Chapter 12: Correlational Research

Correlational Research

Measure and describe relationship between two variables which are observed, not manipulated or controlled

Scatterplot

Pttern of the relationship between X and Y

Correlational Characteristics

Direction of relationship between X and Y

Positive correlation

As X increases, Y increases (same direction)

Negative correlation

As X increases, Y decreases (opposite directions)

Linear correlation

Relationship is a straight line

Curvilinear relationship

Relationship is not a straight line

Correlation coefficient

Based on Cohen’s rule of thumb (R-value) ranges -1 to 1

Strong correlation

Relationship is clear and follows a uniform correlation

Weak correlation

Relationship is unclear and not uniform

Coefficient of determination

Variance in Y explained or “accounted for” by X

Statistical significance of a correlation (p-value)

Probability of observing our (or more extreme) data

Regression

Statistical technique for using one variable to predict another

Strength of correlational research

•Often used for preliminary work in new areas

•Test relations between variables that can’t be manipulated (ethics)

•High external validity

Weaknesses of correlational research

•Low internal validity

•Correlation does not establish which variable affects which (if at all)

•Correlation does not establish causality

Third-variable problem

A third variable accounts for the relation between X and Y

Multiple regression

Use multiple predictor (X) variables to predict one criterion (Y) variable