Explain collinearity and multicollinearity, and why we wish to avoid these in multiple regression.
What will be an ideal response?
Ans: When predictors are too highly correlated we say that they are collinear. When one predictor can be almost perfectly predicted from the remaining predictors we say that the predictors are multicollinear. When predictors are multicollinear then very small changes in the predictors can yield radically different coefficients in the regression equations. The no multicollinearity assumption is that the predictors are not too correlated with each other.
You might also like to view...
Alfred Adler emphasized which type of motivation?
a. The search for scientific understanding b. Adult sexuality c. Drive reduction d. Striving for superiority
Based on the actions of cocaine, one might infer that a toxin that suddenly destroyed synaptic vesicle walls would:
a. also cause the destruction of dopamine. b. cause an increase in dopamine levels in the synaptic cleft. c. cause all dopamine to be degraded by monoamine oxidase. d. cause a decrease in further dopamine synthesis.
The Binet-Simon scale of intelligence yielded a score called the
a. mental age. c. mental quotient. b. intelligence score. d. intelligence scale.
What implication comes from the study on positive affect and better cross-race facial recognition?
A. It shows that positive affect has few connections to facial recognition. B. It can lead to greater positivity in interracial relations. C. It explains that positive affect influences how mixed-race individuals fare in this task. D. It has the potential to explain how positive affect works in other areas of race relations.