Module 6: Correlation and Experimentation — Key Concepts (6.1–6.4)

6-1 Correlation

  • Describes how two things vary together; used to predict one from the other.
  • Scatterplots: each dot represents a pair of values; slope indicates direction; scatter indicates strength.
  • Correlation coefficient rr ranges from −1.00-1.00 to +1.00+1.00; values near the extremes indicate a stronger relationship; near zero indicate little to no relationship.
  • Positive correlation: two sets of scores rise or fall together (e.g., height and weight).
  • Negative correlation: one set rises as the other falls.
  • Perfect positive: r=+1.00r = +1.00; perfect negative: r=−1.00r = -1.00; no relationship: r=0.00r = 0.00.

6-2 Correlation and Causation

  • Correlation enables prediction but does not establish causation.
  • Causation requires experimental control to rule out alternative explanations and directionality.
  • Third-variable problem: a third factor may influence both variables.
  • Therefore, correlation alone cannot explain why a relationship exists.

6-3 Illusory Correlation and Regression Toward the Mean

  • Illusory correlation: perceiving a relationship where none exists, often due to selective attention/recall.
  • Can fuel illusion of control over chance events (e.g., gambling beliefs).
  • Regression toward the mean: extreme or unusual results tend to move back toward the average on subsequent measurements.
  • Misattributing regression to personal actions can fuel superstitions.
  • Examples: unusually good/bad exam scores tend to regress toward the mean; ESP subjects often show reduced effects on retesting.
  • Third factor can explain apparent correlations (e.g., aging factors like golden anniversaries and baldness).

6-4 Experimentation: Variables, Random Assignment, and Validity

  • Experiment: manipulates one or more independent variables to observe effects on a dependent variable.
  • Independent variable (IV): the factor manipulated by the experimenter; what is varied.
  • Dependent variable (DV): the outcome measured.
  • Experimental group: receives the treatment (level of IV).
  • Control group: does not receive the treatment; serves as a baseline for comparison.
  • Random assignment: participants are allocated to groups by chance to minimize preexisting differences (confounds).
  • Confounding variable: another factor that could influence results if not controlled.
  • Operational definitions: precise definitions of how IV and DV are manipulated/measured to allow replication.
  • Validity: the extent to which the study tests what it intends to test.
  • Example (breastfeeding): randomized promotion of breastfeeding showed higher IQ at age 6 in the experimental group, suggesting a causal effect of nutrition.
  • Placebo effect: improvements attributable to expectations rather than the treatment itself; controlled via placebo.
  • Double-blind procedure: neither participants nor researchers know who receives the treatment, reducing bias.
  • AP Exam tip: identifying independent and dependent variables is a highly tested concept in psychology.

6-5 Placebo Effect and Blinding (Procedural Control)

  • Placebo effect: outcomes produced by expectations rather than the active treatment.
  • Blinding (single or double): helps ensure observed effects are due to the treatment, not expectations or researcher bias.
  • Example: placebo pills can produce measurable changes in pain, mood, or performance when participants believe they are receiving active treatment.