肆客足球

• If your query relates to a study which you have already consulted a biostatistician about, please email or telephone the consultant biostatistician.
• A biostatistics clinic is held once a month on the first Wednesday of the month. No appointments are needed for these sessions which are run on a first come first served basis.

Should my supervisor come with me to the appointment?

• For undergraduate or masters level research, it is strongly recommended that the supervisor is present during the consultations, at least for the initial consultation.

What should I bring with me to an appointment?

Any of the below as applicable:

• IRB or Ethics approval
• The approved research protocol or proposal
• The research questions and/or hypotheses should be clearly listed
• The electronically captured data (see next FAQ for details)
• Previously published similar studies

How should I capture my data to ensure that it is able to be analysed?

• Data should be captured electronically in a standard package eg: excel/REDCap/Access
• The format should allow for direct importation of data into a statistical package, ie:
• Variables names in row 1 only
• One column per variable
• Variable names short and unique
• Categorical data coded numerically, and data dictionary/key supplied
• A variable should either be numeric or character- do not put both in the same column. For example, do not use “N/A", “Missing" or “<0.3" in a numeric field.
• Limit text/character variables
• Missing data left as blank cells
• Delete all unnecessary rows/columns including notes and summary statistics
• Any highlighting, colours and special fonts will be ignored
• See this video on the basics of data capture in Excel
• Data should be checked and cleaned by the researcher. Data capture errors cause delays with the analysis and can lead to invalid results if not detected.

Do I need to have a background knowledge of biostatistics?

• To ensure that you gain the most out of your research journey you should have a general understanding of the following if applicable to your study:
1. Hypothesis testing
2. P-values
3. Confidence intervals
4. Some of the common statistical tests

Do I need to have the statistical analysis software on my computer and if so which one/s?

• It is recommended that you obtain a copy of the statistical software package as discussed with the biostatistician to enable sharing of data and output files effectively.
• The biostatistician will try to accommodate any preferences you may have for particular statistical packages.

Will the consultant do all my analysis and interpret the results for me?

• It depends on the level of collaboration which you have negotiated with the biostatistician.
• For student research we encourage students to do their own data analysis with the guidance of a biostatistician.
• Where the biostatistician is a co-investigator in a study, his or her role might be the data management and analysis. In this case, the biostatistician will be responsible for data analysis and interpretation.

Will the consultant write up the results for me?

• It depends on the level of collaboration which you have negotiated with the biostatistician.

• For student research, no, you will be responsible for writing up your own results. The biostatistician may be requested to read and edit your write up as needed.
• Where the biostatistician is a co-investigator in the study, and also a co-author in the publication, he or she will have a large role in writing up the results and will contribute to other sections of the manuscript as well.

Should the biostatistician who assisted me be acknowledged in my publication or a co-author?

• Plans for publication of results should be discussed with the biostatistician.
• If the biostatistician has met the criteria for co-authorship (ICMJE guidelines) then he or she should be offered co-authorship.
• At the very least, acknowledgement of the biostatistics unit and the Dean's fund if applicable should be given in a standardised way. For example, to be discussed with your biostatistician: A consultant at the Biostatistics Unit within the Division of Epidemiology and Biostatistics, Stellenbosch 肆客足球 assisted with the (design and/or analysis) of this study [or experiment or project] through support from the Faculty of Medicine and Health Sciences dean's fund.