|
|
|
||||
|
|
|||||
The examples below use Stata 9. If you are using Stata versions 7 or 8, please see this page.
NOTE: If you want to see the design effect or the misspecification effect, use estat effects after the command.Page 137 in the middle
clear
input person_num cluster gpa
1 1 3.08
2 1 2.60
3 1 3.44
4 1 3.04
1 2 2.36
2 2 3.04
3 2 3.28
4 2 2.68
1 3 2.00
2 3 2.56
3 3 2.52
4 3 1.88
1 4 3.00
2 4 2.88
3 4 3.44
4 4 3.64
1 5 2.68
2 5 1.92
3 5 3.28
4 5 3.20
end
list
+---------------------------+
| person~m cluster gpa |
|---------------------------|
1. | 1 1 3.08 |
2. | 2 1 2.6 |
3. | 3 1 3.44 |
4. | 4 1 3.04 |
5. | 1 2 2.36 |
|---------------------------|
6. | 2 2 3.04 |
7. | 3 2 3.28 |
8. | 4 2 2.68 |
9. | 1 3 2 |
10. | 2 3 2.56 |
|---------------------------|
11. | 3 3 2.52 |
12. | 4 3 1.88 |
13. | 1 4 3 |
14. | 2 4 2.88 |
15. | 3 4 3.44 |
|---------------------------|
16. | 4 4 3.64 |
17. | 1 5 2.68 |
18. | 2 5 1.92 |
19. | 3 5 3.28 |
20. | 4 5 3.2 |
+---------------------------+
tabstat gpa, s(sum) by(cluster)
Summary for variables: gpa
by categories of: cluster
cluster | sum
---------+----------
1 | 12.16
2 | 11.36
3 | 8.96
4 | 12.96
5 | 11.08
---------+----------
Total | 56.52
--------------------
gen pwt = 100/5
svyset cluster [pweight = pwt]
pweight: pwt
VCE: linearized
Strata 1: <one>
SU 1: cluster
FPC 1: <zero>
svy: total gpa
(running total on estimation sample)
Survey: Total estimation
Number of strata = 1 Number of obs = 20
Number of PSUs = 5 Population size = 400
Design df = 4
--------------------------------------------------------------
| Linearized
| Total Std. Err. [95% Conf. Interval]
-------------+------------------------------------------------
gpa | 1130.4 67.16666 943.9154 1316.885
--------------------------------------------------------------
Page 135 at the top
di total = (100/5)*56.52 1130.41130.4
Page 141 at the top
Population A
clear
input cluster value
1 10
1 20
1 30
2 11
2 20
2 32
3 9
3 17
3 31
end
tabstat value, s(mean var)
variable | mean variance
-------------+--------------------
value | 20 84.5
----------------------------------
tabstat value, s(mean var) by(cluster)
Summary for variables: value
by categories of: cluster
cluster | mean variance
---------+--------------------
1 | 20 100
2 | 21 111
3 | 19 124
---------+--------------------
Total | 20 84.5
------------------------------
anova value cluster, regress
Source | SS df MS Number of obs = 9
-------------+------------------------------ F( 2, 6) = 0.03
Model | 6 2 3 Prob > F = 0.9736
Residual | 670 6 111.666667 R-squared = 0.0089
-------------+------------------------------ Adj R-squared = -0.3215
Total | 676 8 84.5 Root MSE = 10.567
------------------------------------------------------------------------------
value Coef. Std. Err. t P>|t| [95% Conf. Interval]
------------------------------------------------------------------------------
_cons 19 6.101002 3.11 0.021 4.071387 33.92861
cluster
1 1 8.628119 0.12 0.912 -20.11225 22.11225
2 2 8.628119 0.23 0.824 -19.11225 23.11225
3 (dropped)
------------------------------------------------------------------------------
Population B
clear
input cluster value
1 9
1 10
1 11
2 17
2 20
2 20
3 31
3 32
3 30
end
tabstat value, s(mean var)
variable | mean variance
-------------+--------------------
value | 20 84.5
----------------------------------
tabstat value, s(mean var) by(cluster)
Summary for variables: value
by categories of: cluster
cluster | mean variance
---------+--------------------
1 | 10 1
2 | 19 3
3 | 31 1
---------+--------------------
Total | 20 84.5
------------------------------
anova value cluster, regress
Source | SS df MS Number of obs = 9
-------------+------------------------------ F( 2, 6) = 199.80
Model | 666 2 333 Prob > F = 0.0000
Residual | 10 6 1.66666667 R-squared = 0.9852
-------------+------------------------------ Adj R-squared = 0.9803
Total | 676 8 84.5 Root MSE = 1.291
------------------------------------------------------------------------------
value Coef. Std. Err. t P>|t| [95% Conf. Interval]
------------------------------------------------------------------------------
_cons 31 .745356 41.59 0.000 29.17618 32.82382
cluster
1 -21 1.054093 -19.92 0.000 -23.57927 -18.42073
2 -12 1.054093 -11.38 0.000 -14.57927 -9.420728
3 (dropped)
------------------------------------------------------------------------------
Page 142 at the top
clear
input person_num cluster gpa
1 1 3.08
2 1 2.60
3 1 3.44
4 1 3.04
1 2 2.36
2 2 3.04
3 2 3.28
4 2 2.68
1 3 2.00
2 3 2.56
3 3 2.52
4 3 1.88
1 4 3.00
2 4 2.88
3 4 3.44
4 4 3.64
1 5 2.68
2 5 1.92
3 5 3.28
4 5 3.20
end
anova gpa cluster
Number of obs = 20 R-squared = 0.4483
Root MSE = .430163 Adj R-squared = 0.3012
Source | Partial SS df MS F Prob > F
-----------+----------------------------------------------------
Model | 2.25568025 4 .563920063 3.05 0.0504
|
cluster | 2.25568025 4 .563920063 3.05 0.0504
|
Residual | 2.77560022 15 .185040014
-----------+----------------------------------------------------
Total | 5.03128047 19 .264804235
Page 149, figure 5.3
use http://www.ats.ucla.edu/stat/stata/examples/lohr/coots.dta, clear graph scatter volume clutch, ylabel( , nogrid angle(0)) ytitle(Egg Volume) xtitle(Clutch Number)
Page 149, figure 5.4
. sort clutch volume by clutch: egen mean = mean(volume) by clutch: gen n = _n drop csize breadth length reshape wide volume, i(clutch) j(n) sort mean gen rank = _n graph twoway rspike volume1 volume2 rank, ylabel( , nogrid angle(0)) /// xtitle(Clutch Ranked by Means) ytitle(Egg Volume)
Page 150, figure 5.5
sort clutch egen sd_dev = rowsd(volume1 volume2) graph scatter sd_dev mean, ylabel( , nogrid angle(0)) xlabel(1(1)5) /// xtitle(Mean Egg Volume for Clutch) ytitle(Standard Deviation for Clutch)
Page 150, table 5.2
use http://www.ats.ucla.edu/stat/stata/examples/lohr/coots.dta, clear sort clutch by clutch: gen n = _n by clutch: egen mean_vol = mean(volume) by clutch: egen sd_vol = sd(volume) gen var_vol = sd_vol*sd_vol by clutch: gen sum_vol = csize*mean_vol gen mi = 2 gen a = (1 - (2/csize))*csize^2*(var_vol/mi) gen yr = 4375.947/1757 gen b = (sum_vol - (csize*yr))^2
NOTE: The extra cases need to be dropped because you can't use if with tabstat.
. drop if n == 2
list clutch csize mean_vol var_vol sum_vol a b in 1/20
+-----------------------------------------------------------------------+
| clutch csize mean_vol var_vol sum_vol a b |
|-----------------------------------------------------------------------|
1. | 1 13 3.864303 .0093972 50.23594 .6719013 318.9229 |
2. | 2 13 4.194183 .0009177 54.52438 .0656166 490.483 |
3. | 3 6 .9162504 .0004814 5.497502 .0057766 89.22637 |
4. | 4 11 2.998335 .000795 32.98168 .0393539 31.19573 |
5. | 5 10 2.495708 .0001574 24.95708 .0062977 .0026303 |
|-----------------------------------------------------------------------|
6. | 6 13 3.98426 .0003304 51.79538 .0236219 377.0529 |
7. | 7 9 1.927069 .0050616 17.34362 .1594406 25.72102 |
8. | 8 11 2.961526 .005123 32.57679 .2535884 26.83677 |
9. | 9 12 3.460579 .0001066 41.52695 .0063963 135.4897 |
10. | 10 11 2.961526 .0223972 32.57679 1.108663 26.83677 |
|-----------------------------------------------------------------------|
11. | 11 12 3.498909 .0035246 41.9869 .2114751 146.4089 |
12. | 12 11 2.999868 .0001694 32.99855 .0083832 31.38445 |
13. | 13 12 3.566441 .012899 42.79729 .7739403 166.6769 |
14. | 14 11 2.986065 .0029402 32.84672 .1455399 29.70632 |
15. | 15 11 2.982998 .0010585 32.81298 .0523944 29.33965 |
|-----------------------------------------------------------------------|
16. | 16 10 2.406982 .0020082 24.06982 .0803278 .6988369 |
17. | 17 9 2.01023 .0005397 18.09207 .0169998 18.68958 |
18. | 18 10 2.437402 .0000289 24.37403 .0011567 .2827725 |
19. | 19 11 2.926252 .0000753 32.18877 .0037259 22.96712 |
20. | 20 11 2.947723 .0415674 32.42496 2.057586 25.28671 |
+-----------------------------------------------------------------------+
list clutch csize mean_vol var_vol sum_vol a b in -10/l
+-----------------------------------------------------------------------+
| clutch csize mean_vol var_vol sum_vol a b |
|-----------------------------------------------------------------------|
175. | 175 10 2.4336 .0008226 24.336 .0329023 .324661 |
176. | 176 12 3.752611 .0026651 45.03133 .1599053 229.3525 |
177. | 177 10 2.527395 .0003213 25.27395 .0128525 .135543 |
178. | 178 11 3.101091 .0127205 34.112 .6296641 45.09971 |
179. | 179 9 1.940416 .0014251 17.46374 .0448905 24.51704 |
|-----------------------------------------------------------------------|
180. | 180 9 1.946576 .0000759 17.51918 .0023906 23.97109 |
181. | 181 12 3.453279 .0017057 41.43934 .1023399 133.4579 |
182. | 182 13 4.219888 .0000367 54.85854 .0026246 505.396 |
183. | 183 13 4.414816 .0088191 57.39261 .6305637 625.7546 |
184. | 184 12 3.484307 6.66e-06 41.81168 .0003997 142.1994 |
+-----------------------------------------------------------------------+
tabstat csize sum_vol a b, s(sum)
stats | csize sum_vol a b
---------+----------------------------------------
sum | 1757 4375.947 42.17445 11439.58
--------------------------------------------------
tabstat sum_vol, s(var)
variable | variance
-------------+----------
sum_vol | 149.5648
------------------------
Page 154 Table 5.3
NOTE: There was an error in the data set. This has been corrected in the data that you can use from the ATS website. However, if you use the data off of the CD in the book or from the publisher's website, your answers may differ slightly. (The problem is with csize for clutch 88.)
use http://www.ats.ucla.edu/stat/stata/examples/lohr/coots.dta, clear
gen relwt = csize/2
gen wtvol = relwt*volume
list clutch csize volume relwt wtvol in 1/8
+----------------------------------------------+
| clutch csize volume relwt wtvol |
|----------------------------------------------|
1. | 1 13 3.795757 6.5 24.67242 |
2. | 1 13 3.93285 6.5 25.56352 |
3. | 2 13 4.215604 6.5 27.40142 |
4. | 2 13 4.172762 6.5 27.12295 |
5. | 3 6 .9317646 3 2.795294 |
|----------------------------------------------|
6. | 3 6 .9007362 3 2.702209 |
7. | 4 11 3.018272 5.5 16.6005 |
8. | 4 11 2.978397 5.5 16.38118 |
+----------------------------------------------+
list clutch csize volume relwt wtvol in -4/l
+----------------------------------------------+
| clutch csize volume relwt wtvol |
|----------------------------------------------|
365. | 183 13 4.481221 6.5 29.12794 |
366. | 183 13 4.348412 6.5 28.26468 |
367. | 184 12 3.486132 6 20.91679 |
368. | 184 12 3.482482 6 20.89489 |
+----------------------------------------------+
tabstat csize relwt wtvol, s(sum)
stats | csize relwt wtvol
---------+------------------------------
sum | 3514 1757 4375.947
----------------------------------------
Page 158 at the top
NOTE: You need to increase the matsize (matrix size) above the default value or you will get an error message.
set matsize 200
anova volume clutch
Number of obs = 368 R-squared = 0.9958
Root MSE = .07697 Adj R-squared = 0.9916
Source | Partial SS df MS F Prob > F
-----------+----------------------------------------------------
Model | 257.417531 183 1.40665317 237.44 0.0000
|
clutch | 257.417531 183 1.40665317 237.44 0.0000
|
Residual | 1.09007809 184 .005924337
-----------+----------------------------------------------------
Total | 258.507609 367 .704380405
Page 167 in the middle
. sort clutch
by clutch: gen n = _n
by clutch: egen mean_vol = mean(volume)
by clutch: gen sum_vol = csize*mean_vol
corr sum_vol csize
(obs=368)
| sum_vol csize
-------------+------------------
sum_vol | 1.0000
csize | 0.9693 1.0000
Page 167, figure 5.11
graph scatter sum_vol csize, ylabel(0(10)60, nogrid angle(0)) ytitle(Estimated Totla of Egg Volumes for Clutch) /// xlabel(6(2)14) xtitle(Clutch Size) msymbol(Oh)
UCLA Researchers are invited to our Statistical Consulting Services
We recommend others to our list of Other Resources for Statistical Computing Help
These pages are Copyrighted (c) by UCLA Academic Technology Services
