Medical Statistics Coursework Assessment

Assessment Requirements

The assessment consists of three equally weighted parts. Grade based assessment will be used to assign an overall mark your work. The accompanying assessment grid gives detail for each section to help you plan your work. While feedback will be given for each section, your grade will be based on an overall view of the coursework, using the weightings shown. Submit your work as a single document Word file to the module dropbox.

Where you use equations you should make use of Microsoft equation editor within word. This coursework is an ‘element of assessment’ which means you need to pass it in order to pass the module. Some seminar time is allocated for you to work on this report. All work will submitted to Turnitin as a means to check for plagiarism. Plagiarism will not be tolerated; students are reminded that this is an individual assignment. You should have covered plagiarism in your tutorial sessions.

Using a one page word document, plan out a clinical trial for a family of three new drugs designed to delay death by slowing the aging process. Consider what selection mechanisms might be involved, what blinding techniques you might require and what power, significance and so on you might use. Detail what tests you will carry out and when. Comment on what trial design you will use (parallel, series, factorial etc.) and why you chose that design over the others. When you use values, justify why they were chosen. (10)

Use Minitab to generate 3 columns of integer (Uniformly) random numbers between 1 and 22. Each column should be 22 rows long. Generate another column (of the same length) of normally distributed integers with a mean of 12 and a standard deviation of 2. Treat this data as lifetime data from people’s 60^th birthday. A value of 7 means that the person died before reaching age 68 (so they survived for 7 years of the study). Put your data in order from lowest to highest. Consider your 11^th and 20^th values for each group to be right censored. 

Using the normally distributed data as your placebo group and the other three columns as your drug groups display the data graphically and comment on the key features of the groups. (3)

Create a survival table for the data and display each group as a Kaplan Meier plot. Generate the Log Rank test for this (as in the lecture). Comment briefly on the results. (7)

1. Write a two page essay detailing one aspect of medical statistics which we have mentioned but not covered in detail. Your essay should be referenced but should not include quotes. Suitable topics include, but are not limited to:

a The ethics of Clinical trial design (and the use of placebos)

b Publication bias (and meta analysis)

c Use of Animals in clinical trials (and possible alternatives)

d Multiple testing approaches (and big data)

e Bayes nets (and medical diagnosis)

For maximum marks, aim to explain in detail one aspect of your topic rather than giving a general overview of a broad area. Your essay should include a brief Mathematical and/or Statistical example relevant to your topic. Marking and feedback comments for the essay will consider it as a section of a dissertation.  This essay will be checked very thoroughly for plagiarism; students are reminded that this is an individual assignment. (10)

Marks listed above are for weighting purposes only – marking will be using Grade Based Assessment using the marking grid shown on now. Returned work will be given a grade rather than a numerical mark.