Session 4 | Regression and T-Test Flashcards
(9 cards)
What does correlation and regression have in common?
Correlation and regression both describe a LINEAR RELATIONSHIP and they describe a RELATIONSHIP between variables.
If I have multiple independent variables, which method for my data analysis should I choose?
If I have multiple independent variables, they can be incorporated in regressions.
Usually a regression equation has a few elements, how can you name them?
Regressions usually have an independent variable (predictor) and a dependant variable (outcome)
Which kind of regression models are there?
Linear (one predictor and one outcome)
Hierarchichal or Multiple Regression : Multiple predictors for one outcome
Logistic Regression (binary outcome)
What is the formula for the Linear Regression?
Yi= B0 + b1x1+ E
yI= outcome
B1= regression coefficient
xI= Predictor (independent variable)
What is one example of Linear Regression? How does it generally work?
An example of Linear Regression is the OLS (ordinary Least Squared Model). It calculates what it the best slope to ensure all points are roughly at the same distance with less possible deviation.
Can you give me an example of question related to the Multiple Regression?
The Multiple Regression model differently from the Linear model can have one outcome but multiple predictors. As such, one example of Multiple Regression model might be whether the size and number of a eyes of a spider affect the amount of fear in the participants of a survey.
How is Hierarchical Regression different from Multiple Regression?
In both Hierarchical and Multiple Regression I have one outcome being impacted by multiple predictors. In this case, Hierarchical Regression has the different that it includes the predictors one step at a time to verify how much each predictor weights. (eg. first how much age and height affect the fear of spiders, then include the size of the spider.)
