Hypothesis Testing Flashcards
observed differences are due to random variation rather than systematic differences
null hypothesis - hypothesis of no difference
there are real systematic differences between observed groups
alternative hypothesis - hypothesis of difference
null vs alternative hypothesis
hypothesis testing
hypothesis test that does not assume direction
two tailed test
hypothesis test that does assume direction
one tailed test
(sample stat - value of population parameter under Ho) / std. error of sample statistic
test statistic
what are the 4 probability distributions
t, z, chi squared, and f test
evidence required to reject the null
the level of significance
reject the null when the null is false
correct decsion
reject the null when the null is true
type 1 error
do not reject the null when the null is false
type 2 error
do not reject the null when the null is true
correct decision
how to define the level of significance
researcher decides
probability of correctly rejecting the null hypothesis
power of the test
formula for power of a test
1-beta or probability of a type 2 error
whats the trade off between significance and power
as the likelihood of a type 1 error increases(decreases) the likelihood of a type 2 error decreases(increases)
how to decrease type I and II errors
increase power
p-value
value of significance below which we would reject the null hypothesis
t test for mean works if: (2)
sample is large or normally distributed
z-test distribution mean testing:
know population variance or not
test concerning difference of sample means
t test where distribution is normal and population variance is equal
test concerning a single variance
chi-squared
test concerning equality or inequality of 2 variances
F-test
tests of correlation
t test
spearman rank correlation
how to calculate spearman rank correlation
calculate the differences in ranks between x and y
parametric vs non-parametric test
parametric - concerned with parameters or specific assumptions
non-parametric - use when don’t know about population distribution