Understanding complete hypothesis tesing for one-way anovas. When follow up tests are needed and their purposes. Ronald aylmer fisher developed the anova technique: also introduced concept of hypothesis tesing and mulivariate staisics, looked to see if things could interact. Aim for equal sample sizes, but even if not, it is okay to use the anova. Assumpioms for the independent measures anova: observaions within each sample are independent, populaions from which the samples are selected must be normal, populaions from which the samples are selected must have equal variances (important) These are the same for the independent measures t test. Rejecing null when it is true: t-tests are limited to only comparing 2 groups at a ime. Using muliple t tests inlates the chance of this error happening: f tests can compare many groups in any experiment simultaneously while keeping type i error at your speciied level (ex.