general inferential statistics formula
effect / chance
inferential statistics
generated from a sample + used to make inferences about characteristics of the population
error
differences between individual scores in a distribution
z test
probability of a sample having a particular mean score
unbiased statistics
never need to correct, good estimates
standard deviation of the sampling distribution of the mean
standard error
central limit theorem
as long as you have a reasonably large sample size (n=30) the sampling distribution will be normally distributed
standard error symbol
sigma (xbar)
parametric statistics
values inferred from a sample stat to a population stat (normal distribution + interval/ratio)
research vs null hypothesis
h1 vs h0
probability that the test will reject the null if there really is a difference in the world
power ( 1 - β )
type 1 error (α)
reject the null when it is in reality true
decrease in type I errors associated with
increase in type II errors
type 2 error (β)
fail to reject null when it is actually false
nondirectional test critical values
+/- 1.96
(looking to see if less than -1.96 or greater than +1.96)
directional test critical values
+/- 1.645
(looking to see if less than -1.645 or greater than +1.645)
proper z APA format
(M = x, SE = x) … z (n=30) = 2.00), p < 0.05, d = 0.40
effect size
a measure of the absolute magnitude of the effect being studied
when to use t test?
small sample, no population standard deviation (sigma)
S (vxbar)
estimated standard error
single sample t test df
n - 1
for directional test, reject null when…
| tcalc | > tcrit, check means in right direction
tcalc | > tcrit (and check if means in right direction)
independent sample t test
between subject designs - random assignment
dependent/related sample t test
related groups - 1 group at two points in time (within subject / repeated measure)
