Week 2 Flashcards
In a study, what are yes/no responses influenced by?
bias and
sensitivity
how do you measure test sensitivity?
calculate d prime (d’)
What statistical information is d prime based on?
the z-scores of the
proportion of “yes” responses
why do we trim data?
to remove outliers
what is a tests power?
the likelihood that the test will produce a statistically significant result
Power is a function of which factors?
- the significance criterion (p-value)
- the sample size
- the population effect size
Can any given test have more than one power value?
yes - power is computed for a potential population effect
what is observed power?
the power of a test assuming the population effect size = observed current sample’s effect size
Why is calculating the observed power not useful?
Because if we find insignificant results, the power will be low
when are post hoc power analyses useful?
when they are based on population effect size of independent interest
meaning, if your estimated power is lower than that of other literature on the topic, you can use the power referenced in that literature
Whay conclusions can you not make if your analysis finds a nonsignificant result?
- the population effect size = 0
- null hypothesis = true
- larger N =significant result
What is the implicit association test (IAT)?
a measure of implicit attitudes
The association between which measures is assessed in the IAT?
the association between target-concept discrimination and attitude discrimination
explain the 5 steps of the IAT
- initial target-concept discrimination
- initial attribute discrimination
- initial combinedtask
- reversed target-concept discrimination
- reversed combined task
how do you measure an IAT effect?
noncompatible task RT - compatible task RT
