Ron Sandland recently wrote about the new phenomenon of 'big data' - weighing up the benefits and concerns. Terry Speed reflected on the same issue in a talk earlier this year inGothenburg, Sweeden noting that this is nothing new to statisticians. So what's all the fuss about? Here's another take on the 'big data' bandwagon.
Social 'Science' - Science No More! 18 March, 2015
This is not a bad dream - the journal Basic and Applied Social Psychology has banned the use of statistical inference!
In their infinite wisdom, the editors of the journal Basic and Applied Social Psychology (BASP) have just announced that "from now on, BASP is banning the NHSTP (null hypothesis significance testing procedure)" because they claim it is "invalid". (Click here to view full editorial)
So, a bunch of social commentators (they have forgone the right to call themselves scientists) believe they have a better understanding of statistical science than professional statisticians. While it may not be perfect, the bedrock of statistical inference is solid. Scholars such as Deborah Mayo at Viginnia Tech. have meticulously teased apart and analysed the logic of hypothesis testing and can point to epistemological deficiencies but the fact remains that no other paradigm (other than Bayesian inference) has stood the test of time, is totally consistent with the scientific method, and has improved the quality of our lives. If you need evidence of the latter, just go to your medicine cabinet, pull out a drug (even an asprin) and ask "how do I know it's safe?" and "how do I know it works?".
BASP's decision is breathtakingly short-sighted and frankly, staggeringly stupid. It is an affront to all of science - not just statistical science and can only result in the dumbing down of their profession. BASP's editors have consigned their journal to scientific irrelevance; this 'statistical fatwa' has been tried before and failed. The following extract is from my paper "Desired and Feared - Quo vadis or Quid agis? " published in The American Statistician (click here to view):
(In 1997) the American Psychological Society was contemplating banning the use of hypothesis testing in its journals (Shrout 1997). As noted by Shrout (1997), this was not the first time such calls had been made; he cited the infamous case of the American Journal of Public Health which advised authors that "all references to statistical hypothesis testing and statistical significance should be removed from the paper" and that you should "delete p-values as well as comments about statistical significance." The journal Epidemiology adopted the same stance under the editorship of Ken Rothman. Rothman's advice to would-be authors was blunt: "you can also enhance your prospects if you omit tests of statistical significance . . . we do not publish them at all. Not only do we eschew publishing claims of the presence or absence of statistical significance, we discourage the use of this type of thinking in the data analysis, such as in the use of stepwise regression" (Rothman 1998). The philosophical debates about null hypothesis significance testing (NHST) have been with us for many years and the attempts of a single misguided journal editor to deny the existence of a well-established mode of statistical inference were inevitably doomed from the beginning. While Nelder (1999) was equally strident in his criticism of sloppy statistical practice, his calls were not to ban p-values per se, but to demolish the culture of uncritical thinking that had developed around the routine application of hypothesis testing and the attendant practice of "asterisk hunting."
And therein lies BASP's problem: rather than address the core issue of sloppy scientific practice, they've demonized the science!
With it's short-sighted action, BASP has rendered itself scientifically neutered - a publication reduced to subjective assessments from observational studies. Good luck with that!
Prof. David Fox Chartered Statistician (Royal Statistical Society, London) Professional Statistician (American Statistical Association)