Basic Biostatistics, fall 2016: Exam
Exam
A successful completion of the course requires a satisfactory
individual solution to the problem posed after the final exercise (Thursday November 3, 2016, at 4 PM),
reported in a single pdffile suitably marked with your name. The solutions should handed in
before Friday November 25, 2016, at 9 AM.
Examples:
Spring 2014: Problem formulation: Spring2014.pdf.
The data files: iq.dta, iq.sav, iq.xls.
A short solution: Spring2014solution.pdf
Fall 2014: Problem formulation: Fall2014.pdf.
The data files: tatsoib2.dta, tatsoib2.sav, tatsoib2.xls, Old Stata version:balanceold.dta.
A short solution: Fall2014solution.pdf
Spring 2015: Problem formulation: Spring2015.pdf.
The data files: balance.dta, balance.sav, balance.xls, Old Stata version:balanceold.dta.
A short solution: Spring2015solution.pdf
Fall 2015: Problem formulation: Fall2015.pdf.
The data files: bweight.dta, bweight.sav, bweight.xls.
A short solution: Fall2015solution.pdf
Spring 2016: Problem formulation: Spring2016.pdf.
The data files: faroe.dta, faroe.sav,
faroe.xls.
A short solution: Spring2016solution.pdf
No other electronic file format will be accepted. That your solution is individual means that your solution
cannot be a copy of a solution made by someone else (not even partially), nor can it be written under
supervision by someone else. While you are allowed to discuss the exam problem with your peers, it is
mandatory that the solution is the result of your own work, and that you have actually written the solution.
If doubts on this aspect are raised and they turn out to be justified, your submitted solution will not count
towards passing the exam.
Your solution must arrive no later than Friday November 25, 2016, at 9 AM., at:
BBEKSAMEN@BIOSTAT.AU.DK
to be considered. When you submit by email, you will automatically receive a confirmation email 
if you do not receive such an email, please contact Erik Parner (parner@biostat.au.dk).
The solution may be in Danish or English at your discretion.
Practical information
 Answer all questions. In particular, be aware that some questions comprise several subquestions that all must be addressed.
 Plot the data whenever possible using scatter plots,
boxplots, histograms, and/or QQ plots.
 Always specify the statistical model used in the analysis.
 Any quantification of the findings of a statistical analysis in terms of estimates should be accompanied by confidence intervals. Any comparative statement should be backed up by a test and a pvalue.
 Model validation is an integral part of any statistical
analysis. It is not necessary to ask for model validation to be
performed, this should always be done.
 Include the Stata commands used in the analysis (the do/syntaxfiles)
and log/outputfiles in appendices.
For the log file it is most convenient to use a monospaced fonts like Courier, Courier New, Lucida Console, Monaco and Consolas.

NO Stata code or output (except graphs) outside the appendix!
 Formulate the conclusions using relevant terms from the
context of the study (it is important to be able to translate the
findings from the statistical analysis into conclusions regarding the
initial scientific question).
Exam formulation fall 2016
Problem formulation: Fall2016.pdf
The data files:
birthweight.dta,
birthweight.sav,
birthweight.xls.
Problems with the data or possible typo in the formulation: Email parner@biostat.au.dk.
In case of errors they will be reported here.