a silly question with a very obvious answer, but what is the usual practice of interpreting multiple dummy variables in a simple OLS linear regression?

Thank you in advance.

[Carlo Lazzaro]

Shruti:
welcome to this forum.
See

-help j_robustsingular-.
In addition, please follow the FAQ and post what you typed and what Stata gave you back. Thanks.

[Shruti Gorsia]

Thank you Carlo,

When I type:

regress sfi durability pricesensitivity influencerimpact environmentalconcern biospheric altruistic hedonic egoistic a5 a4 a3 a2 gen4 gen3 gen2 edu6 edu5 edu4 edu3 edu2 emp7 emp
6 emp5 emp4 emp3 emp2 inc7 inc6 inc5 inc4 inc3 inc2 children, robust

I received these results from Stata:

Linear regression                               Number of obs     =        168
                                                F(31, 132)        =          .
                                                Prob > F          =          .
                                                R-squared         =     0.2629
                                                Root MSE          =     .62045

--------------------------------------------------------------------------------------
                     |               Robust
                 sfi | Coefficient  std. err.      t    P>|t|     [95% conf. interval]
---------------------+----------------------------------------------------------------
           fashionv1 |   -.011832   .0542963    -0.22   0.828    -.1192355    .0955715
           fashionv2 |   .0276481   .0473406     0.58   0.560    -.0659963    .1212925
          durability |   .1021362   .0538454     1.90   0.060    -.0043752    .2086476
    pricesensitivity |  -.0359294   .0440909    -0.81   0.417    -.1231455    .0512868
    influencerimpact |    .031321   .0485923     0.64   0.520    -.0647995    .1274414
environmentalconcern |   .0155045   .0649564     0.24   0.812    -.1129857    .1439946
             hedonic |   .0749375   .0720944     1.04   0.301    -.0676723    .2175473
            egoistic |   .0526443   .0687914     0.77   0.445    -.0834319    .1887205
          biospheric |   .0779409   .0590152     1.32   0.189     -.038797    .1946789
          altruistic |  -.0957764   .0888872    -1.08   0.283    -.2716041    .0800513
                inc7 |  -.0003526   .2271944    -0.00   0.999    -.4497655    .4490603
                inc6 |   .1143295   .2690405     0.42   0.672    -.4178593    .6465182
                inc5 |   .1113412   .2629754     0.42   0.673    -.4088501    .6315326
                inc4 |   .3970283    .272255     1.46   0.147    -.1415189    .9355755
                inc3 |   .0185735   .2742258     0.07   0.946    -.5238722    .5610192
                inc2 |   .2985714   .2356941     1.27   0.207    -.1676548    .7647976
                gen4 |   1.093972   .4412137     2.48   0.014      .221208    1.966737
                gen3 |  -.5062985   .2521965    -2.01   0.047    -1.005168   -.0074289
                gen2 |   .2939569   .1222317     2.40   0.018     .0521705    .5357434
                emp7 |  -.4085834   .5270927    -0.78   0.440    -1.451225     .634058
                emp6 |   -.179046   .5672124    -0.32   0.753    -1.301048    .9429563
                emp5 |   .0681434   .3659653     0.19   0.853    -.6557721     .792059
                emp4 |  -.7533189   .4044519    -1.86   0.065    -1.553365     .046727
                emp3 |   -.039431   .2386615    -0.17   0.869    -.5115271     .432665
                emp2 |  -.1511681   .2783087    -0.54   0.588    -.7016902     .399354
                edu6 |    .178773   .2693026     0.66   0.508    -.3539342    .7114803
                edu5 |   .1670934   .4481752     0.37   0.710    -.7194413    1.053628
                edu4 |   .2535133   .2891159     0.88   0.382    -.3183865    .8254132
                edu3 |   .1148123   .2592631     0.44   0.659    -.3980357    .6276604
                edu2 |   .3883373   .2426893     1.60   0.112     -.091726    .8684007
                  a5 |   .2377038   .5106095     0.47   0.642    -.7723323     1.24774
                  a4 |   .1247279   .2877639     0.43   0.665    -.4444975    .6939533
                  a3 |  -.0471213    .254077    -0.19   0.853    -.5497106    .4554681
                  a2 |   .1098468      .2559     0.43   0.668    -.3963487    .6160423
            children |  -.0210589   .0791178    -0.27   0.791    -.1775618    .1354439
               _cons |   .9640338   .5587247     1.73   0.087     -.141179    2.069247
--------------------------------------------------------------------------------------

Many thanks,
Shruti
(Stata 17.0 SE)

[Carlo Lazzaro]

Shruti:
my previous advice still allplies.
See, under -help j_robustsingular-:
Are you using a svy estimator or did you specify the vce(cluster clustvar) option?
Is there a regressor that is nonzero for only 1 observation or for one cluster?

[Shruti Gorsia]

Carlos:
I have reviewed under -help j_robustsingular-, however I still cannot find the cause. I am not using a svy estimator and I have not specified the vce(cluster clustvar) option. I did have a few variables that were nonzero for only 1 observation, however I dropped these and I am still missing the F-statistic.

Thank you,
Shruti

[Carlo Lazzaro]

Shruti:
see, under -help j_robustsingular-:

You could find out what that number is by reestimating the model parameters without the vce(robust) and vce(cluster clustvar) options

[Shruti Gorsia]

Carlos:
Oh yes, this works, thank you very much!

Many thanks,
Shruti