Multivariate Correlations Flashcards
involve more than two measured variables
multivariate designs
three criteria for establishing causation
- covariance
- temporal precedence
- internal validity
variable 1: sense of belonging
variable 2: self-efficacy
longitudinal designs
why can’t you just do an experiment?
- cannot be assigned to preferences
- unethical to assign participants
indicate if a third variable affects the relationship
regression results
ex: pregnancy risk
criterion variables
ex: sexual TV content
predictor variables
“controlled for”
“taking into account”
“correcting for”
“adjusting for”
regression in popular press articles
_______ ________ is not a foolproof way to rule out all kinds of third variables
multiple regression
- the quality of social ties
- amount of deep talk
- well-being
mediators
mediators vs. third variables
- both involve multivariate research designs
- both can be detected using multiple regression
similarities
mediators vs. third variables
- third variable and external to the bivariate correlation (problematic)
- mediators are internal to the casual variable (not problematic)
differences
mediators ask ____
moderators ask ____ or ____
“why”
“for whom” or “when”
multivariate designs and the four validities
- internal validity
- construct validity
- external validity
- statistical validity
it is similar to r in that they denote the direction and strength of a relationship
beta
