Assignment Task

Objectives of the Assessment

The objective of this assessment is to carry out a simple, desk-based research project relating to the rate of childhood mortality in England in 1851.

You are expected to define a research question with a clear and answerable hypothesis e.g. ‘To what extent does the number of unmarried women affect infant mortality’ or ‘To what extent is the number of doctors a predictor of infant mortality?’ and to think of other variables that you may wish to use as controls.

Your report should be structured in standard report format and include:

• Introduction (including your literature review) This should provide the context for your work based on existing literature. It should identify some possible factors that impact on your dependent variable and should culminate in a research question and/or testable hypotheses.
• Methods This section should be written in prose and not in bullet points or numbered lists. It should provide a succinct description of the methods you used to carry out your analysis. This section can include interim results that helped you determine things like whether or not to transform variables or whether or not to include them in a final model.
• Results This section should contain a detailed description of the results of your final model and the diagnostics you have carried out to determine the effectiveness/reliability of the final model.
• Discussion This section should see you comparing your results to the wider literature and discussing its importance, reliability, similarity to existing literature and it’s limitations.
• Conclusion This needs only to be a couple of small paragraphs that summarises your results with respect to your research question/hypotheses.

At a minimum, your report should provide:

• A description of your dependent variable – including a map
• A review of how independent variables were selected from those available to you within the provided data
• A discussion of any data preparation that was required prior to analysis – e.g. transforming, centering, categorizing
• Informed by the outcomes of the above, a linear regression model should be fitted and a discussion of it’s validity provided that includes discussion of:
  • Coefficient values,
  • R-squared value,
  • Statistical significance,
  • Residuals
  • Outliers

Further modelling should then be carried out to create a minimum adequate model that explains as much of the variation in your dependent variable as possible using only variables that add value to the model.

It is important to note that there will not be one final ‘correct’ model, although there are better and worse ones. The data you are working with is imperfect, and so you must consider that when discussing the model outputs and their meaning.

You are not expected to use every technique you have been taught in this unit. You will be marked on your ability to select appropriate methods and apply them in a meaningful way.

• You can use a maximum of 10 figures and a maximum of 5 tables in this report. You can make use of multi-panel figures (e.g. multiple figures on one page) as long as they are properly aligned, fit onto one page and can be held under one caption similarly to the below. Please also note that incorrect table and figure captions are a pet hate of mine!!!!! Figures should be labelled as figures and with the label and caption belowthe figure. Tables should be labelled as tables and with the label and caption above the table. You were taught this in Y1 (BY ME) so please do this correctly!!!!! It is a terrible pet hate of mine when people get this wrong!
• You should not simply copy and paste the outputs from the R console into your document. They are ugly and largely meaningless so instead you should be placing important values into a suitably designed table to show your reader what you want them to see.
• A discussion board will be set up for questions relating to the assessment. The discussion board should be your first stop for any questions you have about this assessment. If your question has yet to be asked you should post it as a question on the discussion board and a member of the team will answer it. Emailing questions directly to the teaching staff should be a last resort and if the question has already been answered on the discussion board, or if it is a question that should be asked on the discussion board, then you may be redirected to there in the first instance. When we (staff) are learning a new technique or trying to troubleshoot problems we always begin with asking peers through online forums such as stack overflow, so getting used to asking questions of your peer group will help to prepare you for your future career paths.