March 2024
S M T W T F S
 12
3456789
10111213141516
17181920212223
24252627282930
31  

Statistics in the dock? Damned Lies

Statistics in the dock? Damned Lies

Felicity Healey-Benson (BSc, PGCE, MBA, MA (HRM), PG Dip CIPD, Dip. NLP, SAC. Dip, FHEA, Academic Assoc. CIPD)

 

Judge-Court-Gavel-1600by900

 

“Facts are stubborn things, but statistics are pliable” [1]. A face value interpretation of this famous aphorism of disputed origin [1] belies its part in a deeper conceptual battlefield.  On one side, the “Professionals”, a statistics fraternity with codified principles, emphasizing their role as custodians of truth [2].  On the other, the anti-stats brigade, positioning statistics as weapons of mass destruction, utilised to legitimatise false claims, teaching how to lie with data [2, 3].

The Good, the Bad, & the Ugly

In defence of ‘statistics’, bring forth exhibit ‘A’, world-wide research repositories across all disciplines housing research inquiries and investigations that purport quality and rigor through their application of statistical concepts, designs, methods and tests [4]. Representatives abound for a subpoena to bear witness that statistics can, and do improve human welfare, “not by its own ends, but by its contribution in all fields” [5].

Arguably statisticians are more honest in their statements than those who make absolute claims [1]. Yet two key flaws to this testimony exist:

  1. Fundamentally statistics are an abstraction of reality, not the reality itself, not fully substituting for understanding [6], and,
  2. Statistics aren’t always expertly or robustly applied [7]. People tend to cherry-pick that which support their beliefs, ignoring contradictions.

These points seep into counter-claims of statistical unreliability, into the realm of misappropriate use of statistical output; a dark side to its utility.

Culpable Ignorance vs. Intent to Defraud

In my mind, this is not an explicit judgment on the inherent value of statistics, but an observation on the variability in value-sets or intentions behind its employ.

Allegations are duplicitous:

  • Use to bridge raw data and knowledge and understanding [6], skilfully or otherwise, or,
  • Malicious perversion of the truth for a specific agenda [7].

Statistics have been manipulated since time immemorial. Today, in a business context alone, whether its research to advance management knowledge or address policy concerns [8], or bad statistics feeding Fake News [9] or political campaigns [10], data is employed for high impact. Data delivered through the designer’s lens can easily mutate into lies when neutrality is dismissed [11]. Even expert witness found in a statistical professional body, the Royal Statistical Society [12; 13] draw attention to its infallibility. Bias is a key accomplice to statistical crimes.

Motion to dismiss, with conditions

Society needs Statistics, but with guiding conditions. Statistics are not judgment-free facts; they bear high potential to mislead. “Figures will not lie, liars do figure” [14]. Whilst statisticians avoid perversion of the truth, responsibility must also devolve to the user/ recipient, in a form of Caveat Emptor i.e. expectation to investigate claims/arguments hold water.

Rehabilitation

It’s not realistic, to expect bad statistics can be eradicated from the marketplace of ideas [15]. Some statistics are just born bad, based on guesses or dubious data, others go bad after being reworded or mangled [15]. Given the current reproducibility crisis [16] statistical methods require rehabilitation. Weak statistics must get called out, replication must gain respect [17].

I strongly advocate that we all, be that business professional, academic researcher or layman, do our bit for reform, and help clean up the reputation and value of statistics [18], and in doing so, enhance society’s ability to accumulate [19] and take value from knowledge.

https://www.emergentthinkers.com/

References

  1. Wikipedia (2018) ‘Lies, damned lies and statistics’ [online]. Available at: https://en.wikipedia.org/wiki/Lies,_damned_lies,_and_statistics [Accessed 20/10/2018].
  2. Velleman, P. F. (2008) ‘Truth, Damn Truth, and Statistics’, Journal of Statistics Education, 16(2).
  3. Statistics How To (2014) ‘Misleading Statistics Examples in Advertising and The News’ [online]. Available at: https://www.statisticshowto.datasciencecentral.com/misleading-statistics-examples/ [Accessed 20/10/2018].
  4. Rolfe, G. (2006) ‘Validity, trustworthiness and rigour: quality and the idea of qualitative research’, Journal Adv Nurs, Feb; 53(3) pp. 304-10.
  5. Rodriguez, R.N.(2013) ‘Building the Big Tent for Statistics’, Journal of the American Statistical Association,108(501) pp. 1-6.
  6. Pyzdek, T. (2009) Statistics: The Good, the Bad, and the Ugly [online]. Available at: https://www.qualitydigest.com/magazine/2009/may/column/statistics-good-bad-and-ugly.html [Accessed 21/10/2018].
  7. Schenkelberg, F. (2015) The Bad Reputation of Statistics [online]. Available at: http://nomtbf.com/2015/07/the-bad-reputation-of-statistics/   [Accessed 21/10/2018].
  1. Julie L. Ozanne, Brennan Davis, Jeff B. Murray, Sonya Grier, Ahmed Benmecheddal, Hilary Downey, Akon E. Ekpo, Marion Garnier, Joel Hietanen, Marine Le Gall-Ely, Anastasia Seregina, Kevin D. Thomas, and Ekant Veer (2016), “Assessing the Societal Impact of Research: The Relational Engagement Approach.” Journal of Public Policy & Marketing, In-Press. doi: http://dx.doi.org/10.1509/jppm.14.121
  2. Leetaru, K. (2017) Lies, Damned Lies And Statistics: How Bad Statistics Are Feeding Fake News [online]. Available at: https://www.forbes.com/sites/kalevleetaru/2017/02/02/lies-damned-lies-and-statistics-how-bad-statistics-are-feeding-fake-news/ [Accessed 21/10/2018].
  3. Grajales, C.A.G. (2014) How statisticians have changed elections [online]. Available at http://www.statisticsviews.com/details/feature/6931581/How-statisticians-have-changed-elections.html [Accessed 21/10/2018].
  4. Littlejohns, D. & Peterson, M. (2017) Lies, Damned Lies, and Statistics (Data Stories) [online]. Available at: https://design.ncsu.edu/sothen/lies/ [Accessed 21/10/2018].
  5. Royal Statistical Society (2014) Code of Conduct [online]. Available at: http://www.rss.org.uk/Images/PDF/join-us/RSS-Code-of-Conduct-2014.pdf [Accessed 21/10/2018].
  6. American Statistical Association (20180 Ethical Guidelines for Statistical Practice [online]. Available at: http://www.amstat.org/ASA/Your-Career/Ethical-Guidelines-for-Statistical-Practice.aspx [Accessed 21/10/2018O’Toole, G. (2010) Maxim: Figures don’t lie, but liars do figure (1884 February 29)[online]. Available at: http://listserv.linguistlist.org/pipermail/ads-l/2010-April/098260.html [Accessed 21/10/2018].
  7. O’Toole, G. (2010) ‘Maxim: Figures don’t lie, but liars do figure’ (1884 February 29)[online]. Available at: http://listserv.linguistlist.org/pipermail/ads-l/2010-April/098260.html [Accessed 21/10/2018].
  8. Best, J. (2001) ‘Damned Lies and Statistics: Untangling Numbers from the Media, Politicians & Activists’, University of California Press.
  9. Peng, R. (2015) ‘The reproducibility crisis in science A statistical counterattack’, Significance, 12(3), pp. 30-32.
  10. Baker, M. (2016) ‘The Reproducibility Crisis Is Good for Science’ [online]. Available at: http://www.slate.com/articles/technology/future_tense/2016/04/the_reproducibility_crisis_is_good_for_science.html [Accessed 21/10/2018].
  11. Wolfers, J. & Stevenson, B. (2013) Six Ways to Separate Lies From Statistics [online]. Available at: http://www.bloomberg.com/news/2013-05-01/six-ways-to-separate-lies-from-statistics.html [Accessed 21/10/2018].
  12. Schnieder, B. (2004) ‘Building a scientific community: The need for replication, Teachers College Record, 106(1), pp.1471-83.

Keywords. Statistics damned lies pliable stubborn facts knowledge

Comments are closed.