Please enable JavaScript.
Coggle requires JavaScript to display documents.
Preliminary Issues Regarding Research (Bases for knowing things (intuition…
Preliminary Issues Regarding Research
Bases for knowing things
intuition
expertise/ authority
logic
anecdites/ personal impressions
scientific research
Why do formal research
test and justify our claims (hypotheses)
discard propositions that are false- fausitility (things want to test/ can be measured)
claims about the causes of behaviour and the effectiveness of treatment (can be more control in lab)
Should we always use experimental methods
some factor cannot be manipualted
unethical to manipulate
passive observational can have better external validity
some legitimate and important questions do not concerncause and effect
Why published details of method and analyses and results
conclusion can be scrutinised by others (gping smth sensible/ make sense)
can determine the justifiability of others' conclusions
allow us to judge the strength of effects (effect size, manipulation)
make psychologists accountable
Why learn about research
allow us to be discerning consumers of the products of others' researches
Research varies enormously in quality
truthfulness of the results and conclusions depends on:
quality of design (treatment)
quality of measures (tools/ questionnaire accuracy)
appropriateness of the statistical analyses
nature (size of) the effect- power, sample size
match between the results and conclusions
rival hypothesis
= another hypothesis that different from what researcher hypothesized (another explanations
Causality and Internal Validity
inferring causality
experimental (cause and effect conclusions)
Essential conditions for establishing causation
covariation
= 2 events must occur/ changes together
cause must procede the effect
elimination of plausible alternative explanations for the covariation (rival explanations)
non-experimental- suggest cause-and-effect relationships or to demonstrate effects that would be consistent with cause-and-effect relationships
CANNOT
claim to demonstrate causality
External validity: generalize to different population
Problems with interpreting correlation
, cannot be sure whether:
A caused B
B caused A
something else was responsible for the relationship
Good experimental design
hold everything constant except for IV
able to exclude alternative possible explanations