Advice on statistics

1. Statistical advice in the design stage

Statistics is not a set of tricks that you open when you can no longer handle your data or to cover errors in the research design: rather it is an essential part of scientific thinking and your research. In the design stage, important decisions are being made on which variables to be measured, the number of observations, the number of repetitions of experiments. When the final statistical analysis are being performed, many of these decisions are be considered as given and can not be changed anymore.

A statistician from our office can help you with your protocol and submit an advice aimed at the statistical, epidemiological and methodological aspects of the research. Especially when you are not an expert on statistical or methodological issues, such an investment can save a lot of costs and disappointments when the statistical analysis will be performed. It also can increase the quality of the research because final statements from this study can become more certain. Such advice may relate to:

writing the statistical sections of the research protocol,

determining the design of the study,

the sample size calculation (power analysis) or

all statistical aspects (measurement level variables, randomization, dependency of observations).

2. Statistical advice during the analysis stage

Another moment on which statistical advice will be useful, is when the data have to be analyzed. Even if you are fairly familiar with various statistical methods, it is often useful to let an independent statistician check the used methods and statistical analysis in the investigation. This advice may relate to:

gathering information on non-response,

applying the appropriate statistical method justifying the research design,

the choice between parametric methods and non-parametric methods,

the method of reporting the statistical analysis or

answering statistical questions from reviewers or clients.

3. Second opinion on statistical methods

If after an outsourcing research you have doubts on the statistical methods used by the researchers, you can ask us for an independent opinion on the methodology and techniques used. Obviously it is better to ensure in advance that the appropriate statistical methods are actually applied.

In practice, this is not always possible, eg by various interests involved, or ignorance of certain methods of research. In those cases, an independent opinion of a statistician is important to put the conclusions in the proper perspective. Without redoing the research we will attempt to indicate how certain statistical and methodological decisions have affected the conclusions of the investigation.

privacy

disclaimer