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

Flashcards

Stats Lecture 7/8 - Correlations

sarahderveeuw English 10 cards

Stats Lecture 7/8 - Correlations

Cards in this deck (10)

  1. 1Elements of a correlationA correlation coefficient:
    quantifies how well the best fit line explains the data
    gives direction of relationship.
    A correlation has two features
    Magnitude
    direction
    A correlation of zero shows no relationship.
  2. 2Charactersitics of a correlationMagnitude:
    Between 0 and 1.
    A value of ‘0’ means no relationship.
    A value of ‘1’ means a perfect relationship.
    Direction:
    Negative Correlation vs. Positive Correlation.
    For example, -.67 or .56.
  3. 3CovarianceA numerical estimate of how much the two variables (X and Y Covary)
    Like variance formula but with the inclusion of Y
    Covariance
  4. 4Correlation CoefficientA standardized value of the correlation between to vriables
    Most common calculation: Pearson’s correlation coefficient/Pearson’s product-moment correlation
    A score between 0 and 1, and a direction
    Score indicates strength of relationship/ steepness of the line of best fit
  5. 5Calculating the correlation coefficient(∑(
  6. 6What is Coefficient of Determinaionr^2 value tells you exactly how much variance they share
  7. 7Sumary of Correlation coefficientsTells us the strength and direction of a relationship between two variables
    0-1: Closer to 1 indicates stronger link
    Check a scatterplot for outliers/non linear relationships
    Break the formula down when computing Pearson’s by hand
    Correlation does not equal causation
  8. 8Pearson's Coefficient (Paramatretic)There's always a parametric test and a non-parametric equivalent,
    Paramatric - you make certain assumptions about the qualities of your data; needs to be ordinal or ratio, cannot be nominal
    homogenaeity of variance; variance of two variables shouldn't be too different
  9. 9Spearmanh's Rho (the non-paramtetric version of correlation coefficientsIf you have nominal data but still want to determine correlation, use Spearman's Rho.
  10. 10How to Calculate spearman's Rhothe same as the Pearson's correlation coefficient, just using rankings instead of scores
    (∑(

Information

Description
Flashcards on Stats Lecture 7/8 - Correlations, created by sarahderveeuw on 18/05/2015.
Created by
sarahderveeuw
Views
1
Tags
sp300
Rating

Feedback

What is it about?

What do you think of the new look?

How do you like it?