Details
Imagine a clinical trial for a new drug to treat male pattern baldness. Unlike a real clinical trial where initial efficacy against a placebo and effectiveness compared to existing commercial rivals would be considered separately (as phase II and phase III clinical trials respectively), here we will suppose that the new drug (Luxuriant) is considered in the same trial with both a placebo and two existing anti-baldness treatments BaldBeGone and SkinheadNoMore. The patients, all of which were more or less totally bald, were randomly allocated to the groups, shaved and then had any hair growth measured after one month of treatment.
The first four columns represents the hair growth results for Luxuriant, the placebo, BaldBeGone and SkinheadNoMore respectively. The units are in inches but the company want all results to be discussed in mm. The 5-8th columns represent the age of the patients for each treatment group respectively.
The questions that are of interest to the company are
1. is there an effect of Luxuriant above and beyond the placebo
2. Is Luxuriant more effective than the existing treatments on the market
3. is age relevant to any effect?
Instructions
1. Manipulate the file in SAS to render it amenable to analysis. (Hint: you can do it using the methods in the notes or given in the sessions but there may be alternatives)
2. You can include SAS code in your report by enclosing it within three backticks. For example:
“`{r, engine =”sashtm”l}
PROC EXPORT DATA=work.plantdata
DBMS=csv
OUTFILE=reffile;
RUN;
“`
will render as:
PROC EXPORT DATA=work.plantdata
DBMS=csv
OUTFILE=reffile;
RUN;
Although this does not produce a fully reproducible document (as your SAS code is being run externally), it’s good enough for the purpose of this assignment. You can also include figures generated by SAS, using the following knitr function:
knitr::include_graphics(“path/to/your/image”)
3. Undertake an appropriate analysis.
4. Write a technical internal report for the makers of Luxuriant explaining the results of the trial.
This should be written in R Markdown with graphs/tables as necessary from SAS. A small proportion of marks might be available for suggesting any improvements to the protocol in the case of future trials! There should be an appendix containing relevant code. Make sure everything is reproducible i.e. include tidy, commented code files that would allow someone to reproduce the results with effectively one copy and paste.
The report should be limited to 1200 words (this is an absolute maximum not a desired target, it should be possible to write less) and should contain relevant summary plots – however you can put extra output in an appendix at the end (if necessary). Include an introductory section describing the purpose of your investigation and a brief discussion at the end summarising your findings with suggestions for further research. Include an executive summary/abstract at the start of the report.
The report should:
· be ‘professionally’ produced for your target audience in R markdown.
· your choice of analysis should be justified.
· include references in an established academic style e.g. APA.
· be converted to a .pdf file.
Marking Guidance
· Code – readable, logical, reproducible, tidy and appropriately commented in an Appendix 30%
· Report – broken into the following sections:
· Abstract: 5%
· Introduction: 5%
· Analysis description i.e. a clear account of what you did statistically (not the data wrangling, the managers will not care about that and it will be clear from your code) and why: 20%
· Results: are the results clearly explained and presented (includes data visualisation): 30%
· Discussion/conclusions: conclusions along with criticisms of the design if any: 10%