At his blog, Bioinformatician at Large, Ewan Birney lists five statistical concepts he wishes he'd been taught years ago instead of learning as he went along. "To my subsequent regret, no one took [me] aside and said, 'Listen mate, you're going to be doing a lot of statistics, so just get the major statistical tools under your belt now,'" he writes. Among the topics he wishes he'd been taught are non-parametric statistics, R (or S), and the relationships among p-value, effect size, and sample size. On the last, Birney says "we're far too trigger-happy quoting p-values, when we should often be quoting p-values and effect size.
HT: Fejes.ca