Chapter 8: Bivariate Correlational Research Flashcards
Bivariate Correlation/Association
Association that involves exactly two variables
Two or more measured variables
Effect Size
Describes strength of a relationship between 2+ variables
Use weak, moderate, and strong to describe rs of 0.1, 0.3, 0.5
Better to think about in nuanced ways (depends on the context)
Larger effect sizes are more important
Small effect sizes can compile over situations or people
T-Test
Measures difference between means/averages
Confidence Interval (p-value)
Typically use 95% of data
Range where we can expect the data to fall
Confidence interval will contain 95% of all estimates of a certain effect
Don’t want confidence interval to include 0
Should be more than 0.5
Replication
Conduct the study again and find multiple estimates
Compare the findings and across studies
Different researchers and different settings
Outlier
Extreme score
Single or few cases that stand out
Can effect correlation coefficient
Restriction of Range
When there is not a full range of scores on one of the variables in the association, it can make the correlation appear smaller than it actually is
E.g. SAT scores and first year grades (people who got into college, include those who didn’t too)
Curvilinear Association
Relationship between two variables is not a straight line
It might be positive up to a point and then become negative (vice versa)
E.g. use of healthcare system
Moderator
When the relationship between two variables changes depending on the level of another variable
E.g. in cities with high residential mobility there is a positive correlation between sports team success and attendance
E.g. in cities with low residential mobility, there is not a strong correlation between sport team success and attendance
E.g. degree of residential mobility moderates the relationship
Moderators change the relationship between variables but doesn’t change the variables themselves
