Skip to main content

Three Kind of Lies: lies, damned lies and Statistics

Most people would have come across this quote from many statistical basic books especially students who are getting an introduction to statistics, this would be one among the definitions. But remember the words are not the direct meaning for the quote and it is just an sarcastic statement. Let me just pull more of an depth analysis for the quote. There were debates as it was quoted by an non-statistician, however, here i am not getting in to the debate  just thought of sharing what this quote is all about.
When the word 'statistics' is mentioned, automatically the word 'number' pops up in most of our mind. Globally, statistics and numbers are correlated with each other , therefore number is the first thought. Yes! you are right now, Statistics used in the quote refers as 'Data'. When the data is studied, it reaches with an result. Results of any statistical analyses gives out number and these numbers plays an crucial role for interpretation. Most importantly, interpretation should be inferred based on the question which motivated for the analysis. Therefore, interpretation with the framework of the objective would lead to an proper channel of conclusion.
One needs to be cautious while using any statistical techniques or methods before embarking on the results. Data would always give some numbers in return using any statistical analysis or methods. It is our responsibility to scrutinize before paving the way for a conclusion. The truth or lie absolutely depends upon the interpretations which are being inferred with the 'data'. 
Mere numbers(output) without any proper objective or analysis would definitely lead to misleading conclusion . At this point of juncture,  surely the quote would strongly come in to the picture with lies and damned lies. This may not be the actual scenario however due to the false conclusion with the numbers would have lead to misleading results. Therefore, cautious with the numbers would save us from false results. Therefore i end up with one more beautiful quote, where they say " All models(statistical) are false, but some of them are useful" which is tough to digest but this is the truth.

Comments