Download MAT 222 FORMULA CARD

Survey
yes no Was this document useful for you?
   Thank you for your participation!

* Your assessment is very important for improving the work of artificial intelligence, which forms the content of this project

Document related concepts
no text concepts found
Transcript
MAT 222 FORMULA CARD
Standard score: z =
Two-sample t statistic for H0 : ¹1 = ¹2
(¾1 ; ¾2 unknown):
x¡¹
¾
Standard score for sample mean x
¹: z =
x
¹¡¹
p
¾= n
t=
df = minimum of n1 ¡ 1, n2 ¡ 1
Chapter 6: Introduction to Inference
Two-Sample Con¯dence interval for ¹1 ¡ ¹2 :
s
s2
s21
¤
(¹
x1 ¡ x
¹2 ) § t
+ 2;
n1
n2
Con¯dence interval for mean ¹ (¾ known):
¾
x
¹ § z¤ p ;
n
z¤
C
1.645
90%
1.960
95%
2.576
99%
df = minimum of n1 ¡ 1, n2 ¡ 1
Sample size for con¯dence interval for ¹ with
margin of error m:
n=
(¹
x1 ¡ x
¹2 ) ¡ (¹1 ¡ ¹2 )
r 2
;
s22
s1
+
n1
n2
Pooled two-sample estimator of ¾ 2 :
h z ¤ ¾ i2
s2p =
m
(n1 ¡ 1)s21 + (n2 ¡ 1)s22
n1 + n2 ¡ 2
z Statistic for H0 : ¹ = ¹0 (¾ known):
z=
Pooled two-sample t statistic for H0 : ¹1 = ¹2 when
¾1 = ¾2 :
x
¹ ¡ ¹0
p
¾= n
t=
s
Standard error of x
¹: SEx¹ = p
n
Con¯dence interval for mean ¹ (¾ unknown):
Two-sample F statistic for ¾1 = ¾2 :
df= n ¡ 1
F =
x
¹ ¡ ¹0
p ;
s= n
s21
;
s22
df = n1 ¡ 1 (numerator) and n2 ¡ 1 (denominator).
One-sample t Statistic for H0 : ¹ = ¹0 (¾ unknown):
t=
df = n1 + n2 ¡ 2
Pooled two-sample con¯dence interval for ¹1 ¡ ¹2
when ¾1 = ¾2 :
r
1
1
¹2 ) § t¤ sp
+
;
df = n1 + n2 ¡ 2
(¹
x1 ¡ x
n1
n2
Chapter 7: Inference for Distributions
s
x
¹ § t¤ p ;
n
x
¹1 ¡ x
¹2
q
;
1
sp n1 + n12
df = n ¡ 1
Chapter 8: Inference for Proportions
Two-sample z statistic for H0 : ¹1 = ¹2
(¾1 ; ¾2 known):
z=
Sample proportion: p^ = X=n, X = number of \successes"
(¹
x1 ¡ x
¹2 ) ¡ (¹1 ¡ ¹2 )
r 2
¾2
¾1
+ 2
n1
n2
Standard error of p^: SEp^ =
1
r
p^(1 ¡ p^)
n
Chi-square test statistic:
Con¯dence interval for p:
X2 =
¤
p^ § z SEp^
X (observed ¡ expected)2
expected
df = (# of rows { 1)(# of columns { 1)
z statistic for H0 : p = p0
z= r
p^ ¡ p0
Chapter 10: Inference for Regression
p0 (1 ¡ p0 )
n
1 P
(xi ¡ x
¹) 2
n¡1
1 P
Sample variance of y's: s2y =
(yi ¡ y¹)2
n¡1
Sample variance of x's: s2x =
Sample size for desired margin of error m:
n=(
or
z¤ 2 ¤
) p (1 ¡ p¤ )
m
z¤ 2
)
n=(
2m
(p¤ = guessed value)
Sample correlation:
1 X xi ¡ x
¹ yi ¡ y¹
r=
)(
)=
(
n¡1
sx
sy
1
(conservative approach with p = )
2
¤
Di®erence of sample proportions: D = p^1 ¡ p^2
Regression line: y^ = b0 + b1 x,
Standard error of sample di®erence D:
r
p^2 (1 ¡ p^2 )
p^1 (1 ¡ p^1 )
+
SED =
n1
n2
Slope: b1 = r
Mean squared error:
s2 =
(^
p1 ¡ p^2 ) § z ¤ SED
1 X 2
ei ;
n¡2
Standard error of b0 :
s
SEb0 = s
Pooled estimator of p when p1 = p2 :
where ei = yi ¡ y^i
1
x
¹2
+P
n
¹ )2
(xi ¡ x
Level C con¯dence interval for b0 :
X1 + X2
n1 + n2
b0 § t¤ SEb0 ; df = n ¡ 2
Standard error of D under H0 : p1 = p2 :
r
1
1
SEDp = p^(1 ¡ p^)(
+
)
n1
n2
Standard error of b1 :
s
SEb1 = pP
¹) 2
(xi ¡ x
z statistic for H0 : p1 = p2 :
z=
sy
sx
Intercept: b0 = y¹ ¡ b1 x
¹
Con¯dence interval for p1 ¡ p2 :
p^ =
sP
(^
y ¡ y¹)2
P i
(yi ¡ y¹)2
Level C con¯dence interval for b1 :
b1 § t¤ SEb1 ; df = n ¡ 2
p^1 ¡ p^2
SEDp
Test statistic for H0 : ¯1 = 0 :
Chapter 9: Inference for Two-Way Tables
t=
b1
; df = n ¡ 2
SEb1
Expected cell counts:
Estimate for mean response ¹y when x = x¤ :
row total £ column total
expected cell count =
n
¹
^y = b0 + b1 x¤
2
Standard error of ¹
^y when x = x¤ :
s
(x¤ ¡ x
1
¹) 2
SE¹^ = s
+P
n
¹)2
(xi ¡ x
Test statistic for H0 : ¯j = 0:
t=
bj
; df = n ¡ p ¡ 1:
SEbj
Level C con¯dence interval for ¹y when x = x¤ :
¹
^y § t¤ SE¹^ ;
Sum of squares SS: SST = SSM + SSE
df = n ¡ 2
Estimate for future observation of y when x = x¤ :
Degrees of freedom DF:
¤
y^ = b0 + b1 x
DFT = DFM + DFE,
Standard error of y^ when x = x¤ :
s
1
(x¤ ¡ x
¹)2
SEy^ = s 1 + + P
n
¹ )2
(xi ¡ x
DFT = n ¡ 1, DFM = p, DFE = n ¡ p ¡ 1,
SSM
DFM
SSE
Mean square error: MSE =
DFE
Mean square model: MSM =
Level C prediction interval for y when x = x¤ :
y^ § t¤ SEy^ ;
Sum of Squares
X
SST =
(yi ¡ y¹)2 ;
df = n ¡ 2
Test statistic for H0 : ¯1 = ¯2 = ¢ ¢ ¢ = ¯p = 0:
(Total Sum of Squares)
X
(^
yi ¡ y¹)2 ; (Model Sum of Squares)
X
SSE =
(yi ¡ y^i )2 ; (Error Sum of Squares)
Squared multiple correlation: R2 =
SSM =
SST=SSM+SSE
sum of squares
MS =
degrees pf freedoms
One-way ANOVA model:
xij = ¹i + "ij ;
MSM
SSM/DFM
=
; df = (1; n ¡ 2)
MSE
SSE/DFE
for i = 1; : : : ; I and j = 1; : : : ; ni ; and N = n1 + ¢ ¢ ¢ + nI .
p
r n¡2
Test statistic for H0 : ½ = 0: t = p
1 ¡ r2
df = n ¡ 2.
Pooled-sample variance:
s2p =
(n1 ¡ 1)s21 + ¢ ¢ ¢ (nI ¡ 1)s2I
= MSE
(n1 ¡ 1) + ¢ ¢ ¢ + (nI ¡ 1)
Sum of squares (SS): SST = SSG + SSE
Chapter 11: Multiple Regression
SSG
Multiple Regression Model:
=
X
ni (¹
xi ¡ x
¹)2
X
(ni ¡ 1)s2i
allgroups
yi = ¯0 + ¯1 xi1 + ¯2 xi2 + ¢ ¢ ¢ + ¯p xip + "i
SSE
Least squares estimates of ¯0 ; ¯1 ; : : : ; ¯p :
=
allgroups
Degrees of freedom (DF):
b0 ; b1 ; : : : ; bp
Estimate of ¾: s =
SSM
SST
Chapter 12: One-way ANOVA
The ANOVA F test for H0 : ¯1 = 0
F =
MSM
; df = (p; n ¡ p ¡ 1):
MSE
F =
DF T = DF G + DF E;
p
MSE.
where DFT = N ¡ 1, DFG = I ¡ 1, DFE = N ¡ I.
Level C con¯dence interval for ¯j :
Mean square (MS): MSG =
¤
bj § t SEbj ; df = n ¡ p ¡ 1
3
SSG
SSE
, MSE =
DFG
DFE
DFT = DFA + DFB + DFAB + DFE,
Test statistic for H0 : ¹1 = ¹2 = ¢ ¢ ¢ = ¹I :
F =
DFM = DFA + DFB + DFAB ,
MSG
; df = (I ¡ 1; N ¡ I):
MSE
where
DFT = N ¡ 1,
DFA = I ¡ 1,
DFB = J ¡ 1,
DFAB = (I ¡ 1)(J ¡ 1),
DFE = N-IJ.
SSG
SST
P
P
Population contrast: Ã =
ai ¹i , where
ai = 0
Coe±cient of determination: R2 =
Sample contrast: c =
P
ai x¹i
Mean square (MS): For the factors A and B, for the
interaction AB, and for the error E:
Standard error of c :
SEc = sp
s
X a2i
MS =
ni
Test statistic for H0 : Main e®ect of A is zero:
Test statistic for H0 : Ã = 0:
t=
c
; df = N ¡ I
SEc
F =
Level C con¯dence interval for Ã:
c § t SEc ; df = N ¡ I:
F =
Multiple Comparisons t statistic:
tij
MSA
;
MSE
df = (I ¡ 1; N ¡ IJ):
Test statistic for H0 : Main e®ect of B is zero:
¤
x
¹i ¡ x
¹j
= q
;
1
sp ni + n1j
SS
DF
MSB
;
MSE
df = (J ¡ 1; N ¡ IJ):
Test statistic for H0 : Interaction e®ect of A and
B is zero:
df = N ¡ I
F =
Simultaneous Con¯dence Intervals for Mean Differences:
s
1
1
¤¤
(¹
xi ¡ x
¹ j ) § t sp
+
; df = N ¡ I
ni
nj
Chapter 13: Two-way ANOVA
Factors: Two factors A and B, factor A has I levels, and
factor B has J levels
Two-way ANOVA model:
xijk = ¹ij + "ijk ;
for i = 1; : : : ; I; j = 1; : : : ; J and k = 1; : : : nij :
Pooled-sample variance:
P
(nij ¡ 1)s2ij
= MSE
s2p = P
(nij ¡ 1)
Sum of squares (SS): SST = SSA + SSB + SSAB + SSE
Degrees of freedom (DF):
4
MSAB
;
MSE
df = ((I ¡ 1)(J ¡ 1); N ¡ IJ):