sampling distribution via hypothesis testing
Statistics have a predictable distribution called a sampling distribution. The
sampling distribution allows us to quantify the variability in sample statistics.
This allows us to calculate a p-value, which is the probability of observing a test
statistic
to find an increase in y with a given x
ONLY do mx part of equation
ex: given y = 3 + 5x; this indicates that the value of x increases by 4, the predicted value of y would INCREASE BY: 20 (5 x 4)
when want to compare outcomes
ex: changes in cholesterol from day 1 and day 7 on medicine used a paired difference t-test for the change in cholesterol
top 25%
#NAME?
first quartile
lower 25%, take subject number (n) divide by 4 take that number and find where the results are which is your answer
if add one point to data
mean and median both increase, SD unchanged unless score is multiplied or divided
sampling distribution of a stat
the distribution of values taken by a statistic in ALL possible samples of the same size from the SAME population
if do any manipulation of any kind and not NATURAL
experiment
again when ADD # to all subjects
mean and median increase by that # and SD is unchanged
remember with hypothesis testing doing differences
double check greater than or less than if want second variable to be higher than first, will be LESS then NOT GREATER
chance problems
use basic standard z scores,
ex: P (x>10)
(P 13> x <14) --> P(Z<0.5) - P(Z<0.25) {upper value - lower value of p score)
chance problems w/ n included
find equation on sheet that incorporates all variables
same equation but as the divider = SD /square root of n
b1
m in the equation, slope
two sided with z
(1-p value) x 2
LINEAR REGRESSION
NOT APPROPRIATE IF DATA IN SCATTER PLOT HAS A CURVED RELATIONSHIP
population of interest
group of everyone w/ or w/out what you're looking at
parameter of interest
people who have w/e you are looking for
outliers
have a greater impact on the mean than the median
normal distribution
mean falls w/in the middle of the max - min range