The Empirical Project
You are required to carry out an individual empirical project to demonstrate your ability to use the statistical techniques taught in this course in a practical context. You will be expected to collect data, analyse the data and write up a report about this project.
The report should not be more than 8 pages long (not including the title page, bibliography and up to two pages of tables and/or graphs) with word limits of 3000-3500. The report must be word-processed and double-spaced using 12 point font and margins of at least an inch at top/bottom/left/right.
Presentation of your report
Your empirical report should be well organised in terms of clarity and brevity. You may have many different coefficient results and test statistics from your various regression models. But you should decide what information is important and relevant to the issue under investigation and include only the most important and relevant information in the text and report honestly the results that you obtain (even if your results go against your theory).
When reporting summary statistics and estimated results, you should tabulate them in the text and leave the computer print-outs in the appendix if necessary. If you attempt graphical inspection on the data, you should present the graphs in the text.
General guidelines on writing an empirical project
- Introduction. You should begin this report with an introduction that briefly motivates and describes the issue being studied and summarizes the main empirical findings. The introduction should be written in simple non-technical language, with statistical and economic/financial jargon kept to a minimum. A reader who is not an expert in the field should be able to read and understand the general issues and findings of the report or paper. [For this course this can be very brief, 1 page at most] (15 marks)
- Literature Review. This should summarize related work that others have done in terms of their methodology and findings. Note that you should cite the sources of information in the text as well as listing the sources in the references. (10 marks)
- Economic/financial Theory. If your empirical model is derived from a formal theoretical model, then the theory is often described in this section. This section can be more technical than the preceding ones and will typically include some mathematics and economic/financial terminology. In short, you can address this section solely to an audience of experts in your field. (10 marks)
- Data and methodology. Present your empirical model derived from the previous section(s) in this section. Then you should describe your data and discuss how you use the data to estimate the model and carry out diagnostic checking. It is in this section that you should justify the techniques used in the next section. (10 marks)
- Empirical Results. Present your results in this section. You should summarise and tabulate the estimates or statistics relevant to the issue under study and leave the computer printouts in the appendix. That is, tables should be used to present results from a more formal statistical analysis, such as coefficient estimates, together with t-statistics (or P-values), R2s and F-statistics for tests. At this stage you should describe your empirical findings and discuss how they relate to the financial, business or economic issue(s) under investigation. It should contain both statistical and contextual information. By “contextual” information we refer to coefficient estimates and what these estimates may imply for economic/financial theory. In contrast, “statistical” information may include: results from hypothesis tests that show how coefficient estimates are significant; an explanation for the preferred model; a discussion of model fit (e.g. the R2 ); etc. [This is the most important part of the project, perhaps 3-4 pages long] 50 marks (10 marks for presentation of results, 10 marks for estimation, 10 marks for hypothesis testing, 20 marks for interpretation and discussion)
- Conclusion. This should briefly summarize the issues addressed in the paper, specifically, its most important empirical findings. [For this module, this can be brief, probably less than 1 page] 5 marks