代做POLI620041留学生、代写OpenOffice、代写C++，Java，Python编程设计

日期：2019-12-31 07:19

POLI620041

Homework Assignment 2

Fall 2018

Due on Dec.23 , 2019 (Monday)

Important Notes: Type your answers using either OpenOffice, MS Word, or LaTex unless

otherwise instructed. Your work will also be graded on professional appearance, format,

and language choice. If you have any question, feel free to ask your instructor for further

instructions. You can write in Chinese or English.

Bivariate & Multivariate Models (40’)

In the following, you are asked to practice constructing and interpreting bivariate and mul?tivariate models using the data USAid.dta attached. You can run the models in Stata (or

R) .

Suppose we are interested in understanding what factors determine U.S. foreign aid

allocation decisions in 2002. The unit of analysis here is a country. The data set include the

following variables from 2012:

Aid The amount of US aid to the country (Continuous).

AidDummy 1 = Receive US aid, 0 = Does not receive US aid (Dummy)

Dem 1 = Democracy, 0 = Non-Democracy (Dummy)

HR Human rights score (categorical and ordinal)

GDP GDP of the country (in US dollars) (Continuous)

1. If you want to know whether U.S. foreign aid allocation decisions depend on the recip?ients’ regime types or not,i.e., the relationship between AidDummy and Dem, , what

kind of bivariate analysis can you apply? Please implement the model in Stata and

interpret the results. (10’)

2. If you want to know whether U.S. foreign aid allocation decisions depend on the re?cipients’ economy or not , i.e., the relationship between AidDummy and GDP, what

kind of bivariate analysis can you apply? Please implement the model in Stata and

interpret the results. (5’)

1

3. If you want to know whether the amount of U.S. foreign aid correlates the recipients’

economy or not ,i.e., the correlation between Aid and GDP, what kind of test can you

apply? Please implement the model in Stata and interpret the results. (5’)

4. In order to find out what affects U.S. foreign aid allocation, we construct the following

models. Please, firstly, explain what each model aims to do, and then implement the

models in Stata and interpret the results.

Model 1: Aiddi = α (5’)

Model 2: Aiddi = α+βGDPi (5’)

Model 3: Aiddi = α+βDemi (5’)

Model 4: Aiddi = α+β1GDPi +β2Demi (5’)

Regression Diagnostics (60’)

In the following, you are asked to apply the techniques that we have learned about regression

diagnostics into a particular case. Suppose [Model] Aiddi = α + β1GDPi + β2Demi+β3HRi

is

the model that we pick up for test. You are not only asked to conduct tests or create related

graphs. Most importantly, please tell me what conclusions you can draw from the tests/

graphs.

1. Please calculate the student residual, leverage, the cook distance, dfBeta for each

observation. Based on these indexes (you can visualize these indexes as well), can you

detect any unusual observations and influential observations? (10’)

2. Please check the residuals and see whether the assumption of normality of residuals

holds or not, and whether the mean of the residual is zero or not. (10’)

3. Please check whether the assumption of homoscedasticity of residuals holds or not.

(10’)

4. Please check whether there exists multicollinearity among independent variables or

not. (10’)

5. Please check the linearity assumption for the independent variable, GDP, holds or not?

(10’)

6. Please conduct at least a test to examine the possibility of the existence of omitted

variable. (10’)

2