Monday, September 9, 2013
Psych 2C03 Lecture 1: The Scientific Method
Do Scientists "Prove" Stuff?
We do not prove stuff, we give it supporting evidence
Statistics never tell you 100% you're right, it's usually about 95%
We just make it seem very likely that a theory is correct
Observation + Theory --> Hypothesis + Predictions --> Test --> If test confirms, adds to theory and
observations --> Observation + Theory
Observation + Theory --> Hypothesis + Predictions --> Test --> If test fails to confirm, must modify
Ways of Testing Hypotheses: Data Collection Decisions
Experimental control: controlled everything so can make causal statement at the end
Lose generalizability (whether what you found in the experiment would exist in the real world)
Lab and some field experiments has a lot of control
Correlational Studies and field (observational studies) not a lot of control, but doing stuff in the
How are a set of variables related to the other?
o Ie. How is cloud cover related to precipitation? (so a relationship)
How does your score on a survey on extraversion relate to you going to a party?
Some uncontrolled variables may make something look like they're not correlated
Some variables might affect the relationship
o Ie. May have a full time job + school, so does not go to parties
Negative correlation: when one variable goes up, the other variable goes down
Positive correlation: when one variable goes up, the other goes up
Correlations Does Not Equal Causality
There are probably variables you haven't measured Monday, September 9, 2013
The "Third Variable" problem
o As violent crime increases, so does ice cream sales
o Therefore, ice cream makes people violent
o What causes what?
o Maybe it's hot and people turn aggressive and eat more ice cream as well
Experiments (Ss = Subjects)
Experimental controls helps you make causal statements
Use a cover story (deception) for the subjects to establish control on their behaviour
The person who's job is to d