Abstract
Full presentations of many of the entries below have already been distributed to BMS subscribers and RC33 members over the BMS-RC33 distribution list 1
On Thursday, 10 September 2015, the BMS/RC33 list distributed information concerning a new statistics song by our colleague and jazz musician Michael Greenacre. For something completely different, listen to Michael Greenacre’s latest statistical song, The P-Value Song at https://youtu.be/yy4nsEvKh2E. It won an award at the recent US Teaching of Statistics conference. The P-Value Song is a satire on the use and abuse of P-values in statistics. The P-value is a measure of the risk taken in rejecting the so-called “null hypothesis” and thus coming to a substantive conclusion based on empirical data. If the P-value is less than 0.05, or 1 out of 20 (this value is chosen out of pure convention), then the risk of the conclusion being false is low and the result is deemed “statistically significant”. The song satirizes the idea that significance can be achieved merely by increasing the sample size, and that the P-value is often confused with the concept of effect size. “Everything is possible … put your faith in the P!” is an illusion, although many researchers think this is the main objective of a study, to get low P-values! But low P-values are no panacea for small and possibly irrelevant effect sizes! Lyrics by Michael Greenacre, sung by Gurdeep Stephens and accompanied by Lisa Olive, with musical arrangement by Gurdeep and Lisa of Irving Berlin’s “There’s No Business Like Show Business”. Recorded in Victoria, BC, Canada, March 2015.
Interested parties can find other statistical songs by Michael Greenacre, such as It Don’t Mean A Thing (If You Don’t Do Modelling) and It Ain’t Necessarily So, on https://www.youtube.com/user/StatisticalSongs.
Concerning statistical significance and number of cases, you can also see a series of three articles of Louis Guttman published by the BMS in 1984 and which were among the last of Guttman’s published articles: “What Is Not What in Statistics”, BMS, 1984, 4: 3-28; “Statistical Inference Revisited - 1984”, BMS, 4: 28-32; “The Illogic of Statistical Inference for Cumulative Science”, BMS, 1984, 4: 32-34. Guttman often referred to the “star worshippers” who were chasing statistical significance at any price.
BMS-RC33 and AIMSl lists
Since March 2015, AIMSl subscribers have periodically received a message that: For a whole year, you can remain on the BMS-RC33 list for 10 euros paid to the AIMS by Visa or Masters. That can be done by emailing to Pour toute une année, vous pouvez rester sur la liste BMS-RC33 pour 10 euros versés aux AIMS par carte Visa ou Master. Cela peut se faire en envoyant un courriel à
Footnotes
1