Hacker News Reader

Such a controlled trial would not pass ethics review. But for the human side, hopefully we can run statistical analyses to show correlation at least?

I think a mechanism of action in vitro plus a biologically confirmed trial in rats plus correlation in statistical analyses is fairly convincing?

View on HN

With enough accurate data (and time, as these toxins aren't really fast acting) you could get to causality with the right kind of statistical analysis. Typically, counterfactual analysis and some DAGs.

It depends on the audience. Nothing short of RCTs is admissible on HN.

Reply on news.ycombinator.com