chapter 9 Flashcards
What is a multivariate design?
A multivariate design involves more than 2 measured variables and allows us to control for different confounds and brings us closer to causality.
True or False : Using multiple variables gets us a bit closer to establishing causation.
True.
Within a multivariate design, we may see 2 things to help establish causation – what are they and which causal criteria do they each establish?
- Longitudinal designs help address temporal precedence
- Multiple regression analyses help address internal validity
What is a multiple-regression (multivariate regression) analyses?
It is a statistical technique that computes the relationship between a predictor variable (the thing we controlled for) and a criterion variable, controlling for other predictor variables.
What are the 3 types of longitudinal correlation designs?
- Cross-section correlations
- Autocorrelations
- Cross-lag correlations
What is a cross-sectional correlation? Give an example
A cross-sectional correlation shows the relation between the 2 variables over time and whether or not the relation strengthens. Establishes if 2 variables are related at one time.
What is an autocorrelation? Give an example
An autocorrelation looks at the relation of a single variable with itself over time – looks at the stability of the variable.
What is a cross-lag correlation? Give an example
A cross-lag correlation looks at the relation of one variable at time 1, with the second variable at time 2. It helps us get a sense of temporal precedence.
Instead of correlation research, why not just do an experiment?
We do correlation research because some variables can not be manipulated and some aspects of the research would be unethical.
True or False : By using a multivariate design, we can evaluate the correlation between 2 variables while controlling for third variables.
True
What are the two types of variables in a multiple-regression analyses?
Criterion/outcome variables : dependant variables
Predictor variables : Independent variables
What is beta? What is it used for?
Beta is the measure of effect size. Betas denote the direction and strength of a relationship.
Positive beta : positive relationship
Negative beta : negative relationship
Beta near 0 or 0 : no relationship
High beta : strong relationship
Low beta : weak relationship
True or False : Multiple regression is a foolproof way to rule out all kinds of third variables.
False – they can not always establish temporal precedence ; some other variables that were not measured may be taking part in the relation.
What is parsimony?
Parsimony is the idea of simplicity – explaining a phenomenon is the simplest manner.
What is a pattern?
A pattern is when multiple studies point to the same conclusion.
True or False : Journalists do not always fairly represent pattern and parsimony.
True : When journalists report only one study at a time, they are selectively presenting only part of the scientific process.
What is a mediator? What is the cycle of mediation?
A mediator is a variable that helps explain the relationship between two other variables.
A is related to B because A is related to C, which is related to B.
–> p.ex : Amount of deep talk is related to well-being, because the amount of deep talk enhances quality of social ties which in turn, enhances well-being.
How are mediators and third variables similar?
Both involve multivariate research designs and both can be detected using multiple regression.
How can we differentiate mediators and third variables?
Third variables are external to the bivariate correlation (problematic) while mediators are internal to the causal variable (not problematic).
What is a moderator? What is the process behind moderation?
A moderator is a variable that, depending on its level, changes the relationship between two other variables.
A is related to B for one type of C but not for the other type of C. Example : Viewing violent TV is related to aggressive behaviour when there is no parental discussion but not when there is parent discussion. An “it depends” relation.
True or False : Mediation hypothesis are causal statements.
True.
True or False : Mediators do not specify a time sequence for the three variables (i.e : temporal precedence).
False.
What does it mean to say that mediators are a “mechanism”?
It means it is an interesting, measurable reason for the original relationship.
What is multiple regression?
A statistical technique that involves multiple variables allowing to control for potential third variables. A means for dealing with third variable problems.