Correlation and Regression Flashcards
Coefficient of determination
- R squared
- Is the square of r, and reflects shared variance between two continuous variables
Properties of the Correlation Coefficient (r)
A defined range between -1 and +1
1 is a perfect relationshiop
0 is the absence of a relationship (null r=0)
Other correlation coefficients
Spearman’s rho (rank-order correlation fr categorical and ordinal)
Kendall’s Tau (2 ordinal variables and better at handling ties than Spearman’s)
Point biserial (Dichotomous variable)
Pearson correlation coefficient
- r
- Correlation doesn’t indicate causation
- Only tests linear association
- Can be influenced by outliers
- r2 gives true idea of magnitude. Shows how big.
- correlation refers to the strength of the relationship between two or more variables
Cohen scale
- 10 (small relationship)
- 30 (moderate relationship)
- 50 (large relationship)
Types of multiple regression
Simultaneous
Stepwise
Hierarchical
Multiple regression
Involves single dependent variable and multiple independent variables
Simultaneous multiple regression
- All independent variables are entered at the same time
- Huck’s analogy: all thrown in pot of soup (regression model)
Stepwise multiple regression
- All independent variables are considered
- Huck’s analogy: thrown in the pot of soup (regression model) by order of association with dependent variable
- Stop when those without association are considered
Hierarchical multiple regression
- All independent variables are eventually entered
- Huck’s analogy: thrown in the pot of soup (regression model) by theoretical order. Entered in order of previous knowledge and research.
- Researcher has control in the variables
Regression
Refers to a set of techniques for describing the relationship between two or more variables.
It finds out exactly how those two variables are associated
