1. Coursework 2
   This coursework will continue the analysis of the data set relating to 74 European countries which you started in coursework one. A copy of coursework 1 with a full description of the variables is attached. Your marked solution to coursework 1 must be handed in along with your solution to coursework 2 as the two pieces of coursework together form the full analysis of the data. Your marks for the first coursework will not be changed in any way but the second coursework may refer to results in the first coursework. (If you prefer, you may hand in a photocopy of your marked solution to coursework 1).
   Requirements
   This assignment is about multiple linear regression. Your aim is to use correlation and regression techniques to determine how the Index variable depends on all the company characteristics. There will be no single correct answer – you may be able to find equally good models involving different variables. Data transformations may be necessary for some of the explanatory variables but the response (Index) will not need to be transformed.
   Aims
   You will be assessed on the approach you take to the data and not on the goodness of fit of any regression model you propose. Your main aims are:
2. To analyse the data to obtain a multiple regression model expressing the response variable in terms of the other variables.
3. To attempt to simplify this model by reducing the number of explanatory variables.
4. To carry out appropriate checks on your final model (using residual plots and related techniques). These may suggest that you need to reconsider your initial model.
5. To summarise and interpret your final model. This will involve a plot of observed values against fitted (predicted) values, and a discussion of the values of the coefficients in your equation.
   Method
   This is largely down to you but you will need to do at least the following:
6. Explore the data using scatter plots (or a scatter plot matrix). You should also look at summary statistics, including correlations, of the variables. Consider transforming the variables to make the relationships more linear where this seems possible and appropriate. The work you completed for coursework 1 will have begun to explore the data.
7. Create a dummy variable representation of the variable TYPE.
8. Carry out a multiple regression including all of the explanatory variables.
9. Attempt to remove non-significant explanatory variables from your model. You may be able to use an automatic procedure, such as backwards, forwards or best subsets regression, to do this.
10. Carry out a residual analysis and change your model if necessary.
11. Calculate the fitted (or predicted) values and plot the observed values of the response against the fitted.
12. Look at any unusual observations.
13. Consider how to interpret your chosen model.
    Presentation
    Your results should be presented in the form of a report. This should contain:
14. An introduction explaining the objectives of the study. A detailed introduction to the background of the study and the data is not necessary as you did this in coursework 1, which will be attached. If you wish to expand on the introduction of coursework 1 you may do so.
15. A description of the data- definitions of the variables, summary statistics, plots etc. Again you may refer to coursework 1 for any work which is relevant there. It does not need to be repeated but you may wish to expand on it.
16. An explanation of the methods you have chosen to analyse the data supported by your exploratory data analysis.
17. The results of your analysis. This must include your final regression equation, with standard errors, R ² and s , together with plots of observed against fitted values and residuals against fitted values.
18. A discussion including an interpretation of your model, whether it could be improved, any problems with the data and how you overcame these, any recommendations for future work.
19. Conclusion – a summary of what you have achieved.
20. An appendix including additional Minitab output. (This should be edited to remove any mistakes or false starts)
21. A reflection on how the feedback from coursework 1 influenced your report for coursework 2.
    Marking scheme
    A guide to the way that marks will be allocated is as follows:
22. Exploratory analysis of the data, related to choice and explanation of method.
    [20% of marks]
23. Model selection and results – the main regression analysis, which should involve more than 1 fitted model.
    [45% of marks]
24. Interpretation of your preferred model, discussion and conclusions.

[35% of marks]

Price: $31.69

Solution: The downloadable solution consists of 14 pages, 1769 words and 13 charts.
Deliverable: Word Document