The Biostatistics Centre provides opportunities for collaboration and skill development for researchers in the Division of Health Sciences at the University of Otago.
We offer a range of short courses on biostatistics. This includes courses on sample size estimation, basic biostatistics for health researchers, and use of Stata software.
We are also happy to receive suggestions for topics.
Introductory Biostatistics for Health Researchers
This two-day course is a fun introduction to the appropriate application and interpretation of biostatistical concepts. It is for people working in clinical research who have no formal statistical training.
Stata Software workshop
This is a hands-on workshop based in a computer lab, aimed at researchers and research students who need to use statistical software for their research.
The Biostatistics Centre recommends Stata as an ideal software package for many health sciences researchers. It is reliable, easy to use, has a wide range of commonly used statistical analysis options, and produces publication-quality graphics.
Regression Modelling course
This two-day course provides an introduction to regression modelling approaches. The scope of the course runs from basic principles of regression methods, deciphering the output of statistical analyses, and the practicalities of running these various regression methods.
This is an intermediate level statistics course.
Longitudinal Modelling workshop
This two-day course provides an introduction to a range of common approaches for modelling longitudinal and correlated data.
This is an intermediate level statistics course.
HASC413 Biostatistics and HASC415 Regression Methods: Health Sciences Applications
We strongly recommend quantitative research students complete these two papers to learn the skills required to undertake your quantitative analysis. These courses cover the most commonly used statistical methods in health sciences using real world examples.
We invite all who have an interest in biostatistics to attend our Biostatistics Forum. The forum is a regular series of events including seminars and discussion groups which focus on issues in biostatistics.
We usually meet on the last Monday of each month, from 11am to 12pm in room 033/036, ground floor, Adams Building, Frederick Street, Dunedin.
Please contact Andrew if you are interested in presenting a session:
These websites provide useful guidance:
- Sample size calculator:
- Commonly encountered statistical topics and questions (BMJ Statistics Notes):
- Questions of a statistical nature, and from allied disciplines (Cross Validated is part of StackExchange):
- Statistics guide for research grant applicants (St George's Hospital):
UCLA website information
The UCLA website contains extensive information about performing statistical analyses using:
- R: https://cran.r-project.org/
- SAS: https://www.sas.com/en_nz/software/stat.html
- SPSS: https://www.sas.com/en_nz/software/stat.html
- Stata: https://www.stata.com/
The UCLA site also has information about other programs, as well as some non-software specific information about statistics:
For those of you who use SPSS, looking for statistical guidance visit:
You need to pay a subscription (which isn’t that expensive) but they have detailed description of many, many methods and their implementation in SPSS. They also have a Stata section.
Note that SPSS is not routinely used by any of the current biostatisticians, nor are GraphPad, Prism, or Minitab.
Further information on statistical software and useful links are available from these pages by our Christchurch and Wellington colleagues:
- Make each ROW in Excel a different person.
- Try to use numeric codes where you can.
- If you have yes/no answers record them as 0 (no) and 1(yes).
- At the top of each column (if you are using excel) insert a comment that tells you what your codes mean. This file will be the master file and you will be able to refer back to this at any point.
- Do not put any characters (e.g. "n/a" or ".") into a numeric column or it will switch it to character in any stats package you import it into.
- Check your data as you go, just to make sure there are no rogue entries in there. Simple tabulations can help with the data checking.
- If you are recording words (we're not sure why you would need to), make sure the spelling is correct and consistent. Otherwise, when you go to look at it you have many categories of answers that just relate to different spelling of the same word.
If you have some information about why the responses are incomplete use numbers that are out of range to indicate different things. For example, 888 might mean 'refused', 999 might mean 'just didn't respond' etc. those kind of numbers (888, 999, 8888, 9999) are good because they are often the ones that are used for that purpose.
Preparation of your imported data
Import the Excel into Stata (or whichever statistical software you are using), and spend time putting in labels and making sure the formats are what you want. That will help you in the future.
When sending data to our biostatisticians: Is the data anonymised? If not, does the biostatistician need this information?