Regression advanced

8 important questions on Regression advanced

What should one do when there are categorical and quantitative PV's in the analysis?

Make the categorical PV('s) quantitative using DUMMY VARIABLES --> gender, nationality, etc..

What happens with the graph when a dummy variable is added?

Alpha changes (will get higher)

B*0 = 0
B*1 = B

In this case B is a fixed number and can be added to the Alpha

What should be done if the categorical PV that has to be changed to dummy variable has more than two categories?

Make multiple dummy variables (0,0,1) (0,1,0). Reference group should always maintain 0.
  • Higher grades + faster learning
  • Never study anything twice
  • 100% sure, 100% understanding
Discover Study Smart

Is multicollinearity a problem?

Yes and no.

No because it is not a problem for the reliability of the whole model
Yes because it is unclear to what extent a PV has a contribution to the OV.

How van multicollinearity be measured?

Tolerance and VIF

Tolerance: <0,10 = serious collinearity, 0,10-0,20 = collinearity, not to severe, >0,20 = no collinearity

VIF: <5 is no collinearity, 5-10 = collinearity, not to severe, >10 = serious collinearity

By computing Tolerance, one should use Rj^2. This is not the general R^2, but the one that is allocated to the specific PV.

What is a moderator?

One PV has interaction on an other PV (Multicollinearity)

What is a mediator?

When the relation between a PV and an OV run (partially) via a third variable.

For instance: Relation between stress and bodyweight is negative (more stress is less weight)

But, stress causes perhaps sugar consumption, which rises body weight. sugar consumption is a mediator.

Why does one use non-parametric tests?

Because parametric tests (ANOVA, t-test) may be violated (correction calculations or data transformation.

A non-parametric test is assumption free and has fewer assumptions about the distribution because it measures different measures, not the actual data

The question on the page originate from the summary of the following study material:

  • A unique study and practice tool
  • Never study anything twice again
  • Get the grades you hope for
  • 100% sure, 100% understanding
Remember faster, study better. Scientifically proven.
Trustpilot Logo