Question one:

Consider the following model that relates the percentage of a household's budget spent on milk MK to total expenditure TOTEXP, age of the household head AGE, and the number of children in the household NK.

MK = +

This model was estimated using 1400 observations from Bahrain. An incomplete version of this output is provided in the following table.

Regressor

Coefficient

Standard Error

T-Ratio

C

1.4515

2.2019

0.212

2.7648

0.4842

2.206

-1.4541

0.6395

-0.2141

-0.1503

0.0235

-5.5214

0.35

Instructions:

You suspect that random error term is heteroscedastic with a standard deviation that depends on the size of the population. Describe step by step how will go testing for this. Be sure to state (1) the null and alternative hypotheses, (2) the test statistic you will compute and its distribution and (3) the criterion for acceptance or rejection of the null hypothesis.

Question Two:

Consider the following model estimated by OLS with annual data for the years 1963 - 1985 for the United States as a whole:

= -4.759 + 1.873LG - 1.229LLR

(-1.4)(3.8)(-4.0)

= 0.386n= 23DW d= 0.794

Where LH is the logarithm of housing starts, LG is the logarithm of the gross national product, and LR is the logarithm of the mortgage rate. The values in parentheses are t-statistics.

Instructions:

Test the model for first-order serial correlation under the alternative most common in economics. Be sure to state the null and alternative hypotheses. Based on your conclusions, list the properties of the estimates and associated statistics and the validity of the tests of hypotheses.