Correlational Research

Correlational research studies the relationship between two variables that are measured, not manipulated.
Supports association claims, indicating how one variable relates to another.
- Example: Research concluded that MRIs show screen time is linked to lower brain development in preschoolers.

Key Feature: No Manipulated Variables
- Only measured variables are assessed for relationships.
Assessment of Association:
- The extent to which one variable relates to another.

Results of correlational studies are frequently depicted using scatter plots.
Scatter Plot Explanation:
- One variable is plotted on the x-axis and another variable on the y-axis.
- Example of a scatter plot from the MRI screen time study.
- Individual points can represent subjects (e.g., preschoolers).
- Kid A's screen time (4.5 hours per week) corresponds to a score of around 50 on brain development.

Line of Best Fit:
- Closer points to a line of best fit indicate a stronger correlation.
- The correlation coefficient (r) measures the clustering of points around this line.
- It ranges from +1 (perfect positive correlation) to -1 (perfect negative correlation).
- Examples with correlation coefficients:
- Positive ( r = 0.9 )
- Negative ( r = -0.9 )
- Both have equal strength but different directions.

Correlation Coefficient (r):
- Indicates strength and direction of the correlation:
- Positive correlation: high scores in one variable are related to high scores in another variable.
- Negative correlation: high scores in one variable are related to low scores in another.
- Zero association: no relationship exists between variables.
Cohen's Guidelines for Interpreting Correlation Strength:
- Small correlation: ( r = 0.1 ) or ( r = -0.1 )
- Medium correlation: ( r = 0.3 ) or ( r = -0.3 )
- Large correlation: ( r = 0.5 ) or ( r = -0.5 )

Positive Association:
- Example: The relationship between expressed gratitude in relationships and longevity.
- Points clustered high on both axes indicates positive correlation.
Negative Association:
- Example: Frequency of multitasking vs. skill at multitasking.
- As one increases, the other decreases.
Zero Association:
- Example: Time of dinner and childhood weight.
- A scatter plot shows no pattern, just randomness (no correlation).

Analyzing scatterplots to determine correlation magnitude:
- Example scatterplot with ( r = -0.19 ): suggests weak to moderate correlation.
Different scatterplots illustrating various strengths of correlation:
- Zero correlation: cloud of dots.
- Negative 0.3 correlation: slightly clustered points.
- Positive 0.5 correlation: tighter clustering of points in the positive direction.
- Negative 0.7 and positive 0.9 with very tight clustering:
  - ( r = 0.99 ) indicates very strong association.

Correlational studies may include:
- Both variables measured on continuous scales (1 to 100).
- One variable as a categorical variable.
- Example: Sips of beer given to kids is a categorical variable (yes or no).
Scatter Plots vs. Bar Graphs:
- Scatter plots are less common for studies with categorical variables.
- Bar graphs often used to show means of groups (e.g., mean drinks per night for kids given sips vs. those not).

A study is characterized as correlational if it has measured variables (no manipulation involved).
Example of study using sips vs. no sips correlating with teenage drinking elucidates association claims.
Visual representations can include either scatter plots or bar graphs depending on the nature of the data.