correlation
- Created by: Chloe
- Created on: 29-04-15 19:02
View mindmap
- Correlation
- Multivariate: analyses that have more than one DV
- Bivariate: correlation & regression analysis which have two variables that are technically both DV's
- Regression: identifying which variables are effective at predicting other variables of interest
- Correlations show the degree of relationship between variables X & Y
- Strength of a relationship is measured by the correlation coefficient (r) - varies from 0-1 (+/-)
- The Main Assumptions of Pearsons R: (1) Homogenity of variance: the variance of Y is the same at all levels of X (2) Normality: that both X & Y samples come from populations having normal distributions (3) Linearity of regression: that the underlying relationships can be adequately summarised by a straight line & not a curve
- Coefficient of determination: an estimate of the proportion of variability in one variance that can be determined from knowledge of the other - The square of the pearson correlation coefficient (R^2)
- A strong relationship = one which accounts for more than 50% of variation in the criterion variable (R= .7 or more)
- A Moderate relationship = 16%-36% (R= ,4 - .6)
- A strong relationship = one which accounts for more than 50% of variation in the criterion variable (R= .7 or more)
Comments
No comments have yet been made