Skip to content
 

Ethics in statistical practice and communication: Five recommendations.

I recently published an article summarizing some of my ideas on ethics in statistics, going over these recommendations:

1. Open data and open methods,

2. Be clear about the information that goes into statistical procedures,

3. Create a culture of respect for data,

4. Publication of criticisms,

5. Respect the limitations of statistics.

The full article is here.

8 Comments

  1. I think these are all good points, and we would all benefit from following and promoting them.

    At the same time, “Ethics in statistics is about more than good practice” – and more than openness, clarity, etc. A key missing discussion is what statistical analyses are used for. The choice of what to work on matters, and applications can have ethically importantly differences. There are interesting applications of statistics to help maximize repayment rates for sub-prime mortgages, or help insurance companies find which claims are least likely to be provable in court. There are also interesting applications for how to structure nutrition-based interventions in third world countries, or how to maximize success of cancer treatments.

  2. Andrew is a cool dude. I’m so glad I came across this blog. So rich and fun.

  3. Raj says:

    Tukey, J.W. Sunset Salvo, The American Statistician, 40, 1986, 72-76. would have fitted nicely as a reference to the main message conveyed in the article.

  4. Thank you for this pithy and helpful article.

  5. I would like to add that one journal, Judgment and Decision Making, has tried (with some success) for several years to honor these principles. I am an editor, and Andrew is on the board. I hope he agrees.

  6. Francis Gilbert says:

    Nice article. It actually offers suggestions on things we can do better, rather than just finger-wagging about some form or another of purported ethical malfeasance. The quote taken from Loken and Gelman (“maybe we should consider another, less comforting possibility, which is that our fundamental values have been conveyed all too well and the message we have been sending – all too successfully – is that statistics is a form of modern alchemy, transforming the uncertainty and variation of the laboratory and field measurements into clean scientific conclusions that can be taken as truth…”) seems particularly germane in light of the recent NYT articles on the use of genetics research to support fringe groups’ conclusions about ostensible racial differences (https://www.nytimes.com/2018/10/18/insider/science-genetics-white-supremacy.html?module=inline)

    • D Kane says:

      From the article you link to:

      But another reason some scientists avoid engaging on this topic, I came to understand, was that they do not have definitive answers about whether there are average differences in biological traits across populations. And they have increasingly powerful tools to try to detect how natural selection may have acted differently on the genes that contribute to assorted traits in various populations.

      What’s more, some believe substantial differences will be found.

      Given that, how is Loken and Gelman’s (excellent!) article relevant?

Leave a Reply