Question

Suppose we estimate a regression model containing a constant term and two explanatory variables. The analysis is based on a sample size of 25 and produces a Durbin Watson statistic d=1.85.

a. Approximately, what is the correlation coefficient between consecutive error terms?

b. If this model has an R2 = .75, what is the value of the CALCULATED F statistic associated with a GLOBAL test?

c. At the 1% level what is the CRITICAL value associated with a global test of the model?

Answer 1

Here Durbin Watson statistic: d=1.85

Now we know that

where is the correlation coefficient

So 1.85 =

2) The relation between R^2 and F statistic is

Here we have two treatments

so df1 = 2-1 =1

we have 25 observations

so total degrees of freedom = 25 -1 = 24

and error degrees of freedom = df2 = 24-1 = 23

R2 = .75

So

F = 69

So  CALCULATED F statistic = 69

3) At 1% level of significance critical value = 7.881 for df1=1 and df2 = 23

Please give thumbs up if you like the answer.