Correlation Correlation: a measure of the extent to which two events vary together, and thus how well either predicts the other. The correlation coefficient expresses the relationship mathematically, ranging from -1 to +1 Positive correlation, noted as a number from zero (no correlation) to 1, means that variables increase and decrease together, like shoe size and height. Negative correlation, a number from 0 to -1, means that one variable going up predicts the other one going down, like self-esteem and depression scores
Correlation coefficient When we have information of people with different heights and temperament scores, we can plot them on a graph. This graph, a scatterplot, shows a positive correlation (upward slope) but the correlation is low (the dots are not close to forming a line).
Correlation and Causation Correlation does not equal causation!
Illusory Correlation The perception of a relationship where no relationship actually exists. For example, we may notice the times that odd behavior happens with a full moon, and develop the illusion that the two events generally tend to go together.
Exploring Cause and Effect Experimentation Exploring Cause and Effect As in other sciences, experimentation is the backbone of psychological research. Experiments isolate causes and their effects. Preview Question 9: How do experiments clarify or reveal cause-effect relationships?
Exploring Cause & Effect Many factors influence behavior. Experiments (1) manipulate factors that interest us, while other factors are kept under (2) control. When we see an effect from manipulating only one variable at a time, we have isolated and discovered a cause-effect relationship.
Experiments Many factors may influence a phenomenon In an experiment, researchers vary one or more factors (independent variables) to observe the effect on some behavior or mental process (dependent variable)
Experiments If possible, participants should be randomly assigned to control and experimental groups Experimental group: exposed to the treatment (one version of the IV) Control group: contrasts the experimental group, serves as comparison to assess the effects of the treatment
Effects of mother’s milk (Lucas et al., 1992) 424 hospital preterm infants randomly assigned to either formula feedings (control) or breast-milk (experimental) Tested at age 8, the breast-milk group scored higher on intelligence tests Random assignment make sure no other difference between the group (mother’s age, intelligence, etc) explains the difference in intelligence scores
Experimentation
Summary: Value of Experiments Correlational studies uncover naturally occurring relationships Unclear about causality Experiments manipulate factors to determine their effects
Tools for Controlling Bias Placebo effect: Expectations about treatment effects can translate to real effects Control groups may be given a placebo – an inactive substance in place of the experimental treatment Many studies are double-blind – neither participants nor research staff knows which participants are in the experimental or control groups.
Comparing Research Methods Ed. Portal Review