Revision (content from lecture #7) Flashcards
Causal Relationships
What events cause other events to occur? E.g. X causes Y
Need sufficient evidence to conclude causality
Alternative explanations
Y causes X
Z causes X and Y
Simply correlation – merely a relationship between X and Y
Correlation does not imply causation
Experimental design
Goal Figure out why something happened or how it came about Why Predict what will happen in the future Form the basis of our decisions
Three conditions for causality
To show one event causes another to occur (i.e., X causes Y) 3 conditions must be met
- X and Y must be correlated
- X must precede Y in time
- All other factors (Z) must be ruled out
Questionable cause for relationships
A causal relationship for which no real evidence exists
→ superstition
X must precede Y in time
Assuming that because two things occurred close in time to one another, the first event caused the second
Eg. The rooster crowed and the sun came up. Therefore the rooster made the sun come up.
Directionality problem
Does X cause Y or does Y cause X?
E.g. does self esteem determine academic achievement or vice versa
Third variable problem
Does an outside or third factor (Z) cause both X and Y?
E.g., Intelligence causes self-esteem and academic achievement?
Can you eliminate all “third” variables?
Selection Bias
Participant variables
- Biological, behavioural, psychological characteristics
- -> E.g., personality traits
Environmental variables
–> E.g., place of residence
Selection bias occurs when participant variables leads to selection of a particular environment
- Increases chance of finding a spurious correlation between participant variables and environmental variables
Problem: self-selection in clinical trials
- Need random sampling and random assignment
Causal Chains
- Causes and effects usually appear as parts of more complex patterns, or a causal chain.
- Situation in which one thing leads to another, which then leads to another, and so on.
- Which cause is the real cause?
Contributory Causes
- A number of causes can also act simultaneously to produce an effect
- Each cause contributes to the final effect
Interactive Causes
Causes
- Rarely operate in isolation
- Influence (and are influenced by) other factors
Interactive Causes
- Reciprocal influences
Deductive Reasoning
Reasoning from premises (i.e., reasons) known or assumed to be true to a conclusion that follows necessarily from these premises
Deductive reasoning moves down from known general reasons to specific facts
i.e., “top-down” approach
Inductive Reasoning
Reasoning from premises assumed to be true to a conclusion supported (but not logically) by the premises
- Premises provide evidence that makes it more or less likely (but not certain) that the conclusion is true
Moving from specific observations to broader generalizations and theories
i.e., “bottom up” approach
Empirical Generalisation
Drawing conclusions about a target population based on observing a sample population
When generalising from a sample, need to consider:
- Is the sample known?
- Is the sample sufficient?
- Is the sample representative?
Reasoning in Psychological research
Inductive reasoning
- Draw conclusions about target population from observing sample population
- Construct theories based on observations
== Beware Texas sharpshooter fallacy
Deductive reasoning
- Theories used to make specific predictions (i.e., hypotheses)
Texas Sharpshooter fallacy
- A man in Texas fires random shots at his barn wall, then paints targets around the holes. His neighbours see the barn wall, and are amazed by his accuracy…
→ In Psychology, what could be wrong with doing something like this ?
- Fitting theories to observations (i.e., painting circles around the holes) is not good science
- Good science is conducted by first developing a theory and then gathering observations to test the idea
Fallacious Reasoning
- Unsound arguments that can appear logical
- Often persuasive because they appeal to our emotions
- Often support conclusions we want to believe are accurate
Fallacies of false generalisation
Fallacies of false generalisation:
Hasty generalisation
- Making a general conclusion on the basis of a very small sample
Sweeping generalisation
- Applying a generally accepted conclusion incorrectly in a specific instance
- A glass of wine a night is good for your health
False dilemma
- Forced to choose between two extreme examples without being able to consider additional options
- -> Either/or, black-or-white fallacy