BIOL 210 Lecture Notes - Lecture 21: Bayer, John Ioannidis, Reproducibility
Document Summary
Regrade: your grade might end up higher/lower/the same: the answers are there for you to learn from. It"s the yellow parts that you need to study. Independent: because it"s used to predict the dependent one (size of brain/iq: %ebm (expressed breast milk) Any good correlational study - will have confounding variables: this study: has confounding variables, what else could effect it/ Socioeconomic status (living in a more comfortable situation - can afford better food) Social class - operationalized according to 6 categories (in the analysis - a multiple regression: put in these factors, get a relationship that excludes the effects of these confounding variables. Good correlational - variables you wish to associate - the effect is not driven by something else. Descriptive statistics: just tells you what the data looks like, without saying whether these data differ from what you would get by chance. (not by chance: things that summarize your data, mean of a group, median.