The report will be 2400 words long. It will be based on your analysis of a dataset, combining technical output with interpretations and assessments of the results you obtain.
Identify a simple hypothesis that need to be tested with the data. In the theory section, describe why you think this particular question is interesting, what it is and what is going to be tested. Then, do descriptive analysis of the data (creating tables that describe the variables in the dataset
Content
The aim of working on this report is to get you to use the methods and code that was taught in this module in the context of an analysis which you yourself have framed and conducted with data I have made available for the module.
The methods you need to work on the report have all been covered in the module already!
The report you write should focus on a question drawn from the paper by Bloom et al. paper (You do not have to repeat the analysis undertaken in the paper at all.). The data for the analysis is available through the course pages.
Your report should contain the following sections:
1.) Introduction – introducing the analysis, framing it and summarising main results and the structure of the coursework text.
2.) Question – section to set out the problem you are going to analyse and explaining its relevance in the context of the literature and analysis that has been done on the question previously. While you can draw on the literature cited in the paper you are building on it will be important to expand on that.
3.) Hypotheses and Methods – section to set out the way in which the question set out in the previous section will be tested using experimental data. This section should provide some discussion of the merits of the experimental approach being adopted. It should also contain a discussion of the hypotheses you are testing to answer the question you have set yourself.
4.) Data Description – this should contain a graphical and tabular description of the data you will use. You should carefully discuss any implications that you think flow from the data description.
5.) Data Analysis – this should contain results of analysis based on statistical models. You should present model output and your assessment of the findings you present.
6.) Conclusion
Each section of the coursework should contain up to 600 words, which is about 1.5 pages of pure text.
You may wish to make sections 1. & 6. a bit shorter and others longer.
The quality of tables and graphs, as well as their discussion will form part of the marking criteria.
How to tackle this task
You do not have to repeat the analysis undertaken in the paper at all. Anything you can think of that the data may reveal is possible. Sometimes it can be useful to report results duplicating those in the paper as a starting point. Doing only this is not good enough to get a pass grade, as we have covered almost everything in the classes and you would just be replicating the material from the classes.
You should bear in mind that the data you are using have been generated for a specific purpose and some questions you might like to work on cannot be answered in quite the same way as the main questions from the paper. For instance, the working from home experiment is designed to reveal differences between those working from home and those working in the office. You can split those groups up further and still extract useful results: e.g. you can compare the change working from home brings for men to that for women. A direct comparison between men and women working from home will not reveal that much though. The experiment was not set up to randomly assign sex for those working from home!
