Clinical Data Analysis on a Pocket Calculator

Abstract

The time that statistical analyses, including analysis of variance and regression analyses, were analyzed by statistical analysts has gone for good, thanks to the availability of user-friendly statistical software. The teaching department, the educations committee, and the scientific committee of the Albert Schweitzer Hospital, Dordrecht, the Netherlands, are pleased to announce that since November 2009, the entire staff and personnel are able to perform statistical analyses with the help of SPSS Statistical Software in their offices through the institution’s intranet. It is our experience as master’s and doctorate class teachers of the European College of Pharmaceutical Medicine (EC Socrates Project) that students are eager to master adequate command of statistical software for carrying out their own statistical analyses. However, students often lack adequate knowledge of basic principles, and this carries the risk of fallacies. Computers cannot think and can only execute commands as given. As an example, regression analysis usually applies independent and dependent variables, often interpreted as causal factors and outcome factors. For example, gender and age may determine the type of operation or the type of surgeon. The type of surgeon does not determine the age and gender. Yet, software programs have no difficulty to use nonsense determinants, and the investigator in charge of the analysis has to decide what is caused by what, because a computer cannot do a thing like that, although it is essential to the analysis. It is our experience that a pocket calculator is very helpful for the purpose of studying the basic principles. Also, a number of statistical methods can be performed more easily on a pocket calculator, than using a software program.