Get 1 week of unlimited access
Class Notes (1,052,010)
AUS (35,436)
UniMelb (5,194)
MULT (90)
Martin (7)
Lecture 5

# MULT30018 Lecture Notes - Lecture 5: Dependent And Independent VariablesPremium

3 pages98 viewsSpring 2018

Department
MULT
Course Code
MULT30018
Professor
Martin
Lecture
5

This preview shows half of the first page. to view the full 3 pages of the document.
WEEK 5 REGRESSION ANALYSIS
Y = a + bx + e (1 independent variable)
Y = a +bx1 + bx2 + e (2 independent variable) and so on
Every unit increasing in x makes the Y go up or down by that amount.
Tells the magnitude of the effect
Regression with one variable
R^2 tells us the variation percentage
1. Regression Coefficient (b) and p value (sig.)
2. The r squared statistic
3. ANOVA test (esp. check df=N)
4. Remember, multiple regression gives us a way to statistically deal with the 'third factors' we
have been talking about in previous weeks.
Some caveats
Something being 'statistically significant' (sig = 0.05) does not mean the result is 'significant'
Dependent variables must have 3+ categories
And these must make sense numerically
Think before choosing dependent and independent variables
Correlation vs. Regression
Correlation shows the correlation (-1 - +1) between two variables (it is a summary measure)
Regression tells us about the magnitude of the effect of one or more independent variables on
a dependent variable.
Linear Regression
Why do we use regression?
Correlation tells us the degree to which two variables are associated, or how well a line can
describe the relationship between two variables. It shows us how well two variables 'track'
one another
A high r tells us that the values of two variables are closely associated, but it does not allow us
to predict the value of one if we know the value of another
Regression allows us to estimate the value of one variable if we know the value of another
It does this by providing the 'regression line', which allows us to see the expected change in
one variable which corresponds with change in another
Regression line shows us the patterns in our data
Regression tells us how much change in one variable is associated with another, correlation
just tells us how closely they are associated.
The regression equation : Y = a + bx
The regression eqution is simply the formula for a straight line
Y is the value of the DV
X is the value of the IV
'a' is the constant or y-intercept. It is the value of y when x=0 (where the line 'starts').