Testing the null B1=B2

[Julianna Grato]

I need to use a t-statistic to test is the coefficients on education and experience are equal. I need to rewrite the equation. My equation is:
log of hourly wage=B0+B1EDYEARS+B2EXPER+B3EXPER^2+u
Would these commands be correct?
generate EDYEARSexper2= (EDYEARS-exper+exper) (this wold be my new B1)
and then:
generate EDYEARSexper3= (EDYEARS+EXPER) (this would be my new B2)
and then: reg lofhrwage EDYEARSexper2 EDYEARSexper 3 exper^2
I really appreciate any feedback

[Weiwen Ng]

Julianna, if this is a homework question, please see this part of the FAQ.

I would advise you to see the help for testing a linear combination of coefficients:

help lincom

I prefer not to respond in more detail if this is homework for the reasons outlined in the FAQ.

[Julianna Grato]

This is not a homework question I am not in school. This is what I got when I ran the regression.

reg loghrwage EDYEARSexper22 EDYEARSexper2

Source | SS df MS Number of obs = 47,128
-------------+---------------------------------- F(2, 47125) = 6082.63
Model | 4936.82712 2 2468.41356 Prob > F = 0.0000
Residual | 19123.9653 47,125 .405813588 R-squared = 0.2052
-------------+---------------------------------- Adj R-squared = 0.2051
Total | 24060.7925 47,127 .510552177 Root MSE = .63703

--------------------------------------------------------------------------------
loghrwage | Coef. Std. Err. t P>|t| [90% Conf. Interval]
---------------+----------------------------------------------------------------
EDYEARSexper22 | .1321829 .0012152 108.78 0.000 .1301841 .1341817
EDYEARSexper2 | -.0090508 .0002662 -34.00 0.000 -.0094886 -.008613
_cons | 1.098445 .0183896 59.73 0.000 1.068196 1.128694
--------------------------------------------------------------------------------

[Joseph Coveney]

I need to use a t-statistic to test is the coefficients on education and experience are equal.

I'm not sure what you're trying to do with your created variables, for example,

generate EDYEARSexper2 = (EDYEARS - exper + exper)

doesn't make a whole lot of sense to me.

But, building on Weiwen's advice:

.ÿversionÿ15.1

.ÿ
.ÿquietlyÿsysuseÿauto,ÿclear

.ÿ
.ÿgenerateÿdoubleÿlog_wageÿ=ÿlog10(price)

.ÿegenÿbyteÿeducationÿ=ÿcut(weight),ÿgroup(`=22-16')

.ÿquietlyÿreplaceÿeducationÿ=ÿeducationÿ+ÿ16

.ÿegenÿbyteÿexperienceÿ=ÿcut(length),ÿgroup(10)

.ÿ
.ÿ*
.ÿ*ÿBeginÿhere
.ÿ*
.ÿregressÿlog_wageÿc.(educationÿexperience)

ÿÿÿÿÿÿSourceÿ|ÿÿÿÿÿÿÿSSÿÿÿÿÿÿÿÿÿÿÿdfÿÿÿÿÿÿÿMSÿÿÿÿÿÿNumberÿofÿobsÿÿÿ=ÿÿÿÿÿÿÿÿ74
-------------+----------------------------------ÿÿÿF(2,ÿ71)ÿÿÿÿÿÿÿÿ=ÿÿÿÿÿ13.16
ÿÿÿÿÿÿÿModelÿ|ÿÿ.572425397ÿÿÿÿÿÿÿÿÿ2ÿÿ.286212699ÿÿÿProbÿ>ÿFÿÿÿÿÿÿÿÿ=ÿÿÿÿ0.0000
ÿÿÿÿResidualÿ|ÿÿ1.54446445ÿÿÿÿÿÿÿÿ71ÿÿÿ.02175302ÿÿÿR-squaredÿÿÿÿÿÿÿ=ÿÿÿÿ0.2704
-------------+----------------------------------ÿÿÿAdjÿR-squaredÿÿÿ=ÿÿÿÿ0.2499
ÿÿÿÿÿÿÿTotalÿ|ÿÿ2.11688985ÿÿÿÿÿÿÿÿ73ÿÿ.028998491ÿÿÿRootÿMSEÿÿÿÿÿÿÿÿ=ÿÿÿÿ.14749

------------------------------------------------------------------------------
ÿÿÿÿlog_wageÿ|ÿÿÿÿÿÿCoef.ÿÿÿStd.ÿErr.ÿÿÿÿÿÿtÿÿÿÿP>|t|ÿÿÿÿÿ[95%ÿConf.ÿInterval]
-------------+----------------------------------------------------------------
ÿÿÿeducationÿ|ÿÿÿ.0735464ÿÿÿ.0261009ÿÿÿÿÿ2.82ÿÿÿ0.006ÿÿÿÿÿ.0215026ÿÿÿÿ.1255902
ÿÿexperienceÿ|ÿÿ-.0148003ÿÿÿ.0154942ÿÿÿÿ-0.96ÿÿÿ0.343ÿÿÿÿ-.0456949ÿÿÿÿ.0160944
ÿÿÿÿÿÿÿ_consÿ|ÿÿÿ2.459185ÿÿÿ.4179635ÿÿÿÿÿ5.88ÿÿÿ0.000ÿÿÿÿÿ1.625789ÿÿÿÿÿ3.29258
------------------------------------------------------------------------------

.ÿtestÿeducationÿ=ÿexperience

ÿ(ÿ1)ÿÿeducationÿ-ÿexperienceÿ=ÿ0

ÿÿÿÿÿÿÿF(ÿÿ1,ÿÿÿÿ71)ÿ=ÿÿÿÿ4.68
ÿÿÿÿÿÿÿÿÿÿÿÿProbÿ>ÿFÿ=ÿÿÿÿ0.0339

.ÿ
.ÿexit

endÿofÿdo-file


.

[Julianna Grato]

Thank you Joseph.

[Julianna Grato]

I was creating a new variable instead of doing SE(B1-B2)