Chapter 9 - Multivariate Correlational Research Flashcards
define third variable
variable that accounts for the apparent association between your town variables of interest. once taken into account, the variables can no longer be related
define moderating variables
a variable (theoretical interest) that influences or changes the nature of the relation between two other variables. you see a different pattern of association, weak at one level of moderator and strong at another level
what are the 2 things that multivariate research can address?
temporal precedence and internal validity (by identifying thrid variables)
define longitudinal designs
measuring the same varaibles in the same people over a long period of time
what are the 3 correlations that can be made from longitudinal designs?
X1 and X2: autocorrelations, (correlation of one varibles with itself, measured at two different times)
X1 with Y1: only at one time, cross section correlation (correlation between two variables that are measured at the same time)
X1 and Y2, X2 and Y1: cross-lag correlation, time difference/lag between the variables (correaltion between earlier measure of one variables and a later measure of another variable)
how can you rule out third variables?
they can be controlled for by adding the suspected third variables (control variables) into the analysis
there is still only 1 outcome variable, but now you have an infinite amount of predictor variables
what does Beta signify?
it is very similar to r but it does not have a specific range
- it gives an estimate change in relation of y for x
- holds all other predictor variables constant
how can mutliple regression still not establish causation?
- cannot establish temporal precendence: need to have longitudinal studies for this
- can never control everything: still so many variables that are not measured that can still affect your results
define mediation
answers the questions of why X has a causal effect of Y. it is a internal mechanism that is in the middle of x and y
explains the relationship between the 2 varibales
what is teh difference between a medaitor and a moderator?
mediator: caused by x to cause y (why does x influence y)
moderator: changes strength at different levels between x and y (how does the association change when the level of the variables changes)
define modertor
it is an external variables that cahnges the original relationship between x and y, the relationship is different at different levels of the moderator
define multiple regression
statistical technique that computes the relationsip between a predictor and criterion/outcome variables, while controlling for other predictor variables
define controlling a variable
holding a potential third variable at a constant elve while looking at the association between two other variables
define criterion variable
varaibles in teh regression that the researchers are most interested in understanding
define predictor variable
variable that is used to explain variance in the criterion variable