formula for confidence interval
sample mean +or- t critical (SD/sqrtn)
as sample size increases
standard error of the mean decreases
As standard deviation increases...
standard error increases
assumptions for one sample t test
normality, random sampling, independent observations
confidence interval widens as confidence level increases
because t critical value will increase, the higher the confidence level the larger the margin of error.
o
population standard deviation
s
sample standard deviation
the mean of the standard normal distribution is
always 0
the mean of the normal distribution
not always 0
large degree of freedom
thinner distribution
as n gets larger
t critical will get smaller
For a normal model, about what percentage of scores fall within 3 standard deviations of the mean?
99.7%
if extreme values change
it is possible for the mode to change data value occurs more frequently after the change
how to compute new mean after the addition of a value
multiply original n and mean, then add the new value. Then divide by new population size
affect of adding new value on large n vs small n
the addition of the same value to a smaller data set will effect the data more. EX: Dividing by 51 instead of 50 has less impact than dividing by 6 instead of 5
when the mean is greater than the median
skewed right, extreme upper values
When the mean is less than the median
skewed left
does removing extreme values have more effect on the mean or the median
more effect on the mean, extreme values do not effect median greatly
percent score
numeric, continuous
year of birth
discrete, numeric
t stat
compares two means
z score
distance from observation to the mean