Assignment 2:
The data set provided below is 1,000 observations for the model
Y=alpha+beta X + epsilon.
Population parameters are alpha=1 and beta =3. Epsilon is distributed normal with zero mean and unitary variance.
1- Split the data into sizes of 5 observations (first 5 observations of (Y,X), next 5 observations of (Y,X), etc.) for a total of 200 different data sets. Calculate the beta hat of these data sets and plot their sampling distribution. Is your estimator biased/unbiased? Calculate the degree of bias by bias = average (beta hats) - beta. What is the mean squared error (MSE)? Calculate the MSE by MSE = var(beta hat) + bias squared (beta hat).
2- Repeat Question (1) by splitting the data into sizes of 20 observations for a total of 50 different data sets.
3- Repeat Question (1) by splitting the data into sizes of 40 observations for a total of 25 different data sets.
4- Repeat Question (1) by splitting the data into sizes of 50 observations for a total of 20 different data sets.
5- Illustrate consistency of beta hat by using your results from (1)-(4) graphically.
You can use Eviews, Stata, Excel or R as software.
Document Preview:
Assignment 2: The data set provided below is 1,000 observations for the model Y=alpha+beta X + epsilon. Population parameters are alpha=1 and beta =3. Epsilon is distributed normal with zero mean and unitary variance. 1- Split the data into sizes of 5 observations (first 5 observations of (Y,X), next 5 observations of (Y,X), etc.) for a total of 200 different data sets. Calculate the beta hat of these data sets and plot their sampling distribution. Is your estimator biased/unbiased? Calculate the degree of bias by bias = average (beta hats) - beta. What is the mean squared error (MSE)? Calculate the MSE by MSE = var(beta hat) + bias squared (beta hat). 2- Repeat Question (1) by splitting the data into sizes of 20 observations for a total of 50 different data sets. 3- Repeat Question (1) by splitting the data into sizes of 40 observations for a total of 25 different data sets. 4- Repeat Question (1) by splitting the data into sizes of 50 observations for a total of 20 different data sets. 5- Illustrate consistency of beta hat by using your results from (1)-(4) graphically. You can use Eviews, Stata, Excel or R as software. The data set is: AS2data.csv ata, Excel or R as software. The data set is: AS2data.csv del Y=alpha+beta X + epsilon. Population parameters are alpha=1 and beta =3. Epsilon is distributed normal with zero mean and unitary variance. 1- Split the data into sizes of 5 observations (first 5 observations of (Y,X), next 5 observations of (Y,X), etc.) for a total of 200 different data sets. Calculate the beta hat of these data sets and plot their sampling distribution. Is your estimator biased/unbiased? Calculate the degree of bias by bias = average (beta hats) - beta. What is the mean squared error (MSE)? Calculate the MSE by MSE = var(beta hat) + bias squared (beta hat). 2- Repeat Question (1) by splitting the data into sizes of 20 observations...