Hacker News Reader

Comment by continuations | Hacker News Reader

The need to understand basic statistics is a far cry from needing to have "exceptional mathematical fluency."

The point is that to be a great scientist you don't need to be a math genius. Having the math ability of a von Neumann is not a prerequisite to being a great scientist.

View on HN

You should really understand more than basic statistics for most science. How are you even defining basic stats?

Need a modicum of commonsense as well. For some examples see 'Statistical Follies and Epidemiology' http://www.youtube.com/watch?v=C42AwvaZ-04&feature=playe...

no you shouldnt. look at any medical science, exercise physiology, psychology etc. papers. All you see is data gathering and basic statistic analysis (anova, t-test etc.). Basic stats is what they teach at the undergraduate level to scientists and engineers.

Isn't the reason those fields mostly see usage of basic stats that they were never taught anything more sophiscated. For example in exercise physiology, which studies a lot of phenomena over time, the results get reduced to minimum/maximum/average because those points match the required input of their basic ANOVA's. Instead they could be comparing time series and find more subtle difference in when a rise starts to occur or at which rate. Numbers that often get obscured by the massive data reductions. Sure you can get by without more advanced data analysis and statistics, but its not exactly something to strive for.

You should ask a statistician what they think of the quality of the statistical analysis done in some of those papers. Applying a statistical test is easy, picking the right one and being aware of shortfalls and possible problems with the data is the tricky bit.

pfisch's comment refers to "most" sciences. Here you have listed "medical science, exercise physiology, psychology" as examples of fields for which knowledge of basic statistics is sufficient for publishing work. Of course, you have excluded physics where undergraduates should (at the very least) understand geometry and calculus, topics in modern algebra and probability theory. If we consider electrical and mechanical engineers we have to add linear algebra to the mix, as well as somewhat non-elementary material in odes and pdes. In computer science and (modern) biology there will be an expectation that one can solve linear recurrence relations. This list could be expanded in various directions.

Reply on news.ycombinator.com