Quantitative Lecture 9 Flashcards
What is correlation?
A standardised measure of the relationship between 2 variables.
What questions can correlations answer?
- Is change in one variable associated with a change in another variable?
- Is an increase in one variable associated with an increase or decrease in another variable?
- How strong is the relationship between variables?
What is Pearson’s Product Moment Correlation?
A parametric test of relationships/associations that produces a correlation coefficient, or r.
What does a greater r value indicate?
The stronger the relationship between the two variables.
What is bivariate correlation?
The relationship between 2 variables.
What are the generic assumptions for correlation?
- Data must be at interval or ratio level.
- Data must be normally distributed.
- There should be no outliers/extreme scores.
What are the specific assumptions for correlations?
- There should be related pairs.
- Linearity: there should be a linear relationship between the variables.
What does assumption of linearity entail?
The variables should be linearly related, meaning a straight-line relationship between x and y.
How can scatterplots be used in correlation?
Scatterplots can visually plot the relationship between two variables by assigning one variable to the x-axis and one to the y-axis.
What does a positive direction of relationship indicate?
Higher scores on one variable tend to be associated with higher scores on another variable.
What does a negative direction of relationship indicate?
Higher scores on one variable are associated with lower scores on another variable.
What does a zero direction of relationship indicate?
There is no linear relationship between the two variables.
What is the correlation coefficient (r)?
The strength of the relationship between 2 variables, lying between -1 and +1.
What does a correlation coefficient of +1 indicate?
A perfect positive relationship.
What does a correlation coefficient of -1 indicate?
A perfect negative relationship.
What does a correlation coefficient of 0 indicate?
No linear relationship.
What is the null hypothesis in correlation?
There is no relationship between the variables.
What is the alternative hypothesis in correlation?
There is a relationship between the variables.
What is common and shared variance in correlation?
The correlation coefficient (r) is a ratio between the covariance and a measure of the separate variances.
What outputs can be found in SPSS for Pearson correlation?
- Descriptive statistics
- Correlations
- Confidence intervals
What should be included in formally reporting statistical results?
- Type of correlation performed, variables correlated, and direction of relationship.
- Test statistic, df, statistical significance.
- Effect size (Cohen’s r).
What is a caution regarding correlation?
Correlation does not inform us about an underlying causal link.
What is the third variable problem?
There may be other measured or unmeasured variables that affect the results.
What is partial correlation?
Calculates the relationship between 2 variables while controlling for the influence of additional variables.
What must be met to perform a partial correlation?
The assumptions for correlation must also be met.