In: Statistics and Probability
Correlation does not mean causation.
Look through a newspaper, or a magazine or even a journal article that cites correlational data to make a point. Identify other possible explanations for the correlation that don’t fit with the article’s conclusion.
Baseball and fly slack: Correlation does not construe causation
See the article "How stream slack obstructs Major League Baseball execution" in volume 114 on page 1407.
This article has been refered to by various articles in PMC.
My thought was pulled in to the continuous article by Song at al. entitled "How stream slack cripples Major League Baseball execution", by its possibly unpredictable subject and also more altogether in light of the way that I pondered how one could ever truly show the effect of fly slack on baseball execution.
In this paper, Song et al. make sense of how to refine the results gotten by Recht et al. as ideal on time as 1995, showing that eastward travel relates with decreased execution. Specifically, by making use of extend bits of knowledge on the particular broad proportion of Major League Baseball record databases open on the web, Song et al. discovered quantifiably imperative connections of eastward travel with a couple of components related to home-assemble unfriendly execution, and furthermore with "terrific hammers allowed" for both home and away gatherings.
Regardless of the way that I don't scrutinize the considerable proportion of work included and would be well-close unequipped for settling on a choice about the authenticity of the examinations performed, I ought to yield that I was stunned the way Song et al. methodicallly present the connections they perceive as quick proof of causality between fly slack and the affected components. It is extremely outstanding to me that "association" does not appear to be even once in the paper, when this is truly what the journalists have been looking, as I might want to think, to be tentatively correct, the title of the article ought to scrutinize: "How stream slack partners with preventions in Major League Baseball execution."
Given the particular broad proportion of composing on associations between's fly slack and decreased athletic execution in a whole collection of recreations, I am clearly not discussing that fly slack is the most likely reason for the effects recorded. Likewise, looking title of the articles in the reference summary of the Song et al. report, this affinity to amalgamate association with causality is clearly to an extraordinary degree visit in this field of examination. Nevertheless, given the wide readership of PNAS and the subject of this article, I feel that it is most likely going to be given off by the press and to attract the thought of various people, the two specialists and nonscientists.
Considering the present penchant to confuse intelligent data, by methods for the maltreatment of estimations particularly, I feel that a journal, for instance, PNAS should expect to educate by model, and in this way ought to approve more exhaustiveness in the presentation of legitimate articles concerning the differentiation among connections and shown causality.
For anyone holding up be convinced that strong connections don't generally show causality, and to finish on an even more lively note, I invite them to visit the site of Tyler Vigen, which gives some really interesting models of false associations (www.tylervigen.com/misleading relationships).