Null Hypothesis/Significance Testing Flashcards
Define inferential stats
Tell us if the IV is affecting the DV
What is the purpose of stats in psychology?
see if data supports hypothesis
Make conclusions based on empiricism
make inferences about reasoning behind things
What is variability?
the less overlap between groups, the more they differ
What is a key point when selecting sample size?
the bigger, the less error in estimating the population
Relying solely on _________ without sample size is a big mistake.
variability
We use statistics to confirm if observed group differences from a sample are likely to reflect what?
True differences in the population
We have found support for our hypothesis when…?
When the differences in the comparing groups are sufficient enough that it is unlikely that they happened by chance.
What is the probability theory used to determine?
The likelihood that there is no true difference in pop. based upon what we used in sample.
What is the null hypothesis?
Says the IV had no effect on the DV
There is no difference between groups and no association btw variables
What does the alternative hypothesis say?
The opposite of the null
If we reject the null, what does it mean?
There is a true difference
If we fail to reject the null, what does that mean?
It’s unlikely that there is a true difference
What are some limits of NHST?
Cannot prove causality
Cannot say anything about magnitude of an effect, only that there is an effect.
What is the p-value?
The probability of getting the observed effect if the null was true (probability that result occurred by chance)
What is the alpha level?
the p value threshold that needs to be crossed in order to reach statistical significance
For the alpha level, probability should be less than how much?
5 % chance that the observed effect would occur if there were no real effect in the population
In general, the _______ the p value the greater the chance the effect is “real” and will be replicated in another study.
smaller
What can you expect if you set your p at .01 instead of the usual .05?
the more analyses you do, the greater likelihood of catching something significant, even if it’s not a true effect
Alpha level is the threshold p needs to be under to ________ the null hypothesis
reject
P
Reject null, accept alternative hypothesis
P > .05?
Fail to reject null, can’t support the alternative hypothesis
If the difference between groups is likely to be a true difference?
Compare and evaluate group means and SD
if the association between variables is like to be a true association
Compare and evaluate correlations
We are testing the probability that ______ is true.
null hypothesis
Give an example of the test of probability for the null
Probability of finding the difference in our sample if, at the true population level, there is no difference between estimated and actual calories consumed
Define p
Probability less than 5% that the observed effect would occur if there were no real effect in the population.
The _________ the probability, the greater the chance that the effect is real and will be replicated in another study.
smaller
If p
reject
if > .05 you must __________ the null.
fail to reject the null
What most common type of error occurs from running many analyses?
Type 1 error