W1 Flashcards
What are the ways correlation can occur?
- When X causes Y
- When Y causes X
- When Z causes Y – this is when there is a third variable that causes both X and Y
- Spurious correlation
What is reverse correlation?
When it seems that X causes Y, but in reality, it is Y that causes X
What is spurious correlation
When 2 variables seem to be correlated but the relationship is meaningless like cats and cars
What are the 3 necessary conditions for causality?
- There is a relationship between X and Y
- X cannot happen after Y
- All other possible causal factors are eliminated. So this means other possible causes are held cosntant
What is done to the independent variable and what are the 2 ways of applying the independent variable?
The independent variable is manipulated. It can be done in 2 ways:
1. Between subject -> one participant is assigned to one experimental condition of the IV
2. Within subject -> one participant is assigned to several experimental conditions
What is a stimulus?
It is the event/object to which a response is measured.
So sort of like what is changed
e.g visual stimuli, price, verbal stimuli, etc
What are confounding variables?
these are variables that have affected the relationship between IV and DV, on top of what is due to the effects of X on Y
What is internal validity?
it is whether the conclusions about the effects of IVs on DVs are valid in the sample
It can be ensured by randomization, control of extraneous factors, etc
what is external validity
it is whether conclusions can be generalized outside the experiment
Field studies are higher in external validity
What is Randomized control trial
This is an experiment and has 3 types:
RCT in parallel groups -> basically a between subject experiment
RCT in crossover: random allocation into different groups and measured several times so mix of within and pre-post
RCT in cluster -> pre-existing groups of participants are randomly selected
What do experimental data lead to?
Experimental data can infer causality
What are experiments with issues to causal claims
Quasi-experiment where assignemnt to the experimental conditions is not random
Pre-post -> the DV is measured before and after an IV category. it is a non-experimental observational study
what does observational data infer?
It cannot infer causality
why is it good to replicate?
- to see if the effect is really there
- to see whether the effect is robust across situations
- build on previous experiments and see whether your epxlanations make sense
What is A/B testing
This is a randomized experimentation process where 2 versions of something are comapred against each other
