GoConqr has a new look! Let us know what you think.

Flashcards

Regression

marielouise.k English 10 cards

Cards in this deck (10)

  1. 1Simple linear regression modelYi=A+Bxi+Ei
    (dependent= model parameter of mean + effect of one unit change + error)
  2. 2Erroraka. residuals,
    Distance of actual from the mean
  3. 3Regression definitionFits the data given to a prediction model (straight line)
  4. 4R(squared)SSM(regression)/SST
    *100= %
    Shows the percentage of the outcome which can be explained by the model.
    Value between 0 to 1, a higher value= more accurate regression model.
  5. 5Standard error estimateStandard deviation of the distribution of the residuals. Puts confidence intervals around the predicted scores. higher= more vague answers.
  6. 6OutliersMay affect the regression unduley and therefore give misrepresentative results. Leverage point/Influence point. Latter is the worst.
  7. 7Cook's DA figure over one should raise concerns. It is a measure of influence point. (which can seriously affect the results)
  8. 8Multiple regression equation and aimY= a+B1X1+B2X2+B3X3 etc.
    Aims to predict the dependent variable from 2+ independent variables.
  9. 9Variance Inflation Factor (VIF)No multicollinearity then VIF=1, if VIF approaches 10 then there is a problem to be dealt with.
    Comes under collinearity stats. next to tolerance.
  10. 10Tolerance0=BAD
    1=GOOD
    calculation: 1-Rsquared
    in relation to multicollinearity
    and in terms of independence of a predictor.

Information

Description
Flashcards on Regression, created by marielouise.k on 16/05/2013.
Created by
marielouise.k
Views
36
Rating

Feedback

What is it about?

What do you think of the new look?

How do you like it?