28 February 2017: Stephan Conrad

Just a few clicks to reveal the absolute truth or rather “do not trust any statistics…”?

`Evaluation strategy´ is the magic word in Q-DAS software products qs-STAT, solara.MP, destra and the CAMERA tools. This article tells you what an evaluation strategy actually is and how you can design it.

Before dealing with the selected topic, I first want to give a warning. Please note that I wrote parts of the text with a slight twinkle in my eye but only parts of it, not the whole text.

<mode sporadic irony on> 

Be aware of the fact that any software evaluation strategy is nothing but a representation of a quality philosophy, of course, i.e. a philosopher develops it in his head, puts it on paper and calls it corporate strategy. Then he implements the strategy in the software to ensure a uniform approach throughout the company. The entire company will then apply the same correct methods.

Even though the creation of an evaluation strategy caused a lot of trouble, you will finally do everything right as long as you follow this strategy. Or more precisely: Whoever pursues the strategy of company Smith, will do everything the right way – for company Smith. However, if you work for company Taylor, you will have to use Taylor’s strategy. This leads us to a decisive question. If Smith’s strategy obtains results that are different from the ones calculated by Taylor, but both results are correct, which result can be considered to be more correct? We can even carry this question to the extremes: Which one of the many evaluation strategies in qs-STAT and solara.MP is the most correct?

Critical fellows only know a single answer to this question: My own strategy is the most correct.

Even in `pre´factual times, men already regarded statistics with deep suspicion. Starting a conversation with `Statistically speaking…´ usually made your opponent frown. This is no surprise because statistics is a branch of mathematics. `The cognition of the reason through construction of conceptions is mathematical´, said Immanuel Kant. He also stated that no natural science contained more truth than `pure mathematics´. But why did M. L. H. Kessel alias Hans Bruehl dare to say statistics was `the fairy tale of rationality´? And there are even more irreverent quotes we avoid here for reasons of youth protection!

Where do these issues we have with statistics actually come from?

The first aspect you should know is that there are two different types of statistics – descriptive ...