# Regression

Regression
I will undertake a stepwise linear regression
1 of 17
Descriptives Statistics
The case to variable ratio is _ = _:1 .. which is less/greater than our minimum requirement of 40:1 ... confidence/caution
2 of 17
Correlations
ignore
3 of 17
Variables entered/removed
X number of variables were entered into the equation which were _ and _. X variables were excluded which were_.
4 of 17
Model summary
For the model R=_, R2=_, AdjR2=_. Thus, using the R2 the model account for _% of the variance in DV
5 of 17
ANOVA
ignore
6 of 17
Coefficients
Tolerances for (IVs) are _, greater than min of 0.1 therefore there is no evidence of multicollinearity or singularity. The model is predicting DV better than the mean. The model is _=constant+(IV1*constant)+(IV2*constant) For increase in 1 of...etc
7 of 17
Excluded variables
ignore
8 of 17
Collinearity diagnostics
ignore
9 of 17
casewise diagnostics
report after tests of normality
10 of 17
residual statistics
ignore
11 of 17
The residual plot suggests that the distribution is/not random. Curved loess line suggests heteroscedasticity so a non linear regression might be more appropriate
12 of 17
Explore: case processing summary and descriptives
ignore
13 of 17
Explore: tests of normality
Testing the residuals for normality W(_) =_, p=_. Significant, therefore the residuals are not normally distributed
14 of 17
Exclude outliers
There were X number of outliers, cases _. (Run the regression again (lots of output to ignore!)
15 of 17
Model Summary
For the final model after exclusion of outliers, R=_, R2=_, AdjR2=_, thus using the R2 the model account for _% of the variance of DV
16 of 17
Coefficients
The final model is consumption = constant+(IV1*constant)+(IV2*constant)
17 of 17

## Other cards in this set

### Card 2

#### Front

Descriptives Statistics

#### Back

The case to variable ratio is _ = _:1 .. which is less/greater than our minimum requirement of 40:1 ... confidence/caution

Correlations

### Card 4

#### Front

Variables entered/removed

Model summary