Location via proxy:   [ UP ]  
[Report a bug]   [Manage cookies]                
Scribd
Upload
35 views

Practical Lab 2

This document outlines four problems related to multiple regression analysis for a practical lab assignment. Problem 1 involves estimating a demand function for pork and testing hypotheses about elasticities. Problem 2 uses dummy variables to model consumption with data from the US from 1929-1970, during and after World War II and the Vietnam War. Problem 3 examines multicollinearity in a model using income and assets to predict consumption. Problem 4 tests for heteroscedasticity in a model predicting debt from GDP using country-level data, and suggests remedies.

Uploaded by

crazyfrog1991
Copyright
© Attribution Non-Commercial (BY-NC)
Available Formats
Download as DOC, PDF, TXT or read online on Scribd
35 views

Practical Lab 2

This document outlines four problems related to multiple regression analysis for a practical lab assignment. Problem 1 involves estimating a demand function for pork and testing hypotheses about elasticities. Problem 2 uses dummy variables to model consumption with data from the US from 1929-1970, during and after World War II and the Vietnam War. Problem 3 examines multicollinearity in a model using income and assets to predict consumption. Problem 4 tests for heteroscedasticity in a model predicting debt from GDP using country-level data, and suggests remedies.

Uploaded by

crazyfrog1991
Copyright
© Attribution Non-Commercial (BY-NC)
Available Formats
Download as DOC, PDF, TXT or read online on Scribd
You are on page 1/ 1

Practical lab 2

Problem 1: Review of multiple regression


The data set in file CH3_BT5 consists of variables: q (demand for pork, 10kg/person), Y (Income per capital USD) and P (price of pork) 1. Estimate the demand function for pork (Using logarithm form function). What is the elasticity of demand w.r.t income and price. Explain the meaning of estimated coefficients. 2. Test the hypothesis that elasticity w.r.t price is -1 3. Test the hypothesis that elasticity w.r.t price is -1 and w.r.t income is 1 4. When both price and income increase 1%, how much the demand for pork change. 5. Test that if both price and income increase 1%, the demand for pork will not change. 6. What are the confidence intervals for elasticity of demand for pork

Problem 2: Regression on Dummy variables


Data in file CH4BT1 contains the information on consumption (CS) and Income (Y) from 1929 to 1970 in US. There are two events happening within this period: World war II (41-46) and Vietnam war (60-70). D 1 is the dummy (equals to 1 in the period 41-46), D2 is dummy (equals to 1 in the period 60-70) 1. Produce scatter plots of CS on Y for three period and for the whole data and compare 2. Estimate consumption function for the whole period and for the period affected by WWII 3. Estimate model: CS= b1+b2Y+b3D1+b4(D1Y)+u and explain the results 4. Estimate model: lnCS= b1+b2 lnY+b3D1+u and explain the results 5. Estimate model: CS= b1+b2Y+b3D1+ b4D2+u - Did Vietnam war have effect on consumption - Did World war II have effect on consumption - Did both wars have effect on consumption

Problem 3: Multicollinearity
Data set CH5BT4 includes variables Y (family consumption), X2 ( Income) and X3 (transferable assets) 1. Compute the correlation amongst variables and comment 2. Estimate the linear model of Y on X2 and X3 3. Based on the results in 2, comment on the multicollinearity in model in 2 4. Estimate model between independent variables and make a test for multicollinearity 5. Suggest a remedial measure, Which variable we should drop out?

Problem 4: Heteroscedasticity
Workfile CH6BT3 contains two variables on debt (D88) and GDP (Y88), both variables are recorded by USD and converted to year of 1988. Data is for 73 developing countries by World Development Report. Use the data set, do the followings: 1. Estimate the model of D88 on Y88 and explain the results. Record residuals with name "e" and fitted values with name "fv". Produce scatter plot of squares of residuals on Y88. 2. Conduct Park test to detect if model has heteroscedasticity 3. Conduct Gleijer test to detect if model has heteroscedasticity 4. Conduct White test to detect if model has heteroscedasticity 5. Using the log form model to reestimate the model and do the test for heteroscedasticity for the new model 6. Assume that in the first model, the variance of disturbances is proportional to squares of independent variable. Suggest the method to remedy the heteroscedasticity and estimate the new model 7. Similar to question 6 but now assume that the variance of disturbances is proportional to squares of fitted values.

You might also like