![Show Menu](styles/mobile-menu.png)
![Page Background](./../common/page-substrates/page0088.png)
МОНГОЛЫН ХҮН АМЫН СЭТГҮҮЛ Дугаар (367) 20, 2011
87
For the dummy variable
sex
: holding all other
variables at their means, being male increases
the both probability of living in
Ger
and House
by 3.3 per cent and 5.3 per cent, respectively
and decreases the probability of living in
apartment by 8.6 per cent.
It is easy to see from the Figure 3, the
average absolute change for a standard
deviation change in
totrev
is 11.4 per cent,
the maximum value from the all of average
absolute change. Thus
household income
contributes more relatively to the explanation
of the other (numerical) variables. The effect
of
totrev
is largest on the probability of living
in apartment, where the expected change for
a standard deviation in
totrev
is 17.1 per cent.
Furthermore, the average absolute change
for a standard deviation change in
childnum
is -9.5 per cent, the minimum value from the
all of average absolute change. The effect
of
childnum
is smallest on the probability of
living in apartment, where the expected change
for a standard deviation in
childnum
is -14.3
per cent.
CONCLUSIONS
This study has attempted to determine
the factors which influence to live in a
particular type of dwelling. The sample is from
the Urban Poverty and in-Migration survey
in 2004, which covers 1500 households in
Ulaanbaatar. Ten plausible factors were used
in analysis but during the estimation stage two
of them are excluded. Using both ordered logit
and multinominal logit models, significant ML
estimates and tests were reported.
The factors, namely age, sex and education
level of the head of family, number of
person in household, percentage of migrants
in household, number of children in
household, total revenue of household per
year and percentage of food expenses in total
Figure 3:
Change in Predicted Probability for type of dwelling ( )
-.14
-.09
-.04
.01
.07
.12
.17
G H
A
G
H
A
G H
A
G H
A
GH
A
G H
A
totper-std
mig_percent-std
age-std
childnum-std
totrev-std
food-std
expenditure per year have significant effect
on probability to live in particular type of
dwelling. The excluded two factors have
strong logical statements to be considered as
a factor in this type of analysis. Therefore, it
leaves us further study.
REFERENCE
Damodar N. Gujarati, D. C. (2009). Chapter 15. In D.
C. Damodar N. Gujarati,
Basic Econometrics
(pp. 541-
591). Singapore: McGraw-Hill.
Greene, W. H. (2003). Chapter 21. In W. H. Greene,
Econometric Analysis
(pp. 719-729). New Jersey:
Prentice Hall.
Government of Mongolia, UNDP, (2007).
Mongolia
Human Development Report 2007: Employment and
Poverty.
Ulaanbaatar: Admon LLC
J.Scot Long, J. F. (2001).
Regression Models for
Categorical Dependent Variables Using Stata.
Texas: A
Stata Press Publication.
Long, J. S. (1997).
Regression Models for Categorical
and Limited Dependent Variables.
California: Sage
Publications, Inc.
NSO, 2000-2010.
Mongolian Statistical Yearbook.
(2000-2009). Ulaanbaatar: National Statictical Office of
Mongolia, Ulaanbaatar.
Statistical Office of Ulaanbaatar City.
(n.d.). Retrieved
2010 оны 12 22 from Website of Statistical Office of
Ulaanbaatar City:
http://www.statis.ub.gov.mn/Wikipedia.
(n.d.). Retrieved 12 22, 2010, from
Wikipedia, the free encyclopedia:
http://en.wikipedia.
org/wiki/Zud
Wooldridge, J. M. (2009). Chapter 17, Chapter 19. In J.
M. Wooldridge,
Introductory Econometrics: A Modern
Approach
(pp. 575-612, 668-687). South-Western
Cengage Learning.