Regression use if condition from stata

[Dong Shen]

Can someone help me to explain the inside meaning of using regress with if condition? For example, run the code:
sysuse auto
reg price mpg turn if turn>40

I can get the estimates. I thought it is the same if i run:
sysuse auto
drop if turn > 40
reg price mpg turn

But they produce different estimates. What is the reason?

Thanks in advance!

[Joseph Coveney]

Try

sysuse auto
keep if turn > 40
regress price c.(mpg turn)

and see if that helps.

[Carlo Lazzaro]

Dong:
as an aside to Joseph's helpful code, your two codes asks Stata to consider different groups of observations, as you can see from the following toy-example:

. sysuse auto
(1978 Automobile Data)

.
.  reg price mpg turn if turn>40

      Source |       SS           df       MS      Number of obs   =        35
-------------+----------------------------------   F(2, 32)        =      5.55
       Model |   105537647         2  52768823.7   Prob > F        =    0.0085
    Residual |   304173759        32  9505429.96   R-squared       =    0.2576
-------------+----------------------------------   Adj R-squared   =    0.2112
       Total |   409711406        34  12050335.5   Root MSE        =    3083.1

------------------------------------------------------------------------------
       price |      Coef.   Std. Err.      t    P>|t|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
         mpg |  -610.6069   226.6445    -2.69   0.011    -1072.267   -148.9471
        turn |    25.3587   290.1651     0.09   0.931    -565.6884    616.4057
       _cons |    16480.7   15265.45     1.08   0.288       -14614    47575.41
------------------------------------------------------------------------------

.  reg price mpg turn if turn<=40 /// In this second code the -if- qualifier asks Stata to consider only the observations with -turn-<=40

      Source |       SS           df       MS      Number of obs   =        39
-------------+----------------------------------   F(2, 36)        =      5.70
       Model |  45783659.4         2  22891829.7   Prob > F        =    0.0071
    Residual |   144495372        36  4013760.35   R-squared       =    0.2406
-------------+----------------------------------   Adj R-squared   =    0.1984
       Total |   190279032        38  5007342.94   Root MSE        =    2003.4

------------------------------------------------------------------------------
       price |      Coef.   Std. Err.      t    P>|t|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
         mpg |  -208.8436   63.88249    -3.27   0.002    -338.4033   -79.28385
        turn |   -104.624   154.1457    -0.68   0.502    -417.2461     207.998
       _cons |   14447.08   6426.082     2.25   0.031     1414.379    27479.77
------------------------------------------------------------------------------