experimental design Flashcards
external validity
how well findings can apply to real world. Sampling bias can affect this
internal validity
confounders, Hawthorne Effect can affect this
predictive validity
how well test predict future performance
Hawthorne Effect (also called demand characteristics)
people adapt responses based on social norms or what they think researchers’ expectations are. Comes from lack of blinding
Lack of reliability
stems from error in measurement tools
sampling bias
basically selection bias (sample not representative of population)
attrition effects
participants drop out of study
demand characteristics
participants guess what study is about; subconsciously respond to be consistent with hypothesis
situational effects
the research environment changes outcome (e.g. MRI machine is claustrophobic)
statistical power
can be lacking when sample is too small
pearson correlation
from -1 to +1, expressing both +/- of correlation, as well as strength (close to zero is weak)
ethnographic studies
qualitative method, you immerse yourself in the lives of the people you study
construct validity
do the tools measure what they purport to measure?
response bias
respondents might not have perfect insight into their internal state and reply falsely
type 2 error
false negative
type 1 error
false positive
p-value
probability that the observed difference was due to chance or fluke rather than true correlation (ought to be less than 0.05)
between subjects design
compares between 2 groups of subjects; whereas within-subjects design may follow one subject before and after
phenomenological studies
what individuals report about their personal experiences
values vs beliefs
values are a culture’s standards for evaulating good and bad; beliefs are principles that people hold
sociobiology
study of how biology and evolution determine behaviour