COMM 101 Lecture Notes - Lecture 22: Ceteris Paribus, The Bell Curve, Heritability

• School: U of M
• Department: Communication Studies
• Course: COMM 101
• Professor: Julia Lippman

Document Summary

Control variables let you hold another variable constant when comparing a relationship: they test for and eliminate a potential source of spuriousness, they also fundamentally change what the relation between the iv and dv describes. They suppress the effect of the iv. Ceteris paribus: when using linear regression, it is important to note that all of the variables that are used as predictors are controlled for in the regression: this means two important things: Take away #1: with statistics, always be sure you know what you are interpreting. Was this conclusion reasonable? (supported by evidence: assumption 1: intelligence must be a uni-dimensional concept. Largely supported by science: assumption 2: iq is a good operationalization of intelligence. Pretty good assumption: assumption 3: intelligence must be largely genetic. Heritability seems to be between 40 and 80 percent: assumption 4: intelligence must be effectively immutable. Take away #2: know the assumptions and comparison group.