Correlation and Experiments

1
Correlation and Experiments

• (Cont. of Psychology is a science.)
• Chapter 1 Myers
• Chapter 2 Barrons

2
Review Last class

• Descriptive Methods
• Case Study
• Survey
• Naturalistic Observation

3
Correlation

• Correlation measures the relationship between
  two variables
• Correlation coefficient a statistical measure of
  the extent to which two factors vary together,
  and thus of how well either factor predicts the
  other. A correlation means that there is an
  association between two factors.

4
Examples of Correlations

• Ex. of some correlations
• Thin people live longer
• Overweight people earn less money than their
  peers
• School uniforms decrease school violence
• Handedness (L or R) and driving skills
• Juvenille Delinquincy and growing up in a single
  parent home
• The Marshmallow Experiment Time
• Oh, The Temptation

5
Correlation

• The correlation coefficient shows two aspects of
  a relationship
• Direction (Positive or negative)
• Strength (how close to 1.00 OR -1.00)
• Correlation coefficients range between
• -1.00 and 1.00
• Number shows Strength
• or shows Direction
• Positive direct relationship
• Negative inverse relationship
• Exercise on this AND Live!Psych

6
Correlation Coefficient
Correlation Coefficients Explanation
1.00 Perfect positive correlation an increase in one factor will always be matched by an equal increase in the other factor.
0.50 Positive correlation as one factor increases the second factor tends to increase. (Can range from 0.99 to 0.01)
0.00 Zero correlation no relationship between the two factors. This is the weakest possible correlation.
- 0.50 Negative correlation as one factor tends to increase, the second factor tends to decrease.
- 1.00 Perfect negative correlation an increase in one factor will always be matched by an equal decrease in the other factor.
7
Correlation and Causation

• Correlation does NOT prove causation
• Correlation indicates the possibility of a cause
  and effect relationship
• Correlation may be due to coincidence
• Ex. Ice Cream consumption and rapes.
• Correlation may be due to other factors
• (confounding variables).
• Q Correlation between juvenille delinquincy and
  growing up in a single parent home?

8
Illusory Correlations

• Illusory Correlations the perception of a
  relationship where none exists
• When I wash my car, it always rains
• As the number of ice cream sales increases, the
  number of murders increases
• Those who adopt a child are more likely to
  conceive.

9
Experiments

• Experiment a research method is one which an
  investigator manipulates one or more factors (IV)
  to observe/measure the effect of some behavior or
  mental process (DV). Only experiments can
  identify cause and effect relationships

10
Experimental Conditions

• Experimental condition to expose a participant
  to treatment (Ex. pill increase focus when taking
  tests).
• Control condition contrast to the experimental
  condition. Serves as a comparison in that this
  subject does not receive treatment OR receives a
  placebo.
• Confounding variable any difference between the
  experimental and control conditions, except for
  the independent variable. One can control for
  confounding variables by random assignment.
  Exercise Ruler Experiment

11
Random Assignment

• Random Assignment (ONLY Used in the experimental
  method) is the process by which subjects are
  placed into an experimental or control group.
  Random assignment limits the effect of
  subject-relevant confounding variables. (outside
  factors of the subject other than the IV).

12
Controlling for all other variables

• Equivalent environments control for
  situation-relevant confounding variables.
• Room temperature, test administrator with a cold
• Random assignment controls for subject-relevant
  confounding variables.
• What if each student got to pick the group in
  which he/she wanted to take part?
• Experimenter bias (when the experiment wants or
  expects certain results) can be eliminated by
  using double blind procedure.

13
Variables

• Independent Variable the experimental factor
  that is manipulated. This is the variable being
  studied.
• Dependent variable the experimental factor that
  is being measured. This is the behavior or the
  mental process. This variable may change
  depending on manipulations in the IV.
• Live Psych slides 3 (IV and DV) and 5.(Op def.)

14
IV and DV

• If one receives more sleep before a test he/she
  performs better
• IV Sleep
• DV Test Scores
• Does the use of color of a football jersey affect
  perception of referees.
• IV watching football plays with a team wearing
  black
• DV amount of penalties called.
• Activity Hunting for Causes

15
Example Experiment

• Hypothesis This pill will lower hyperactivity
  in children with ADHD
• Independent Variable receiving the pill
• Dependent Variable level of hyperactivity
• Experiment Design - Students with ADHD are
  randomly assigned to a group. Exp. Group receives
  pill while the control group receives a placebo
  (fake pill). A month later students level of
  hyperactivity is measured

16
Double Blind

• Blind unaware of what treatment is received
• Double Blind - participants and research staff
  are ignorant about the treatment/placebo given to
  each subject. Double blind procedures can
  eliminate subject and experimenter bias.
• Why could this be beneficial in a study?
• Research expectations can also affect how a
  patient acts/reacts to certain treatment
• Essential in drug research to control for
  self-fulfilling prophecies, placebo effects, or
  biases from the experimenters.

17
Methods of Eliminating Bias

• Random Sample
• Random Assignment
• Placebo (if a experiment is drug related)
• Double Blind
• Replication

18
Placebo Effect

• Placebo Effect - results caused by expectations
  alone.
• Using a Placebo - One group receives the
  treatment (drug being tested) the other receives
  a pseudo treatment (placebo).
• Ex. In the Korean War patients were given
  placebos when they ran out of morphine. Most
  patients reported a reduction in pain.
• Ex. College freshman were given 6 non-alcoholic
  beers to drink in 2 hours. Most exhibited some
  intoxicating behaviors.
• Penn and Teller Placebo Effect

19
Hawthorne Effect

• Research in which workers were monitored to see
  if the amount of light in a room would affect
  worker productivity.
• Under both conditions, performance increased
  because subjects were being watched.
• The Hawthorne Effect states that people tend to
  act atypical having been chosen for an
  experiment.
• The Hawthorne Effect complicates experiment
  design

20
Experiments aim to .

• MANIPULATE the IV
• MEASURE the DV
• CONTROL for all other factors through random
  sampling
• Prove a CAUSE AND EFFECT relationship

21
Correlation vs. Experiment

• In pairs (or by yourself), state whether research
  is correlational or experimental. If its
  correlational state whether its positive or
  negative. If its experimental state the IV and
  the DV.