Statistics in Experimental Stroke Research: From Sample Size Calculation to Data Description and Significance Testing

Abstract

Experimental stroke researchers take samples from populations (e.g., certain mouse strains), and make inferences about unknown parameters (e.g., infarct sizes, outcomes). They use statistics to describe their data, and they seek formal ways to decide whether their hypotheses are true (“Compound X is a neuroprotectant”). Unfortunately, experimental stroke research at present lacks statistical rigor in designing and analyzing its results, and this may have negative consequences for its predictiveness. This chapter aims at giving a general introduction into the do’s and don’t’s of statistical analysis in experimental stroke research. In particular, we will discuss how to design an experimental series and calculate necessary sample sizes, how to describe data with graphics and numbers, and how to apply and interpret formal tests for statistical significance. A surprising conclusion will be that there are no formal ways of deciding whether a hypothesis is correct or not and that we should focus instead on biological (or clinical) significance as measured in the size of an effect and on the implications of this effect for the biological system or organism. “Good evidence” that a hypothesized effect is real comes from replication across multiple studies; it cannot be inferred from the result of a single statistical test.