• Join over 1.2 million students every month
  • Accelerate your learning by 29%
  • Unlimited access from just £6.99 per month

Linear regressions.

Extracts from this document...


Linear regressions

Problem 1

You have estimated the linear regression model

yt = a + b1x1t + b2x2t + b3x3t + et

using annual data for the period 1960-94. Explain briefly how you would construct a test of the model’s forecasting performance using additional data for the period 1995-98.


First, estimate the model and obtain estimations for the coefficients a, b1, b2, b3 using 1960-94 data.

Then, obtain forecast for the 1995-98 based on estimated model together with errors of forecast. So, the intervals for the values for the 1995-98 will be obtained.

After this one can check whether actual data fit into the obtained intervals or not. Also, one can check how far are the actual values from the forecast.

Problem 2

Describe briefly how you would test whether the OLS residuals from the linear regression model

Yt = a + bXt + ut

are serially correlated. Outline how you would modify the specification of your model, or

the estimation procedure, if your test revealed showed significant serial correlation.


One may estimate the initial regression and obtain regression residuals:


then one should estimate the regression of et on its lag:


If the coefficient ρ appeared to be significant, then there is serial correlation in residuals.

...read more.


Data period 1991Q1-1997Q4 (28 observations) SSR = 0.0010334

(iii) Test the hypothesis that the coefficients of Model A are constant over the period 1979-1997 against the alternative that there is a structural break after 1990Q4.


Let us call constant in the regression as β1, coefficients on D4Ct-1 - D4Ct-4as β25, coefficients on D4Yt-1 - D4Yt-4  as β69 correspondingly.


The test is simply a Fisher test i.e. β23=…=β9=0.

One has to calculate F-statistic (according to model A):


The critical value for F-statistic is F0.95(k-1,n-k)=F0.95(8,56)=2.1

Since 35>2.1 then null hypothesis on all the coefficient are zeroes has to be rejected, i.e. hypothesis that the four quarter change in the logarithm of consumption is unaffected by any lagged variables is rejected.

An economist is interested in this hypothesis to check whether consumption in current period is affected by previous periods or is determined by current income only.


The tested hypothesis is H0:  β6=…=β9=0

To test the hypothesis one has to calculate Fisher statistic (model A is unrestricted regression, model B is restricted regression i.e. regression in which β6=…=β9=0):


Critical value for the F-statistic is F0.95(q,n-k)=F0.95(4,56)=2.536

Since 1,69<2.526 then  the hypothesis is accepted


...read more.


(a) A fellow-student, B, suggests to A that he can increase the size of his dataset, while avoiding the effort involved in collecting more data, just by duplicating each of his existing observations the appropriate number of times. Explain why a dataset which is extended in this way fails to satisfy the assumptions of the Gauss-Markov theorem, and identify the assumption which will not hold.

(b) Suppose A follows B’s advice, and constructs an extended dataset which consists of a number of copies of observation 1, followed by a number of copies of observation 2, and so on. Explain which standard diagnostic test is likely to reveal his deception.

<Think how residuals may reveal the problem>



The dataset obtained in this way fails to satisfy the assumptions of the Gauss-Markov theorem since errors et for duplicated observations are the same, so they are not independent, as required by Gauss-Markov Theorem. To be more precise, the condition that E[εtεs]=0 when t ≠s is not satisfied for s and t are numbers of the pair of duplicated observations

(b) In this case errors are correlated with order 1 (and probably there is higher order correlation). So, the deception is to be revealed by Darbin-Watson test which will show high positive autocorrelation in errors (i.e. DW will be close to 4).

...read more.

This student written piece of work is one of many that can be found in our AS and A Level Probability & Statistics section.

Found what you're looking for?

  • Start learning 29% faster today
  • 150,000+ documents available
  • Just £6.99 a month

Not the one? Search for your essay title...
  • Join over 1.2 million students every month
  • Accelerate your learning by 29%
  • Unlimited access from just £6.99 per month

See related essaysSee related essays

Related AS and A Level Probability & Statistics essays

  1. I am going to design and then carry out an experiment to test people's ...

    11 46 13 9 13.5 13 11 47 15.5 15 14.5 16.5 11 48 15 17 15 15 11 49 15 13.5 7 12 11 50 7 5 16 10 11 51 15 4 8 5 11 52 15 21 18.5 18 11 53 8 7 11 10 11 54

  2. Study of the height/diameter ratio of limpets inhabiting the middle shore region of exposed ...

    would also be existent, but low in number. A normal distribution for this shore would look like this: The sheltered shore sustains limpets with a higher ratio (increased tallness, reduced diameter, 0.589 being the mean). This suggests that the limpets do not need to compromise their tallness and the size of the foot does not have to be large to survive.

  1. Driving test

    the better you become at it hence the saying practice makes perfect. Therefore I anticipate that there is a negative correlation between number of lessons and number of mistakes. To investigate this hypothesis, I will draw a scatter graph for the number of mistakes against the number of lessons Scatter

  2. Statistics: Survey of Beijing and China during the SARS storm

    Population Sample 18 MIN 18 MIN 39 43 41 62 43 64 43 75 53 76 55 77 61 83 LQ 62 83 63 84 64 84 75 87 76 99 77 99 83 103 LQ 83 104 MEDIAN 83 106 84 109 84 112 84 117 87 117 91

  1. Chebyshevs Theorem and The Empirical Rule

    For several years Dr. Gill has kept a record of the total number of work hours volunteered by s student in the program each semester. For students in the program, for each semester the mean number of hours was 29.1 hours with a standard deviation of 1.7 hours.

  2. &amp;quot;The lengths of lines are easier to guess than angles. Also, that year 11's ...

    For the year 11 estimates for the length of line 2, the deviation of the estimates was 1.07, which means that the year 9 estimates for the length of line 2 were less spread out than the year 11's. For the year 9 estimates of the angle, the spread from

  1. Comparative Newspaper Project

    are more consistent in length; which is further evidenced in the diagram below. Box and Whisker Plots: (using medians, upper quartiles and lower quartiles calculated from separate stem and leaf diagrams) Telegraph: Sun: Stem and Leaf Diagrams: See separate sheet - Stem and Leaf Diagrams These diagrams, again, show the

  2. Estimating the length of a line and the size of an angle.

    easy method it will not be suitable for choosing a sample from a population such as this. In my opinion stratified sampling would be better than systematic sampling. I have chosen stratified sampling over convenience sampling. This is because convenience sampling could be biased because of the way in which the sample is collected.

  • Over 160,000 pieces
    of student written work
  • Annotated by
    experienced teachers
  • Ideas and feedback to
    improve your own work