Correlational Designs

1
Correlational Designs

• Assessing Relationships Between Variables

2
Correlational Designs

• A type of research that tries to discover and
  describe relationships between naturally
  occurring variables.
• A type of descriptive research describes
  relationships.
• Variables are measured, not manipulated
• Is there a relationship between height and
  weight?
• Is obesity related to depression?
• What are the risk factors for ADHD?
• Surveys are one type of correlational research.

3
Correlational Designs

• Allow psychologists to detect when variables vary
  in relation to one another, or co-vary.
• Enable predictions to be made
• Allow us to predict the value of a variable based
  on knowing the other (correlated) variable.
• Utilize a correlation coefficient to express
  the strength and direction of the relationship
  between two variables.
• Can be used to explore relationships and patterns
  among multiple variables.

4
The Correlation Coefficient

• The correlation coefficient statistic r
  indicates the strength (and direction) of a
  relationship between two variables.
• r ranges from 0.00 to or - 1.00
• 0.00 indicates no correlation, or 1.00
  indicates a perfect relationship (strongest
  possible).

5
(No Transcript)
6
Examples of Research Questions that would need a
Correlational Design

• Is amount of time spent watching commercial TV
  associated with increased aggression in children?
• Is hormone replacement therapy associated with
  lowered risk of Altzheimers Disease?
• Is access to positive role models in the
  community associated with better teens
  adjustment?

7
Misinterpreting Correlations

• Note causation is not implied by a correlation!

8
Incorrectly Assuming Causality

• A Common Error assuming that if A and B are
  correlated, then there is a causal relationship
  between the two.
• However, the presence of a correlation between
  variables does NOT imply that a causal
  relationship exists.
• Brown hair and brown eye color are strongly
  correlated. Does one cause the other?

9
Problem with Assuming Directionality in
Correlational Research

• Example Research found literacy and drug use in
  high school students to be strongly inversely
  correlated. A literacy program used this finding
  as follows
• Lets stop drug use in high school students by
  making sure they can all read.
• Is there a problem with this conclusion?

10
Correlation and Causation

• If 2 variables co-vary, there are three
  possibilites about why the co-variation exists
• Variable A causes changes in variable B.
• Variable B causes changes in variable A.
• A 3rd variable is responsible for the association
  seen between A and B.

11
Causation and the Third-Variable Problem

• When two variables are correlated, the reason for
  the relationship may lie with a third variable
  that exists outside the study.
• Taiwanese researchers looked for variables that
  predicted contraceptive use. Among a number of
  behavioral and environmental variables measured,
  the one that correlated most strongly with
  contraceptive use was the number of electrical
  appliances in the household!
• How might this be explained?

12
Grumpy Old Man Example(Why correlation does not
imply causation)

• A psychologist working at the Tranquil Rest
  Senior Residence hypothesized that residents who
  had many visits from family and friends would
  adjust better than those who had few or no visits
  during the prior 3 months.
• Male residents were given a set of questionnaires
  to measure their coping, adjustment, positive
  mood, and other variables.
• Scores on these measures were indeed correlated
  with the number of visits a resident had in the
  previous three months.

13
Grumpy, cont.

• The psychologist concluded that visits from
  family friends functioned to improve residents
  mood, coping, and adjustment to nursing home
  life.
• What do you think of this conclusion? Do the
  data support it? _____
• Why or why not??