L07 Regressions

Question 1

Regression

Answer 1

Used to predict the value of a dependent variable based on the value of one/more explanatory variables
Describe the characteristics of the line of best fit
Slope of the line and y intercept

Question 2

Types of variables in regression

Answer 2

Outcome/Dependent/Criterion variable

Predictor/Independent/Exploratory variable

Question 3

Multiple regression

Answer 3

If there are more than one predictor variable/IVs: how they contribute to DV
Plane of best fit
z = ax + by + c

Question 4

Equation for regression

Answer 4

y = ax + b
a = slope of the line
b = y intercept

Question 5

Slope of regression line

Answer 5

a in y = ax + b
The number of units that the regression line moves up the vertical axis for each unit of movement along the horizontal axis

Question 6

Assumptions for regression

Answer 6

1. Continuous level of measurement for both variables - ratio/interval
2. Linear relationship between the two variables
3. No significant outliers (checked with scatterplot)
4. Lack of autocorrelation of residuals: Durbin Watson test ideally between 1.5-2.5
5. Outcome variable comes from a normally distributed population

Question 7

Residual

Answer 7

Difference between observed value of dependent variable (y) and predicted value (y hat)
Residual is denoted as e
Sum and mean of residuals is 0
A good line of best fit gives very small residuals for each data point

Question 8

Autocorrelation

Answer 8

Correlation of errors/residuals on each point - residuals are not independent of each other
No autocorrelation for a regression test
Checked with Durbin-Watson Test
Value should be between 1-3 
ideally between 1.5 to 2.5