In 1971 two psychologists wrote a paper titled “Belief in the Law of Small Numbers”. Their thesis is that individuals (both trained scientists and laypeople) consider a sample (a set of data) drawn from a population to be more representative of the population than they should, especially when the sample is small. In other words, when dealing with small data sets, we will erroneously assume that our “mean” value is very representative of the population mean. ...

When I am not working at work, I am usually thinking about (and tinkering with) ways to adopt things from other industries to make us better. Most of these just sit on the shelves in my mind with little to no action taken (my excuse is that they would require too much behavior change from other people, the hardest thing in the world). Some of those: Use latex (through Overleaf) to collaborate on and produce engineering reports. Use GIT to comment / track / update the reports mentioned above. Use jupyter notebooks (with the help of connerferster’s handcalc library) for calculation packages instead of excel. Use documentation structures similar to software packages or wiki’s to manage project documents. Something like bookstack for a quick start, but it would eventually need a little more functionality. Better adoption of existing industry data “standards” (like DIGGS and AGS). And of course, there’s the never-ending effort to automate the boring stuff.