Please enable JavaScript.
Coggle requires JavaScript to display documents.
Correlation and Regression - Coggle Diagram
Correlation and Regression
Correlation
measures the relationship between two variables
Correlational design
observes variables without manipulating them
Positive correlation
X and Y increase or decrease together
Negative correlation
X increases while Y decreases
Linear correlation
measures straight-line relationships between variables
Perfect correlation
exact, perfectly predictable relationship
Zero correlation
no consistent relationship between variables
Purpose
determines whether variables change together
Pattern idea
points forming a trend show strength/direction of relationship
Direction
shows whether relationship is positive or negative
Pearson correlation
measures linear relationship between two variables
Sum of Products of Deviations
SP
measures covariability between X and Y
Prediction
uses one variable to predict another
Imperfect prediction
correlations allow prediction but not perfect accuracy
Regression link
prediction based on correlation is called regression
Validity
shows whether a test measures what it claims to measure
Construct validity
supported when test correlates with related measures
Reliability
measures consistency of scores over time or conditions