Test 3 Flashcards
Multivariate correlational designs
measuring more than 2 variables
longitudinal designs
multivariate correlational studies where the same variables are measured over time to help establish temporal precedence
cross-sectional correlation
the relationship between 2 variables measured at the same time
Autocorrelation
correlation between a variable and itself at another point in time
cross-lagged correlation
the correlation between one variable at one time and another variable at another time
bi-directional relationship
the relationship goes both ways (can’t establish temporal precedence because they both cause each other)
identifying temporal precedence in longitudinal designs
look for a statistically significant and strong r value for the cross-lagged correlation
statistical control
hold constant in analyses to measure the unique effect of a variable
multiple regression
statistical analysis that can be used to rule out third variables (does the relationship persist when controlling for the third variable)
beta
isolating a variable and measuring it to see if it impacts our correlation then when we control for this variable and see if our correlation still remains
experimental control
hold a construct constant across participants in an experiment
moderation
used to test if the direction and/ or strength of an association between two variables changes based on another variable
mediation
used to test why two variables are related (explanation)
complete mediation
proportion mediated ≥ .80
partial mediated
proportion mediated < .80
when can researchers say that a correlation is statistically significant
when the CI does not include 0
random assignment
all participants have an equal and known chance of being assigned to either condition
establishing covariance
comparison group
establishing temporal precedence
longitudinal study
establishing internal validity
multiple regression
confound
a variable that could also be related to the DV and varies systematically with the levels of the IV
noise
a variable that could also be related to the DV, but does not vary systematically with the levels of the IV
selection effect
when the type of person systematically differs between conditions (especially a problem when you let participants pick their own condition)
matched group design
pairing participants based on some variable and then randomly put into separate groups