The Effect - - Describing Relationships (except

10 important questions on The Effect - - Describing Relationships (except

What is a downside with a scatterplot?

They are hard to read if you have a lot of data points.
They tempt you into thinking that the x axis causes the y axis

Why use the conditional mean and not other features, such as the Standard Deviation?

The mean behaves better in small samples
It helps us weight prediction errors.

How can you describe the relationship between two variables using the conditional mean?

If the mean of Y is higher conditional on a higher value of X, then Y and X are positively related
  • Higher grades + faster learning
  • Never study anything twice
  • 100% sure, 100% understanding
Discover Study Smart

Through what ways can you condition a continuous variable?

  1. Us a range of values for the variable we're conditioning on, instead of a single value.
  2. Use a shape or line to fill the gap with no observations

What is the downside of using ranges for continuous variables when trying to condition them?

It is arbitrary due to numerous reasons, i.e.;

number of bins picked is arbitrary, evenly sized bins, difference between top of the one bin and bottom end of next is not representative.

4.4 Line fitting, what does it mean and how does it differ from conditional means?

Instead of thinking locally and producing estimates of the mean of Y conditional on values of X, we can assume that the underlying relationship between Y and X can be represented by some sort of shape

What are some benefits of the line fitting approach?

  1. It gives you conditional mean of Y for any value of X we can think of, as it is a straight line and fills in all the gaps, as opposed to the conditional mean method.
  2. It lets you cleanly describe the relationship between the variables (if the slope coefficient on X is posititive, then the relation between x and y is positive)
  3. Results are more precise than when only using local mean data
  4. The line can easily be extended to include more than one variable

How does OLS work?

OLS picks the line that gives the lowest sum of squared residuals. A residual is the difference between an observation’s actual value and the conditional mean assigned by the line.
It uses the information encoded in the covariance

When is nonlinear regression often used?

When Y can only take a limited number of values (such as binary variables)

4.5 Controlling for a Variable, what does this entail?

By taking the residuals of your x to z and y to z, and then use the mean of the x residual conditional on the y residual. If the x mean does not change much on different values of y, then the entire relationship is explained by z.

One way to take conditional conditional means is with regression, and multivariate regression.

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