Week 11 Flashcards
research is designed to establish cause and effect
if we don’t use an experimental design to research a hypothesis, it may not be possible to establish cause and effect
causation
does one variable cause the change of another
correlation
two variables, one alters the other
why cant a correlation be a causation?
may be due to other factors
results may be a fluke -random correlates
a causes b or b causes a
reverse causality
when either direction is possible
don’t know if a causes b or b causes a
therefore more research is needed for causation to be established
sample bias
we generally rely on people willing to take part
if people don’t have time or aren’t willing our sample may bot be generalisable and may lead to bias
low population validity can be improved
change recruitment strategies
compensation
key measures survey for those that declined to take part
non-engaging client forms
anonymous
collects no personal data
offers an alternative for those that may not be as willing to participate
no consent
no data collection
data protection act and ethics
convince samples
usually psychology students
cross cultural studies
some behaviours are universal, others are culturally different
these studies show similarities and differences
representative samples
should consider
- age
- socio-economic groups
- gender
- income
- education levels
nominal data
categorical data
categorical data
putting observations into discrete categories
cannot be ranked, they are just different
chi squared goodness of fit
used when single nominal variables with 2 or more categories
we have sorted them
