The name is absent



Stata Technical Bulletin

15


. replace xy = xy + (-xbar)*mhiθ if mhiθ~=.

(1879 changes made)

. replace xy = xy + (.25-xbar)*mhi3mo if mhi3mo~=.

(3415 changes made)

. replace xy = xy + (l-xbar)*mhilyr if mhilyr~=.

(1891 changes made)

. replace xy = xy + (2-xbar)*mhi2yr if mhi2yr~=.

(1822 changes made)

. replace xy = xy + (4-xbar)*mhi4yr if mhi4yr~=.

(1455 changes made)

. gen slope = xy∕x2 if mhimiss<4 & (mhilyr~=. ∣ mhi2yr~=. ∣ mhi4yr~=.)

(1850 missing values generated)

We can now use slope as the subject of our analysis. The regression below is weighted by x2 (which is proportional to the
reciprocal of the variance of the slope) to account for the unequal sample sizes over which the slopes were calculated.

. reg slope married imale Iagecont ages* Inonwht =x2
(sum of wgt is 1.3459e+04)

(obs=1925)

Source I SS         df MS

Number of obs =    1925

---------+-

Model I

Residual ∣

3090.71465

47690.8921

7

1917

—

441.530664

24.877878

F( 7, 1917)

Prob > F
R-square

=   17.75

= 0.0000

= 0.0609

— А ЛК7Л

---------+-

—

Adj R-square

— U .UO t⅛

Total I

50781.6067

1924

26.3937665

Root MSE

= 4.9878

Variable ∣

Coefficient

Std. Error       t

Prob > It I

Mean

—^^ ^^—^^ ^^ ^^+-

—

slope I

.8629254

—^^ ^^—^^ ^^ ^^+-

—

married I

-.3264564

.2480897    -1.316

0.188

.5914187

imale I

-.6309712

.2438765    -2.587

0.010

.409048

iagecont I

-.0243468

.0301197    -0.808

0.419

55.76643

ages45 I

-.0271835

.0727632    -0.374

0.709

13.50555

ages55 I

-.1099776

.0904884    -1.215

0.224

7.085644

ages65 I

.107903

.0701864     1.537

0.124

2.46074

inonwht I

-.7529708

.3063881    -2.458

0.014

.1713754

_cons I
---------+-

3.681739

1.106394     3.328

0.001

1

—

. test ages45 ages55 ages65

( 1) ages45 = 0.0

( 2) ages55 = 0.0

( 3) ages65 = 0.0

F( 3, 1917) =    2.36

Prob > F =    0.0686

Since the F test is insignificant, I will dispense with the spline terms:

. reg slope married imale iagecont inonwht =x2

(sum of wgt
(obs=1925)

Source I

is 1.3459e+04)

MS

Number of obs =    1925

T? /  Λ    HACΛ'∣ —    OA AO

SS

df

Model I

Residual ∣

2914.63057

47866.9762

4

1920

728.657642

24.9307167

P X

Prob > F
R-square
Adj R-square
Root MSE

Prob > It I

—   ZJ.ZJ

= 0.0000

= 0.0574

= 0.0554

= 4.9931

Mean

Total I

Variable ∣

50781.6067

Coefficient

1924 26.3937665

Std. Error

t

slope I

.8629254

married I
imale I

iagecont I

inonwht I
_cons I

---------+_.

-.2398817

-.6003657

-.073299

-.678705

5.454313

.2423235

.2437838

.0073438

.3049193

.4483571

-0.990

-2.463

-9.981

-2.226

12.165

0.322

0.014

0.000

0.026

0.000

.5914187

.409048

55.76643
.1713754

1

—

As we found when we ran the specific change regressions, males relative to females have declining mental health over
time. We started by wondering whether marriage might improve mental health over time (real effect) or if instead persons with



More intriguing information

1. Educational Inequalities Among School Leavers in Ireland 1979-1994
2. Confusion and Reinforcement Learning in Experimental Public Goods Games
3. AGRICULTURAL PRODUCERS' WILLINGNESS TO PAY FOR REAL-TIME MESOSCALE WEATHER INFORMATION
4. CURRENT CHALLENGES FOR AGRICULTURAL POLICY
5. An institutional analysis of sasi laut in Maluku, Indonesia
6. Perfect Regular Equilibrium
7. Estimating the Economic Value of Specific Characteristics Associated with Angus Bulls Sold at Auction
8. The name is absent
9. Robust Econometrics
10. Determinants of U.S. Textile and Apparel Import Trade