Enhancing your research: collaboration with the Department of Biostatistics

The Department of Biostatistics offers a complete statistical consultation service to researchers of the Faculty of Health Sciences, from undergraduate students to staff. The department can contribute throughout the research process, from the planning of the project to the publication of the findings.

Planning of studies

It is important that you discuss your project with a biostatistician early in the planning. This not only ensures that the data which you will collect is analysable, but the biostatistician is also able to give valuable inputs with regard to research design, sampling and measurement. Guidelines are available online - Protocol guidelines; Sample size guidelines.

Ideally a biostatistician should be approached once your project idea has been approved by your own department/supervisor, and you have formulated some ideas in writing. 

Ethics Committee

If the department is to be involved in the analysis of the project data, the protocol must be submitted to the department for input and approval timeously (4 weeks) before submission to the Health Sciences Research Ethics Committee (HSREC) of the UFS. A study may not commence before the HSREC has given approval. The department will therefore not accept data for studies before the researcher has provided the department with a copy of the approval letter from the Ethics Committee.

Research for study purposes

If researchers plan to use the research for a qualification (undergraduate or post-graduate), we prefer that the study leader should be present, at least for the initial discussion between our department and the student, so that clarity and agreement can be found between all the parties concerning each person’s responsibilities and the research methods to be used. The department’s contribution must also be given the necessary acknowledgement in the thesis or dissertation.

Data analysis

Data must be presented electronically (typically MS Excel) for analysis. We give advice on drawing up and coding of questionnaires, and the format of electronic data files. Should you want us to analyse the information, the final questionnaire or coding form must be submitted to us before data collection. We will advise how the information should be entered onto the computer.

The first step of the analysis is the checking of the data to identify any errors which occurred during collection, coding or data entry and if possible to rectify them. This can be a time-consuming process, especially in the case of large data sets. Summary values and highest and lowest values will be given to the researcher on a computer printout to check whether the values are possible. The researcher must also assist the biostatistician by identifying information which can be checked by relating it to other information.

After the data errors have been corrected, the analysis usually starts with descriptive statistics (percentages, means, medians). Further analyses will be done depending on the aims and objectives of the project. The analysis is done with interaction between the researcher and the biostatistician during which results and statistical methods are discussed.

Handling of projects

Projects are handled as they come in. At a given time a biostatistician may be working on more than 10 projects. Sometimes it is difficult to estimate the duration of an analysis beforehand, as a seemingly simple data set may lead to all manner of complicated analyses. Please discuss your time schedule with the biostatistician and refrain from requesting that your analyses be done at the last minute (e.g., a week before the cut-off date for the submission of abstracts for a conference). We always try to complete projects quickly, but must also ensure that the work is done thoroughly.

Co-authorship / Acknowledgement

If the biostatistician has made a noteworthy contribution to the planning and/or execution of your project (in accordance with the Faculty guidelines on authorship), we consider the offering of co-authorship of a publication or presentation as recognition of the professional service delivered. The biostatistician will play an active role in the compilation of the publication or presentation. A final manuscript must please be submitted to the co-author before you submit the article for publication or submit an abstract for a conference. If the biostatistician performed the analysis but did not qualify as co-author, the Department’s contribution must be noted in Acknowledgements.

Should you publish or present something from your research project (whether the biostatistician is co-author or not), please notify us, so that we can note the information on the project file.