STAT 217 Lecture Notes - Lecture 4: Statistical Process Control, Correlation And Dependence, Standard Deviation

• School: University of Calgary
• Department: Statistics
• Course: STAT 217
• Professor: Bingrui Cindy Sun

Document Summary

Limitations of bivariate analysis: several factors may influence one outcome, may be overlap, what may appear to be the effect of one predictor may be due to a second predictor, running several regressions is messy. Is a correlation matrix for pearson correction: -1 meaning weak correlation, +1 meaning strong correlation. Prove a hypothesis: 1. chronological ordering (x occur before y, 2. x is related to y best used for bivariate, 3. rule out alternative explanations for multivariate. Spurious, two or more events are not causally related to each other but mistaken that they are, or that there is a third factor. Antecedent: a variable that helps explain relationships between other variables that are nominally in cause/effect relationship (z is responsible for x and y) Redu(cid:374)da(cid:374)t varia(cid:271)le: do(cid:374)"t know what is the dv or iv, only know one direction relationship while the other might not be related to the third variable.