Multiple Regression Flashcards
What does the linear regression model describes?
It describes a linear function and it is very general. the model y(hat) = a+bX
For what kind of analysis is the Multiple Linear Regression used?
It analyses multiple kind of data - continuous, categorical, likert scale. It’s also used to figure out which IVs are the most important.
What is a Multiple Regression?
Like a simple regression but now you have multiple predictors of an outcome variable
What is the standard/Simultaneous linear regression?
All variables are entered into the equation at the same time without specific order. Each variable is assessed as if it were the last variable entered, this controls for the other IVs.
It takes together all of these variables explain his amount of variance in the data and takes all slopes which result into an average of slope across all independent variables.
What kind of problem arises with standard/simultaneous linear regression?
If there are two highly correlated IVs, the one with the biggest semi-partial correlation gets all the variance therefore the other IV will get very little variance associated with it and look unimportant (suppression).
What is a semi-partial correlation or sr?
The variance from only that IV over the total variance. Tells you how much variance overall that variable accounts for. The unique contribution of that variable to R^2 - increase in proportion explained Y when that variable is added to the equation.
What is Hierarchical/sequential linear regression?
When predictor variables are entered in as sets or steps. Variance gets assigned at each step to the first variables entered. You can assign order according to theoretical importance, you can also control for nuisance variables in the first step.
What are the steps in testing the IVs in the hierarchical/sequential linear regression?
- The first Iv is basically tested against r or sr since there’s nothing else in the equation it gets all the variance.
- the next IVs are tested against pr, they only get the left over variance.
What is partial correlation (pr)?
The variance from only that IV over the variance not accounted for (error). Tells you how much variance accounts for when you only look at variance that you can explain. Proportion of variance in U not explain by other predictors.
If you have a group of IVs that are super highly correlated but you don’t know how to combine them or want to eliminate them, how can you use the hierarchical regression?
You can process each step as a SET and you don’t have to care about each individual predictor.
What is statistical/stepwise linear regression?
Predictor variables are entered in steps, based on some statistical cutoff. Entry into the equation is solely based on statistical relationship and nothing to do with theory or your experiment.
What are the different types of stepwise regression?
- Forward: the biggest IV is added first, then each IV is added as long as it accounts for enough variance. You’d also need to create a cutoff.
- Backward: all are entered in the equation at first, and then each one is removed if it doesn’t for enough variance. It’s similar to simultaneous LR.
- Stepwise: it’s a mix between the two (adds them but then may later delete them if they are no longer impt.
What particularity does the stepwise regression have?
It’s the most common regression used but it’s a hot mess because of the constant back-and-forth approach. This approach is also not theory driven and it is not to be used in psych.
What interpretations can we use with multiple linear regression?
How good is the equation, if we can predict people’s scores better than chance. We can also use the adjusted r-squared for effect size as well as knowing which IV is the most important, which one contributes the most to prediction. We can use the coefficient statistics (t values) and adding predictors will always increase the explained variance.
What is predictor level MLR?
It includes statistical control, mediation and moderation.
