Null Hypotheses, Rejection Regions. and Formulas,
Null hypotheses:
One proportion-- H0 : p = p0 .
Two proportions-- H0 : p1 = p2
Three or more proportions: H0 : p1 = p2 = p3 = . pk
One mean: H0 : mu = mu 0.
Two means: H0 : mu1 = mu2
Three or more means: H0 : mu1 = mu2 = . = muk
Rejection Regions (given for the three possible alternatives in the order
a. ' < ' ; b. ' > '; and c. ' ¹ ' )
1. One proportion p:
a. phat < c (that is, if phat is 'small')
b. phat > c (that is, if phat is 'large')
c. phat < c1 or phat > c2 (either much
smaller or much larger than p0)
*phat is the sample proportion.
2. Two proportions:
a. p1hat - p2hat < c
b. p1hat - p2hat > c
c. p1hat - p2hat < c1 or p1hat - p2hat > c2
*p1hat and p2hat are the two sample proportions.
3. One population mean mu :
a. xbar < c
b. xbar > c
c. xbar < c1 or xbar > c2
4. Two population means mu1 and mu2 :
a. xbar - ybar < c
b. xbar - ybar > c
c. xbar - ybar <c1 or xbar - ybar > c2
5. Three or more proportions: reject null
hypothesis if p-value is
less than alpha = level of significance.
6. Three or more means: reject null hypothesis if
p-value is less than
alpha = level of significance.
Sample mean: xbar = (sum x ) / n ; IQR = Interquartile
Range = Q3 - Q1
5# summary: min Q1 median Q3 max
Sample median: Arrange observations in order. If n is odd, median is
middle observation. If n is even, median is the average of the middle two.
Sample standard deviation s: First calculate s2 = [sum(
x2) - (sum x
)2 / n ] / (n - 1). Then s = sqrt (s2 )
x2: square the x's and then sum them. (sum x) 2: add up
the x's and get
the sum. then square the sum.
Population mean m : m = sum[ x p(x)]
where p(x) is the probability
of x.
Population standard deviation sigma :
First calculate sigma squared:
sigma squared = sum [ x2 p(x)] - (mu- squared) = E(X2) - mu squared
.
Then sigma = sqrt(sima squared ).
x2: x squared