Week 1 - methods Flashcards
Hindsight bias
People’s tendency to be overconfident about whether they could have predicted an outcome
what do we test
In social psychology, we collect data to systematically test intuitions, possible causes and explanations of thought and behaviour.
Social psychology is not a science
Science is defined by the METHOD
Scientific method
1) Form a theory:
Propositions that describe or explain the problem or phenomenon
2) Form a hypothesis:
Prediction about what will happen under particular circumstances
3) Test the hypothesis
4) Accept or Reject the hypothesis
Correlation designs
- Examines the relationship between 2 or more variables
- All variables are measured as they naturally occur in the real world
Correlations Do Not Imply Causation!
A correlation is a necessary but insufficient condition for establishing causality
Correlations Do Not Imply Causation because?
- Reverse causation
- Third variable
- Self-selection (people in the sample might already exhibit the desired specific traits that affect the result)
Value of Correlational Designs
- Suggests possible causal relationships that exist in the world
- It is the best alternative when an experimental study is difficult or unethical to conduct
- Longitudinal studies that measure changes in responses (e.g., behaviour, thoughts, emotions) over time can rule out reverse causation
Experimental designs
Requires an independent and dependent variable
Independent variable
Proposed causal variable
(1) Manipulated by the researcher to create conditions
(2) Participants are randomly assigned to different conditions
Dependent variable
Outcome variable that is measured AFTER the IV
Manipulation of independent variable
- Experimental vs. Control condition
- Control condition: A condition comparable to the experimental condition in every way except in one key characteristic that is expected to result in differences in the dependent variable
- Rules out reverse causality
Random assignment of participants to conditions
- Participants are put into different conditions based on chance, which ensures equal likelihood of being placed into any of the conditions.
- Remove the effect of individual differences (age, gender, mood, personality) between participants that are not of interest
- Rules out third variables
Quasi-Experiments
- Quasi-experiments are similar to experiments, except that you rely on existing group memberships, and there is no random assignment
- Because random assignment is not used, third variables/confounds cannot be ruled out in quasi-experiments
When can you make a causal claim?
Conducted a study that uses an experimental design, which includes:
Manipulation of independent variables & Random assignment