Statistical inference from sample to population: Sample Statistics and Population Parameters
Stats in hypothesis testing, decision making, and experimental design
What are the odds of getting this outcome by chance alone? Statistics tables: calculate or look-up P
Alpha error (falsely rejecting the null hypothesis), Statistical signficance and what is so magical about P=0.05 anyway ...?
A Tale of two tails: one-tailed vs. two-tailed rejection region in directional and non-directional null hypotheses
Beta error (failing to reject the null hypothesis when it actually should be rejected): Power of a test is the probability that we reject the null hypothesis when the null hypothesis should be rejected.
With a (biological) reason we can compare a sample distribution to a particular null distribution (e.g., testing for randomness in a spatial or temporal distribution)
comparison-wise error rate and experiment-wise error rate: "... we then used t-tests to compare the 6 treatments with each other ..." Dunn-Sedak: a'=1-(1-a)1/k
a-priori set of comparisons: orthogonal comparisons (non-overlapping variances)
a-posteriori set of comparisons:
Parametric vs. nonparametric
What we can't do with stats ...
Houdini the Statistician - not: Stats as an afterthought and the statistician as pathologist: You can only assess what the experiment died of ...
a'=1-(1-a)1/k
