Pixel International Conferences

With statistics we can prove everything! Due to misuse of statistics, this statement can become true. But it is a prominent duty of teachers to demonstrate the correct use of statistics and to convince students of the value of this discipline. For example the notion of correlation is often misused. Showing that two factors A and B are correlated, is sometimes inaccurately described as ‘A causes B’, although it is only spurious correlation. Another possible danger is hidden in graphics: pictures can mislead the reader due to inadequate scaling, inappropriate use of two-dimensional representations, etc. In everyday life, headlines of newspapers are full of astonishing conclusions based on ‘scientific research’. If you are lucky, you will find more detailed information inside the article that reveals some information about the way the results were obtained. But in many cases not even the sample size is mentioned, nor is any confidence interval of the measured variable, so no notion of variability or statistical significance is included. Moreover, formulations of comparisons can be misleading, e.g. when a value is reduced by 50 %, we can say as well that we need an increase of 100 % to reach the original value… It is of major importance that we make our students, who are the future users of statistics, aware of these traps. Therefore teaching statistics consists of more than just proving formulas. This paper describes how we can ensure deeper insights into statistics to prevent the cultivation of the abuse of statistics by our students due to ignorance.