Criminal Justice

Correlation Studies & Correlation Statistic

Correlation Studies-represent a type of research in which we measure two or more variables to see whether they are related to each other

– Example: height and weight

• Co-related – taller you get more you weigh

• Example: TV Violence and Aggression

• Correlation study

– Must involve at least two measurements

• Amount of TV violence watched

• Amount of aggression

Types of Correlation • Positive Correlation

– A relationship between two sets of measurements in which an increase in one variable is accompanied by an increase in the other.

– Example: As the amount of TV violence increases the amount of aggressive behavior also increases.

• Positive Correlation

• Negative Correlation

– Whenever an increase in one variable is accompanied by a decrease in the other

• Also referred to as an inverse relationship

– Example: As the amount of work in your life increases the amount of leisure will decrease

• Negative Correlation

• No Correlation

– When there is no relationship at all between two variables

– Example: As amount of violent TV increases aggressive behavior stays the same.

• The variables are not correlated in any way

• No Correlation

Correlation Statistic • Statistical measure of how closely two things are related • The value computed to assess the significance of a correlation

– Runs between –1.0 to +1.0 – Stronger relationship

• closer to +1.0 or -1.0 • “r” represents the Pearson correlation or correlation coefficient – the statistic

– Negative –1.0 to 0 – Positive 0 to 1.0 – No correlation 0, or close to 0

• Common Errors with Correlation • Only tells us if variables are related

– Does not tell us why something happens – Example: TV violence & aggression

• Spurious relationship(false correlation) – When two variables are significantly related but are really not related – Example: Home heating bills & Suicide

• Third variable(confound)

– When something is not measured but is responsible for the correlation