Moderation
ANOVA numerical notation for:
Two Factors, 3 Levels Each
What is a 3 x 3 ANOVA
Effect that represents the direction and magnitude of mediation.
What is the Indirect Effect?
Assumption specific to repeated measures ANOVA when time points are 3 or more.
What is sphericity?
Additional source of variation considered when using a LMM.
What is a random effect?
A type of Y variable using a GLMM with a binary response.
What is a Bernoulli Variable?
How moderation is statistically tested in models.
What is an interaction?
Type of mediation model where all mediators are examined at the same time.
What is parallel mediation?
Minimum number of within-subjects factors to use a Mixed Design ANOVA.
What is one?
LMM are needed to address violation of this assumption of the general linear model.
What is independence of observations?
Link function used for a multi-level Poisson regression model.
What is the log link?
Mean Square Model divided by Mean Square Residual
What is an F-statistic?
The direct effect plus the product of the a and b paths.
What is the Total Effect?
Effect of one factor on the dependent variable.
What is a main effect?
What is an unstructured covariance structure?
Number of levels within a data structure for a traditional repeated measures design where time-points are clustered within individuals.
What is a 2-level data structure?
Difference in regression slope coefficients is represented by this.
What is regression coefficient for interaction?
Type of mediation model where a mediator variable can predict another.
What is a serial mediation model?
Type of ANOVA follow-up test that are planned and compared time-point to another using simple or adjacent methods.
What is a contrast?
Type of Figure that shows individual level trajectories across time.
What is a Spaghetti Plot?
Within a spreadsheet, each higher level of a data structure must have this.
What is an identifier number?
No moderation is graphically represented in a clustered line graph showing this.
Methods of obtaining variance (error) estimates when the sample size is small or distribution highly skewed.
What is bootstrapping?
Violation to sphericity can be addressed using a correction factor on this?
What are the degrees of freedom?
These "effects" are estimated and interpreted within an LMM just like ordinary least squares (traditional) linear regression.
What are fixed effects?
The random effects are represented by this within the following mathematical representation:
g {E(Y|X, 𝛍)} = 𝛃X + 𝙕𝛍, Y ∼ F
What is 𝛍 or "mu"?